Re: Should this be filed as grave? Gcc-2.95

[Steve Lamb]

On Tue, 5 Aug 2003 00:25:27 -0400
Daniel Jacobowitz <[EMAIL PROTECTED]> wrote:
> I fail to see how 2.95 installing both 3.3 and 2.95 somehow equates to
> a problem!

A failed kernel compile when trying to bring stability to a machine
constitutes as a problem in my book.  

> I build kernels with alternate compilers all the time.  Did you check
> the log to see which compiler the kernel actually built with?

Given that I told it to build with 2.95 and it failed in the same manner
as with 3.3 but when I installed 2.95 from Woody which ONLY installs 2.95 it
succeeded I, quite frankly, don't care if it compiled with 1.10.0.101.10.2. 
2.95 should install what it says it installs, 2.95.  Debian has version
numbers in the names for a reason and that reason being when people NEED the
previous version and not to upgrade to the current one.

See the whole thread about exim vs exim4 as reference.

-- 
 Steve C. Lamb | I'm your priest, I'm your shrink, I'm your
   PGP Key: 8B6E99C5   | main connection to the switchboard of souls.
   |-- Lenny Nero - Strange Days
---+-


pgpObY4Tf4W0w.pgp
Description: PGP signature

Re: libraries being removed from the archive

[Matthias Urlichs]

Hi, Peter Mathiasson wrote:

> "[...] distcc sends the  complete preprocessed source code across
> the network for each job."

Hmm, OK, but that would just speedup the actual compilation. Granted,
that's the largest chunk, but cpp/asm/ld could do with a speed-up too.

Anyway, thanks for the pointers to existing software everybody.

-- 
Matthias Urlichs   |   {M:U} IT Design @ m-u-it.de   |  [EMAIL PROTECTED]
Disclaimer: The quote was selected randomly. Really. | http://smurf.noris.de
-- 
If Columbus had had an advisory committee he would probably still be at
the dock.
-- Justice Arthur Goldberg

Debconf in Porto Alegre (Re: debconf 2005 in Vienna, Austria)

[Benj. Mako Hill]

On Wed, Jul 30, 2003 at 07:37:29PM +, Michelle Ribeiro wrote:
> Em 30 Jul 2003 15:09:51 -0400
> Joe Drew <[EMAIL PROTECTED]> escreveu:
> 
> > I've been waiting for them to speak up as they have said they
> > would. I am in favour of holding Debconf 4 in South America, but
> > will organise it in Vancouver if necessary.
> 
> Yeah, we would like to host the next Debconf in Brazil, Porto
> Alegre, some days before the International Free Software
> Forum. Thus, devels can join us at this event, if they like.

With a list set up, discussion planning, and funding research well
underway and with the most prominent Debconf4 organizer up until now
(Joe Drew) in support, I think the Brazilians have their foot in the
door on this one. There was a bit of discussion about doing it again
next year at Brazil at Debconf3 with a lot of positive feedback. Doing
it with the IFSF will cut down on travel costs and increase
sponsorship possibilities.

Personally, I really like to see South America host an official
Debconf. With a bit of searching the tickets can be reasonable from
both the North American and Europe. Once you're there, you will make
it up quickly because you won't spend much while you are there. 

It would be nice to give the folks in South America, who have been
largely unrepresented at the past Debconfs, a chance to
participate. Brazil has been extremely active in the Free Software
arena and much of the rest of the world hears remarkably little.

> Before make a "official proposal", we are checking for free food and
> accommodations, plane tickets and some government support.

This seems extremely reasonable. I hope this all works out well and am
looking forward to hearing more about this.

> As the dollar is 1,00 to 2.89 real (local money) this travel should
> be very cheap. :)

From what I've been told, this means that in practical terms you can
go out to dinner and eat and drink as much as you can and have trouble
dropping more than 12USD.

But IMHO, the best part about having Debconf in Brazil that is those
those us that go to Debconf also get to go to Brazil. :)

Regards,
Mako

-- 
Benj. Mako Hill
[EMAIL PROTECTED]
http://mako.yukidoke.org/



pgpxRVHX0GcYl.pgp
Description: PGP signature

Re: Should this be filed as grave? Gcc-2.95

[Aaron Lehmann]

On Mon, Aug 04, 2003 at 09:14:08PM -0700, Steve Lamb wrote:
> Uh, no.  I am aware of that.  That, however, did not prevent it from
> running the wrong GCC.  v2.4.21 of the kernel had a problem with 3.3.  It
> would die repeatedly on the same line in ide-cd.h.  I did tell make to use
> gcc-2.95 and it failed on the exact same line.  Removing gcc, which is 3.3,
> gcc-2.95 which depended on 3.3 (this is NOT 2.95 in my eyes) and then
> installing the packages from woody did allow me to recompile that version of
> the kernel.

What exactly does gcc-2.95 -v say? It could be a different version on
2.95 than what's in woody.

Re: Should this be filed as grave? Gcc-2.95

[Keith Dunwoody]

Steve Lamb wrote:

I build kernels with alternate compilers all the time.  Did you check
the log to see which compiler the kernel actually built with?

Given that I told it to build with 2.95 and it failed in the same manner
as with 3.3 but when I installed 2.95 from Woody which ONLY installs 2.95 it
succeeded I, quite frankly, don't care if it compiled with 1.10.0.101.10.2. 
2.95 should install what it says it installs, 2.95.  Debian has version
numbers in the names for a reason and that reason being when people NEED the
previous version and not to upgrade to the current one.

I have managed to compile using gcc-2.95, using the gcc-2.95 package from 
testing a few weeks ago.  I'm sure that it was actually gcc-2.95 that was used, 
since the program I was compiling has a bug which manifests itself on gcc 3.x, 
but not 2.95.  And it worked when I used gcc-2.95, (from the debian package), 
but not with gcc-3.x (also from the debian packages).

-- Keith

Re: libraries being removed from the archive

[Aaron Lehmann]

On Tue, Aug 05, 2003 at 01:07:42AM +0400, Nikita V. Youshchenko wrote:
> So buidd + distcc on a slow m68k/arm/whatever, and distccd on a fast P4 or
> Athlon, or even on several of those. This is expected to reduce the compile
> time to almost the same as it is on x86 :).

I'm not sure that's true; but it would be interesting to know for
sure. Many packages spend a lot of time running tests and generating
documentation. These would still take forever. Also, when using
distcc, the m68k is still responsible for preprocessing source,
assembling compiler output, and linking objects. These operations take
up a nontrivial amount of time on m68k. Plus, the buildd's currently
install every non-essential build dependency before building a
specific package, and remove them after. Running dpkg is slow on m68k,
especially if you're installing a giant chain of build dependencies
like GNOME or KDE.

I'm sure distcc would help though ;).

Bug#204177: ITP: sec-rpc -- rpc-proxy that allows tunneling of nis and nfs over ssh

[Fabian Franz]

Package: wnpp
Version: N/A; reported 2003-08-05
Severity: wishlist

* Package name: sec-rpc
  Version : 1.5.2
  Upstream Author : John Bowman <[EMAIL PROTECTED]>
* URL : http://www.math.ualberta.ca/imaging/snfs/
* License : GPL
  Description : rpc-proxy that allows tunneling of nis and nfs over ssh

ecure versions of the Network File System (SNFS) and Network Information
System (SNIS) have now been implemented via SSH2 tunneling of UDP
datagrams, as described in the SSH FAQ. This tunneling software is
available for download under the GPL License: sec_rpc-1.52.tar.gz. This
is a major enhancement of the original sec_rpc package developed by
Holger Trapp. 

-- System Information
Debian Release: testing/unstable
Architecture: i386
Kernel: Linux voyager 2.4.20-xfs #1 SMP Mit Jan 29 18:47:59 CET 2003 i586
Locale: [EMAIL PROTECTED], [EMAIL PROTECTED]

Re: How to install X-Chat in five hours (or more)

[Ian Hickson]

Hey guys,

I was amused to see my blog post [1] made it to this list. I figured
I'd clarify a few points which were omitted from that blog in the
interests of brevity and humour.


Mike Hommey <[EMAIL PROTECTED]> wrote:
>
> Unfortunately, his main problem is "Having not used Debian for about
> 8 years".

Most of your new users (all of them in fact, by definition!) will
never have used Debian before. This should not stop them using your
product.

Part of the problem I had was that I had a vague understanding that
there was something called "apt", but that I didn't know what it was
or how to do anything with it. The man page said to see apt-get's;
apt-get's man page suggested the tool was a back-end but didn't really
give any clues as to what front end to use.


> The strange thing is that he has been able to apt-get install
> aptitude, but tryed something else for xchat at first...

I did actually try using apt-get originally when I had my original
libgtk1.2 problem:

   # apt-get libgtk1.2
   E: Invalid operation libgtk1.2

I tried to understand this:

   # man apt-get
   /E: Invalid
   Pattern not found (press RETURN)

After reading the whole man page, I did try apt-get install libgtk1.2,
but then I got the whole conflict problem I mention later in my blog.
I decided to get the source and do it myself, but since I didn't know
where that would be, I switched to trying to get X-Chat, and since I
was already in source mode, it didn't even cross my mind to use
apt-get for that. (After all, apt-get had just failed me, why would it
succeed now?)

To the end user (me), apt-get is arbitrarily verbose. "Selecting
previously deselected package libbla3.2"? "Get:1 ftp://apt sid/main
libbla3.2 3.2.10-9 [827kB]"?

Look at operating systems used by less intelligent users. They just
see:

   [# ] 60%  2 minutes remaining

Admittedly, this is usually followed by a crash, but the user is not
intimidated. (Having said that, personally, I quite like seeing the
verbose output, it's useful for debugging. Most users don't.)


> And another thing : it seems that the pre-installed Debian he got
> was configured with both testing/unstable in the sources.list file.
> Pinning is not the easiest thing to catch when you are (alone)
> beginner with Debian...

Matt Zimmerman <[EMAIL PROTECTED]> wrote:
>
> It's also a really stupid trick to pull on someone for whom you are
> installing a system where they hope to get actual work done.

Yeah. Unfortunately, to support my radeon chipset I have to use the
unstable version apparently.


> (though it's equally possible that he did this himself, making random
> changes without understanding them in the hope of making things work)

Nope, I wouldn't even have known where to start with respect to using
unstable releases if it wasn't for the kind people on #mozilla. :-)


Nikolai Prokoschenko <[EMAIL PROTECTED]> wrote:
>|
>| I used apt-get to get aptitude. I fired up aptitude.
>
> The writer is obviously a moron if he did this with such ease and it
> never occured to him he could've done the same for any of the other
> packages he needed.

While I wouldn't like to dispute my moronity, as I described above, I
had tried apt-get earlier with little success. I used it for aptitude
possibly because Eddy suggested I do that, or possibly because having
had a break, I was no longer locked on to the idea of getting the
source for the packages I wanted.


