sawfish: removing it switches alternatives to auto mode???
I had set up fvwm2 as x-window-manager (using update-alternatives) but sawfish gets repeatedly set up as default window manager which is very annoying becuase it causes window manager to exit when I try to restart it. while testing the problem I have found out that when I apt-get remove sawfish somebody sets the x-window-manager to auto mode, I get following message: Removing sawfish ... Removing manually selected alternative - switching to auto mode dpkg - warning: while removing sawfish, directory `/var/lib/sawfish' not empty so not removed. I guess it is caused by the following command in prerm script: update-alternatives --remove x-window-manager /usr/bin/sawfish is it a bug in update-alternatives? IMO it shouldn't switch back to auto mode when I remove alternative (which is not currently selected). Should I file a bug? Or should sawfish use something else? I checked few other WMs (wmaker, fvwm) and they use the same command... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Mozilla Mail & Microsoft Outlook
Ron Johnson wrote: On Thu, 2003-10-16 at 15:51, Alan Chandler wrote: On Thursday 16 October 2003 02:45, Frederico Rodrigues Abraham wrote: Hi. Has anyone tried/succeeded in importing messages from Microsoft Outlook (.pst files) to Mozilla Mail? Does anyone have any idea on how to do this? The prefered way is to set up an imap server on a linux box, and set outlook up to talk to it. You can then copy all the contents of a pst file (including directories) onto the imap server. Note that configuring an IMAP server isn't trivial, and documentation is sketchy, unless you want to spring for the O'Reilly book. depending on which one. cyrus is pretty complicated to set up (mostly because of very weird or missing error messages) but uw-imap is trivial to set up - you install it and that's pretty much it. So if it's IMAP just to get stuff out of outlook I'd say uw-imap makes sense... (in fact IMO there's really no excuse to not have imap:-) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Mozilla Mail & Microsoft Outlook
Ron Johnson wrote: On Thu, 2003-10-16 at 16:51, Alan Chandler wrote: ... When I did it, "apt-get install courier-imap" just about did all I needed. I certainly don't remember having to do any other playing about to get it to work. Obviously if you want to use it to receive mail into the Maildirs that it uses you have to do something with an MTA, but thats a different issue. But isn't that what make IMAP really useful (for a "home LAN" user)? I.e., getting fetchmail & SpamAssassin & exim|postfix to feed the mail to IMAP, so that you can use Evo, WebMail, KMail, etc. exactly, I'd never go back (to not using imap)! also it's easy to check email remotely (imaps), no worries about conflicts (when using multiple clients at the same time) etc... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: OT - Programming Languages w/o English Syntax
Monique Y. Herman wrote: On Fri, 17 Oct 2003 at 19:01 GMT, Ron Johnson penned: On Fri, 2003-10-17 at 12:29, Monique Y. Herman wrote: You're right; the anglo-centric nature of most programming languages is distressing. It would be fun to code in a language based on a totally Distressing What an over-reaction. Guess what? When French/German/Chinese/Spanish/Portuguese/Japanese Computer Scientists decide to write a programming language in their own native language, there will be programming languages in those languages. But then, why did Niklaus Wirth use English key words, even though he is Swiss/German? Distressing was the wrong word. I personally find variety interesting, and using a language with a different natural-language origin would be entertaining. Then again, I'm pretty good with languages, so I might be in the minority there. Of course you're right that languages based on various natural languages will exist when people write them. That's a tautology. I was mourning the fact that it seemingly hasn't happened. Maybe if I had half a clue about what my ideal programming language would contain, I'd go about writing one with non-English keywords, just to entertain myself. english has a fairly simple a regular grammar so it's fairly easy to create english based programming language - the basic control structures are pretty much english sentences. This would be fairly hard todo in other languages that has more irregular grammar (the ones I know anything about have a lot more complicated/irregular grammar). On a personal level, I find "local color" interesting, and the ever-more-prevalent assumption that everyone can or should speak English saddens me, for the same reason that it saddens me when a Walmart or a Starbucks puts a local storefront out of business. IMO it's not the same. I don't think I would like to program in slovak based programming language (even though it's my native language). So, in the spirit of "think globally, act locally," I try not to assume that everyone speaks English, and I try to buy stuff from friendly, helpful local stores rather than saving five bucks by buying from Walmart. sometime it's better to do what makes sense, instead of trying to push one or another agenda that doesn't really have too much to do with the issue. I am all for local stores but I want my phone to work everywhere (i.e. not everything should be globalized, not everything should be localized (IMO)) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: OT - Programming Languages w/o English Syntax
Monique Y. Herman wrote: On Fri, 17 Oct 2003 at 22:37 GMT, Erik Steffl penned: english has a fairly simple a regular grammar so it's fairly easy to create english based programming language - the basic control structures are pretty much english sentences. This would be fairly hard todo in other languages that has more irregular grammar (the ones I know anything about have a lot more complicated/irregular grammar). Hrm. German and Latin are much more regular than English. French is, too, iirc. English has a *lot* of irregularity. german is regular? with each word changing depending on how it's used in sentence (case)??? gender being pretty much random? what are you talking about??? in english there are few cases of irregularity (past tense/past participle of some verbs, few words have non-standard way to create plural and that's pretty much it). each words has at most few forms, easily recongizable (as in: the forms are created in same way for almost all the words). and the structure of the sentence is pretty simple as well. compare that to german where each words has number of forms (depending on what it relates to), and these forms are created in different ways for different words. example: in english, if I know the verb (one word) I can pretty much use it in a sentence. how many forms of each verb in german do you need to know to be able to use it in a sentence? erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: OT - Programming Languages w/o English Syntax
Don Werve wrote: On Fri, Oct 17, 2003 at 03:37:33PM -0700, Erik Steffl wrote: english has a fairly simple a regular grammar so it's fairly easy to create english based programming language - the basic control structures are pretty much english sentences. Actually, English grammar is a nightmare to behold; there is no nightmare? try to learn (if you don't know already) e.g. slovak (or any of the slavic languages). consistent method of handling verb conjugations, and the structure of a sentence is integral to its meaning; you can't just randomly move words around in an English sentence and expect things to work. The way a and that's what's simple about it: you have vocabulary and structures of sentences and that's it. very simple. vocabulary: for each word you only need the word, possibly few irregularities sentence structures: relatively few compare to e.g. slovak: each words is used in number of forms, depending on relationships to other words (case, gender) and what you want to express (in other words: a lot more than the basic meaning is/can be encoded in a word). sentence structure is very flexible, words can be in different order which gives sentence slightly or very different meaning... you have to know gender for each noun (well, unless the noun means something that is explicitly male or female) computer works at the low level (e.g., assembler and/or machine code) is actually much more similar to Japanese, where you have an action and the associate data stapled together in pairs, much like Japanese words are (nominally) paired with particles. it's possible that there are natural languages that are even better suited for computerization... The only reason that English-esque languages are prevalent is that, in the early days, most of the programmers were native English speakers, and as such, wrote tools and compilers that best fit their native linguistic models. If computerdom had started in Germany, then I'd wager that we'd see more languages which used a German grammatic style. not so sure... there was significant computer related research/business done in non-english speaking countries and yet there was no push towards computer languages that are not english based... (and take a look at how many programming languages there are). even people who do not speak english prefer english based computer languages (as far as I can tell). a lot of it is inital momentum but I don't think that's the only reason... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: way-OT: regularity of german v. english [was: Re: OT - Programming Languages w/o English Syntax]
Nori Heikkinen wrote: on Sun, 19 Oct 2003 04:10:38AM -0700, Erik Steffl insinuated: Monique Y. Herman wrote: On Fri, 17 Oct 2003 at 22:37 GMT, Erik Steffl penned: english has a fairly simple a regular grammar so it's fairly easy to create english based programming language - the basic control structures are pretty much english sentences. This would be fairly hard todo in other languages that has more irregular grammar (the ones I know anything about have a lot more complicated/irregular grammar). Hrm. German and Latin are much more regular than English. French is, too, iirc. English has a *lot* of irregularity. german is regular? more so than english, yes. with each word changing depending on how it's used in sentence (case)??? that's quite regular -- it's called declension, and is well-documented in any introductory german text. ok, let me give you a random word, let's say 'xxx' how much information do you need to use it in german? how much information do you need to use it in english? gender being pretty much random? that has nothing at all to do with the grammar -- you're talking about it doesn't matter what it is. I was claiming that german language is a lot more complex than english... grammar is part of it... genders are part of it... the lexicon. the gender of german nouns is as arbitrary as the phonemes that make up english words -- both have some historical background, but none may make any sense. both are just items to be memorized when learning the language -- just as we map "fork" to our concept of that thing with tines we use to eat broccoli, germans map "die Gabel" onto the same thing -- a word, and a gender to go with it. same deal. or, in other words, in german you need more information about the word in english there are few cases of irregularity (past tense/past participle of some verbs, few words have non-standard way to create plural and that's pretty much it). each words has at most few forms, easily recongizable (as in: the forms are created in same way for almost all the words). again, lexicon. this point has nothing to do with the "regularity" of language. ok, it makes understanding the language a lot harder, because there are a lot more rules that you need to apply, each (most?) having exceptions etc. so for each word you not only need the word but all its forms (some of them can be derived based on rules, but how do you know which ones?) and the structure of the sentence is pretty simple as well. clearly, you've never tried to map it out. go on, then, i dare you -- write me a regular grammar that can express the grammar of english. of course, you can create various complex and ambiguous sentences in english, the point is that you can take few forms of sentences and have a working language (that's pretty much what BASIC (talking about programming language) is). compare that to german where each words has number of forms (depending on what it relates to), declension naming it doesn't make it simpler and these forms are created in different ways for different words. all part of the lexicon. all making language a lot harder because you need a lot more infromation about each word. example: in english, if I know the verb (one word) I can pretty much use it in a sentence. how many forms of each verb in german do you need to know to be able to use it in a sentence? a root form (lexical); a knowledge of its behavior (also lexical); the basic rules for declension (a regular part of grammar). answer: one. not true. you're joking. you need to know the word, and depending on the word you need to various bits of info: gender, which rules of declesion to use (or specific forms for words then do not follow general rules)... and when using words you need to know how they related to other words... you need to know gender of those other words... etc. in english the words stay pretty much unchanged and the grammar is defined by structure of the sentence. in german the grammar is defined by changing the words, often according to general rules but fairly often not following the rules... think about it: when learning english the only challenge is to learn how to pronounce words (and learn irregular verbs). you built vocabulary by learning words, where you pretty much only need to remember the word itself (in its basic form). while when learning german... I don't even want to think about it. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: way-OT: regularity of german v. english [was: Re: OT - Programming Languages w/o English Syntax]
csj wrote: On Sun, 19 Oct 2003 12:38:45 -0700, Erik Steffl wrote: [...] think about it: when learning english the only challenge is to learn how to pronounce words (and learn irregular verbs). you built vocabulary by learning words, where you pretty much only need to remember the word itself (in its basic form). while when learning german... I don't even want to think about it. Because everybody from the poor war orphan "Hey, Joe, eat!" to the UN Secretary General speaks it, English has become a rather tolerant language. But if the same standard for proper German is applied to what one considers proper English, then yes, German is easier to learn. It's a purer, therefore more consistent language, than the French-infected English. purity has nothing to do with it (not sure what you mean by pure). not sure what your agenda is. english is a a lot simpler than german, the usage of words is simple, the grammar is simple. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: way-OT: regularity of german v. english [was: Re: OT - Programming Languages w/o English Syntax]
Nori Heikkinen wrote: on Mon, 20 Oct 2003 01:40:19PM +0200, David Jardine insinuated: On Mon, Oct 20, 2003 at 12:56:37AM -0700, Erik Steffl wrote: csj wrote: [...] Because everybody from the poor war orphan "Hey, Joe, eat!" to the UN Secretary General speaks it, English has become a rather tolerant language. But if the same standard for proper German is applied to what one considers proper English, then yes, German is easier to learn. It's a purer, therefore more consistent language, than the French-infected English. purity has nothing to do with it (not sure what you mean by pure). good point -- languages by definition evolve, and the notion of a "pure" language is utterly ridiculous and meaningless. not sure what your agenda is. english is a a lot simpler than german, in what sense? to learn? to master? to write basic sentences in? to write novels in? to read novels in? in all of these senses. but my original claim was that english is better suited to be a computer language than lot of other natural languages. the two are apples and oranges, my friend, especially when you're dealing with something that no one can have an objective point of view on, given different native languages. ??? you can measure how much information you need to understand/parse each of the languages... that doesn't have anything to do with language being native or not... and, contrary to popular belief, you can easily compare apples and oranges... the usage of words is simple, the grammar is simple. Depends what you mean by purity. By European language standards it's fairly pure in the sense of not being cluttered up with things like redundant inflections, but this is probably because it is impure in the sense of having been knocked around by neighbouring languages and dialects until there's not much left of it apart from what's really necessary to communicate. you're kidding, right? if i read you right, you're stating that note that the 'you' above and the 'you' far above refer to different people "there's not much left of [English] apart from what's really necessary to communicate"? on the contrary -- it's one of the richest, least threadbare languages there is! erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: update in sid has killed gnome-terminal
Marc Wilson wrote: On Mon, Oct 20, 2003 at 05:09:19PM -0400, TR wrote: I just did an upgrade in a machine running sid and after that can't star a gnome terminal anymore. Yes, and certainly you're going to get LOTS of help with that problem, given this EXTREMELY informative report you've made. Why, I'll just bet that your problem jumps right out and begs to be solved. Feh. Why do these people think they should be running unstable? perhaps because it's the only usable debian distro/version? the stable is too old for desktops... testing is even worse than unstable... orignal posters: hope you get the hint - provide us with enough information about your problem (exact error message etc.) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: way-OT: regularity of german v. english [was: Re: OT - Programming Languages w/o English Syntax]
Nori Heikkinen wrote: on Mon, 20 Oct 2003 11:53:34AM -0700, Erik Steffl insinuated: Nori Heikkinen wrote: the two are apples and oranges, my friend, especially when you're dealing with something that no one can have an objective point of view on, given different native languages. ??? you can measure how much information you need to understand/parse each of the languages... that doesn't have anything to do with language being native or not... my point was that depending on your native language, different aspects will seem natural to you, whereas others will appear radically different and therefore harder. depending on how different one's native language is from the target language, certain people can find a language very difficult to learn, whereas others with closely related languages could master it much faster than others. for instance, it is easier for speakers of english to master a romance language than something much more remotely related, like mandarin. 1 - the thread is about why english is used as base for programming languages so it does not matter that much how easy it is for particular person with particular background to learn (I was claiming that subset of english is easier for computers to process (than lot of other languages) since grammar is very simple) 2 - e.g. russian is fairly close to slovak so it was fairly easy for me to learn yet I wouldn't say it's easy/simple to learn. it's not hard for me to recognize that english is much simpler than russian even though english is very different from slovak. you can measure the information needed to define/understand grammar/language erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: way-OT: regularity of german v. english [was: Re: OT - Programming Languages w/o English Syntax]
Nori Heikkinen wrote: on Sun, 19 Oct 2003 12:38:45PM -0700, Erik Steffl insinuated: ... of course, you can create various complex and ambiguous sentences in english, the point is that you can take few forms of sentences and have a working language (that's pretty much what BASIC (talking about programming language) is). you can do that in both languages. let's say you have a function called isRed(x) (returns true if x is red). Now how would you call this function in german? it would never be in agreement with all possible x (grammatically). not sure if this is the best example - perhaps in this case it would be acceptable to use istRot, regardless of gender of x. point is you would run into problems like this trying to use german, you would very rarely come up with problems of this nature in english... ... remember the word itself (in its basic form). while when learning german... I don't even want to think about it. have you ever? ein bisschen... but what does it matter? it's just a random example, I could as well use slovak (but then less people would understand examples). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Gender in language (was Re: way-OT: regularity of german v. english [was: ])
Ron Johnson wrote: ... Being a native speaker of American, I've always wondered - What is the purpose of "gender" in grammar/language? as far as I can tell there's no purpose (not a linguist but my native language has genders, can't find any reason other then that it has genders:-) - Is it only the European/Latinate languages that have the gender concept? lot of them has the same gender concept: all slavic languages (russian, ukrainian, polish, slovak, czech, bulgarian, most (all?) languages in former yugoslavia etc.), I know spanish (more relevant since you ask about latin languages) has genders... btw your european/latin seems to imply that european languages are latin languages, that's not the case... - Why English doesn't have gender, since it's predecessor, German, does have gender? looks like a lot of unneccessary stuff was removed from english language (last century or two?), as far as I can tell it's because it's used as a non-native language for pragmatic purposes (i.e. as long as the message gets accross it's all good:-) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: classic deficiancy in both windows and linux ?