I think Debian's package system is remarkably nice. Unfortunately,
it's UI leaves a lot to be desired. The biggest problem is probably
the package names: "freetype", "pango", "libgtk2.0", etc, mean
absolutely nothing to me, as a user, and I really shouldn't ever have
to even see these packages.

I really think aptitude should only show end user packages with
decent, readable, localised names ("Apache Web Server", "x Chat (IRC
Client)", "Infrared Control for XMMS"). At the moment the user is
completely overwhelmed by the list of packages, which is not helped by
the fact that each one comes with a dozen or more libraries,
extensions, and so forth.

Anyway. Just a few thoughts. :-)

-- References --
[1] http://ln.hixie.ch/?start=1060025253&count=1

-- 
Ian Hickson  )\._.,--,'``.fL
"meow"  /,   _.. \   _\  ;`._ ,.
http://index.hixie.ch/ `._.-(,_..'--(,_..'`-.;.'

Re: How to install X-Chat in five hours (or more)

[Mike Hommey]

On Tuesday 05 August 2003 10:33, Ian Hickson wrote:
> Hey guys,
>
> I was amused to see my blog post [1] made it to this list. I figured
> I'd clarify a few points which were omitted from that blog in the
> interests of brevity and humour.
>
> Mike Hommey <[EMAIL PROTECTED]> wrote:
> > Unfortunately, his main problem is "Having not used Debian for about
> > 8 years".
>
> Most of your new users (all of them in fact, by definition!) will
> never have used Debian before. This should not stop them using your
> product.

What I meant with that is that your main problem is that you have an old linux 
user approach. A typical new user won't even try to build from sources.

[snip]
> After reading the whole man page, I did try apt-get install libgtk1.2,
> but then I got the whole conflict problem I mention later in my blog.

This is the problem with pinning. It is not easy to handle for beginners 
(especially if they are by themselves), and it is IMHO a very bad idea from 
whoever installed a Debian system on your laptop to have put different 
releases sources in the sources.list file.

[snip]
> (After all, apt-get had just failed me, why would it succeed now?)

Because Dr. Murphy sometimes forgets to annoy you ? ;)

[snip]
> Yeah. Unfortunately, to support my radeon chipset I have to use the
> unstable version apparently.

I think the "testing" version works at least in 2D for every radeon chip. What 
can't be denied, is that X is not the easiest part to configure on a Debian 
system...

> > (though it's equally possible that he did this himself, making random
> > changes without understanding them in the hope of making things work)
>
> Nope, I wouldn't even have known where to start with respect to using
> unstable releases if it wasn't for the kind people on #mozilla. :-)

When trying to solve debian issues, try #debian ;)

Mike

-- 
"Do you know what's the best thing about being me ? There's so many me ! "
-- Agent Smith Reloaded

Re: How to install X-Chat in five hours (or more)

[Frederik Rousseau]

> I think Debian's package system is remarkably nice. Unfortunately,
> it's UI leaves a lot to be desired. The biggest problem is probably
> the package names: "freetype", "pango", "libgtk2.0", etc, mean
> absolutely nothing to me, as a user, and I really shouldn't ever have
> to even see these packages.
>
> I really think aptitude should only show end user packages with
> decent, readable, localised names ("Apache Web Server", "x Chat (IRC
> Client)", "Infrared Control for XMMS"). At the moment the user is
> completely overwhelmed by the list of packages, which is not helped by
> the fact that each one comes with a dozen or more libraries,
> extensions, and so forth.

Some people like Ian Hickson don't want to see package names like libgtk2.0. 
But _I_ want to see this, I am sys admin and want to know exactly what I am 
putting on _my_ systems.

Just one of the reasons why I like GNU/Linux ... I know what I'm doing!

Anyway, does this mean we need something like a GNU/Linux Debian and a 
GNU/Linux Debian For Dummies showing only icons?


Fred
--
http://www.linox.be

L I N U X   .~.
   The  Choice  /V\
of a  GNU  /( )\
   Generation  ^^-^^

商务之家－网站建设咨询热线QQ：230414701

[商务之家]

商务之家网站：http://www.35home.com
商务之家论坛：http://bbs.35home.com

值班ＱＱ：230414701


-
注册国际域名：６５元（免费代办）
注册国内域名：１８０元（名费代办）

虚拟主机 
30M  asp 空间：60元
100M asp 空间：100元（９月１日前） 
50M  php 空间：70元（有5Mmysql空间，９月１日前） 
100  php 空间：100元（有mysql空间，９月１日前） 
组合餐1：国际域名+50M(ASP空间)+10邮箱===120元。(推荐） 
组合餐2：国际域名+20M(html空间)+10邮箱==100元。（推荐） 

商务中国的各项服务，请查看本站：http://www.35home.com，请注册试用空间。; 

如果其它问题直接和站长QQ：230414701联系。 

－

buildd probs sbcl/common-lisp-controller

[Neil Roeth]

What's the status of the sbcl/common-lisp-controller problems that were
blocking builds on sparc and other platforms?

-- 
Neil Roeth

Re: libraries being removed from the archive

[Nikita V. Youshchenko]

> On Tue, Aug 05, 2003 at 01:07:42AM +0400, Nikita V. Youshchenko wrote:
> > So buidd + distcc on a slow m68k/arm/whatever, and distccd on a fast
> > P4 or Athlon, or even on several of those. This is expected to reduce
> > the compile time to almost the same as it is on x86 :).
>
> I'm not sure that's true; but it would be interesting to know for
> sure. Many packages spend a lot of time running tests and generating
> documentation.

Maybe this can be compensated by running distccd on several fast x86 
machines, and run parallel make while package building? Distcc is known 
to scale almost lineary for small number of hosts. I use distcc at 
regular basis and can confirm this.

Perhaps someone who has access to both m68k and a fast x86 connected by a 
fast network should try this.

Re: Should this be filed as grave? Gcc-2.95

[Matt Zimmerman]

On Mon, Aug 04, 2003 at 09:14:08PM -0700, Steve Lamb wrote:

> On Mon, 4 Aug 2003 23:37:32 -0400
> Matt Zimmerman <[EMAIL PROTECTED]> wrote:
> > What you meant to do was to run "make CC=gcc-2.95" instead of make.  There
> > is no need to futz around with the default gcc version; just ask for what
> > you want.
> 
> Uh, no.  I am aware of that.  That, however, did not prevent it from
> running the wrong GCC.

Works fine for me.

*** End of Linux kernel configuration.
*** Check the top-level Makefile for additional configuration.
*** Next, you must run 'make dep'.

mizar:[.../linux/kernel-source-2.4.21] make CC=gcc-2.95
make[1]: Entering directory 
`/space/tmp/mdz/linux/kernel-source-2.4.21/arch/i386/boot'
make[1]: Nothing to be done for `dep'.
make[1]: Leaving directory 
`/space/tmp/mdz/linux/kernel-source-2.4.21/arch/i386/boot'
rm -f .depend .hdepend
make _sfdep_kernel _sfdep_drivers _sfdep_mm _sfdep_fs _sfdep_net _sfdep_ipc 
_sfdep_lib _sfdep_crypto _sfdep_arch/i386/kernel _sfdep_arch/i386/mm 
_sfdep_arch/i386/lib _FASTDEP_ALL_SUB_DIRS="kernel drivers mm fs net ipc lib 
crypto arch/i386/kernel arch/i386/mm arch/i386/lib"
make[1]: Entering directory `/space/tmp/mdz/linux/kernel-source-2.4.21'
make -C kernel fastdep
make[2]: Entering directory `/space/tmp/mdz/linux/kernel-source-2.4.21/kernel'
gcc-2.95 -D__KERNEL__ -I/space/tmp/mdz/linux/kernel-source-2.4.21/include -Wall 
-Wstrict-prototypes -Wno-trigraphs -O2 -fno-strict-aliasing -fno-common 
-fomit-frame-pointer -pipe -mpreferred-stack-boundary=2 -march=i686 
-malign-functions=4  -nostdinc -iwithprefix include -E -D__GENKSYMS__ signal.c
[...]

Yes, I know that's 2.4.21, but I'm not going to unpack a whole 2.4.20 tree
to demonstrate that it works the same way.  It does.

> v2.4.21 of the kernel had a problem with 3.3.  It would die repeatedly on
> the same line in ide-cd.h.  I did tell make to use gcc-2.95 and it failed
> on the exact same line.  Removing gcc, which is 3.3, gcc-2.95 which
> depended on 3.3 (this is NOT 2.95 in my eyes) and then installing the
> packages from woody did allow me to recompile that version of the kernel.

gcc-2.95 doesn't depend on 3.3; it depends on the "gcc" package, which
happens to be version 3.3 in unstable.  That package doesn't contain any
compilers; it just sets the default compiler and related tools, e.g.
/usr/bin/gcc.

> I fail to see how 2.95 installing 3.3 somehow equates to 2.95.

It doesn't.

-- 
 - mdz

Re: Should this be filed as grave? Gcc-2.95

[Steve Lamb]

On Tue, 5 Aug 2003 08:56:50 -0400
Matt Zimmerman <[EMAIL PROTECTED]> wrote:
> Yes, I know that's 2.4.21, but I'm not going to unpack a whole 2.4.20 tree
> to demonstrate that it works the same way.  It does.

I never said it didn't work.  What I said was that when I did it 2.4.20
had the same error as if it were compiled with 3.3.  So without compiling
2.4.20 w/ide-cd in it you're not replicating what I did and providing any
basis of comparison whatsoever.

-- 
 Steve C. Lamb | I'm your priest, I'm your shrink, I'm your
   PGP Key: 8B6E99C5   | main connection to the switchboard of souls.
   |-- Lenny Nero - Strange Days
---+-


pgpzCqOoIThkC.pgp
Description: PGP signature

Re: Should this be filed as grave? Gcc-2.95

[Matt Zimmerman]

On Tue, Aug 05, 2003 at 06:00:27AM -0700, Steve Lamb wrote:

> On Tue, 5 Aug 2003 08:56:50 -0400
> Matt Zimmerman <[EMAIL PROTECTED]> wrote:
> > Yes, I know that's 2.4.21, but I'm not going to unpack a whole 2.4.20 tree
> > to demonstrate that it works the same way.  It does.
> 
> I never said it didn't work.  What I said was that when I did it 2.4.20
> had the same error as if it were compiled with 3.3.

Then perhaps this particular problem was not with gcc 3.3.  I think some
additional investigation would be prudent before any talk about grave bugs.