Joyce, Matthew wrote: Hey, Linux is Perfect!!! You must be an Evil Windows Troll!!! But seriously: On Thu, 2003-10-23 at 17:08, James D. Freels wrote: Can you figure out a way to get a listing of a directory (folder in Windows) and print it, without resorting to command prompt ? What's wrong with the command line? Is "ls -l" too geeky looking for PHBs, or are you nervous/unsure at the command line? Or, heaven forbid, will Untrained Users have to do it, and Linux is too difficult? I have to agree, a simple 'export listing' on the right click or tools menu would be nice, check boxes for what attributes to include, sort order, recurse y/n, humanise units y/n, include totals y/n. one could also set up a web server (e.g. apache), root document directory would be root of file system, indexes auto-generated in all directories... (NO default index like index.html etc.) then just browse and print from browser:-) it's kinda funny that while e.g. konqueror makes it possible to browse file:// it doesn't allow printing (at least the version I just tried it in) erik (yes, I'm just kidding...) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Gender in language (was Re: way-OT: regularity of german v. english [was: ])
David Palmer. wrote: ... First in were the Gaels (Irish) through Skandinavia, then the Icenii Brythonics (which is where 'Briton' and then 'Britain' came from) and some lesser tribes, such as the Manx. But none of these spoke German, either high, middle or low. Germany as a territory was defined later. The term 'Germanic' is a term applied later to describe the general area of issue, but it is only approximate, Czechoslovakian would be just as appropriate. not sure what you mean but: 'germanic' is at least somewhat relevant, (former) czechoslovakia on the other hand is completely different beast (inhabited (mostly) by slavic people (czechs and slovaks)) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Gender in language (was Re: way-OT: regularity of german v. english [was: ])
Pigeon wrote: On Wed, Oct 22, 2003 at 10:54:24PM -0500, Ron Johnson wrote: Being a native speaker of American, I've always wondered - What is the purpose of "gender" in grammar/language? Argh, this does my head in too. Especially when you come across things like all the words for female genitals in lots of languages having the masculine gender. Work that one out. Also, what do the advocates of "gender-neutral" language do in German? And what do they do in French? not sure about those countries but in slovakia (with 'genderic' language) there is no such thing as gender neutral. since everything has gender it doesn't really stand out when something has gender when related to person - in english it feels 'special' when you refer to something as he or she, it sticks out so you find people who are bothered by it. same in much less words: mu erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Gender in language (was Re: way-OT: regularity of german v. english [was: ])
Nori Heikkinen wrote: on Fri, 24 Oct 2003 12:52:10AM +0100, Pigeon insinuated: ... Argh, this does my head in too. Especially when you come across things like all the words for female genitals in lots of languages having the masculine gender. Work that one out. yeah, or a fork being feminine, spoon being masculine, and knife being neuter in german. lots of weird ones like that. why do you find it weird? genders are pretty much random, any gender for any word is not more or less weird... (except for when it refers to something that has explicit gender and grammatical gender doesn't agree with real gender, so the example Pigeon gave is somewhat weird, but not much - body parts are treated as things and have random gender) I mean in slovak the genders are: fork - feminine, spoon - feminine, knife - masculine. is that more weird? less weird? Also, what do the advocates of "gender-neutral" language do in German? And what do they do in French? what do you mean by advocates of gender-neutral languages? people who think that nothing in english should be gendered? i know in italian, in which a male profesorr is "il professore" and a female one has traditionally been "la professoressa", there's a movement to stop appending the "-essa" affix, in an effort to ungender the language a bit. however, there's a very important distinction to be made here -- that of linguistic gender -- which has NOTHING TO DO with human sex (m/f) -- and human gender, or sex. there are only two genders in lots of latinate languages (french, spanish, italian, portuguese, i think catalan, and likely rumanian -- so, i guess all of them), and they were at one point termed "masculine" and "feminine" by some asshole who wished to confuse all future students of language wondering why a girl is neuter in German. really, these genders are just categories that words fit into -- they could just as easily be called the reverse, or called "red" and "green" or something. not really because pretty much everything that really is female is referred to using feminine gender and what is male is reffered to using masculine gender. You can find examples where females or males are referred to using neuter gender but I don't think there are examples where oposite gender is used (feminine for males or masculine for females) but you're right that for things that do not have explicit sex the gender is pretty much random and doesn't _mean_ anything. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: way-OT: regularity of german v. english [was: Re: OT - Programming Languages w/o English Syntax]
Nori Heikkinen wrote: On Wed, Oct 22, 2003 at 06:47:13PM -0700, Erik Steffl insinuated: Nori Heikkinen wrote: on Sun, 19 Oct 2003 12:38:45PM -0700, Erik Steffl insinuated: ... of course, you can create various complex and ambiguous sentences in english, the point is that you can take few forms of sentences and have a working language (that's pretty much what BASIC (talking about programming language) is). you can do that in both languages. let's say you have a function called isRed(x) (returns true if x is red). Now how would you call this function in german? it would never be in agreement with all possible x (grammatically). how so not? istRot( dieGabel ) istRot( dasMesser ) istRot( derLoeffel ) how are any of those less grammatical than their english equivalents: isRed( theFork ) isRed( theKnife ) isRed( theSpoon ) ?? not sure if this is the best example nope, guess not. try again if you're trying to make a coherent argument. - perhaps in this case it would be acceptable to use istRot, regardless of gender of x. in fact, in german, adjectives only agree with nouns if they IMMEDIATELY PRECEDE the noun. so you have "die Gabel ist rot" (the fork is red), but "die rote Gabel" (the red fork) -- note that the feminine "-e" only applies to the adjective "rot" in the case in which it immediately precedes the noun. that's what I was confused about. doesn't Rot (in function name) immediately precede the noun? shouldn't the function name be istRote(gabel)? I guess you could argue that even though it's written istRote(gabel) it really is ist gabel rot (because that would be proper question) now, think of an example in which you encounter anything remotely like full sentence structure in code, and try to apply this. good luck. what about if the function name is verb? what if it's not a question? e.g. getRed(apple), or make(red, apple) etc. english is like lego, yes there are some pieces that change shape etc. but it consists mostly of bricks and brick like pieces. german (and lot of other languages) is more like putty - you mold things together. the lego-like structure of english makes it easier to create a computer language... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: way-OT: regularity of german v. english [was: Re: OT - Programming Languages w/o English Syntax]
Nori Heikkinen wrote:
on Fri, 24 Oct 2003 12:11:49PM -0700, Erik Steffl insinuated:
Nori Heikkinen wrote:
On Wed, Oct 22, 2003 at 06:47:13PM -0700, Erik Steffl insinuated:
Nori Heikkinen wrote:
on Sun, 19 Oct 2003 12:38:45PM -0700, Erik Steffl insinuated:
...
of course, you can create various complex and ambiguous sentences in
english, the point is that you can take few forms of sentences and
have a working language (that's pretty much what BASIC (talking
about programming language) is).
you can do that in both languages.
let's say you have a function called isRed(x) (returns true if x is
red). Now how would you call this function in german? it would never
be in agreement with all possible x (grammatically).
how so not?
istRot( dieGabel )
istRot( dasMesser )
istRot( derLoeffel )
how are any of those less grammatical than their english equivalents:
isRed( theFork )
isRed( theKnife )
isRed( theSpoon )
??
not sure if this is the best example
nope, guess not. try again if you're trying to make a coherent
argument.
- perhaps in this case it would be acceptable to use istRot,
regardless of gender of x.
in fact, in german, adjectives only agree with nouns if they
IMMEDIATELY PRECEDE the noun. so you have "die Gabel ist rot" (the
fork is red), but "die rote Gabel" (the red fork) -- note that the
feminine "-e" only applies to the adjective "rot" in the case in which
it immediately precedes the noun.
that's what I was confused about. doesn't Rot (in function name)
immediately precede the noun? shouldn't the function name be
istRote(gabel)? I guess you could argue that even though it's
written istRote(gabel) it really is ist gabel rot (because that
would be proper question)
you just answered your own question here. the implied sentence from
"istRot( gabel )" is "ist die Gabel rot?" -- and as you noted, in the
implied sentence, 'rot' does not immediately preced 'Gabel'. note
that the exact same word-shuffling is necessary from *english* code to
an *english* grammatical sentence -- "isRed( fork )" --> "is the fork
red?".
you just illustrated the point i've been trying to make here -- in the
above two examples, neither english nor german code makes a
grammatical sentence. both imply one, sure -- but that doesn't mean
either language is better suited to code than the other.
now, think of an example in which you encounter anything remotely like
full sentence structure in code, and try to apply this. good luck.
what about if the function name is verb? what if it's not a question?
e.g. getRed(apple), or make(red, apple) etc.
boolisch bekommRot( Apfel ) { }
mach( rot, Apfel );
what's different from English about those two? again, the implied
sentences are clear; the underlying code is equally as ungrammatical
in both.
shoudn't rot in these cases follow gender of subject?
are you arguing that while in german language there has to be
agreement in gender of verb/adjective and noun when you use german in
computer language this need suddenly disappears?
erik
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: way-OT: regularity of german v. english [was: Re: OT - Programming Languages w/o English Syntax]
Nori Heikkinen wrote:
on Sat, 25 Oct 2003 11:05:22AM -0700, Erik Steffl insinuated:
Nori Heikkinen wrote:
on Fri, 24 Oct 2003 12:11:49PM -0700, Erik Steffl insinuated:
Nori Heikkinen wrote:
now, think of an example in which you encounter anything remotely like
full sentence structure in code, and try to apply this. good luck.
what about if the function name is verb? what if it's not a question?
e.g. getRed(apple), or make(red, apple) etc.
boolisch bekommRot( Apfel ) { }
mach( rot, Apfel );
what's different from English about those two? again, the implied
sentences are clear; the underlying code is equally as ungrammatical
in both.
shouldn't rot in these cases follow gender of subject?
no. the implied sentence from the latter example above, were one to
translate it into german prose, would be "mach der Apfel rot" -- "make
the apple red." note that the adjective follows the noun, not
precedes, so the rule does not apply.
are you arguing that while in german language there has to be
agreement in gender of verb/adjective and noun when you use german
in computer language this need suddenly disappears?
no.
well, so far that's what you did. I admit my examples are not very
good since my knowledge of german is fairly limited but YOU should be
able to come up with better examples, instead you're just arguing for
sake of argument (it seems)
i'm arguing that _neither_ english _nor_ german is perfectly suited to
code, since one needs to do some translation to get the sentence into
the form in which a human would say it.
this is a weird thing to argue (it's obvious). and certainly not
relevant to the discussion. what I said is that english with its lego
like nature is _better_ suited as a starting point...
in other words - just because neither is _perfectly_ suited to code
doesn't mean that one of them is not better suited as base for
programming language...
on top of that, i'm arguing that _no_ language fits this bill. think
about it -- if there were a human language that could be described as
easily as a computer language, we would be able to express that human
language as a finite state automaton, thereby solving the language
problem, and the whole of AI with it. there's a reason we write in a
formal, simplified language when we code!
that's obvious but the question is - why is english used so often as
base for programming languages? I argued that the nature of english is
responsible, in part, for this...
erik
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: netiquette: CCing on lists
Richard Lyons wrote: On Tuesday 28 October 2003 20:30, Monique Y. Herman wrote: [...] Hrm.. Does debian-user not set the reply-to to the list, or is this my [...] Apparently not. I wonder why not. It would surely be a good idea - for those using simpler mail clients. I use kmail and filter lists direct to their own folders, where I set the reply-to-list address to try to prevent myself making mistakes... À propos, I've been thinking of giving mutt a try: can it do that too? yes, you can set up a list of list email addresses and then hit (IIRC) L to reply to the list erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
mozilla doesn't open last page visited (new window)
I have the preferences set up to open last page visited but for some reason it always opens the home page. this is happening for quite some time, not sure which version, but definitely few 1.4 and current 1.5-2 pacakge (unstable). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
mozilla mail freezes on start
last mozilla 1.4 worked OK, now that I upgraded to 1.5-2 when I open mail window it just freezes (all mozilla windows are completely frozen, not updated at all). I have an IMAP server (imaps) and few nntp servers configured. I can connect to IMAP server (cyrus) using other clients - so far I tried mutt and evolution, both work. any ideas on how to troubleshoot this or what's the problem? TIA, erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
SOLVED Re: mozilla doesn't open last page visited (new window)
On Sat, 2003-11-01 at 00:38, Erik Steffl wrote: > I have the preferences set up to open last page visited but for some > reason it always opens the home page. > > this is happening for quite some time, not sure which version, but > definitely few 1.4 and current 1.5-2 pacakge (unstable). right after sending email I noticed that mozilla now has different start option for mozilla start, new window and new tab, I was only setting the one for mozilla start, not for new window... of well... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [ot] Linux gender in French
On Sat, 2003-11-01 at 18:52, Tom wrote: > On Sat, Nov 01, 2003 at 07:42:03PM -0700, Wesley J Landaker wrote: > > > (I have a theory, but I don't want to influence what you say). > > > > But, now I'm curious... what is your theory? > > Funny timing, I just said it in my previous email a couple of minutes > ago. My (rather facile) theory is that it's freudian. > > My guess: > Round things seem feminine, angular things seem masculine. > Things that jut out (gas nozzle) seem masculine; things that recede (gas > tank) seem feminine (okay, that's TOO facile). > Things that receive actions seem feminine (button); things that cause > action (lever) seem masculine. > > I'm just guessing :-) I don't know about french but in other languiages that have gender it seems pretty random (unless the subject has explicit gender), often you can use words of different genders to name same entity, there is no feminine or masculine feel to the gender of things (see exception above). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Stable Debian == obsolete??