-- 
 - mdz

Re: Should this be filed as grave? Gcc-2.95

[Daniel Jacobowitz]

On Mon, Aug 04, 2003 at 10:51:41PM -0700, Steve Lamb wrote:
> On Tue, 5 Aug 2003 00:25:27 -0400
> Daniel Jacobowitz <[EMAIL PROTECTED]> wrote:
> > I fail to see how 2.95 installing both 3.3 and 2.95 somehow equates to
> > a problem!
> 
> A failed kernel compile when trying to bring stability to a machine
> constitutes as a problem in my book.  
> 
> > I build kernels with alternate compilers all the time.  Did you check
> > the log to see which compiler the kernel actually built with?
> 
> Given that I told it to build with 2.95 and it failed in the same manner
> as with 3.3 but when I installed 2.95 from Woody which ONLY installs 2.95 it
> succeeded I, quite frankly, don't care if it compiled with 1.10.0.101.10.2. 
> 2.95 should install what it says it installs, 2.95.  Debian has version
> numbers in the names for a reason and that reason being when people NEED the
> previous version and not to upgrade to the current one.
> 
> See the whole thread about exim vs exim4 as reference.

And you're missing the point.

Installing gcc-2.95 does install GCC 2.95.

What else it installs is irrelevant.  It installs a particular version
of glibc that isn't gcc 2.95, too.

Please stop crusading, and find out what your kernel build actually
did.  Because it works just fine for all the rest of us.

-- 
Daniel Jacobowitz
MontaVista Software Debian GNU/Linux Developer

Re: How to install X-Chat in five hours (or more)

[David Z Maze]

Frederik Rousseau <[EMAIL PROTECTED]> writes:

>> I really think aptitude should only show end user packages with
>> decent, readable, localised names ("Apache Web Server", "x Chat (IRC
>> Client)", "Infrared Control for XMMS"). At the moment the user is
>> completely overwhelmed by the list of packages, which is not helped by
>> the fact that each one comes with a dozen or more libraries,
>> extensions, and so forth.
>
> Some people like Ian Hickson don't want to see package names like libgtk2.0. 
> But _I_ want to see this, I am sys admin and want to know exactly what I am 
> putting on _my_ systems.
>
> Just one of the reasons why I like GNU/Linux ... I know what I'm doing!
>
> Anyway, does this mean we need something like a GNU/Linux Debian and a 
> GNU/Linux Debian For Dummies showing only icons?

It probably means there should be a configuration option in aptitude
to hide sections like devel, libdevel, libs, and interpreters, since
end users typically don't care (and 90% of the time I find myself not
caring, too).  Perhaps suggesting to new users that they look in the
"tasks" section of aptitude would help reduce the package overload,
too.  I don't think we need to abandon the power of our current
infrastructure, just have ways of making it less visible for people
who don't want it.

-- 
David Maze [EMAIL PROTECTED]  http://people.debian.org/~dmaze/
"Theoretical politics is interesting.  Politicking should be illegal."
-- Abra Mitchell

Re: [Debconf] Re: The slides for my talk

[Andreas Tille]

On Wed, 23 Jul 2003, Branden Robinson wrote:

> Well, poor Joey Hess didn't notice it, and he's a native speaker.  ;-)
This does not necessarily help!  In Germany we all know that best English
is spoken by native German spaekars - at least we understand them best (because
we do mostly the same mistakes). ;-)
I guess this works for Italian people, too.

> I picked up on it, I think, because I'd been hanging around Enrico a
> fair amount and could tell he had a good sense of humor.
Definitely!

Kind regards

  Andreas, who just placed a symlink to my talks directory to the 
apropriate place.

gluck:/home> ls -l -d */public_html/talks
drwxr-xr-x3 dopeyDebian   4096 Jan 10  2002 dopey/public_html/talks
drwxr-xr-x5 enrico   Debian   4096 Jul 22 06:46 enrico/public_html/talks
drwxr-xr-x2 joerglan Debian   4096 Jul 14 07:55 
joergland/public_html/talks
drwxr-xr-x3 kelbert  Debian   4096 Jul 25 17:00 
kelbert/public_html/talks
drwxr-xr-x4 lolando  Debian   4096 Jun 12 15:00 
lolando/public_html/talks
drwxr-x---2 maxx Debian   4096 Mar 11 14:40 maxx/public_html/talks
drwxr-xr-x3 mjb  Debian   4096 Jul 22 09:53 mjb/public_html/talks
lrwxrwxrwx1 tilleDebian 16 Aug  5 08:03 tille/public_html/talks 
-> debian-med/talks

Re: How to install X-Chat in five hours (or more)

[Ross Burton]

On Tue, 2003-08-05 at 15:04, David Z Maze wrote:
> I don't think we need to abandon the power of our current
> infrastructure, just have ways of making it less visible for people
> who don't want it.

Just a random off-the-wall idea, but *maybe* there could be a new
package tag added which means that a package is something that Joe User
is likely to use, i.e. a good word processor/web browser/etc.  Then a
minimal graphical interface could be built to show just these.

Of course, the flames when deciding what packages to tag would be
huge...

Ross
-- 
Ross Burton mail: [EMAIL PROTECTED]
  jabber: [EMAIL PROTECTED]
 www: http://www.burtonini.com./
 PGP Fingerprint: 1A21 F5B0 D8D0 CFE3 81D4 E25A 2D09 E447 D0B4 33DF


signature.asc
Description: This is a digitally signed message part

Re: Should this be filed as grave? Gcc-2.95

[Steve Lamb]

On Tue, 5 Aug 2003 09:25:38 -0400
Matt Zimmerman <[EMAIL PROTECTED]> wrote:
> Then perhaps this particular problem was not with gcc 3.3.  I think some
> additional investigation would be prudent before any talk about grave bugs.

Which is why I asked here first before just filing.

-- 
 Steve C. Lamb | I'm your priest, I'm your shrink, I'm your
   PGP Key: 8B6E99C5   | main connection to the switchboard of souls.
   |-- Lenny Nero - Strange Days
---+-


pgpw6uX16u2UM.pgp
Description: PGP signature

Re: Should this be filed as grave? Gcc-2.95

[Steve Lamb]

On Tue, 5 Aug 2003 09:33:53 -0400
> Please stop crusading, and find out what your kernel build actually
> did.  Because it works just fine for all the rest of us.

Who's crusading?  I am pointing out what I see as an apparent problem for
discussion.  Crusading would be to file the damned bug without discussion and
defending it to the last breath.  Calm down!

-- 
 Steve C. Lamb | I'm your priest, I'm your shrink, I'm your
   PGP Key: 8B6E99C5   | main connection to the switchboard of souls.
   |-- Lenny Nero - Strange Days
---+-


pgpClj4e8XXDx.pgp
Description: PGP signature

Re: How to install X-Chat in five hours (or more)

[Hamish Moffatt]

On Tue, Aug 05, 2003 at 01:33:19AM -0700, Ian Hickson wrote:
> Part of the problem I had was that I had a vague understanding that
> there was something called "apt", but that I didn't know what it was
> or how to do anything with it. The man page said to see apt-get's;
> apt-get's man page suggested the tool was a back-end but didn't really
> give any clues as to what front end to use.

Not directly, but the "SEE ALSO" list does include dselect(8) which is
what you really should have used.

[...]
> To the end user (me), apt-get is arbitrarily verbose. "Selecting
> previously deselected package libbla3.2"? "Get:1 ftp://apt sid/main
> libbla3.2 3.2.10-9 [827kB]"?
> 
> Look at operating systems used by less intelligent users. They just
> see:
> 
>[# ] 60%  2 minutes remaining

I don't see how some extra verbosity hurts. apt-get still displays a
percentage and a time estimate.

Frankly if verbosity loses us some users, too bad. I'm sure we pick up
more users because of the same. To rant a bit, the thing that bugs me
the most about MS Windows is how when it breaks randomly you can't fix
it because it runs on smoke and mirrors and doesn't give helpful
information on what went wrong. With UNIX/Linux you get details
and you can fix it.

> I think Debian's package system is remarkably nice. Unfortunately,
> it's UI leaves a lot to be desired. The biggest problem is probably

Which UI did you use? We have a few. apt-get is not an interface for the
Debian newbie. dselect and aptitude are GUI tools if that's what you
need.


Hamish
-- 
Hamish Moffatt VK3SB <[EMAIL PROTECTED]> <[EMAIL PROTECTED]>

Re: Should this be filed as grave? Gcc-2.95

[Matt Zimmerman]

On Mon, Aug 04, 2003 at 09:20:21PM -0700, Steve Lamb wrote:

> On Mon, 4 Aug 2003 21:14:08 -0700
> Steve Lamb <[EMAIL PROTECTED]> wrote:
> > Uh, no.  I am aware of that.  That, however, did not prevent it from
> > running the wrong GCC.  v2.4.21 of the kernel had a problem with 3.3.
> 
> Correction, 2.4.20.  For some reason 2.4.21 seems to be crashing my system
> hard every few minutes/hours/days (pick one).

Don't compile your kernel with gcc 3.3.  I don't know whether the bugs lie
in the kernel or in gcc (or both), but this combination does not work
correctly.

-- 
 - mdz

Re: Should this be filed as grave? Gcc-2.95

[Steve Lamb]

On Tue, 5 Aug 2003 10:54:38 -0400
Matt Zimmerman <[EMAIL PROTECTED]> wrote:
> Don't compile your kernel with gcc 3.3.  I don't know whether the bugs lie
> in the kernel or in gcc (or both), but this combination does not work
> correctly.

   Yeah.  That was the whole reason I was trying to get a copy of 2.4.20
compiled with gcc 2.95.  I didn't know if it was the compiler or the newer
version of the kernel that had the problem.  I just knew that my problems
started with the newer version.  If 2.4.20 is stable for 2 weeks I'll
move it to my "stable" boot option, compile 2.4.21 w/gcc 2.95 and install it
as the current and give it a whirl.

-- 
 Steve C. Lamb | I'm your priest, I'm your shrink, I'm your
   PGP Key: 8B6E99C5   | main connection to the switchboard of souls.
   |-- Lenny Nero - Strange Days
---+-


pgpAa52aqzGqG.pgp
Description: PGP signature

Re: Should this be filed as grave? Gcc-2.95

[Matt Zimmerman]

On Tue, Aug 05, 2003 at 07:59:20AM -0700, Steve Lamb wrote:

>Yeah.  That was the whole reason I was trying to get a copy of 2.4.20
> compiled with gcc 2.95.  I didn't know if it was the compiler or the newer
> version of the kernel that had the problem.  I just knew that my problems
> started with the newer version.  If 2.4.20 is stable for 2 weeks I'll
> move it to my "stable" boot option, compile 2.4.21 w/gcc 2.95 and install it
> as the current and give it a whirl.

2.4.21 built with gcc 2.95 has been rock solid for me.