Chema wrote: ... So the network installation of Sarge is my new bet. > But I want to know, how really unstable is it? I don't > think most people could live with Woody, so is it test the most used distro? server: I'd go with stable desktop: I'd go with unstable (that's what I use) testing, unless it changed dramatically, is not for people to use (not sure what it is for (I know what it is claimed to be for but in reality it does not work that way at all)). perhaps less bugs get to testing than to unstable but they also take longer (sometime a lot longer) to fix... in the end you get the worst of both worlds unstable is generally very stable, there are some problems when big changes are being made (major gcc, X, kde, gnome upgrades etc.). If something breaks it is usually fixed fairly soon. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Progress meter on copying
Mike Dresser wrote: On Mon, 3 Nov 2003, Rus Foster wrote: Hi All, Does anyone know of a nice way of being able to show a progress meter on copying a large file from one part of the disk to another. I tried scp localfile localfile2 but scp calls cp. you can call it like this: scp someFile hostname:scpedFile then you have progress meter erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Hacked: .bash_history linked somewhere
Mike Egglestone wrote: Hi, My server was trojaned recently, not sure how. It looks like /bin/ps was modified or replaced with a trojan. The /root/.bash_history file is set to this: chsslx1:~# ls -la .bash_history -rw-r--r--1 root root0 Nov 7 05:31 .bash_history and I can't edit it or delete it. It looks like its linked somewhere: chsslx1:~# rm .bash_history rm: remove write-protected file `.bash_history'? y rm: cannot unlink `.bash_history': Operation not permitted First off, nothing to much was compromised. Only /etc/samba/* was wiped. (There may be more stuff but haven't detected yet) It seems that the only way to recover is to re-install? Is there a way to find out why the .bash_history is linked in someway? it wasn't linked in a way you think. (generally) every file has at least one hard link to it, it's a hard link and that's what you think of as the file. when you remove the file you call unlink and it removes the link, if it was last link the file is removed. hard links are the ones you see as files, soft links are the ones you see as link (when you do e.g. ls -l). for more info man ln erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: What is the password of "root" when first run after the installation of the base system!
Otto Wyss wrote: On Tue, Nov 11, 2003 at 10:48:16PM +0100, Otto Wyss wrote: ... Failing that, boot with the 'init=/bin/sh' parameter (so if you'd normally type 'linux' to LILO, say, type 'linux init=/bin/sh' instead) and run 'passwd root'. Unfortunately there is no boot prompt where I could enter anything. It just runs through until the login prompt. I guess I have to start it over. hit the left shift during boot process (right after POST, before kernel loads, I think you see LILO at the screen). You should get boot prompt. If you're not sure when to do this exactly try to hold down the shift or hit it repeatedly (holding it down does not work with some keyboards). once you have the boot prompt you can use Tab key to get the list of boot choices. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
openGL cannot do fullscreen in dual monitor config
just tried dual monitors with two card (geForce FX 5600 Ultra, nvidia drivers and ATI Radeon 9800, ati drivers), debian unstable, X 4.2.1: one monitor: openGL works fullscreen or windowed two monitors: openGL works in window, not fullscreen tested with xscreensaver hacks considering that these are two fairly different implementations of openGL (I think, they both provide openGL libs and their own way to set up two monitors) - is there something that I should set up to be able to run fullscreen openGL apps fullscreen? or is some way to set up dual monitors better than other when using fullscreen openGL apps? TIA erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: openGL cannot do fullscreen in dual monitor config
Marc Wilson wrote: On Tue, Nov 11, 2003 at 08:44:11PM -0800, Erik Steffl wrote: just tried dual monitors with two card (geForce FX 5600 Ultra, nvidia drivers and ATI Radeon 9800, ati drivers), debian unstable, X 4.2.1: The important thing... are you using dual-head, or are you using Xinerama? If you're using dual-head, you should have full OGL acceleration on both heads, subject to whatever their respective driver(s) provide or don't provide. If, on the other hand, you're using Xinerama or something that fakes X into *thinking* it has Xinerama (nVidia's TwinView does this, I dunno what ATi does), then you only get accelerated OGL on the primary head. Effects on the second head vary from simply being unaccelerated, to windows that render properly on the primary head being empty black rectangles on the second head. well, openGL worked in window on both monitors, just fullscreen wouldn't work (all black, no errors). I used TwinView with nvidia (and it seems that fvwm knew there are two monitors - when moving windows it would snap on the side of monitor as if it was edge), the ATI says they have single framebuffer (and it really looks like nobody's aware that there are two monitors), at lteast the mode I was testing. I guess it does not matter though, I just found out that I'll probably only use one monitor setup since it's 1920x1200 and openGL on Radeon only works if overall size is less then two thousand and something... Maybe I'll try the dual head setup for ATI (I think the two thousands limitation is for single frame buffer only). BTW in one monitor setup the openGL works both windowed and full screen. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: window manager recomendation
Micha Feigin wrote: Hello, Hoping this won't turn into a flame war, I am looking for recommendations for a window manager. I tried quiet a few but none seem to fit the bill yet. I need a window manager with the following - As lite as possible on memory (I heavily stress my laptop so I don't have much to spare). - Multiple desktops - A pop up menu application (don't need a panel) that has support for both the debian menus and a custom menu. - Hotkeys (mainly for maximize/minimize/desktop switch) - multiple desktops I like fvwm, it is fairly lightweigth (lot of stuff is in modules so if you don't need it just don't use it), not easy to configure (text files, but simple config is fairly simple), it has very good virtual screen support (you can drag windows from pager to current screen, move windows in pager, switch to different screens by moving mouse (if you want), by hotkeys or clicking on appropriate screen in pager, drag windows from one screen to another etc.), you can have keys for pretty much everything. You can have a panel or pop-up menu, your choice (and if you don't use panel it does not use any memory since it'sa module). Default debian config has menus (click root menu with each mouse button), you can easily change them and/or create your own menus and display them using either hotkeys or some mouse action etc. You can also define all the window decorations and what they do - how to minimize/maximize/iconify window, how to move resize window, how to lower/raise window etc. (e.g. I generally use no borders and resize windows by right clicking the title) the default look kinda sucks though... it can be changed completely (take a lok at www.fvwm.org) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Step by Steps??
Jeffrey W. Pearson wrote: Does anyone have step by steps for setting up a LAMP environment with Debian? Ive just moved from Red Hat to Debian. I don't seem to be able to find the locations of certain files needed. Right now Im stuck at trying to find the libphp4.so file. I compiled php from source and am not finding that file. if you know the name of the file and want to know which package this file is in goto debian.org, go to packages (left side menu) and use the last form to search for a file in debian packages. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Social Engineering. {was: Re: Opium [was: Re: freelance sysadmining -- superlong -- [WAS: "Red Hat recommends Windows for consumers"]]
Ron Johnson wrote: On Fri, 2003-11-14 at 15:12, David Palmer. wrote: ... These potentially highly dangerous individuals are confined to institutions known as 'research centres', and if non conforming are seen as a disruptive and undesirable element by the established social order, and are further relegated to the classification of 'terrorist'. Jeez, *I* got good grades in school, yet still (I think) think out- side the box, yet don't live in a "research centre", the established social order doesn't think I'm a disruptive and undesirable element, and hasn't yet classified me a "terrorist". conclusion: you have found a good place to hide :-) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
ATI Radeon 9800 freezes after using X :1
if I run another instance of X the system freezes when I try to switch back to the first one (completely, does not respond to keyboard, mouse or network connections). This only happens if I have two monitors (VGA and DVI). Any ideas what's the problem? system: debian unstable kernel 2.4 21-ac4 intel D865PERL MB ATI Radeon 9800 + ATI Radeon drivers additional info: I had similar experience with GeForce FX 5600 Ultra card - after I connected another monitor (to DVI connector) the system would freeze after switching to another X server. possibly related problem: when I switch to text console and back to X the colors are screwed up (not all of them). when I run X :1 there are some error messages about not being able to initialize DRI (X :0 can initialize DRI without problems, openGL programs work): (II) fglrx(0): VisualConfigs initialized (EE) fglrx(0): Failed to initialize UMM driver. (EE) fglrx(0): [drm] failed to remove DRM signal handler (II) fglrx(0): [drm] removed 1 reserved context for kernel DRIUnlock called when not locked (II) fglrx(0): [drm] unmapping 8192 bytes of SAREA 0xf8987000 at 0x40028000 (WW) fglrx(0): *** (WW) fglrx(0): * DRI initialization failed! * (WW) fglrx(0): * (maybe driver kernel module missing or bad) * (WW) fglrx(0): * 2D acceleraton available (MMIO) * (WW) fglrx(0): * no 3D acceleration available* (WW) fglrx(0): * * (II) fglrx(0): FBADPhys: 0xd000 FBMappedSize: 0x0800 (II) fglrx(0): -- (II) fglrx(0): | panel native mode is 1920x1200 | (II) fglrx(0): -- TIA erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Installation Help: SATA Drive
On Fri, 2003-11-14 at 17:15, Justin Burke wrote: > Hi All, > > I could use some help getting Debian installed on a new machine with a > SATA drive. I've downloaded CD images for both sarge and sid, and both > installation methods hang at the same point: > > "Loading kernel modules > > Detected module 'ide-probe-mod' for 'Linux IDE probe driver'" > > I have reason to believe that the system is hanging because of the SATA > drive. I think that I need to use a 2.6.0 kernel. Is this right? How do > I create an installation CD with a different kernel? you need a fairly new kernel for SATA, 2.4.21 iwth ac4 patches, I think. If you have a large SATA drive (>130GB) you also need libata5 patches from Jeff Garzik. I have 2.4.21-ac and manually applied libata5 patch (and 250 GB Matrox on intel D865PERL motherboard) Some of the newer kernels might work without ac or Jeff's patches, hard to tell. Check the linux kernel mailing list archives (search for SATA) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: ATI Radeon 9800 freezes after using X :1
On Sat, 2003-11-15 at 07:49, TR wrote: > Is the last version of Xfree in sid supporting ati radeon 9800? Last I > had checked it was not yet, so I have fireglx from the ati site running, > which of course has created a lot of conflicts everytime that I run an > update. no it doesn't. I downloaded latest X from xfree86.org (xserver and x modules) and it doesn't recogninze the card either. I downloaded the driver from ati website. It works fairly well except of when I try to start another X (:1), then it freezes machine when I try to go back to :0 erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
cyrus21-imapd: imapd killed by signal 7 (SIGBUS)
when I rebooted yesterday (first time after recent dist-upgrade) I got number of messages about cyrus db being currupted (and there was a huge number of cyr_* processes running), the messages recommended to recover the db, so I stopped cyrus21 and ran db3_recover (output below). then I started cyrus again, it does not complain about db anymore but when I try to read email (from mozilla or mutt) the imapd process dies, here's what syslog says: Jan 23 09:09:00 jojda cyrus/imapd[8145]: login: robota.dhs.org[207.214.64.208] erik plain+TLS User logged in Jan 23 09:09:09 jojda cyrus/imapd[8145]: skiplist: recovered /var/lib/cyrus/user/e/erik.seen (121 records, 12352 bytes) in 9 seconds Jan 23 09:09:09 jojda cyrus/imapd[8145]: seen_db: user erik opened /var/lib/cyrus/user/e/erik.seen Jan 23 09:09:10 jojda cyrus/imapd[8145]: open: user erik opened INBOX Jan 23 09:09:11 jojda cyrus/master[8139]: process 8145 exited, signaled to death by 7 here's the db3_recover output: jojda:/home/erik# db3_recover -v -h /var/lib/cyrus/db db_recover: Finding last valid log LSN: file: 17 offset 4636212 db_recover: Checkpoint at: [17][4633823] db_recover: Checkpoint LSN: [17][4633823] db_recover: Previous checkpoint: [17][4633244] db_recover: Checkpoint at: [17][4633823] db_recover: Checkpoint LSN: [17][4633244] db_recover: Previous checkpoint: [17][4633093] db_recover: Recovery starting from [17][4633244] db_recover: Recovery complete at Thu Jan 23 08:38:46 2003 db_recover: Maximum transaction ID 8013 Recovery checkpoint [17][4636358] db_recover: Recovery complete at Thu Jan 23 08:38:46 2003 db_recover: Maximum transaction id 8000 Recovery checkpoint [17][4636358] jojda:/home/erik# and here are the original errors: Jan 23 07:48:22 jojda cyrus/imapd[2990]: DBERROR db3: region error detected; run recovery. Jan 23 07:48:22 jojda cyrus/imapd[2990]: DBERROR: dbenv->open '/var/lib/cyrus/db' failed: DB_RUNRECOVERY: Fatal error, run database recover y Jan 23 07:48:22 jojda cyrus/imapd[2990]: DBERROR: init : cyrusdb error no similar bugs found... any ideas on what's going on? TIA! btw I'd appreciate if you'd cc me on reply this time, I can't read my account where I receive debian-user (because of the problems described above:-) erik __ Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now. http://mailplus.yahoo.com -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
FIXED, but why? WAS: Re: cyrus21-imapd: imapd killed by signal 7(SIGBUS)
it looks like the problem described below was fixed by running /usr/sbin/cyrreconstruct (I ran it after I noticed that imapd only died when trying to open inbox, not other folders). any ideas about what could have caused the problem? erik Erik Steffl wrote: when I rebooted yesterday (first time after recent dist-upgrade) I got number of messages about cyrus db being currupted (and there was a huge number of cyr_* processes running), the messages recommended to recover the db, so I stopped cyrus21 and ran db3_recover (output below). then I started cyrus again, it does not complain about db anymore but when I try to read email (from mozilla or mutt) the imapd process dies, here's what syslog says: Jan 23 09:09:00 jojda cyrus/imapd[8145]: login: robota.dhs.org[207.214.64.208] erik plain+TLS User logged in Jan 23 09:09:09 jojda cyrus/imapd[8145]: skiplist: recovered /var/lib/cyrus/user/e/erik.seen (121 records, 12352 bytes) in 9 seconds Jan 23 09:09:09 jojda cyrus/imapd[8145]: seen_db: user erik opened /var/lib/cyrus/user/e/erik.seen Jan 23 09:09:10 jojda cyrus/imapd[8145]: open: user erik opened INBOX Jan 23 09:09:11 jojda cyrus/master[8139]: process 8145 exited, signaled to death by 7 here's the db3_recover output: jojda:/home/erik# db3_recover -v -h /var/lib/cyrus/db db_recover: Finding last valid log LSN: file: 17 offset 4636212 db_recover: Checkpoint at: [17][4633823] db_recover: Checkpoint LSN: [17][4633823] db_recover: Previous checkpoint: [17][4633244] db_recover: Checkpoint at: [17][4633823] db_recover: Checkpoint LSN: [17][4633244] db_recover: Previous checkpoint: [17][4633093] db_recover: Recovery starting from [17][4633244] db_recover: Recovery complete at Thu Jan 23 08:38:46 2003 db_recover: Maximum transaction ID 8013 Recovery checkpoint [17][4636358] db_recover: Recovery complete at Thu Jan 23 08:38:46 2003 db_recover: Maximum transaction id 8000 Recovery checkpoint [17][4636358] jojda:/home/erik# and here are the original errors: Jan 23 07:48:22 jojda cyrus/imapd[2990]: DBERROR db3: region error detected; run recovery. Jan 23 07:48:22 jojda cyrus/imapd[2990]: DBERROR: dbenv->open '/var/lib/cyrus/db' failed: DB_RUNRECOVERY: Fatal error, run database recover y Jan 23 07:48:22 jojda cyrus/imapd[2990]: DBERROR: init : cyrusdb error no similar bugs found... any ideas on what's going on? TIA! btw I'd appreciate if you'd cc me on reply this time, I can't read my account where I receive debian-user (because of the problems described above:-) erik __ Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now. http://mailplus.yahoo.com -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Some myths regarding apt pinning
Adrian 'Dagurashibanipal' von Bidder wrote: On Fre, 2003-01-24 at 14:59, Adrian Bunk wrote: Since some people seem to thing apt pinning can solve all problems with outdated packages in stable I want to explain why this is wrong: apt pinning is good if you are running testing but need a package (e.g. a security update) from unstable. There are people that use apt pinning to install packages from unstable on a woody system. This is bad because nearly every installation of a package from unstable pulls a new libc6 and it's also possible that it pulls a new Perl and Python. Then some _very_ essential components of your system are upgraded to the potentially more buggy versions in unstable. apt-get tells you beforehand exactly what it's going to do. apt-listchanges even shows you the changelogs so you have a very late point of no return. I claim everybody who accidently upgraded perl deserves it. The only thing that could be better is perhaps that apt-get should display what it's going to install in terms of ... NEW packages ... perl/unstable or so. I often recommend apt pinning if somebody asks about installing woody but wanting newre packages. I'd expect that reading a man page and thinking about what one is going to do is something that everybody learns to do on a unixy system. but the point is that pinning is not very good because you either bring a number of important packages from unstable (libc6, perl etc) or you simply cannot use it. reading of the manual page and checking the apt-listchanges does not solve the problem. i.e. you recommend pinning, person reads the manpage, tries pinning and finds out that it was pretty much pointless excercise because it would upgrade large part of the system to unstable. or yet another wording: Adrian Bunk wasn't complaining about system actually upgrading packages but about system trying to upgrade packages. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Using VIA C3 and Woody?
Bill Moseley wrote: I want to build a very quiet and stable machine. Anyone using a VIA C3 based system with Woody? If so, what motherboard are you using? Any hardware issues? not sure which chipset it will be, some chipsets have problems with IDE, namely audio cd ripping, I know mine has the issue (lost interrupt, cd audio ripping slow, with BIOS upgrade some MBs work somewhat better): jojda:~> lspci 00:00.0 Host bridge: VIA Technologies, Inc. VT82C693A/694x [Apollo PRO133x] (rev c4) 00:01.0 PCI bridge: VIA Technologies, Inc. VT82C598/694x [Apollo MVP3/Pro133x AGP] 00:07.0 ISA bridge: VIA Technologies, Inc. VT82C686 [Apollo Super South] (rev 40) 00:07.1 IDE interface: VIA Technologies, Inc. VT82C586/B/686A/B PIPC Bus Master IDE (rev 06) 00:07.4 Host bridge: VIA Technologies, Inc. VT82C686 [Apollo Super ACPI] (rev 40) 00:09.0 SCSI storage controller: Artop Electronic Corp ATP865 (rev 02) 00:0b.0 Ethernet controller: D-Link System Inc RTL8139 Ethernet (rev 10) 00:0d.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8139/8139C/8139C+ (rev 10) 00:0f.0 VGA compatible controller: 3Dfx Interactive, Inc. Voodoo 3 (rev 01) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: voodoo3 troubles
Cameron Matheson wrote: Hi, I'm trying to get my voodoo3 working but i am having some serious problems... The computer boots up fine w/ the card in, and X works great w/ the "vesa" driver (other than no acceleration of course). If i try the "tdfx" driver, however, X goes crazy (the screen is bright shifting colors and the cursor is a gigantic black rectangle). After i kill X w/ ctrl+alt+backspace the console is ruined (looks a lot like X except in b&w) so i have to reset the computer. I have been trying to figure this out but so far no luck, i have attached my XFree86.log and my XF86Config-4 (and also output from dmesg). Here is the stuff that is probably significant although i am not sure: (II) LoadModule: "int10" (II) Loading /usr/X11R6/lib/modules/linux/libint10.a (II) Module int10: vendor="The XFree86 Project" compiled for 4.2.1.1, module version = 1.0.0 ABI class: XFree86 Video Driver, version 0.5 (II) TDFX(0): Softbooting the board (through the int10 interface). (II) TDFX(0): Primary V_BIOS segment is: 0xc000 (II) TDFX(0): Softbooting the board succeeded. Why is it softbooting my board? Isn't that only supposed to be done when you have two cards and one of isn't started at bootup by the bios? not sure what it is but I have the same stuff in my log (and I have only one voodoo 3 video card and it works) I think this part right here is the major bad stuff: (II) TDFX(0): Failed to set dac value, bypassing CLUT (II) TDFX(0): Failed to set dac value, bypassing CLUT (II) TDFX(0): Failed to set dac value, bypassing CLUT (II) TDFX(0): Failed to set dac value, bypassing CLUT don't have this though. are you sure you have tdfx kernel module compiled? here's my config file: ### BEGIN DEBCONF SECTION # XF86Config-4 (XFree86 server configuration file) generated by dexconf, the # Debian X Configuration tool, using values from the debconf database. # # Edit this file with caution, and see the XF86Config-4 manual page. # (Type "man XF86Config-4" at the shell prompt.) # # If you want your changes to this file preserved by dexconf, only make changes # before the "### BEGIN DEBCONF SECTION" line above, and/or after the # "### END DEBCONF SECTION" line below. # # To change things within the debconf section, run the command: # dpkg-reconfigure xserver-xfree86 # as root. Also see "How do I add custom sections to a dexconf-generated # XF86Config or XF86Config-4 file?" in /usr/share/doc/xfree86-common/FAQ.gz. Section "Files" FontPath "unix/:7100" # local font server # if the local font server has problems, we can fall back on these FontPath "/usr/lib/X11/fonts/misc" FontPath "/usr/lib/X11/fonts/cyrillic" FontPath "/usr/lib/X11/fonts/100dpi/:unscaled" FontPath "/usr/lib/X11/fonts/75dpi/:unscaled" FontPath "/usr/lib/X11/fonts/TrueType" FontPath "/usr/lib/X11/fonts/Type1" FontPath "/usr/lib/X11/fonts/Speedo" FontPath "/usr/lib/X11/fonts/100dpi" FontPath "/usr/lib/X11/fonts/75dpi" EndSection Section "Module" Load "GLcore" Load "bitmap" Load "dbe" Load "ddc" Load "dri" Load "extmod" Load "freetype" Load "glx" Load "int10" Load "pex5" Load "record" Load "speedo" Load "type1" Load "vbe" Load "xie" EndSection Section "InputDevice" Identifier "Logitech Cordless iTouch" Driver "keyboard" Option "CoreKeyboard" Option "XkbRules" "xfree86" Option "XkbModel" "logicordless" Option "XkbLayout" "us" EndSection # don't forget xmodmap -e "pointer = 1 2 3 6 4 5" # so that scrolling works (4 & 5 are the scrolling buttons in X) Section "InputDevice" Identifier "Logitech Cordless MouseMan Wheel" Driver "mouse" Option "CorePointer" Option "Protocol" "MouseManPlusPS/2" Option "Device" "/dev/psaux" Option "Emulate3Buttons" "false" Option "Buttons" "6" Option "ZAxisMapping" "5 6" Option "SendCoreEvents" "true" EndSection Section "Device" Identifier "3Dfx Voodoo3" Driver "tdfx" EndSection Section "Monitor" Identifier "SONY Multiscan 100sx" HorizSync 30-65 VertRefresh 50-120 Option "DPMS" EndSection Section "Screen" Identifier "Default Screen" Device "3Dfx Voodoo3" Monitor "SONY Multiscan 100sx" DefaultDepth 16 SubSection "Display" Depth 1 Modes "1280x1024" "800x600" "640x480" EndSubSection SubSection "Display" Depth 4 Modes "1280x1024" "800x600" "640x480" EndSubSection SubSection "Display" Depth 8 Modes "1280x1024" "800x600" "640x480" EndSubSection SubSection "Display" Depth 15 Modes "1280x1024" "800x600" "640x480" EndSubSection SubSection "Display" Depth 16 Modes "1280x1024" "800x600" "640x480" EndSubSection SubSection "Display" Depth 24 Modes "1280x1024" "800x600" "640x480" EndSubSection EndSection Section "ServerLayout" Identifier "Default Layout" Screen "Default Screen" InputDevice "Logitech Cordless iTouch" InputDevice "Logitech Cordless MouseMan Wheel" EndSection Section "DRI" Mode 0666 EndSection ### END DEBCONF SECTION erik -- To UNSUBSCRIBE, email to [EMAIL
Re: Suggest CD RW app ?
David selby wrote: Hello all, I have just installed a SAMSUNG DVD - CD RW drive, done the config I tried xcdroast, I like it but it insists on a 1024x768 screen which is a bit of a problem with my eyes / monitor. I just tried it and the window is about 800x570. I can make it bigger but not smaller, not sure what it depends on... font size? You might want to investigate this, xcdroast seems to be one of the better (best?) cd burners... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Suggest CD RW app ?
David selby wrote: Erik Steffl wrote: David selby wrote: Hello all, I have just installed a SAMSUNG DVD - CD RW drive, done the config I tried xcdroast, I like it but it insists on a 1024x768 screen which is a bit of a problem with my eyes / monitor. I just tried it and the window is about 800x570. I can make it bigger but not smaller, not sure what it depends on... font size? You might want to investigate this, xcdroast seems to be one of the better (best?) cd burners... erik I think I will investigate as you said, wierd, most apps arn't fussy on screen size. even weirder is that when I try to resize it it says 0x0 and grows from that (but cannot get smaller). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Install stalled
[EMAIL PROTECTED] wrote: Hi, I have spent several hours studying various Linux webpage and trying to install the OS, but have not had any successes. There is lots of good information and software, but I do not know where to begin. Can you help me install the Linux OS? If so what information is needed? you can start with knoppix (google for it), then install it on harddrive... that's probably the easiest way to install a debian based distro... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: XFree86 Config
Rob VanFleet wrote: On Wed, Sep 17, 2003 at 04:44:39PM -0400, Rishikesh wrote: I bought a new monitor and I want to reconfigure X. How can I get the debconf menus to configure X I got when I was installing debian. dpkg-reconfigure xserver-xfree86 should do it. depends on whether the XFree config is managed manually or not (it asks during install/upgrade) if it's managed manually: look for monitor section, update the frequencies (most probably no modelines required), it's very simple. the config filename is /etc/X11/XF86Config-4 or, if it does not exist, /etc/X11/XF86Config erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: ssh + X11
Rodrigo Valenzuela wrote: hi everyone: When i connect via ssh to my computers it goes fine, but when i tried 'startx ' it complains with 'X: user noy authorized to tun the x server'. Obviously the user can startx with no complain when sitting in front of the computer, so it is authorized. What is missing? Any comments? Pointers? ps: is this the best option in order to access my desktop remotely, and do all the stuff I'm used to? thanks in advance there's no need to start X on remote comupter, as long as you have X on local computer you can use ssh X forwarding to run X programs from remote machine and display them on local machine (ssh -X host.com, make sure that both client and server have X forwarding enabled) another option is vnc (it's basically an X server that you can view remotely, using vncviewer, the advantage is that if your connection goes down you can reconnect or if you want to connect from different location you can do tht as well) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: tar vs dpkg --get-selections for backup scheme
Benedict Verheyen wrote: Hi, my server setup is nearing completion and of course i want to save my hard work in case something goes wrong. Since i have problems in the past with various backup programs like mondo to name one, i want to keep it simple this time so a moronic me can understand whats happening :) 1) Method 1. Tar /etc, /root, /home, /boot, /var, /usr, ... and write it to cd's. Not sure how you can fit several tar to one cd but maybe one can use mkisofs to do this? To restore, you would have to boot with a rescue cd like Knoppix, partition the hd and restore the tars or install a base system and then restore the tars. I can see a problem with the restore from the cd's when a tar file, for instance var.tar.gz, spans more than 1 cd. I had some problems restoring my system once like that and i ended up with copying the different tars from the cd that made up the 1 big tar, joined them and then untarred them. Or isn't that a problem and is there another way to do that? 2) Method 2. First save the system status with dpkg --get-selections. Next backup /etc, /root, /home, /boot, /usr. I think you only need /etc for the system and then any user data (probably in /home), if you have custom kernel you need that as well (kernel + kernel config so that you can rebuild that). Then you can reinstall the system from scratch, install all the packages that were installed before, restore /etc (or most of /etc) and you're up and running... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: LILO problems,no boot, can I use Knoppix rescue to fix?
J F wrote: Repost with modifications, first one did make it to the list: I did an aptitude and did an update, and a few "g" keystrokes to download and install. Aptitude said something like: Warning, it was deleting modules for old version of kernel. Now the system won't reboot. THe screen says: LILO 22.5.6.1 Loading Linux ... This is from a Knoppix CD hard disk install. Any way, I can reboot using the Knoppix CD as a rescue disk and then edit /mnt/hda1/etc/lilo.conf . But when I run lilo, it says: # sudo su # lilo -C /mnt/hda1/etc/lilo.conf Fatal. Sorry, don't know how to handle device 0xf000 How can I use Knoppix to fix LILO boot? AFAIK you have to create new version of your /etc/lilo.conf where you replace the old paths by new ones (just like/etc/lilo.conf becomes /mnt/hda1/etc/lilo.conf, you have to map all of them in the same way) of course, do not change your original lilo.conf, you will need it in future... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Help Please.....