-- 
 - mdz

Re: Should this be filed as grave? Gcc-2.95

[H. S. Teoh]

On Tue, Aug 05, 2003 at 07:59:20AM -0700, Steve Lamb wrote:
> On Tue, 5 Aug 2003 10:54:38 -0400
> Matt Zimmerman <[EMAIL PROTECTED]> wrote:
> > Don't compile your kernel with gcc 3.3.  I don't know whether the bugs lie
> > in the kernel or in gcc (or both), but this combination does not work
> > correctly.
> 
>Yeah.  That was the whole reason I was trying to get a copy of 2.4.20
> compiled with gcc 2.95.  I didn't know if it was the compiler or the newer
> version of the kernel that had the problem.  I just knew that my problems
> started with the newer version.  If 2.4.20 is stable for 2 weeks I'll
> move it to my "stable" boot option, compile 2.4.21 w/gcc 2.95 and install it
> as the current and give it a whirl.
[snip]

Did you check your compile logs to see if it actually compiled with
gcc-2.95 or with just gcc (==3.3) ? It happened to me several times that
when building 2.4.21, it would use gcc-2.95 for the initial configuration
and cleanup targets (since I specified CC=gcc-2.95), but revert to gcc for
the actual build. 

I had to hand-edit kernel makefiles to stop it from using gcc by default
and use gcc-2.95 instead. Or perhaps try setting CC=gcc-2.95 in your
environment before running the build. 


T

-- 
We are in class, we are supposed to be learning, we have a teacher... Is it
too much that I expect him to teach me??? -- RL

Re: How to install X-Chat in five hours (or more)

[Ian Hickson]

On Wed, 6 Aug 2003, Hamish Moffatt wrote:
>
> On Tue, Aug 05, 2003 at 01:33:19AM -0700, Ian Hickson wrote:
> > Part of the problem I had was that I had a vague understanding that
> > there was something called "apt", but that I didn't know what it was
> > or how to do anything with it. The man page said to see apt-get's;
> > apt-get's man page suggested the tool was a back-end but didn't really
> > give any clues as to what front end to use.
>
> Not directly, but the "SEE ALSO" list does include dselect(8) which is
> what you really should have used.

The term "dselect" means nothing to me. It isn't a usable name. That's
another example of the problem I mentioned.

Would it not be possible for debian to have a command "setup" or "install"
or something similarly named?

Note that, if for some reason the user knew about the command
"apropos", even that wouldn't help him -- none of dselect, aptitude,
and apt-get come up for "apropos install" or "apropos setup".


>> To the end user (me), apt-get is arbitrarily verbose. "Selecting
>> previously deselected package libbla3.2"? "Get:1 ftp://apt sid/main
>> libbla3.2 3.2.10-9 [827kB]"?
>>
>> Look at operating systems used by less intelligent users. They just
>> see:
>>
>>[# ] 60%  2 minutes remaining
>
> I don't see how some extra verbosity hurts.

It hurts because it scares users. My dad would take one look at the
text, and give up. (And 15 years ago he was a VMS administrator, so
it's not like he's computer illiterate.) My mum wouldn't even give the
text a chance, she'd just see a wad of text with odd punctuation and
run for the hills.


> Frankly if verbosity loses us some users, too bad. I'm sure we pick
> up more users because of the same.

You will lose many more than you will gain, since there are many more
computer illiterate users than geeks.


> To rant a bit, the thing that bugs me the most about MS Windows is
> how when it breaks randomly you can't fix it because it runs on
> smoke and mirrors and doesn't give helpful information on what went
> wrong. With UNIX/Linux you get details and you can fix it.

Just to clarify, I've nothing against verbosity itself. /var/log, for
instance, is great (although "var" is a historical name that really
should be replaced by something more user friendly, but that's another
story). The problem is verbosity when things don't go wrong.

I'm all for a "tell me what is going on" feature for debugging.

Even then, though, it would be nice if the verbose messages were
consistently formatted, and used plain english instead of jargon.
Error messages like "E: Invalid operation foo" are not helpful.


>> I think Debian's package system is remarkably nice. Unfortunately,
>> it's UI leaves a lot to be desired. The biggest problem is probably
>
> Which UI did you use? We have a few. apt-get is not an interface for the
> Debian newbie. dselect and aptitude are GUI tools if that's what you
> need.

I used aptitude. It is not easy to use. It's fine for me, as I'm a
geek. But if I told my mum to load aptitude and install X, she
wouldn't have a clue how to do it.

I just tried using deselect, to see if it is any better.

The first option I'm faced with is:

   * 0. [A]ccessChoose the access method to use.

I have no idea what that means. I tried using it (not logged in as
root) and I got the following message:

   dselect: requested operation requires the superuser privilege

Yet another example of an obscure error message. :-)

-- 
Ian Hickson  )\._.,--,'``.fL
"meow"  /,   _.. \   _\  ;`._ ,.
http://index.hixie.ch/ `._.-(,_..'--(,_..'`-.;.'

Re: How to install X-Chat in five hours (or more)

[Steve Lamb]

On Tue, 5 Aug 2003 08:14:23 -0700 (PDT)
Ian Hickson <[EMAIL PROTECTED]> wrote:
> Note that, if for some reason the user knew about the command
> "apropos", even that wouldn't help him -- none of dselect, aptitude,
> and apt-get come up for "apropos install" or "apropos setup".

I do believe they are mentioned several times in the manual.  Y'know, that
thing collecting dust over yonder.
 
> It hurts because it scares users.

And?

> You will lose many more than you will gain, since there are many more
> computer illiterate users than geeks.

And?  There are a slew of other OSs and Linux distributions as well.  

This is what always gets me in discussions that are based on the lowest
common denominator in computer users; the presumption that everyone wants to
cater to them.

It has been my experience that packages that cater to the lowest common
denominator are packages I don't care to use because I find them *hard to use*
even though they purport to be "easy to use".  I have almost always been on
the side of the scale where I preferred the "harder" package because in the
long run, once I got over the nominal learning curve, it was easier to use. 
However there was always that drive to go for the lowest common denominator,
the computer illiterate.  It has ruined more programs that I care to list
because they would add too much, dumb things down, make the program too large
and hard-to-use in the quest to get people who have to ask "Left or right
click" after the first time you tell them to right-click something.  IE,
people with no concept of "default behaviors/actions".

I have never, EVER understood why anyone would want to take a package
which is beloved by the niche geek market and destroy it for the illiterate
market.  That is triply so for commercial packages.  

You're wrong in saying that we (in general) would gain more by making our
package (program, OS, etc) more palatable to the neophyte.  As I said above,
there are a slew of OSs and distributions that cater to that segment.  To move
Debian into that realm would be to lose what it has and compete with
established entities.  100% of this slice is better than <1% of *that* slice.
 
> Just to clarify, I've nothing against verbosity itself. /var/log, for
> instance, is great (although "var" is a historical name that really
> should be replaced by something more user friendly, but that's another
> story). The problem is verbosity when things don't go wrong.

Erm, no, it should not be.  While it is a historical name it is a name
that should remain because every person who's ever worked on a Unix-like
system during that history knows where /var is, why it is there and what is in
it.  It is up to those new people to catch up, not for us to ruin what works
and works well in the vain attempt to catch more of a market which, in the
end, doesn't really matter as this is not a commercial venture.

> I'm all for a "tell me what is going on" feature for debugging.

Which is why you need verbosity when nothing is going wrong.  Let's see a
show of hands on this situation.

Ya boot up a Windows box post after '95.  Here's the sum of it letting you
know something is going on: a rotating palette for the bar at the bottom.  The
palette stops rotating.  So, uh, what's wrong?  Oh, wait, it started rotating
again.  No, wait, it stopped again... for 30 seconds.  No, there it goes, it's
fine.

> Even then, though, it would be nice if the verbose messages were
> consistently formatted, and used plain english instead of jargon.
> Error messages like "E: Invalid operation foo" are not helpful.

No, that's a bad idea.  Take a look at IE's 404 message sometime.  It's a
dumbed down version which doesn't explain jack or shitte.  Error messages are
there for people who know what they need to do.  People who don't know what
they need to do will not have that knowledge suddenly imparted upon them by a
"plain english" error message because, without the jargon to point you in the
right direction, there would be absolutely no place to start.

> The first option I'm faced with is:
 
>* 0. [A]ccessChoose the access method to use.
 
> I have no idea what that means. I tried using it (not logged in as
> root) and I got the following message:

Did you choose it to find out?

>dselect: requested operation requires the superuser privilege
 
> Yet another example of an obscure error message. :-)

Uh, no, it isn't.  Superuser, aka, root.  But not always root.  sudo can
grant superuser access w/o root.  Also any account with a uid of 0 has
superuser access but that doesn't mean it is called root.  I recall one job
where we had root and jfroot.  Both were uid 0 but they had different
passwords.  Don't ask me why we had to 0s w/different passwords.  Didn't make
sense to me then, doesn't make sense to me now.

But the really ironic part about all this is that the above message is
more of the "plain english" message you want.  Root i

Re: Should this be filed as grave? Gcc-2.95

[Steve Lamb]

On Tue, 5 Aug 2003 11:06:26 -0400
"H. S. Teoh" <[EMAIL PROTECTED]> wrote:
> Did you check your compile logs to see if it actually compiled with
> gcc-2.95 or with just gcc (==3.3) ? It happened to me several times that
> when building 2.4.21, it would use gcc-2.95 for the initial configuration
> and cleanup targets (since I specified CC=gcc-2.95), but revert to gcc for
> the actual build. 

That is most likely what happened.  I didn't check logs.  Didn't care.  It
didn't work, fuggit, I needed the machine stable and was mighty pissed that I
couldn't just rip 3.3 off the damned system to force the issue without
resorting to a serious downgrade to woody packages just to do it.

> I had to hand-edit kernel makefiles to stop it from using gcc by default
> and use gcc-2.95 instead. Or perhaps try setting CC=gcc-2.95 in your
> environment before running the build. 

Might have worked but forcing the issue is better IMHO.  Without 3.3
present there is absolutely no chance of some process down the line sanitizing
the environment and monkeying up the works.

-- 
 Steve C. Lamb | I'm your priest, I'm your shrink, I'm your
   PGP Key: 8B6E99C5   | main connection to the switchboard of souls.
   |-- Lenny Nero - Strange Days
---+-


pgpmQil8My3gr.pgp
Description: PGP signature

Re: Should this be filed as grave? Gcc-2.95

[Adam Heath]

On Tue, 5 Aug 2003, Steve Lamb wrote:

> On Tue, 5 Aug 2003 11:06:26 -0400
> "H. S. Teoh" <[EMAIL PROTECTED]> wrote:
> > Did you check your compile logs to see if it actually compiled with
> > gcc-2.95 or with just gcc (==3.3) ? It happened to me several times that
> > when building 2.4.21, it would use gcc-2.95 for the initial configuration
> > and cleanup targets (since I specified CC=gcc-2.95), but revert to gcc for
> > the actual build.
>
> That is most likely what happened.  I didn't check logs.  Didn't care.  It
> didn't work, fuggit, I needed the machine stable and was mighty pissed that I
> couldn't just rip 3.3 off the damned system to force the issue without
> resorting to a serious downgrade to woody packages just to do it.

If you need the machine stable, then why are you running testing or unstable?
gcc 3.3 isn't in stable/woody.

Plus, this sounds like a kernel bug not honoring your CC cmdline var.
Bitching to debian will not help you, and just annoys us.

> > I had to hand-edit kernel makefiles to stop it from using gcc by default
> > and use gcc-2.95 instead. Or perhaps try setting CC=gcc-2.95 in your
> > environment before running the build.
>
> Might have worked but forcing the issue is better IMHO.  Without 3.3
> present there is absolutely no chance of some process down the line sanitizing
> the environment and monkeying up the works.

Forcing the issue?  What are you forcing?  We can't fix this.

Besides, what stability issues are you having?  Hardware?  If so, test it
better before deploying, or if it has gone flaky after deployment, buy
replacement hardware.

Re: Should this be filed as grave? Gcc-2.95

[Steve Lamb]

Adam, where does it say anywhere in my sig or headers that I want a CC?  I
read the list just fine, you can reply to the list and only the list just
fine.  I don't appreciate replying to what I think is a private message only
to see a copy of it in the public area and have to resend the message for
others to see.

-- 
 Steve C. Lamb | I'm your priest, I'm your shrink, I'm your
   PGP Key: 8B6E99C5   | main connection to the switchboard of souls.
   |-- Lenny Nero - Strange Days
---+-


pgpwYxfZdMQ89.pgp
Description: PGP signature

Re: Should this be filed as grave? Gcc-2.95

[H. S. Teoh]

On Tue, Aug 05, 2003 at 08:43:36AM -0700, Steve Lamb wrote:
> On Tue, 5 Aug 2003 11:06:26 -0400
> "H. S. Teoh" <[EMAIL PROTECTED]> wrote:
> > Did you check your compile logs to see if it actually compiled with
> > gcc-2.95 or with just gcc (==3.3) ? It happened to me several times that
> > when building 2.4.21, it would use gcc-2.95 for the initial configuration
> > and cleanup targets (since I specified CC=gcc-2.95), but revert to gcc for
> > the actual build. 
> 
> That is most likely what happened.  I didn't check logs.  Didn't care.  It
> didn't work, fuggit, I needed the machine stable and was mighty pissed that I
> couldn't just rip 3.3 off the damned system to force the issue without
> resorting to a serious downgrade to woody packages just to do it.

Downgrading sounds like overkill in this situation. I only had to edit
/usr/src/linux/Makefile to change HOSTCC to gcc-2.95 and export
CC=gcc-2.95 in the environment, and it worked fine for me. This is on
2.4.21, of course, but I suspect the same holds for 2.4.20.

> > I had to hand-edit kernel makefiles to stop it from using gcc by default
> > and use gcc-2.95 instead. Or perhaps try setting CC=gcc-2.95 in your
> > environment before running the build. 
> 
> Might have worked but forcing the issue is better IMHO.  Without 3.3
> present there is absolutely no chance of some process down the line sanitizing
> the environment and monkeying up the works.
[snip]

ln -s /usr/bin/gcc-2.95 /usr/bin/gcc

ln -s /usr/bin/gcc-3.3 /usr/bin/gcc

Problem fixed.


T

-- 
Nothing in the world is more distasteful to a man than to take the path that
leads to himself. -- Herman Hesse

Re: How to install X-Chat in five hours (or more)

[Colin Watson]

On Tue, Aug 05, 2003 at 08:14:23AM -0700, Ian Hickson wrote:
> On Wed, 6 Aug 2003, Hamish Moffatt wrote:
> > On Tue, Aug 05, 2003 at 01:33:19AM -0700, Ian Hickson wrote:
> > > Part of the problem I had was that I had a vague understanding that
> > > there was something called "apt", but that I didn't know what it was
> > > or how to do anything with it. The man page said to see apt-get's;
> > > apt-get's man page suggested the tool was a back-end but didn't really
> > > give any clues as to what front end to use.
> >
> > Not directly, but the "SEE ALSO" list does include dselect(8) which is
> > what you really should have used.
> 
> The term "dselect" means nothing to me. It isn't a usable name. That's
> another example of the problem I mentioned.

Tools have names, and they don't really have to be generic. I think it's
quite acceptable for the installation manual to tell you the names you
need to know to get started, and it does: see sections 8.11 to 8.15 of
.

"mozilla" isn't a particularly obvious name, when you get down to it;
why not "web-browser"? (Because it's not the only web browser on the
planet, of course. Likewise, it's important not to confuse people
migrating between distributions by having the generic name "setup" do
something completely different wherever you go because people were
scared of using program names.)

> Would it not be possible for debian to have a command "setup" or "install"
> or something similarly named?

I think that's a good example of why generic names may not be a good
thing to encourage people to expect; /usr/bin/install has existed for a
long time, can't feasibly be changed, and is definitely not what you're
looking for.

> Note that, if for some reason the user knew about the command
> "apropos", even that wouldn't help him -- none of dselect, aptitude,
> and apt-get come up for "apropos install" or "apropos setup".

They all show up for 'apropos package', along with a bunch of other
stuff, but yes, that would probably be a useful enhancement to at least
one of those man pages.

> >> To the end user (me), apt-get is arbitrarily verbose. "Selecting
> >> previously deselected package libbla3.2"? "Get:1 ftp://apt sid/main
> >> libbla3.2 3.2.10-9 [827kB]"?
> >>
> >> Look at operating systems used by less intelligent users. They just
> >> see:
> >>
> >>[# ] 60%  2 minutes remaining
> >
> > I don't see how some extra verbosity hurts.
> 
> It hurts because it scares users. My dad would take one look at the
> text, and give up. (And 15 years ago he was a VMS administrator, so
> it's not like he's computer illiterate.) My mum wouldn't even give the
> text a chance, she'd just see a wad of text with odd punctuation and
> run for the hills.

However, it's better for a command-line tool to be verbose up-front,
because if it crashes or blows up or just goes slightly wrong, at least
the last of which is frequent with buggy packages, we need the
information for bug reports (at any rate until such time as we have
logging in dpkg, but since that's bug #957 I wouldn't hold your breath
...). With a graphical front-end it's much easier to hide the verbosity
and have a "show me the installation log" option in case of error. I've
seen graphical front-ends for the Debian package management system that
do exactly this.

I wouldn't suggest that your mum use apt-get in any case, no more than
I'd try to get my parents to do everything on the command line. I'd
suggest a point-and-click package manager instead, if I were having them
do routine package maintenance at all.

> > Frankly if verbosity loses us some users, too bad. I'm sure we pick
> > up more users because of the same.
> 
> You will lose many more than you will gain, since there are many more
> computer illiterate users than geeks.

That hasn't historically been a good argument on Debian mailing lists.
:-) Fundamentally, we're trying to produce the best, most stable, most
reliable, etc. system we can, not get as many users as we can. The two
goals aren't always entirely synonymous. In fact, computer-literate
users may often be better from our perspective because they often
produce better bug reports!

That's not to say that the goal is user-hostility, just that
user-friendliness isn't always the all-defeating trump card when
discussing relatively low-level tools like dpkg and apt-get.

> > To rant a bit, the thing that bugs me the most about MS Windows is
> > how when it breaks randomly you can't fix it because it runs on
> > smoke and mirrors and doesn't give helpful information on what went
> > wrong. With UNIX/Linux you get details and you can fix it.
> 
> Just to clarify, I've nothing against verbosity itself. /var/log, for
> instance, is great (although "var" is a historical name that really
> should be replaced by something more user friendly, but that's another
> story).

You can't get there from here, I think. Unix admins coming to Debian
wil

Re: Should this be filed as grave? Gcc-2.95

[Adam Heath]

On Tue, 5 Aug 2003, Steve Lamb wrote:

> Adam, where does it say anywhere in my sig or headers that I want a CC?  I
> read the list just fine, you can reply to the list and only the list just
> fine.  I don't appreciate replying to what I think is a private message only
> to see a copy of it in the public area and have to resend the message for
> others to see.

lart, killfile.

Please NMU dovecot

[Jaldhar H. Vyas]

Could someone please NMU dovecot adding the patch in bug #203892?

Either gcc 3.3.1 sucks or I'm having another hardware problem,

...

Making all in lib
make[4]: Entering directory
`/home/jaldhar/src/dovecot/dovecot-0.99.10/src/lib'
i386-linux-gcc -DHAVE_CONFIG_H -I. -I. -I../.. -g -O2 -Wall -W
-Wmissing-prototypes -Wmissing-declarations -Wpointer-arith
-Wchar-subscripts -Wformat=2 -Wbad-function-cast  -c alarm-hup.c
alarm-hup.c: In function `alarm_hup_init':
alarm-hup.c:72: internal compiler error: Bus error
Please submit a full bug report,
with preprocessed source if appropriate.
See http://gcc.gnu.org/bugs.html> for instructions.
make[4]: *** [alarm-hup.o] Error 1
make[4]: Leaving directory
`/home/jaldhar/src/dovecot/dovecot-0.99.10/src/lib'
make[3]: *** [all-recursive] Error 1

...

-- 
Jaldhar H. Vyas <[EMAIL PROTECTED]>
La Salle Debain - http://www.braincells.com/debian/

Re: How to install X-Chat in five hours (or more)

[Gustavo Noronha Silva]

Em Tue, 5 Aug 2003 10:49:17 +0200, Frederik Rousseau <[EMAIL PROTECTED]> 
escreveu:

> Anyway, does this mean we need something like a GNU/Linux Debian and a 
> GNU/Linux Debian For Dummies showing only icons?

Yes, I think so... Debian-Desktop should be that, probably. I would like
to see a package manager like the one suggested by Ian Hickson you
don't need to mess with all of Debian, you just need a better UI, really.

The short description, together with a good selection of 'what is for
the users, what is for the developers and what is for system admins'
could be used to achieve that.

[]s!

-- 
[EMAIL PROTECTED]: Gustavo Noronha 
Debian:   *  
Dúvidas sobre o Debian? Visite o Rau-Tu: http://rautu.cipsga.org.br

Re: How to install X-Chat in five hours (or more)

[Ian Hickson]

On Tue, 5 Aug 2003, Steve Lamb wrote:
>>
>> Note that, if for some reason the user knew about the command
>> "apropos", even that wouldn't help him -- none of dselect, aptitude,
>> and apt-get come up for "apropos install" or "apropos setup".
>
> I do believe they are mentioned several times in the manual.  Y'know,
> that thing collecting dust over yonder.

What manual?

I receieved the machine with Debian preinstalled and no offline
documentation except a post it note with the root username and password.
On other systems (Mac OS X, Windows XP, etc) I am clearly shown where to
look for more information (on Windows, in fact, the OS goes to the other
extreme and tries to ram help down your throat), but on Debian, there was
no clear path to the documentation.