[EMAIL PROTECTED] wrote: * ***Hey, I'm about to buy *VIA C3M266 Mainboard *for Linux Debian Woody. Like you said all VIA motherboards, well almost all, should work fine. if you have any thing you wanna add to help me install debian on this motherboard please do, Little things can help too you know. * ***Thanks you so much for your help not sure what's the relationship, the MB with following via chipset wasn't able to do audio cd ripping (everything else worked) ABIT VH6-II (with VT82C693A/694x chipset) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Executing a Script every hour
Christof Hurschler wrote: Hi, I've scanned the cron man pages, but it seems that cron is only set up to do daily, weekly, and monthly jobs in Debian. Is there a simpe way to have a script execute at shorter time intervals. I'd you can specify the values separated by commas, example for minutes: 00,15,30,45 * ... would be executed every 15 minutes there are other ways to specify when it should be run, read man 5 crontab carefully (not man crontab), 5 is section traditionally used for file formats erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: ps axu|grep grep doesn't output always
Dan Jacobson wrote: Just curious, why doesn't this make output every run: $ ps axu|grep grep jidanni 1163 0.0 0.2 2376 608 pts/1R02:26 0:00 grep grep $ ps axu|grep grep $ ps axu|grep grep $ ps axu|grep grep $ ps axu|grep grep jidanni 1171 0.0 0.2 2376 608 pts/1R02:26 0:00 grep grep Is it some kind of race condition? I think so, it depends when exactly is grep started, the process snapshot that ps outputs might be from time right before grep is written to it... it might also depend on shell (how much time is there between starting ps and grep) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Please Help with Serial ATA!
Jon Earle wrote: I'm having serious trouble getting Debian loaded on my new system. The specs are: AMD Athlon XP 2700+ 333FSB MSI K7N2 Delta-ILSR NForce 2 512MB Corsair XMS 3200 C2 36GB Western Digital Raptor SATA 10K RPM MSI GFX 5600VTDR 128 Antec Performance II SX635BII LG 52x24x52 CDRW LG DVD Panasonic 1.44 Floppy Drive After my initial attempt at installing Debian 3.0r1 failed, I eventually found a Promise Fastrack 376 driver that I was able to feed to the debian installer so that it saw the drive and allow me to partition and install to it. The problem now is, the first reboot. Lilo doesn't even begin to show itself. No L, no errors, nothing. The system just sits there blinking at me. There has to be a way to get this going. I'd hate to have to return the system and settle for something that's a few years old, just to get Linux to work. Is it even possible to get Debian installed and running on a SATA drive? you might need some cutting edge kernel, the first kernel that supports my sata (on intel D865PERL motherboard) is 2.4.21-ac4 (SCSI_ATA). When I boot kernel that doesn't have CONFIG_SCSI_ATA enabled the system only boots up to detection of harddrives, then freezes. Only when I use CONFIG_SCSI_ATA (you need at least 2.4.21-ac4) the system boots and recognizes drive properly (as scsi drive, /dev/sda or something like that) you might also be able to make it look like ata ide drive in bios (legacy mode or something like that). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Executing a Script every hour
Pigeon wrote: On Wed, Oct 01, 2003 at 07:55:30PM -0700, Paul Johnson wrote: On Wed, Oct 01, 2003 at 12:15:19PM -0400, Derrick 'dman' Hudson wrote: anacron. cron and anacron solve similar problems but with a different approach based on different requirements. The two are not mutually exclusive. Ah. Well, that's for stuff that has to be run even when the machine's down. I'd love to be able to run stuff when my machine's down. :-) What did you actually mean? cron runs program at specified times. if machine is down at the specified time the job does not run. anacron runs jobs with specified frequency, e.g. daily. So if the machine is up at least sometime the job is run... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: To be safe don't use shift key...
[EMAIL PROTECTED] wrote: ...or else the riaa might sue you. http://www.cnn.com/2003/TECH/ptech/10/08/bmg.protection.reut/index.html quote from article: "Computers running Linux and older versions of the Mac operating system are unable to run the software and are able to copy the disc freely, he said." ROFL erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
postfix: very long startup time (few minutes)
for some reason the postfix thatkes a LONG time to startup (during boot). any ideas why? It seems like it's getting worse over time. this is on debian unstable. I don't have too many emails and I am not aware of any long queues waiting, this is on my personal workstation, I get few hundred emails a day (mostly mailing lists), send only few emails a day and normal operation of computer seems to be unaffected by email, the only problem is when the postfix service is started. I am not 100% sure it's postfix, at the time when it's started the system is not up yet. restarting postfix doesn't take too long - less than 1 second (/etc/init.d/postfix stop; /etc/init.d/postfix start). I think the problem is postfix because that's what it says on the screen but I guess at that time it might be something else (that's doesn't have output on the screen). any ideas? TIA! erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: postfix: very long startup time (few minutes)
Matthew Weier O'Phinney wrote: > -- Erik Steffl <[EMAIL PROTECTED]> wrote > (on Sunday, 06 October 2002, 08:43 PM -0700): > >> for some reason the postfix thatkes a LONG time to startup (during >>boot). any ideas why? It seems like it's getting worse over time. >> >> this is on debian unstable. > > I've used postfix on unstable and testing, and don't see this. > > >> I don't have too many emails and I am not aware of any long queues >>waiting, this is on my personal workstation, I get few hundred emails a >>day (mostly mailing lists), send only few emails a day and normal >>operation of computer seems to be unaffected by email, the only problem >>is when the postfix service is started. >> >> I am not 100% sure it's postfix, at the time when it's started the >>system is not up yet. restarting postfix doesn't take too long - less >>than 1 second (/etc/init.d/postfix stop; /etc/init.d/postfix start). > > It *probably* isn't postfix, then. Is this the last daemon to start > during boot? Even if it isn't, it's possible that a previous daemon or how would the other deamon stop the postfix from continuing? I see the postfix start line, then nothing moves for few minutes, disk is busy, then I get next daemon's start line. I mean no matter how busy the system would be doing whatever else it might be doing the postfix would STOP for few minutes (when I restart it once the system is booted up it takes less then a second to start). > the next daemon in the boot order is causing the delay. I know, for > instance, that on my machine, the USB routines will be showing > on-screen, and I'll have a delay before the last item shows... but that > this is due to my webserver taking a while to load (on average, about > 10-15 seconds!). You might try going through each of the scripts in your > default runlevel to see which ones have delays. If postfix doesn't take > long to start when you do it manually, my bet is you'll find another > script in there that does. I went over all of them (the ones that seemed relevant). Nothing takes longer that about a second to start, except of postgresql that takes maybe 2 - 5 seconds. the next step was running top -b (I added it on the beginning of the postfix), I see the echo I added so I assume the top starts running at the beginning of the delay. however top doesn't show anything interesting, no high load, no program (even low load) stays near top too long... the only suspicious process running is find - the only script that runs find is fetchmail and it starts quite fast... not sure what to do... I'll probably start removing the startup scripts one by one and see what happens... any better ideas? btw here's the top output (filtered out a bit) 21:41:23 up 0 min, 0 users, load average: 1.33, 0.33, 0.11 35 processes: 34 sleeping, 1 running, 0 zombie, 0 stopped CPU states: 3.5% user, 11.5% system, 0.0% nice, 85.0% idle Mem:385460K total,63436K used, 322024K free, 2856K buffers Swap: 243104K total,0K used, 243104K free,42984K cached PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME COMMAND 418 cyrus 14 0 1408 1408 1216 D 1.9 0.3 0:00 ctl_deliver 422 root 16 0 1856 1856 1428 S 0.9 0.4 0:00 omniNames 1 root 8 0 476 476 408 S 0.0 0.1 0:04 init 21:41:28 up 0 min, 0 users, load average: 1.46, 0.37, 0.12 39 processes: 37 sleeping, 2 running, 0 zombie, 0 stopped CPU states: 5.2% user, 2.0% system, 0.0% nice, 92.8% idle Mem:385460K total,65320K used, 320140K free, 3008K buffers Swap: 243104K total,0K used, 243104K free,43788K cached PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME COMMAND 418 cyrus 9 0 1460 1460 1268 D 0.7 0.3 0:00 ctl_deliver 451 root 13 0 1180 1180 928 S 0.3 0.3 0:00 postfix-script 6 root 10 0 00 0 DW0.1 0.0 0:00 kupdated 21:41:33 up 1 min, 0 users, load average: 1.58, 0.42, 0.14 39 processes: 38 sleeping, 1 running, 0 zombie, 0 stopped CPU states: 2.2% user, 1.4% system, 0.0% nice, 96.4% idle Mem:385460K total,66136K used, 319324K free, 3124K buffers Swap: 243104K total,0K used, 243104K free,44292K cached PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME COMMAND 417 cyrus 19 0 1224 1224 928 D 1.9 0.3 0:00 ctl_cyrusdb 217 root 12 0 780 780 660 S 0.5 0.2 0:00 syslogd 418 cyrus 12 0 1532 1532 1340 D 0.5 0.3 0:00 ctl_deliver 424 root 10 0 920 920 728 R 0.1 0.2 0:00 top 21:41:38 up 1 min, 0 users, load average: 1.78,
Re: postfix: very long startup time (few minutes)
Kourosh wrote: > Could it be that somewhere in the script is a FQDN, > possibly that of your machine, and that the delay > is a DNS lookup that times out? Do you have localhost > and your machines domain name in the hosts file to avoid > unneccesary DNS lookups? yes, the name is in /etc/hosts. in addition to that when the system stops there is continuous HD activity. So I guess something must be going on. Just can't figure out what... erik > > Regards. > > Kourosh > > On Wed, Oct 09, 2002 at 10:09:00PM -0700, Erik Steffl wrote: > >>Matthew Weier O'Phinney wrote: >> >>>-- Erik Steffl <[EMAIL PROTECTED]> wrote >>>(on Sunday, 06 October 2002, 08:43 PM -0700): >>> >>> >>>>for some reason the postfix thatkes a LONG time to startup (during >>>>boot). any ideas why? It seems like it's getting worse over time. >>>> >>>>this is on debian unstable. >>> >>>I've used postfix on unstable and testing, and don't see this. >>> >>> >>> >>>>I don't have too many emails and I am not aware of any long queues >>>>waiting, this is on my personal workstation, I get few hundred emails a >>>>day (mostly mailing lists), send only few emails a day and normal >>>>operation of computer seems to be unaffected by email, the only problem >>>>is when the postfix service is started. >>>> >>>>I am not 100% sure it's postfix, at the time when it's started the >>>>system is not up yet. restarting postfix doesn't take too long - less >>>>than 1 second (/etc/init.d/postfix stop; /etc/init.d/postfix start). >>> >>>It *probably* isn't postfix, then. Is this the last daemon to start >>>during boot? Even if it isn't, it's possible that a previous daemon or >> >> how would the other deamon stop the postfix from continuing? I see >>the postfix start line, then nothing moves for few minutes, disk is >>busy, then I get next daemon's start line. I mean no matter how busy the >>system would be doing whatever else it might be doing the postfix would >>STOP for few minutes (when I restart it once the system is booted up it >>takes less then a second to start). >> >> >> >>>the next daemon in the boot order is causing the delay. I know, for >>>instance, that on my machine, the USB routines will be showing >>>on-screen, and I'll have a delay before the last item shows... but that >>>this is due to my webserver taking a while to load (on average, about >>>10-15 seconds!). You might try going through each of the scripts in your >>>default runlevel to see which ones have delays. If postfix doesn't take >>>long to start when you do it manually, my bet is you'll find another >>>script in there that does. >> >> I went over all of them (the ones that seemed relevant). Nothing >>takes longer that about a second to start, except of postgresql that >>takes maybe 2 - 5 seconds. >> >> the next step was running top -b (I added it on the beginning of the >>postfix), I see the echo I added so I assume the top starts running at >>the beginning of the delay. however top doesn't show anything >>interesting, no high load, no program (even low load) stays near top too >>long... >> >> the only suspicious process running is find - the only script that >>runs find is fetchmail and it starts quite fast... >> >> not sure what to do... I'll probably start removing the startup >>scripts one by one and see what happens... >> >> any better ideas? >> >> btw here's the top output (filtered out a bit) >> >> 21:41:23 up 0 min, 0 users, load average: 1.33, 0.33, 0.11 >>35 processes: 34 sleeping, 1 running, 0 zombie, 0 stopped >>CPU states: 3.5% user, 11.5% system, 0.0% nice, 85.0% idle >>Mem:385460K total,63436K used, 322024K free, 2856K buffers >>Swap: 243104K total,0K used, 243104K free,42984K cached >> >> PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME COMMAND >> 418 cyrus 14 0 1408 1408 1216 D 1.9 0.3 0:00 ctl_deliver >> 422 root 16 0 1856 1856 1428 S 0.9 0.4 0:00 omniNames >>1 root 8 0 476 476 408 S 0.0 0.1 0:04 init >> >> 21:41:28 up 0 min, 0 users, load average: 1.46, 0.37, 0.12 >>39 processes: 37 sleeping, 2 running, 0 zombie, 0 stopped >>CPU states: 5.2% user, 2.0% system,
cyrus doesn't work after upgrade (could not find auxprop plugin)
after I dist-upgraded system yesterday the cyrus21 stopped working (I used aptitude ugg) when I try to connect to imapd I get the following message in syslog: Oct 27 12:18:59 localhost cyrus/imapd[4339]: badlogin: localhost[127.0.0.1] plaintext erik SASL(-13): user not found: checkpass failed and this message in auth.log: Oct 27 12:18:59 localhost cyrus/imapd[4339]: could not find auxprop plugin, was searching for '[all]' why it's not able to find any plugins? I tried to specify sasldb plugin explicitly but then the message just changes so that instead of [all] it says [sasldb]. I have some plugins installed (including sasldb): jojda:~> ls /usr/lib/sasl2/ libanonymous.solibgssapiv2.solibotp.so libanonymous.so.2 libgssapiv2.so.2 libotp.so.2 libanonymous.so.2.0.9 libgssapiv2.so.2.0.9 libotp.so.2.0.9 libcrammd5.so liblogin.so libplain.so libcrammd5.so.2liblogin.so.2 libplain.so.2 libcrammd5.so.2.0.9liblogin.so.2.0.9 libplain.so.2.0.9 libdigestmd5.solibmysql.so libsasldb.so libdigestmd5.so.2 libmysql.so.2 libsasldb.so.2 libdigestmd5.so.2.0.9 libmysql.so.2.0.9 libsasldb.so.2.0.9 here's what I set in /etc/imapd.conf: sasl_pwcheck_method: auxprop in /etc/default/saslauthd: MECHANISMS=sasldb saslauthd is running: jojda:~> ps ax|grep sasl 4285 ?S 0:00 gvim /etc/default/saslauthd 4326 ?S 0:00 /usr/sbin/saslauthd -a sasldb 4327 ?S 0:00 /usr/sbin/saslauthd -a sasldb 4328 ?S 0:00 /usr/sbin/saslauthd -a sasldb 4329 ?S 0:00 /usr/sbin/saslauthd -a sasldb 4330 ?S 0:00 /usr/sbin/saslauthd -a sasldb 4370 pts/9S 0:00 grep sasl any ideas on what could be a problem? this is on debian unstable, here's a list of sasl related packages that I have installed (and not installed): jojda:/home/erik# dpkg -l \*sasl\* Desired=Unknown/Install/Remove/Purge/Hold | Status=Not/Installed/Config-files/Unpacked/Failed-config/Half-installed |/ Err?=(none)/Hold/Reinst-required/X=both-problems (Status,Err: uppercase=bad) ||/ NameVersion | Description +++-===-===-== un libauthen-sasl-cyrus-perl (no description available) pn libsasl-bin (no description available) ii libsasl-dev 1.5.27-3.3 Development files for authentication abstraction library. pn libsasl-digestmd5-des (no description available) ii libsasl-digestmd5-plain 1.5.27-3.3 DIGEST-MD5 module for SASL pn libsasl-gssapi-heimdal (no description available) pn libsasl-gssapi-mit (no description available) pn libsasl-krb4-mit (no description available) un libsasl-modules (no description available) pn libsasl-modules-krb5 (no description available) un libsasl-modules-nonus (no description available) ii libsasl-modules-plain 1.5.27-3.3 Basic Pluggable Authentication Modules for SASL ii libsasl22.1.9-3 Authentication abstraction library. pn libsasl2-dev (no description available) un libsasl2-digestmd5-des (no description available) pn libsasl2-digestmd5-plain (no description available) pn libsasl2-gssapi-mit (no description available) pn libsasl2-krb4-mit (no description available) ii libsasl2-modules2.1.9-3 Pluggable Authentication Modules for SASL ii libsasl2-modules-gssapi-heimdal 2.1.9-3 Pluggable Authentication Modules for SASL pn libsasl2-modules-plain (no description available) ii libsasl7
sawfish makes itself x-window-manager
not sure when it happened, I guess it was during dist-upgrade, but suddenly sawfish was the x-window-mananger (I definitely did not run update-alternatives). is this a bug or intented behaviour? If it's a "feature, can it be disabled? This is pretty annoying, while it's not that hard to re-run update-alternatives and reset x-window-manager to whatever I like it also causes more serious error: when I restarted fvwm it did not restart but instead X windows was restarted. All unsaved data gone:-) [btw I was lucky this time, nothing lost, leisure suite larry tought me well: save early, save often:-)] should I file a bug? BTW: IIRC it happened before with scwm or some other WM... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Xlib: connection to ":0.0" refused by server ???