Typing "help" at the command line, which to most new users would seem like
a sensible starting point, gives very terse information about bash and
shell commands. (The bit at the top of this does mention 'man', but it
scrolls off the top of my screen so it's not that helpful to a new user
who doesn't know about Shift + Page Up.)

The only manual I was aware of was "apropos" and "man". But as I said in
another message, neither "apropos install" nor "apropos setup" lead the
user to look at any of apt-get, dselect, or aptitude.


>> It hurts because it scares users.
>
> And?

>From a usability point of view, scaring users is a bad thing.


>> You will lose many more than you will gain, since there are many more
>> computer illiterate users than geeks.
>
> And? There are a slew of other OSs and Linux distributions as well.

And I'm a geek, one who has been using GNU-based distributions on multiple
machines on a daily basis for at least 3 years, and Sun for 6 years before
that, and I _still_ have difficulty. This should be ringing usability
alarm bells.


> This is what always gets me in discussions that are based on the lowest
> common denominator in computer users; the presumption that everyone
> wants to cater to them.

Being usable does not mean catering for the lowest common denominator; I
fully agree that other distributions are more adequately positioned to
target computer illiterate users.


>> although "var" is a historical name that really should be replaced by
>> something more user friendly, but that's another story
>
> Erm, no, it should not be.  While it is a historical name it is a name
> that should remain because every person who's ever worked on a Unix-like
> system during that history knows where /var is, why it is there and what
> is in it.  It is up to those new people to catch up, not for us to ruin
> what works and works well in the vain attempt to catch more of a market
> which, in the end, doesn't really matter as this is not a commercial
> venture.

Without meaning offense, that is a very selfish attitude. The number of
future debian users is *significantly* larger than the number of existing
users, unless something drastic happens to either humanity or debian
itself. Why should everyone who will use debian in future be forced to
learn archaic commands, paths, and deal with other historical holdbacks,
instead of the few who already use it being taught easier conventions?


>> I'm all for a "tell me what is going on" feature for debugging.
>
> Which is why you need verbosity when nothing is going wrong.  Let's see
> a show of hands on this situation.
>
> Ya boot up a Windows box post after '95.  Here's the sum of it letting
> you know something is going on: a rotating palette for the bar at the
> bottom.  The palette stops rotating.  So, uh, what's wrong?  Oh, wait,
> it started rotating again.  No, wait, it stopped again... for 30
> seconds.  No, there it goes, it's fine.

Right. That's poor UI. However, it's not _that_ much better on Debian:

   Incomprehensible status message #1.
   Incomprehensible status message #2.
   Incomprehensible status message #3.

   (long pause)

   Incomprehensible status message #4.
   Incomprehensible status message #5.
   Incomprehensible status message #6.

By "Incomprehensible status message" I mean things like:

   bootlogd.
   Activating swap.
   fsck 1.35-WIP (01-Aug-2003)
   Running 0dns-down to make sure resolv.conf is ok...done.
   Please contribute if you find this software useful.
   DHCPDISCOVER on eth0 to 255.255.255.255 port 67 interval 5
   Starting Xprint servers: Xprt.

...all of which are taken directly from my boot log.

Why can't we instead have nice friendly messages? e.g.:

   Startup logging has begun. Log will be stored in '/var/log/boot'.

...instead of "bootlogd".


>> Even then, though, it would be nice if the verbose messages were
>> consistently formatted, and used plain english instead of jargon.
>> Error messages like "E: Invalid operation foo" are not helpful.
>
> No, that's a bad idea.  Take a look at IE's 404 message sometime.  It's
> a dumbed down version which doesn't explain jack or shitte.

Again, IE's 404 message is terrible UI. It is much, much too long and is
n

Re: Fw: Re: Ruby 1.8 transition plan; debian-ruby

[Fumitoshi UKAI]

At Mon, 4 Aug 2003 18:02:52 +0300,
Dmitry Borodaenko wrote:
> 
> As rightfully pointed out by Fumitoshi UKAI, this discussion belongs to
> the wider audience of debian-devel, especially since Ruby 1.8.0 was
> released today.
> 
> NB: some points raised here can be of interest not only to Ruby
> developers, but also to developers from other scripting languages.
> Generic questions are:
> 
> Is there or shouldn't there be policy or guidelines on major version
> transition for scripting languages?

Currently, there are no such policy or guidelines.  There are policies 
for each specific languages such as perl, python, java, emacsen,
however, it seems that they have somewhat different approaches.
I think it is impossible or very difficult to make general policy.
Anyway, it would be very helpful to make a guideline that
suggests the answers for the following questions.
 
> What considerations should be taken into account when deciding whether
> to keep several different major language versions co-installable?
> 
> Where should libraries written in a scripting language go? (That one is
> my pet question: FHS seems to point to /usr/share/, while in Debian,
> most languages use /usr/lib/, and I have arguments against both :)
> 
> Ruby-specific question is, of course, how do we deal with Ruby 1.8
> transition?

Since ruby 1.8.0 was released recently, ruby developers will go to 
ruby 1.8.x, so that we, ruby maintenance team (akira, tagoh, ukai), 
are discussing about how to deal with Ruby 1.8 transition and trying to make
debian ruby policy soon.

For now, ruby package provides /usr/bin/ruby of ruby 1.6.x, and
ruby1.8 package provides /usr/bin/ruby1.8 of ruby 1.8.x.  Someone
want to use /usr/bin/ruby of ruby 1.8.x, so we're considering to use 
alternatives for /usr/bin/ruby to choice either ruby1.6, ruby1.8 (or any
other version of ruby in future).

I wonder we should rename ruby package to ruby1.6 package and ruby package
for meta package for default version of ruby now.  It requires package renaming
which is somehow troublesome, so I suspect it's worth to do it.
Anyway, I don't think we'll change ruby1.8 to ruby in the future, and
ruby becomes meta package for default version of ruby after ruby 1.6.x
is removed from unstable.

The default include paths of each ruby version are as follows:

% ruby -e 'puts $:'
/usr/local/lib/site_ruby/1.6
/usr/local/lib/site_ruby/1.6/i386-linux
/usr/local/lib/site_ruby
/usr/lib/ruby/1.6
/usr/lib/ruby/1.6/i386-linux

Re: How to install X-Chat in five hours (or more)

[Theo Cabrerizo Diem]

I agree with every word  

There are lot's of packages to do Debian more user friendly, they are
available at install time and after by running tasksel.

Maybe the problem was the way that Debian was "pre-installed" I
think they only installed base. This isn't suitable for a "user". Maybe
if you installed Debian by your own you could feeled more confortable
with dselect and tasksel (called from debian installer) among other
words/commands.

> You will lose many more than you will gain, since there are many more
> > computer illiterate users than geeks.
> 
> And?  There are a slew of other OSs and Linux distributions as well.  

I think many people choose Debian because it's a "geek" distro, without those 
RH/Mandrake
annoyances that are good for users and bad for "us".

After all .. all this problem happened not just because libgtk1.2 .. but to 
install a
browser, right ? ... just apt-get install mozilla   . =)

C´ya and sorry for poor english .. just my 2 c.
[]'s

On Tue, 2003-08-05 at 12:41, Steve Lamb wrote:

> > You will lose many more than you will gain, since there are many more
> > computer illiterate users than geeks.
> 
> And?  There are a slew of other OSs and Linux distributions as well.  
> 
> This is what always gets me in discussions that are based on the lowest
> common denominator in computer users; the presumption that everyone wants to
> cater to them.
> 
> It has been my experience that packages that cater to the lowest common
> denominator are packages I don't care to use because I find them *hard to use*
> even though they purport to be "easy to use".  I have almost always been on
> the side of the scale where I preferred the "harder" package because in the
> long run, once I got over the nominal learning curve, it was easier to use. 
> However there was always that drive to go for the lowest common denominator,
> the computer illiterate.  It has ruined more programs that I care to list
> because they would add too much, dumb things down, make the program too large
> and hard-to-use in the quest to get people who have to ask "Left or right
> click" after the first time you tell them to right-click something.  IE,
> people with no concept of "default behaviors/actions".
> 
> I have never, EVER understood why anyone would want to take a package
> which is beloved by the niche geek market and destroy it for the illiterate
> market.  That is triply so for commercial packages.  
> 
> You're wrong in saying that we (in general) would gain more by making our
> package (program, OS, etc) more palatable to the neophyte.  As I said above,
> there are a slew of OSs and distributions that cater to that segment.  To move
> Debian into that realm would be to lose what it has and compete with
> established entities.  100% of this slice is better than <1% of *that* slice.
>  
> > Just to clarify, I've nothing against verbosity itself. /var/log, for
> > instance, is great (although "var" is a historical name that really
> > should be replaced by something more user friendly, but that's another
> > story). The problem is verbosity when things don't go wrong.
> 
> Erm, no, it should not be.  While it is a historical name it is a name
> that should remain because every person who's ever worked on a Unix-like
> system during that history knows where /var is, why it is there and what is in
> it.  It is up to those new people to catch up, not for us to ruin what works
> and works well in the vain attempt to catch more of a market which, in the
> end, doesn't really matter as this is not a commercial venture.
> 
> > I'm all for a "tell me what is going on" feature for debugging.
> 
> Which is why you need verbosity when nothing is going wrong.  Let's see a
> show of hands on this situation.
> 
> Ya boot up a Windows box post after '95.  Here's the sum of it letting you
> know something is going on: a rotating palette for the bar at the bottom.  The
> palette stops rotating.  So, uh, what's wrong?  Oh, wait, it started rotating
> again.  No, wait, it stopped again... for 30 seconds.  No, there it goes, it's
> fine.
> 
> > Even then, though, it would be nice if the verbose messages were
> > consistently formatted, and used plain english instead of jargon.
> > Error messages like "E: Invalid operation foo" are not helpful.
> 
> No, that's a bad idea.  Take a look at IE's 404 message sometime.  It's a
> dumbed down version which doesn't explain jack or shitte.  Error messages are
> there for people who know what they need to do.  People who don't know what
> they need to do will not have that knowledge suddenly imparted upon them by a
> "plain english" error message because, without the jargon to point you in the
> right direction, there would be absolutely no place to start.
> 
> > The first option I'm faced with is:
>  
> >* 0. [A]ccessChoose the access method to use.
>  
> > I have no idea what that means. I tried using it (not logged in as
> > root) 

Re: How to install X-Chat in five hours (or more)

[Ian Hickson]

On Tue, 5 Aug 2003, Colin Watson wrote:
>>
>> The term "dselect" means nothing to me. It isn't a usable name. That's
>> another example of the problem I mentioned.
>
> Tools have names, and they don't really have to be generic. I think it's
> quite acceptable for the installation manual to tell you the names you
> need to know to get started, and it does: see sections 8.11 to 8.15 of
> .

Ok, that's fair enough.


>> Note that, if for some reason the user knew about the command
>> "apropos", even that wouldn't help him -- none of dselect, aptitude,
>> and apt-get come up for "apropos install" or "apropos setup".
>
> They all show up for 'apropos package', along with a bunch of other
> stuff, but yes, that would probably be a useful enhancement to at least
> one of those man pages.

I'm glad you agree. :-)


> However, it's better for a command-line tool to be verbose up-front,
> because if it crashes or blows up or just goes slightly wrong, at least
> the last of which is frequent with buggy packages, we need the
> information for bug reports [...].

That's probably true. It would still be nice if the verbose messages were
more consistent, though. For example, 'apt-get' error messages start with
'E:' instead of the more standard 'apt-get:'. Similarly, when you do an
apt-get update, you get some messages of the form:

   Hit ftp://apt sid/mail Packages

...some of the form:

   Get:1 ftp://apt ./ Packages

...and some of the form:

   Reading Package Lists... Done

...which is odd: why three different kinds of messages? What does "Hit"
mean, as opposed to "Ign" or "Get:1"? And so on.

This isn't only a problem with apt-get, of course. Error and status
messages throughout the industry and in particular throught the free
software world are often obscure, obtuse, and unclear. Indeed, Mozilla has
its share of such problems! :-)


> With a graphical front-end it's much easier to hide the verbosity and
> have a "show me the installation log" option in case of error. I've seen
> graphical front-ends for the Debian package management system that do
> exactly this.

That's cool. (aptitude doesn't, as far as I can tell.)


> :-) Fundamentally, we're trying to produce the best, most stable, most
> reliable, etc. system we can, not get as many users as we can.

That's fair enough! :-)


> That's not to say that the goal is user-hostility, just that
> user-friendliness isn't always the all-defeating trump card when
> discussing relatively low-level tools like dpkg and apt-get.

I think that user-friendliness, even at such a low level, should still be
important -- just because the user is an expert doesn't mean he wants to
have to decode messages.


>> although "var" is a historical name that really should be replaced by
>> something more user friendly, but that's another story.
>
> You can't get there from here, I think. Unix admins coming to Debian
> will scream blue murder if it starts being "/My Variable Data/Logs", and
> that group is important to us.

Note that there is at least one project which is looking at doing exactly
that while retaining backwards compatability (GoboLinux). It may be worth,
on the long term, looking at how it may be possible to migrate from
obscure paths like "/opt", "/bin", "/sbin", "/usr/bin", etc, to more
sensible names, in that way.

And I would scream if you called it "/_My_ Variable Data/" too... :-P

-- 
Ian Hickson  )\._.,--,'``.fL
"meow"  /,   _.. \   _\  ;`._ ,.
http://index.hixie.ch/ `._.-(,_..'--(,_..'`-.;.'

Should MUA only Recommend mail-transfer-agent?

[Artur R. Czechowski]

Hello

I've found your bugreport:
http://bugs.debian.org/202869

I see no issue to not depending mutt on mail-transfer-agent.

Mutt as is, is a software for reading, writing and sending emails.
And to provide a full functionality it needs a kind of transfer-agent.

I am not convinced to only Recommend on mail-transfer-agent. I rather
tend to closing this wishitem or tag it as wontfix.

OTOH this case concerns not only mutt but also other MUA's. Feel free
to discuss it on debian-devel mailing list or propose a changes
to Debian Packaging Policy. I will leave this wishitem open until
an agreement is reached.

Regards
Artur
-- 
Artur R. Czechowski <[EMAIL PROTECTED]>

Re: Should MUA only Recommend mail-transfer-agent?

[Gunnar Wolf]

Artur R. Czechowski dijo [Tue, Aug 05, 2003 at 07:21:00PM +0200]:
> Hello
> 
> I've found your bugreport:
> http://bugs.debian.org/202869
> 
> I see no issue to not depending mutt on mail-transfer-agent.
> 
> Mutt as is, is a software for reading, writing and sending emails.
> And to provide a full functionality it needs a kind of transfer-agent.
> 
> I am not convinced to only Recommend on mail-transfer-agent. I rather
> tend to closing this wishitem or tag it as wontfix.
> 
> OTOH this case concerns not only mutt but also other MUA's. Feel free
> to discuss it on debian-devel mailing list or propose a changes
> to Debian Packaging Policy. I will leave this wishitem open until
> an agreement is reached.

I agree with Andreas. On one hand, I have used mutt to read a mbox file
I have lying around while flying - it does not require a MTA to work. On
the other hand, mutt can work in a more mundane environment with a
remote IMAP server. Yes, a MTA is required to send mail, and is thus
strongly recommended - but not having one does not render the package
unusable, so it is not really depending on it.

Greetings,

-- 
Gunnar Wolf - [EMAIL PROTECTED] - (+52-55)5630-9700 ext. 1366
PGP key 1024D/8BB527AF 2001-10-23
Fingerprint: 0C79 D2D1 2C4E 9CE4 5973  F800 D80E F35A 8BB5 27AF


pgpIIvqr3YCcD.pgp
Description: PGP signature

Re: How to install X-Chat in five hours (or more)

[Daniel Burrows]

On Wed, Aug 06, 2003 at 12:33:18AM +1000, Hamish Moffatt <[EMAIL PROTECTED]> 
was heard to say:
> > I think Debian's package system is remarkably nice. Unfortunately,
> > it's UI leaves a lot to be desired. The biggest problem is probably
> 
> Which UI did you use? We have a few. apt-get is not an interface for the
> Debian newbie. dselect and aptitude are GUI tools if that's what you
> need.

  aptitude is neither a GUI tool nor a tool for the Debian newbie.  The
GUI tool I hear mentioned most frequently is synaptic; maybe the writer
would have better luck with that?

  Daniel

-- 
/ Daniel Burrows <[EMAIL PROTECTED]> ---\
|  Will the last person to leave the Universe please  |
|  turn off the lights and close the door?|
\-Evil Overlord, Inc: planning your future today. http://www.eviloverlord.com-/

Re: Should MUA only Recommend mail-transfer-agent?

[Bernd Eckenfels]

On Tue, Aug 05, 2003 at 07:21:00PM +0200, Artur R. Czechowski wrote:
> OTOH this case concerns not only mutt but also other MUA's. Feel free
> to discuss it on debian-devel mailing list or propose a changes
> to Debian Packaging Policy. I will leave this wishitem open until
> an agreement is reached.

There are enough SMTP/POP3 MUAs which do not need any MTA infrastructure on
the local host, whatsoever. Mutt can fetch by pop-3, but I think it has no
smtp support build in, or?

Greetings
Bernd
-- 
  (OO)  -- [EMAIL PROTECTED] --
 ( .. )  [EMAIL PROTECTED],linux.de,debian.org} http://home.pages.de/~eckes/
  o--o *plush*  2048/93600EFD  [EMAIL PROTECTED]  +497257930613  BE5-RIPE
(OO)  When cryptography is outlawed, bayl bhgynjf jvyy unir cevinpl!

Re: How to install X-Chat in five hours (or more)

[Josef Spillner]

On Tuesday 05 August 2003 18:55, Ian Hickson wrote:
> Without meaning offense, that is a very selfish attitude. The number of
> future debian users is *significantly* larger than the number of existing
> users, unless something drastic happens to either humanity or debian
> itself. Why should everyone who will use debian in future be forced to
> learn archaic commands, paths, and deal with other historical holdbacks,
> instead of the few who already use it being taught easier conventions?

More comfort requires more system resources (except for a few really clever 
hacks).
Beginners are most likely willing to accept that a few CPU cycles are used for 
making their computing time easier - you'll see that they like icons 
(animated ones even), tooltips, first-time wizards and so on.

Now, the unix file system layout was never designed to be end-user friendly in 
the beginning (the short ASCII-style file and directory names take care of 
that), yet development didn't stop.
What Linux and other OSes lack in the kernel space, has started to exist in 
the user space. With KDE, you've got a "Home" folder already and a "Trash" 
folder. With GNOME, you've got a "Preferences" container.

You can easily write an implementation which fits your needs, e.g. a KIO slave 
for directory name translation, or a Hurd translator for system-level 
bindings.
Of course, in order to not be flamed from time to time, you should come up 
with a nice concept first. You don't have to write the code yourself, but you 
do have to invent the user-friendly layer and make it public.

Josef

-- 
Play for fun, win for freedom.
Hurd^H^H^H^HLinux-Info-Tag Dresden 2003: http://www.linux-dresden.de

Re: Should MUA only Recommend mail-transfer-agent?

[Artur R. Czechowski]

On Tue, Aug 05, 2003 at 08:00:03PM +0200, Bernd Eckenfels wrote:
> There are enough SMTP/POP3 MUAs which do not need any MTA infrastructure on
> the local host, whatsoever.
But there are some important packages which depends on MTA directly, like:
at, cron, debconf, logrotate, mailx.

I can imagine a workstation without those packages but it is, IMO,
mutilated box.

> Mutt can fetch by pop-3, but I think it has no
> smtp support build in, or?
Mutt has no support for SMTP.

BTW, there is no need for exim4-daemon-heavy. There are other lightweight
MTA's.

Another solution is to prepare a dummy-mta package, which only
provides mail-transfer-agent and required by policy /usr/sbin/sendmail
and /usr/bin/newaliases binaries to do nothing[1].

Advanced Debian users has another opportunity to solve this problem: equivs.

I would like to know Md's opinion, but for me there are no reasons to relax
dependencies for mutt (and other MUA). I would not like to do it without
policy requirements because it concerns also other MUA's.

I, personally, like the dummy-mta solution, however nullmailer also looks
good.

Cheers
Artur

[1] maybe logging to syslog will be good.
-- 
http://hell.pl/arturcz/


pgpdTSwUxStgb.pgp
Description: PGP signature

Re: How to install X-Chat in five hours (or more)

[Emile van Bergen]

Hi,

On Tue, Aug 05, 2003 at 09:55:59AM -0700, Ian Hickson wrote:

[SNIP]

> Why can't we instead have nice friendly messages? e.g.:
> 
>Startup logging has begun. Log will be stored in '/var/log/boot'.
> 
> ...instead of "bootlogd".

[SNIP]

> > Error messages are there for people who know what they need to do.
> 
> So if I do something wrong (like get the command line arguments to
> 'apt-get' wrong, as I did), then I don't deserve to be helped by the
> program? What would be wrong with a helpful message, such as:
> 
>apt-get: the first argument should be one of 'install', 'remove', 
> 'update', or another operation
> 
> ...instead of just "E: Invalid operation foo"?

[SNIP] 

> How about a message such as:
> 
>dselect: to select an installation source, superuser privileges are 
> required (try logging in as root)
> 
> It's still accurate, but now it's helpful as well, and uses a more
> friendly voice. (This also changes "access method" to "installation
> source", which makes more sense to me.)