daves debian wrote: When i am logged in as a user, I want to execute an X program as root, I type su ... The X library has been refused by the X server, because root is not autherised to connect to the server ?? ... only authorized user can connect to X server, just because you are root doesn't mean you can connect to X server on given machine (just like you cannot connect to other services - e.g. you'd still need password for database, ...) you have few options: 1) xhost (see man xhost), don't do that though! 2) as root run: xauth merge ~userThatRunsX/.Xauthority option 2 will give you access to X server without compromising security (while it's running, if you restart X you have to run xauth again) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: sawfish makes itself x-window-manager
Jerome Acks Jr wrote: On Thu, Nov 28, 2002 at 03:09:33AM -0800, Erik Steffl wrote: not sure when it happened, I guess it was during dist-upgrade, but suddenly sawfish was the x-window-mananger (I definitely did not run update-alternatives). is this a bug or intented behaviour? If it's a "feature, can it be disabled? Yes, if set to automatic, update-alternatives will set alternative with the highest priority. If you run "update-alternative --config x-window-manager" and select your preferred window manager, update-alternatives will go into manual mode and will not change x-window-manager again until you manually change it or reset update-alternatives to auto mode. it happened few times already - after each incident I set fvwm as x-window-manager (using update-alternatives --config x-window-manager and pipcking fvwm). looks like a bug in update-alternatives (dpkg), right? Is there any way to verify what happened? I mean is there any history of changes or something that I can check AFTER the fact (obviously, I cannot check before since I don't know it's goingto happen, and even if I knew there's nothing really to check). right now it looks ok: jojda:/home/erik# update-alternatives --display x-window-manager|head x-window-manager - status is manual. link currently points to /usr/bin/fvwm2 /usr/bin/xfwm - priority 20 slave x-window-manager.1.gz: /usr/share/man/man1/xfwm.1.gz ... did anybody else noticed the same problem? erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: sawfish makes itself x-window-manager
Craig Dickson wrote: > > Erik Steffl wrote: > > > looks like a bug in update-alternatives (dpkg), right? Is there any > > way to verify what happened? I mean is there any history of changes or > > something that I can check AFTER the fact (obviously, I cannot check > > before since I don't know it's goingto happen, and even if I knew > > there's nothing really to check). > > > > did anybody else noticed the same problem? > > Yes. Every time the X packages are updated in Sid, x-window-manager gets > set to fluxbox (which is installed, but not what I want the default to > be), and x-terminal-emulator gets set to gnome-terminal (also not what I > want by default). I run "update-alternatives --config" to change them to > the defaults I want, and the next time X is updated, the same thing > happens again. It's kind of annoying. in case of most programs it's kind of annoying, but in case of x-window-manager it can cause data loss - that makes it IMO a fairly serious bug (though not very likely to be triggered, I guess most people (including me) almost never restart WM). > Bug #164214, in its second message, describes this problem, and in its > third message, has a proposed patch for update-alternatives that is > claimed to be a fix for it. Whether it is a valid fix or not, I do not > know, though it looks reasonable at first glance. Since this proposed > patch is nearly two months old already, it's probably not safe to assume > that a fixed package is imminent. The dpkg team seems to have a lot > going on currently, to judge from a few messages I've seen on > debian-devel over the last few months. it's not exactly the same, it mostly deals with --remove not switching back to auto mode and the patch only fixes this particular problem (switching back to auto). as far as I can tell that would have no effect on problem of update-alternatives behaving as if it was in auto mode when it is actually in manual mode. anyway, I guess I'll just be careful restarting WM and hope it'll get fixed eventually... considering the number of old bugs (few years old) it doesn't seem to make any sense to file bugs... are they re-impplementing most/all of dpkg? erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: sawfish makes itself x-window-manager
Colin Watson wrote: On Mon, Dec 02, 2002 at 08:26:28AM -0800, Erik Steffl wrote: anyway, I guess I'll just be careful restarting WM and hope it'll get fixed eventually... considering the number of old bugs (few years old) it doesn't seem to make any sense to file bugs... are they re-impplementing most/all of dpkg? No. I believe dpkg-dev is being reimplemented, but that's about it. dpkg is one of those packages that has a lot of bugs but by and large works pretty well. Looking purely at the number of bugs and ignoring the fact that it's mostly very solid, tested code is not really rational. depends, doesn't it? what sense does it make to file a bug that is going to be ignored? (I am not saying dpkg doesn't work well, mostly). anyway, are you saying I should file a bug (I am asking since you (and others) apparently know about the problem and you didn't file a bug yet) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
problem with voodoo 3 and multiple X servers
I don't use multiple X servers too often but I think that since X 4.x there are some problems when running multiple X servers on same machine (voodoo 3 card). I have dri enabled, it works ok. as long as I run only one instance of X it works without any problems. the problem as manifested now is: I am running one X server (0) fvwm, it works without problems. I started another X server (1), runing kde and openoffice (started using startx from virtual console). from time to time it almost stops reacting, it doesn't update screen properly, only fonts are redrawn, the rest of the screen seems to be frozen (the effect is that the screen gets overwritten by text that seems to be slightly relevant to what should be going on, e.g. when I go to menu and this happens the menu is not updated anymore but near menu the menu item labels are redrawn (text only)). when I switch to another vt and back it works ok, for some time. it seems like everytime this happens the following message is printed on the console where X was started from: Resetting FIFO googling revealed that some people had the same message printed when using xinerama, didn't find any resolution for that but not sure it's rlated. I also foound somewhat similar debian bug 164303, that one was closed because submitter said he found out that it's a hw problem... any ideas on what's going on or how to troubleshoot? system: debian unstable kernel 2.4.18 (with tdfx) ii xserver-xfree8 4.2.1-4 00:0f.0 VGA compatible controller: 3Dfx Interactive, Inc. Voodoo 3 (rev 01) TIA! erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: sawfish makes itself x-window-manager
Colin Watson wrote: On Wed, Dec 04, 2002 at 10:23:04PM -0800, Erik Steffl wrote: Colin Watson wrote: dpkg is one of those packages that has a lot of bugs but by and large works pretty well. Looking purely at the number of bugs and ignoring the fact that it's mostly very solid, tested code is not really rational. depends, doesn't it? what sense does it make to file a bug that is going to be ignored? (I am not saying dpkg doesn't work well, mostly). anyway, are you saying I should file a bug (I am asking since you (and others) apparently know about the problem and you didn't file a bug yet) I think there are already bugs filed about glitches in u-a ... didn't find anything that would resemble this problem (updating of alternative that is in manual mode) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
mozilla start (multiple times versus -remote)
mozilla wrapper used to start up a new mozilla or mozilla -remote, depending on whether mozilla was already running. Which made it easy to open new browser window, regardles of whether one was already running. However now mozilla starts profile wizard and doesn't allow to use default profile (says profile is already used) when mozilla is already running. which makes it inconvenient to use in scripts (or various launchers etc.) is there any trick I can use to easily open a new mozilla window (regardles of whether one is already open? I mean something simpler then (pseudocode): if mozilla -remote 'ping()' then mozilla else mozilla -remote 'xfeDoCommand (openBrowser)' fi TIA- erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Debian Can't See All of Large Disks?
Karsten M. Self wrote: on Thu, Jul 31, 2003 at 08:44:12PM -0700, Mike Hunt ([EMAIL PROTECTED]) wrote: Hi all, I had recently tried to install Debian on a computer containing a 200 GB harddrive. However, during cfdisking of the Debian install process cfdisk fails to see past approximately 130 GB of hard disk space. I'm suspecting this is because it is an old version of cfdisk (later versions of cfdisk seem to see the whole drive). So, does anyone here know how to work around this problem so I can use all of my hard disk space with Debian? Thanks for your help! Please set your mailer/editor linewrap to 68-75 characters. I strongly recommend 72 as a good default. Thank you. See the Large Disk HOWTO. There's a history of successivly larger disk sizes which require different mechanisms for handling what was initially a 528 MiB limit for HD sizes. I have a similar problem so I read the large disk howto (again) but it's pretty much irrelevant today... For disks beyond 137 GiB, the HOWTO doesn't provide specific guidance, though a good overview is here, in general, you want 2.4.18 or better kernels: http://www.nber.org/sys-admin/maxtor-160.html in addition to that: I have SATA disk (250GB), seen as scsi device (kernel scsi ata configuration option enabled), for which you need at least 2.4.21-ac4, most tools see the whole disk but can't read above 137GB, here's the latest from lkml: from [EMAIL PROTECTED] I just realized, 2.4 kernels don't support scsi's READ_CAPACITY_16, > nor 64-bit sector_t on a 32-bit processor. Can you test Alan Cox's 2.6.0-test-ac tree? I bet the 137GB > limitation may disappear there. Jeff I didn't try it yet so I don't know whether it's going to work... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Changing window managers The Debian Way (was Re: how NOT to workwith debian)
Ron Johnson wrote: On Sun, 2003-08-10 at 21:11, Damien Solley wrote: Looks like X is running fine. You just need a window manager other than twm! from a console, edit your ~/.xinitrc file. For simplicity, this file can have just one line, reading: startkde That'll start kde (if it's installed correctly). Otherwise you can specify: gnome-session #(To start gnome 2) xterm #(to start an xterm) Isn't it The Debian Way to do: # update-alternatives --config x-window-manager You may also have to do: # update-alternatives --config x-session-manager I had really BAD experience with this. Once you manually set it, theoretically it shouldn't be changed automatically. Unfortunately it is. In case of lot of other programs it is not a big deal, but if window manager is changed and you restart it the X goes down. Very unpleasant experience. You can check archives, I already asked about this and it looks like it's not really clear who is at fault (I filed bug against (IIRC) sawfish which was the offending WM at the time, maintainer closed it (instead of reassigning, which IMO would be a better choice), maybe I should open another one against update-alternatives?). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: ISA network cards
jleclair wrote: I'm trying to set up an old 486 as router/nat/firewall server for home network. The 2 nics are dlink 220 isa. When I run modconf, I choose the NE net module with this parameter: io=0x300 This works and the module loads for eth0. How do I load the module now for the second nic? I dont know the io address of the 2nd nic and the module will only load once. I guess I could do something like: io=0x240,0x300 when loading the module through modconf. BTW, I guessed at the io address that DID respond and have since unloaded the module and tried every other io address to make the second nic work. Are there tools to probe for hardware responding on io addresses not listed in /etc/proc? ... I think you need isapnptools erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: root login How ???
charlie derr wrote: It's probably not a full and complete answer to your question, but I find what works for me is to log in to the graphical user environment of your choice as a regular user and then execute "su" in one of the terminals (xterm, gnome-terminal, konsole, whatever...) inside the gui (not "su -"). I can then execute a graphical program (as root) by initiating it from that terminal. hth, ~c Victory wrote: I found out that after debian 3.0r1 installed, try to login as root and passwd at the GNOME Desktop Manager and it said " The system administrator is not allowed to login form this screen" How to login with root account to run some utility from graphic mode ??? some additional info: su (or su -) to root and run xauth merge ~user/.Xauthority (where user is the name of the user that is running X server), yuou can use xauth to give other user access to X server as well (they have to have read access to .Xauthority) add user to sudoers and run sudo programThatRequiresRoot DO NOT DO THIS: xhost + (everybody can use the X server), man xhost for more info, but really, don't use this unless you really want it that way (security) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: OT: Why is C so popular?
Steve Lamb wrote:
On Wed, 27 Aug 2003 23:52:35 -0600
Jacob Anawalt <[EMAIL PROTECTED]> wrote:
Was this code on a Unix system, or did you have one nearby? Did you know
about the indent program at the time? (man indent)
It _seems_ to work for me to convert someone elses sytle (or lack of it)
in coding C into a format that I like (K&R).
That's just it, I don't like C indenting any more, period. It isn't
"someone else's style" that is the problem, it is the fact that it is the
antithesis of how I've grown to like to code. Lemme put it this way:
C:
if foo
bar;
Python:
if foo:
bar
Want to extend it?
Python:
if foo:
bar
baz
C:
if foo
bar;
baz;
Bzt, wrong
if foo {
bar;
baz;
}
what do you do if the indentation is lost? you e.g. copy and paste or
email piece of code etc. and it changes spaces into tabs or vice versa
or removes newlines or somebody accidentally joins two lines together or
otherwise changes formatting... what do you do then? how would you know
whether baz was supposed to be part of if or not?
erik
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: OT: Why is C so popular?
Joey Hess wrote:
Steve Lamb wrote:
That's just it, I don't like C indenting any more, period. It isn't
"someone else's style" that is the problem, it is the fact that it is the
antithesis of how I've grown to like to code. Lemme put it this way:
C:
if foo
bar;
That's bad style unless you have a reason to know for sure that you will
only ever need exactly one statement in the if. Since there's no benefit
to leaving out the braces and you'd have to think hard every time you
did, it's better to just avoid this C misfeature entirely..
not sure whether it is fair to call it misfeature. which brings us
back to topic: IMO part of the popularity of C is that it has bunch of
pretty simple but consistent rules that are taken to extreme (you can
really apply them any way you want, if you can live with absurdity of it
i[3] and 3[i] :-). that way you have a system (language) that can be
used in ways undreamt of by inventors (authors?).
e.g. everywhere you have a statement you can have compound statement
(because, after all, it's just a statement), everywhere you can have
variable you can have expression etc.
so it would be quite inconsistent to suddenly say that in some
special cases you can have compound statement but not a simple statement
- it would screw the consistent application of simple rules.
that said I also consider it better style to use if(something) { f(): }
erik
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [OT] Why does X need so much CPU power?
Neal Lippman wrote: ... Well, most replies to my posting have pinned the "blame" on KDE and Gnome rather than X per se. I'll have to reinstall on the laptop and see how it looks with a more minimal WM. I hope you're not reinstalling just to change the WM... This does still beg the question of how Win95/98/Me/NT, etc, managed to provide a reasonable "desktop" when KDE/Gnome could not, however. It both KDE and Gnome are under development, more effort is spent on having things actually working, adding/changing features etc. than on performance improvements. as others said if you don't have resources to waste just use something else, there's number of other WMs. You can definitely have more eye candy per buck in X than you can have in windows (because you have different types of eye candy available). hey, sco unix had pretty good X back on 40MHz 386 (certainly a lot better than win 3.0 or 3.1 or whatever version was out in '90 - '91). ... From what little I know of X, I'd tend to agree that X is being overtaxed supporting a desktop environment that it was never designed to do. Aside from the present market penetration of X (which could also be used to argue to stick with Windows instead of ever having adopted Linux), what would be the obstacle (other than, of course, the time/effort for development) for a new graphics paradigm to sit atop Linux? [Yes, I know there'd be a lot of apps to redo and so forth as well, although if there were a Gtk+ compatibility layer...) X is GREAT. just because a particular combination of software/hardware doesn't work well (too slow) doesn't mean there's a need to throw out the baby with the... that's not to say that X is perfect, far from it, but it's being worked on, it's very flexible and extensible and there is nothing better, at least now. btw there's a relevant slashdot.org article about Xr/Cairo (Xr was renamed Cairo), you can read something about how they plan to make better support for eye candy (vectors instead of bitmaps, because vectors are cheaper to transfer, easier to scale) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [OT] Why does X need so much CPU power?
Nicos Gollan wrote: On Tuesday 02 September 2003 02:56, Erik Steffl wrote: X is GREAT. just because a particular combination of software/hardware doesn't work well (too slow) doesn't mean there's a need to throw out the baby with the... X is really good at what it was built to be. It provides an interface to create contexts applications can use to draw to and to manage those contexts. Because it was designed for environments with semi-dumb terminals this will even work over a network. It does NOT provide a generalized toolkit for building UIs. All libraries like Xaw, QT, GTK and whatever just grab a window context and draw to them via a networked interface. Windows OTOH was designed (and please don't start arguing whether "designed" is the right term... we all know what we think about that ;-) ) to provide a nice UI on a relatively powerful workstation without the whole overhead of a client/server concept which allowed to do a lot of nice things easily (transparency with an updated background, native 3D acceleration, font management that actually works, etc.). To be honest, were it not for the way Microsoft handles things, I'd be using it right now. I'd probably keep the servers in the basement running Linux, but for the desktop use Win would win. why? it's true that in _some_ cases X isn't the _best_ performer but in general I find it much better than windows, mostly because of flexibility. btw the overhead of client/server isn't anything that one needs to be concerned about even on 386 (X with reasonable WM performs same/better as windows) I admit that it's somewhat inconsistent (huge number of different widget libs), not trivial to setup (that's mostly a thing of the past though) etc. what do you mean native 3D acceleration? you need directX or openGL for 3d in windows, openGL (with DRI) in X. how is one more or less native? font management: not sure what you mean. I have some fonts, I can pick which one I want to use (based on app). that describes both win and X. yeah, the real transparency is what we need. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [OT] Why does X need so much CPU power?
Nicos Gollan wrote: On Tuesday 02 September 2003 20:00, Erik Steffl wrote: why? it's true that in _some_ cases X isn't the _best_ performer but in general I find it much better than windows, mostly because of flexibility. You've made better experiences than I did, then... On each and every system I had the pleasure to work on so far (no, not all configured by myself...), Windows beat X in terms of subjective responsiveness. It just *feels* quicker to me. It's nothing to get really excited about. if you're using KDE and/or Gnome than yes, they are slow. Too slow. But we can expect them to get at least somewhat better as they mature, both of these systems are fairly complex and fairly early into development. compare the speed of mature WMs - pretty much all of them (except of enlightenment:-) are about the same or even faster than windows. provided you have 2d acceleration working, but that shouldn't be an issue on most cards. what do you mean native 3D acceleration? you need directX or openGL for 3d in windows, openGL (with DRI) in X. how is one more or less native? OK, "native" was a bad choice of words. Let's just put this in the "bad hardware vendor support" corner... I agree on that, you have to choose your HW if you want 3D (openGL + DRI) support. I used to have voodoo III (fairly good), now I've got nvidia (I use nvidia drivers, so far no problems, but binary only) font management: not sure what you mean. I have some fonts, I can pick which one I want to use (based on app). that describes both win and X. - One word: -adobe-courier-medium-r-*-*-12-*-*-*-*-*-*-15 most apps provide somewhat better interface to this, I mean you don't have to remember what that 12 stands for... - QT seems to have some serious issues (google for "kde konsole fonts"). that's possible but is it X problem? I checked few email that google found, seems like kde bug. not a reason to throw X out. - Mozilla looks like something unspeakable unless you get it running with FreeType. And then there's a lot of guesswork to do with the min, max and gain settings until it's looking good. At least it looks better than anything else I've seen so far. on solaris the netscape looked ugly. on debian mozilla looks great. BTW it's IE (or is it windows?) that has very strange font settings and then lot of stuff is too small in mozilla (netscape). don't remember the issue. but yes, you can set the minimum size of fonts to get rid of this fonts-too-small problem. but again, this is not an X issue but application issue and it's being sorted out. There is no central font management. For some time now, X seems to support what do you mean? you have font server (standalone or just use X server). how much more central can you get? BTW AFAIK there's no way to have standalone fontserver for windows. truetype fonts and antialiasing, but for some strange reason nobody either knows how or wants to use it. I've been using true type fonts since forever, it got much simpler over time (it used to be that you had to install special font server, now it's more or less out of the box) yeah, the real transparency is what we need. But for efficiency reasons, the clipped content of windows isn't transmitted so that's not so easy. (Disclaimer: I'd never have thought I would write such stuff some day. I've been using Linux/X for several years now and it has replaced my Windows installation for productivity purposes, but as it is now, I would *not* recommend it to a "standard user".) neither would I:-) as long as by "standard user" you mean somebody who doesn't know and doesn't care to know much about computers (not a bad thing by itself). But the reason is not X - the reason is installation/setup/maintenance/configuration - and it's being worked on. All major distros are getting easier to install - Lindows is supposedly pretty much as easy to use as windows (=similar to windows), knoppix can be booted straight from CD, both redhat and madrake have very nice setup/install... reminder: I was arguing that X is good, not that linux is for everybody or that all guis on top of X are good... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [OT] Why does X need so much CPU power?
Yves Goergen wrote: On Tuesday, September 02, 2003 8:00 PM CET, Erik Steffl wrote: btw the overhead of client/server isn't anything that one needs to be concerned about even on 386 (X with reasonable WM performs same/better as windows) Could be, yes (I don't know). Just as a note, Windows has the same feature. OK, a similar one. You can connect to a remote desktop and see all yep, similar. ... But by the way: 2 questions on that: I have set up a debian Linux box and would like to run X applications on it. I haven't installed nor run the X server on the Linux machine itself, but I'd like to tunnel the X connection through SSH. That works fine for my account at university. I can run my cygwin X server locally (with a window manager running from local, too. think it's blackbox or so) and run xclock on the SSH shell. But when I do this on my own computer, it says it "cannot connect to the display ". I actually don't know what this variable is for nor what would be the right value for it. I've tried the value from university, the one I entered in PuTTY (for X forwarding) and some others, but it just didn't work. So what libraries do I have to install (I guess I already have them all) and what's the correct value for $DISPLAY ? $DISPLAY is what is the default display used by apps, if you don't specify display on command line (lot of the X apps use -display command line options). it should be set to hostname:n.m (n is display number, m is screen number) if you are using ssh X forwarding (in case you say it does not work) - you have to make sure that ssh server allows it, that you specify it on command line (or config dialog with putty) and that you have rights to use X on local host (and that, after doing ssh you don't do su). If ssh sets up X forwarding it should sets the $DISPLAY, if it doesn't work it's probably problem with local X. You need to provide more specific info if you need more specific help (what you run and how and what are the error messages) And a question just of interest: Is there something like a global clipboard in Linux as we know it from Windows? I mean not only per application, but shared by the entire system (or maybe user, in this case). yes there is. it is somewhat more complicated then windows clipboard. generally you mark by left mouse button, adjust existing selection by right mouse button, paste using middle button. Some applications support keyboard shortcuts (shift-insert for paste, ctrl-x/c/v for cut/copy/paste etc.). If you have problems to copy&paste between apps try xcutsel (select, click on one of the buttons on xcutsel, try to paste, if it didn't work try the other button). also take a look at xclipboard. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [OT] Why does X need so much CPU power?
Yves Goergen wrote: On Tuesday, September 02, 2003 8:51 PM CET, Yves Goergen wrote: So what libraries do I have to install (I guess I already have them all) and what's the correct value for $DISPLAY ? Ha! *big-grin* I got it... Just looking around in Webmin to find the Samba config and - zak - I found the SSH config ;) There was a switch that said it would disable X connection forwards Just changed that, and it works!! Cool... But I guess I'd never found it otherwise... another *big-grin* you would, because I have just sent you an email saying that ssh server has to allow X forwarding:-) sometime you can't avoid your fate... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [OT] Why does X need so much CPU power?
Derrick 'dman' Hudson wrote: On Mon, Sep 01, 2003 at 08:09:53PM -0400, Neal Lippman wrote: | what would be the obstacle [...] for a new graphics paradigm to sit | atop Linux? You already listed the obstacles. Anyways, FWIW here are some projects attempting to redesign how graphics are handled : http://www.directfb.org/ http://www.ggi-project.org/ < uhh, some other project working on an entire graphics architcture but I don't remember the name and a quick google search didn't reveal it > are you thinking of berlin? erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Mail from M$ Outlook Express -> something
Derrick 'dman' Hudson wrote: On Sun, Jun 22, 2003 at 11:58:23AM +0200, JZidar wrote: | I'm archiving my mail inWindoze and as I'll be switching to Debian soon I | would like to ask if there's any converter that would convert my Outlook | mail to the format used by kmail or some other linux mail client. One reliable method is to move the mail from your local windows system to a remote IMAP server. Then use your preferred linux client to pull the mail back from the IMAP server. or stay with IMAP (possibly local on your system). the beauty of it is that you can try different email clients (most of the modern ones support imap) without worries about file format, you can access email at the same time from different places, access email remotely (over SSL) etc. IAMP servers in debian: a very simple one to setup is uw-imap, somewhat more flexible but harder to install is cyrus (there are other, I think, these are the only ones I have experience with) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: memory leaks
matt zagrabelny wrote: On Mon, 2003-06-23 at 17:45, Shyamal Prasad wrote: "matt" == matt zagrabelny <[EMAIL PROTECTED]> writes: matt> hi, after boot my system runs out of memory in ~36-48 hours. What precisely happens when you "run out of memory"? Also, what version of Debian are you running. the normal scenario is i leave my computer running at night and then i come down in the morning and find applications (usually the big boys, evolution, opera) to be gone. i check top and find i have roughly 4M of memory left (no swap is left). i reboot and start over. sometimes i will do not reboot. try to figure out who uses a lot of memory. use e.g. top or some other program (lavaps:-) to see which program uses most memory. also, when you boot set up top to run in batch mode and redirect it to file (man top), that way you can see what happened over time (some offenders might not be running anymore). erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
sourceforge - why it depends on exim and proftp?
does sourceforge really need to depend on exim and proftp? wouldn't other mail resp. ftp servers be enough? since both exim and proftp conflict with other mail resp. ftp servers it means that sourceforge cannot be installed if I want to use e.g. postfix for my mail server... TIA erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
tora - anybody got it working woth postgress?
when I try to connect to postgress using tora it always says that authentification failed, even though I can log in into database using the same username/password from other program (pgaccess), I can also see the user in pg_shadow. related Q: why does tora only include postgresql connector? docs say that it can connect to mysql and oracle, but debian readme says that oracle and mysql native interfaces are built in - why and how to add them on? my system: linux 2.4.18, debian unstable, latest versions of relevant packages. any ideas? TIA erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
postgres: No pg_hba.conf entry for host
I get the following error message when trying to connect to postgres using tora: No pg_hba.conf entry for host... I think that the settings are ok since I can loging using psql: jojda:~> psql pokus -U erik -W Password: Welcome to psql, the PostgreSQL interactive terminal. here's my pg_hba.conf (uncommented lines): host all 0.0.0.0 0.0.0.0 reject localall md5 host all 127.0.0.1 255.0.0.0 md5 host pokus127.0.0.1 255.0.0.0 md5 erik it doesn't work with or without last line. any ideas? this is on debian unstable, kernel 2.4.18 TIA erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
postgers: how to use both md5 and passwordless access
I would like to use both passwordless access to user's own db as debian default config allows: localall ident sameuse host all 127.0.0.1 255.0.0.0 ident sameuse as well as being able to connect to other DBs using md5: localall md5 host all 127.0.0.1 255.0.0.0 md5 is it possible to combine both approaches? I tried to add sameuse after md5 but it didn't help (in fact I wasn't able to login at all) subquestion: when using the first option (ident) I cannot login using username/password, I can only autologin to user's own db - any ideas on how to fix that? (this is on debian unstable) TIA erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
postgres: Password authentification failed for user... when usingtora
when trying to login using psql the following works: psql -h localhost -d pokus -U erik -W however the tora (unstable package tora) gives me the following error (possible typos, it's retyped because I can't copy&paste from the dialog window where the message is displayed): Unable to connect to the database Fatal 1: Password authentification failed for user "erik" the only three uncommented lines in /etc/postgresql/pg_hba.conf: localall md5 host all 127.0.0.1 255.0.0.0 md5 host all 0.0.0.0 0.0.0.0 reject why would tora give such an error? (btw I tried to type everything few times to make sure there's no typo/capslock on etc.) this is on debian unstable, postgres 7.2.2-2, tora 1.3.7-4. any ideas? TIA! erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
LDAP address book - where to start?
I'd appreaciate an advice on where to start when I want to have LDAP based address book that can be accessed from various mail clients, at least mozilla mail client (mutt, evolution a plus). a web interface would be nice. any hints on where to look, which one to use etc. TIA! (I use unstable) btw I checked howtos (searched www.linuxdoc.org), I found some info there but it's kinda scattered so I would appreciate some pointers... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
pimppa: cannot access db during install
when trying to install pimppa I get the following errors (I let debconf configure pimppa): Unpacking pimppa (from .../pimppa_0.5.6-1_i386.deb) ... Setting up pimppa (0.5.6-1) ... ERROR 1045: Access denied for user: 'root@localhost' (Using password: NO) ERROR 1045: Access denied for user: 'root@localhost' (Using password: NO) Warning: pimppa MySQL database seems to exist. It WON'T be recreated Refer to the documentation on how to eventually upgrade However the MySQL user's password WILL be changed ERROR 1045: Access denied for user: 'root@localhost' (Using password: NO) dpkg: error processing pimppa (--configure): subprocess post-installation script returned error exit status 1 Errors were encountered while processing: pimppa E: Sub-process /usr/bin/dpkg returned an error code (1) mysql is running, mysql -u root works fine, with or without -h localhost (I have /root/.my.cnf with username and password: [mysqladmin] user = root password = [client] user = root password = (no, is not a password:-) what is it trying to access? TIA! erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Installation Help: SATA Drive
On Sun, 2003-11-16 at 00:47, Elie De Brauwer wrote: > On Saturday 15 November 2003 22:33, Erik Steffl wrote: > > On Fri, 2003-11-14 at 17:15, Justin Burke wrote: > > > Hi All, > > > > > > I could use some help getting Debian installed on a new machine with a > > > SATA drive. I've downloaded CD images for both sarge and sid, and both > > > installation methods hang at the same point: > > > > > > "Loading kernel modules > > > > > > Detected module 'ide-probe-mod' for 'Linux IDE probe driver'" > > > > > > I have reason to believe that the system is hanging because of the SATA > > > drive. I think that I need to use a 2.6.0 kernel. Is this right? How do > > > I create an installation CD with a different kernel? > > > > you need a fairly new kernel for SATA, 2.4.21 iwth ac4 patches, I > > think. If you have a large SATA drive (>130GB) you also need libata5 > > patches from Jeff Garzik. I have 2.4.21-ac and manually applied libata5 > > patch (and 250 GB Matrox on intel D865PERL motherboard) > > > > Some of the newer kernels might work without ac or Jeff's patches, > > hard to tell. Check the linux kernel mailing list archives (search for > > SATA) > > > > I'm using a fresh from kernel.org 2.4.22 without any patches for my 80 gig > sata drive. do you use it as: - legacy IDE (setting in BIOS so that itr looks like regular IDE drive [didn't try this one] - IDE (looks like SATA but handled by IDE driver) [system freeze during boot, when probing disks] - SCSI (CONFIG_SCSI_ATA) [works but needs libata5 patches for >133GB support] thanks, erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: pasting into vi inserts many indents
Justin Guerin wrote: On Tuesday 18 November 2003 01:27 pm, Brian wrote: Hi, Lately, if I open a file with vi from inside a gnome-terminal and then select a large chunk of text, (it doesn't matter where, e.g., Mozilla or the same vi session, both do it), it corrupts the formatting with many indents. The pasted output inserts more indents on each line than there were on the previous line to create a "cascading formatting". Can someone point me to how I can stop this? I know I can fix past problems with cat file | tr -d '\t' , but how can I stop the bug in the first place? However, I can't get it to do it now for an example. Brian -- This is due to a vi setting, autoindent. Try typing :set noai before pasting text. If you want ai back on, :set ai. :help ai for more info. if you run gvim you can paste using mouse (middle button), it doesn't do autoindenting then erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: pasting into vi inserts many indents
Jonathan Dowland wrote: On Tue, Nov 18, 2003 at 02:04:57PM -0800, Erik Steffl wrote: if you run gvim you can paste using mouse (middle button), it doesn't do autoindenting then erik You can paste in console-vims, too, depending on your terminal, by setting mouse=a . However I have the indent-problem when doing this too. AFAIK gvim recognizes mouse input and treats it as paste (no ai, maybe something else), text mode vim makes no difference between keyboard input, keyboard pasting and mouse pasting. erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Logitech MX310 mouse: Dead extra button?
Tom wrote: On Wed, Nov 19, 2003 at 04:08:12PM -0600, Nick Welch wrote: So.. Anyone out there have this mouse? It's fairly new, so there's nothing on google except sites selling it and reviews and things of that nature. I would really like to get this button working! I'm still pondering what exactly I should bind all of these buttons to. :D I bought one after reading the review of it on Toms Hardware. Here's my settings: Identifier "Mouse1" Driver "mouse" Option "Protocol" "ImPS/2" Option "ZAxisMapping" "4 5" Option "Device" "/dev/input/mice" The wheel works fine. The additional button on the right side appears to behave identically to the right mouse button. The additional button on the left appears to behave like clicking the wheel button (i.e., the middle button) in Mozilla and appears to have bizarre, unpredictable effects in gnome-terminal (if I press it while in "less" then "nano" gets started and some text gets pasted in). The small button behind the wheel does nothing. I wouldn't even know what's useful to map them to in GNU/Linux + X. In windows they act as browser-back and browser-forward for Web Pages and the shell. Basically I don't need them. if your window manager supports it you can use the extra buttons to move/resize/raise/lower windows by clicking in the windows instead of using the title/borders area only (since you are pretty sure that application in window doesn't use the extra buttons). I had it configured like that with older version of fvwm, unfortunately it doesn't work with newer version :-( erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Logitech MX310 mouse: Dead extra button?
Nick Welch wrote: On Wed, Nov 19, 2003 at 10:01:59PM -0800, Tom wrote: Wow, that was something I would have never thought of. However.. no dice. xev reports nothing but MotionNotify events, and hexcat'ing the device produces the same things when I move the mouse, regardless of whether I'm holding the button or not. :/ You can probably install the Windows DDK and use a firewire or serial cable with the Kernel debugger to learn more. In my experience at M$ it is possible to do most of the things Linux kernel hackers do without any source. Obviously with full .pdb's you get the best stack traces, but M$ releases "stripped pdbs" that give you maybe 80% of what you need. Hm, I don't have any windows installations, but I've been thinking about hacking into the linux kernel to check whether this button truly is silent, or if it's just sending malformed data of some sort. But then I run into the weird kernel/hardware problems I've been having lately, which kinda prevent me from doing any kernel experimentation. Sigh. I should try to work that problem out more and see if I can get it solved, so I can work on this one. :) Err.. Woah! Before I hit send on this message, I decided to plug the mouse into my NetBSD machine (can't believe I didn't think of that already). 'cat /dev/wsmouse0', and garbage spews out. Hit the little middle button and... garbage! Woohoo! So the mouse isn't broken or insanely flawed - that's GOOD. Now, to find out why Linux doesn't recognize this button, and attempt a fix. :D I noticed that xmod-ing the buttons (to make the wheel works as 4,5 instead of whatever the mouse 'naturally' sends) can have strange consequencies. If you didn't already: try the mouse without using xmod to remap the buttons, or, better yet, without any mouse using software (no X, no gpm etc.). If you still see nothing when you press the problem button it's time for kernel fun (or post to linux kernel mailing list). You might also get some answers from manufacturer's support, but you'd have to be fairly smart about how to ask questions (they usually don't support linux much but if you leave linux out and ask about protocol or something they might give you some info) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Experiment: Neophyte versus Windows XP & Debian Woody
Paul E Condon wrote:
On Thu, Nov 20, 2003 at 01:36:56PM -0500, David Z Maze wrote:
Paul E Condon <[EMAIL PROTECTED]> writes:
On Wed, Nov 19, 2003 at 10:24:57PM -0500, M. Kirchhoff wrote:
Two months later, I--like so many others before me--came crawling back
to Debian, my hands weary from long hours spent fighting RPM dependency
^^
hell, instability, package conflicts, and a general lack of consistency.
I left RH long ago, when I was far less knowledgable. I was more
successful at install than you, but never felt I had any chance of
gaining control of my computer within the RH environment. I have no
desire to go back. So, out of curiosity, what is RPM dependency hell?
I'm not interested enough to find out for myself. It sort of sounds like
'what does it feel like to hit your thumb really hard with a hammer?'
i.e. the sort of question for which direct personal knowledge is best
avoided. So, what is it?
Try this for an experiment: pick a package. The 'openbox' window
manager is probably a good pick. Find it on your favorite Debian
mirror; take, for example,
ftp://ftp.debian.org/debian/pool/main/o/openbox/. Hmm, lots of files
there. Bet you want the newest one. So download
openbox_3.0-1_i386.deb, and try installing it with 'dpkg --install'.
Uh-oh, you're missing lots of dependencies. So return to the FTP site
and try to download and install those. Lather, rinse, repeat. The
experience with Red Hat is fundamentally the same, except you're using
rpm instead of dpkg, rufus.w3.org is hosed into the ground, and for
any package there are six subtly different versions built for every
Linux distribution but your own.
Debian has always dealt better with this particular case; if you
installed something in dselect, even before there was APT, you'd get
all of the dependencies. It now also happens to be easy to do this
from the command line ('aptitude install openbox').
Could this happen to .deb packaging? Suppose people started producing .deb
packages that were specifically designed for use with Knoppix, or Libranet,
or whatever? It seems that .rpm in and of itself is not the problem.
Also, I wonder why the RH powers never tried to copy dselect, apt-get,
etc. Surely, they were aware of their existence.
I think you're right in saying that it's not the packaging system but
how it's used - the good thing about debian is that all debian packages
are debian packages, working well with the system, some debian
maintainer taking care of them.
packages that are external to debian should never ever provide .deb
that would install itself into system directories (/etc, /usr ...), they
should religiously use /opt, possibly creating links in /usr/local.
Otherwise debian becomes redhat-like hell of screwed up packages.
that's why I don't like alien (or other ways to install rpms) - I'd
much rather have tarball (either binary that goes into
/opt/package-version or 'normal' source tarball that I can ./configure
--prefix=/opt/package-version && make && make install)
erik
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: xterm - selecting URL
Nick Welch wrote: On Thu, Nov 20, 2003 at 03:24:07AM +, Antony Gelberg wrote: Hi all, I miss one thing from gnome. gnome-terminal let me ctrl-click on a url, and open mozilla. Any other xterm replacements that can do this? But I don't want to install all the gnome-terminal dependencies if I can use something else... I'm not aware of any other terminal that "sees" urls and lets you simply click on them other than ... er.. I *thought* konsole, but I'm checking it out right now and it doesn't seem to do it. So I know of no others. :) But what you *can* do, is map a keybinding to load the currently selected url. Install 'xsel', a small command line utility that lets you interact with the clipboard, and you could map a keybinding to do something (I believe) like this: mozilla -remote `xsel` As far as creating that keybinding.. it depends on your window manager / desktop environment. you can also middle click into mozilla window, it will open the new window with URL being whatever is in current selection. It's not exactly the same as Antony wants but I find it good enough for me... erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Experiment: Neophyte versus Windows XP & Debian Woody
csj wrote: On November 20, 2003 at 11:53AM -0800, Erik Steffl <[EMAIL PROTECTED]> wrote: [...] that's why I don't like alien (or other ways to install rpms) - I'd much rather have tarball (either binary that goes into /opt/package-version or 'normal' source tarball that I can ./configure --prefix=/opt/package-version && make && make install) I haven't come across packages that install into /opt. For source packages, I use ./configure --prefix=/usr/local/stow/packagename and have "stow" handle the symlinking to /usr/local/bin, etc. exactly shouldn't that be a requirement? if the package is not signed by a debian maintainer it can only touch /opt (and possibly /usr/local) or something like that? (debian package system would create a directory in opt, make it writable by some user, sudo into that user, run install of particular package or something like that) at this point my personal rule is to install either official debian package or force the install into /opt (nicely or not so nicely:-) erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: mozilla mail freezes on start
Erik Steffl wrote: last mozilla 1.4 worked OK, now that I upgraded to 1.5-2 when I open mail window it just freezes (all mozilla windows are completely frozen, not updated at all). I have an IMAP server (imaps) and few nntp servers configured. I can connect to IMAP server (cyrus) using other clients - so far I tried mutt and evolution, both work. any ideas on how to troubleshoot this or what's the problem? am I the only one with this problem? I also filed a bug, no response either... I guess that means there's something offending in my prefs (I've seen it before, mozilla seems to be fairly senstive to prefs) ? tia erik -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
evolution usability (somewhat OT)
(unstable debian, evolution)
I find evolution to be fairly nice MUA but:
- is there a way for it to move deleted mails to trash?
- is there a way for it not to show deleted emails (it shows them
striked out)
- is there a way to go to next unread mail, even if it happens to be
in next folder (']' goes to next unread email in current folder)
tia
erik
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: evolution usability (somewhat OT)
On Thu, 2003-11-20 at 20:39, Greg Folkert wrote:
> On Thu, 2003-11-20 at 17:58, Erik Steffl wrote:
> >(unstable debian, evolution)
> >
> >I find evolution to be fairly nice MUA but:
> >
> >- is there a way for it to move deleted mails to trash?
>
> CTRL-E will expunge. It is also under the Actions menu.
I meant for the delete action to move the messages to trash folder.
when I hit ctrl-e the messages are gone (didn't find them in trash
folder)
> >- is there a way for it not to show deleted emails (it shows them
> > striked out)
>
> Under the View Menu: Hide Deleted Messages
aha! and I thought I went through menu and preferences:-)
>
> >- is there a way to go to next unread mail, even if it happens to be
> > in next folder (']' goes to next unread email in current folder)
>
> Um... look at setting up a virtual Folder (VFolders)
not sure how that could do what I want but I'll try it.
> Hope this helps. Explore a bit and you might be surprised.
I hope so!
erik
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: evolution usability (somewhat OT)
On Thu, 2003-11-20 at 22:50, David Palmer. wrote:
> On Thu, 20 Nov 2003 14:58:24 -0800
> Erik Steffl <[EMAIL PROTECTED]> wrote:
>
> >(unstable debian, evolution)
> >
> >I find evolution to be fairly nice MUA but:
> >
> >- is there a way for it to move deleted mails to trash?
> >
> >- is there a way for it not to show deleted emails (it shows them
> > striked out)
> >
> >- is there a way to go to next unread mail, even if it happens to
> >be
> > in next folder (']' goes to next unread email in current folder)
> >
> >tia
> >
> > erik
> >
> Hello Erik,
>
> Go to the 'view' menu, and click 'hide deleted messages.'
> That will solve your first two points.
they don't seem to go to trash (I would like the emails from IMAP
server go to trash on the same server, just like it can be set in
mozilla)
> I'm not sure what you mean exactly, on the third point.
let's say I have new messages in inbox and in debian-user folders. So
I go to inbox and view the messages, ']' goes to next unread message.
After I read the last message in inbox I would like to go to the next
unread message in debian-user folder, unfortunately ']' doesn't do it
and I didn't find any shortcut/menu item that would do that (compare to
mozilla that asks whether I want to go to the next unread message in
debian-user folder).
erik
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