You know, I think these are actually good suggestions. I think there's a
lot to be gained *not* by dumbing down, *not* by losing any information
that might be useful to a geek or to a new user as (s)he's learning, but
by phrasing texts so that they appeal more to generally intelligent
human beings, rather than to people that just happen to have some
specific knowledge.

Apple has a great way of doing that. They don't dumb down, they don't
belittle you, they assume an intelligent being who can grasp reasonably
complex English sentences, but who has less knowledge of computer
idiom.

I think this is an area in which software can be improved without
scaring the geeks with offensive 'my first Sony' UIs, teletubby
landscapes, informationless error messages, and stupid attempts to fix
things behind the users back (simply displaying expired pages from a
cache for instance).

Indeed, Microsoft gets it all wrong, from a UI standpoint. But instead
of assuming that we are pushed in that scary direction when someone
complains about our UI, we may also realise that there are things that
can actually be improved, without harming Unix's strong points, UI-wise.

I fully agree with the poster that increasing the number of intelligent
and intelligible English sentences that we output is one of those things
that can be done without harming the hacker in any way.

Of course, like any enhancement, it needs a volunteer who can scratch
his or her itch by doing this work. This may actually be one of the
bigger problems. Most developers who can do the work have gotten used to
the idiom and badly worded error messages.

Cheers,


Emile.

-- 
E-Advies - Emile van Bergen   [EMAIL PROTECTED]  
tel. +31 (0)70 3906153   http://www.e-advies.nl


pgprXSw5bIZGU.pgp
Description: PGP signature

Bug#204263: ITP: gngeo -- NeoGeo emulator

[Julien Delange]

Package: wnpp
Version: reported 2003-08-05
Severity: wishlist

* Package name: gngeo
  Version : 0.5.9a
  Upstream Author : M. Pepone <[EMAIL PROTECTED]>
* URL : http://m.peponas.free.fr/gngeo/
* License : GPL
  Description : NeoGeo emulator

gngeo is an emulator for NeoGeo games, it runs with SDL library.
Gamepad can be use to play the games.
You MUST have a neo-geo bios rom stored on your system (in
/usr/share/gngeo)


-- System Information:
Debian Release: testing/unstable
Architecture: i386
Kernel: Linux evira 2.6.0-test1 #2 Mon Jul 14 18:56:06 CEST 2003 i686
Locale: [EMAIL PROTECTED], [EMAIL PROTECTED]

Re: How to install X-Chat in five hours (or more)

[Emile van Bergen]

Hi,

On Tue, Aug 05, 2003 at 10:19:53AM -0700, Ian Hickson wrote:

> And I would scream if you called it "/_My_ Variable Data/" too... :-P

I would even scream at 

/Variable Data/

simply because it encourages slow and RSI-inducing click and drag
behaviour, because such path names are impossible to type in (and this
one even requires escaping the space to distinguish between the argument
separator). 

Don't underestimate the typing convenience of TLAs! They may not be the
most descriptive, but at least they're blindingly fast to work with when
you get used to it.

In short, such path names are definitely no improvement from a UI
standpoint, even though it may seem that way from a casual observer.
Even shorter: UIs are rarely obvious. And there must be a reason why
geeks like working with Unix.

One thing for me is that the command line allows me to work almost as
quickly as I think. I really hate it when the UI drags me down ('open
folder, open subfolder, click on file, type Ctrl-C, open other folder,
type Ctrl-V', it's just horrible). The computer should wait for me, not
the other way around!

Cheers,


Emile.

-- 
E-Advies - Emile van Bergen   [EMAIL PROTECTED]  
tel. +31 (0)70 3906153   http://www.e-advies.nl


pgpX3JoSsrW5E.pgp
Description: PGP signature

Re: Should MUA only Recommend mail-transfer-agent?

[Erik Steffl]

Artur R. Czechowski wrote:
On Tue, Aug 05, 2003 at 08:00:03PM +0200, Bernd Eckenfels wrote:
There are enough SMTP/POP3 MUAs which do not need any MTA infrastructure on
the local host, whatsoever.
But there are some important packages which depends on MTA directly, like:
at, cron, debconf, logrotate, mailx.
I can imagine a workstation without those packages but it is, IMO,
mutilated box.
...
  it's strange for MUA to require MTA, lot of them support IMAP (for 
viewing email) and SMTP (for sending email), both of which can be on 
remote servers. So why MTA on local box?

  Yes, there are other reasons to have MTA on box (non related to MUA) 
but that's irrelevant.

erik

Re: Should MUA only Recommend mail-transfer-agent?

[Milan P. Stanic]

On Tue, Aug 05, 2003 at 08:00:03PM +0200, Bernd Eckenfels wrote:
> There are enough SMTP/POP3 MUAs which do not need any MTA infrastructure on
> the local host, whatsoever. Mutt can fetch by pop-3, but I think it has no
> smtp support build in, or?

I just (actually few hours ago) find patch which adds smtp support to
mutt. 
http://www.deez.info/sengelha/projects/mutt/libesmtp/patch-1.5.3.sde.libesmtp.3

It depends on libesmtp.

I personally always liked MUAs with smtp and my main objection to mutt
was lack of it. Now, I have it for my (loved) MUA. :-)

-- 
Milan

Re: How to install X-Chat in five hours (or more)

[Carlos E. Pruitt, Jr.]

Debian should not change its attitude or methods to "meet the end
user's  needs".  Think of Debian as a the painter's palette.  All of the
tools you need are available to you.  It installs a base system and you
customize from there.  Would I recommend Debian for John Doe user? As a
base raw install (even with X), I would not.  However, Debian is a
wonderful tool.  I can pick and choose what packages I wish to install
and how I wish those packages to be configured.  If I so desire, I can
setup a Debian to meet the needs of an end user.  An end user has
different needs such as a router, an OpenMosix cluster, a
desktop machine, and the list can go on and on.
Debian is a generalist.  Debian sub-projects and Debian based distros
takes the Debian tools to meet the needs of the end user.  Without the
painter's palette, the pictures for the end user  cannot be created.
By the way, I like the verbose messages.
Also, there are mechanisms to submit bug reports to fix code and
documentation bugs.
That's my two bits.
carlosP
--
QOTD:
Recursion n.:
  See Recursion.
   -- Random Shack Data Processing Dictionary
Carlos E. Pruitt, Jr.
National Center for Computational Hydroscience and Engineering
102 Carrier
University, MS 38677
USA
phone: 662-915-7786

 On Tue, 5 Aug 2003, Colin Watson wrote:

 The term "dselect" means nothing to me. It isn't a usable
 name. That's another example of the problem I mentioned.
>
>>
>> Tools have names, and they don't really have to be generic. I
>> think it's quite acceptable for the installation manual to tell
>> you the names you need to know to get started, and it does: see
>> sections 8.11 to 8.15 of
>> .
 Ok, that's fair enough.
 Note that, if for some reason the user knew about the command
 "apropos", even that wouldn't help him -- none of dselect,
 aptitude, and apt-get come up for "apropos install" or
 "apropos setup".
>
>>
>> They all show up for 'apropos package', along with a bunch of
>> other stuff, but yes, that would probably be a useful enhancement
>> to at least one of those man pages.
 I'm glad you agree.
>> However, it's better for a command-line tool to be verbose
>> up-front, because if it crashes or blows up or just goes slightly
>> wrong, at least the last of which is frequent with buggy
>> packages, we need the information for bug reports [...].
 That's probably true. It would still be nice if the verbose messages
 were more consistent, though. For example, 'apt-get' error messages
 start with 'E:' instead of the more standard 'apt-get:'. Similarly,
 when you do an apt-get update, you get some messages of the form:
 Hit ftp://apt sid/mail Packages
 ...some of the form:
 Get:1 ftp://apt ./ Packages
 ...and some of the form:
 Reading Package Lists... Done
 ...which is odd: why three different kinds of messages? What does
 "Hit" mean, as opposed to "Ign" or "Get:1"? And so on.
 This isn't only a problem with apt-get, of course. Error and status
 messages throughout the industry and in particular throught the free
 software world are often obscure, obtuse, and unclear. Indeed,
 Mozilla has its share of such problems!
>> With a graphical front-end it's much easier to hide the verbosity
>> and have a "show me the installation log" option in case of
>> error. I've seen graphical front-ends for the Debian package
>> management system that do exactly this.
 That's cool. (aptitude doesn't, as far as I can tell.)
>> Fundamentally, we're trying to produce the best, most stable,
>> most reliable, etc. system we can, not get as many users as we
>> can.
 That's fair enough!
>> That's not to say that the goal is user-hostility, just that
>> user-friendliness isn't always the all-defeating trump card when
>> discussing relatively low-level tools like dpkg and apt-get.
 I think that user-friendliness, even at such a low level, should
 still be important -- just because the user is an expert doesn't mean
 he wants to have to decode messages.
 although "var" is a historical name that really should be
 replaced by something more user friendly, but that's another
 story.
>
>>
>> You can't get there from here, I think. Unix admins coming to
>> Debian will scream blue murder if it starts being "/My Variable
>> Data/Logs", and that group is important to us.
 Note that there is at least one project which is looking at doing
 exactly that while retaining backwards compatability (GoboLinux). It
 may be worth, on the long term, looking at how it may be possible to
 migrate from obscure paths like "/opt", "/bin", "/sbin", "/usr/bin",
 etc, to more sensible names, in that way.
 And I would scream if you called it "/_My_ Variable Data/" too...
 -- Ian Hickson )\._.,--,'``. fL "meow" /, _.. \ _\ ;`._ ,.
 http://index.hixie.ch/


--
QOTD:
Recursion n.:
   See Recursion.
-- Random Shack Data Processing Dictionary
Carlos E. Pruitt, Jr.
National Center for Compu

python 2.2 -> python 2.3 transition

[Matthias Klose]

Last weekend, python 2.3 was released.  For an overview see

http://python.org/2.3/highlights.html

With the next python2.3 upload, python2.3 becomes the default python
version.  Some packages become uninstallable until they are converted
to the new version.  In this time you should not update python or
remove all packages depending on python2.2 as the default.  The
upgrade itself should not make major problems.  Many packages are
already built for python2.3, those which do not build/work with
python2.3 (I'm not aware of any) can still explicitely use python2.2.

You can find packages with python2.3 as the default python version at
http://ftp-master.debian.org/~doko/python/.  I'm going to upload these
on Friday. As there are "new" binary packages involved, these won't
hit the archives before Saturday.

Some packages as python-bz2 or python-optik become obsolete, because
these modules are included in the python2.3 standard library.

Re: How to install X-Chat in five hours (or more)

[Tollef Fog Heen]

* Emile van Bergen 

| Hi,
| 
| On Tue, Aug 05, 2003 at 10:19:53AM -0700, Ian Hickson wrote:
| 
| > And I would scream if you called it "/_My_ Variable Data/" too... :-P
| 
| I would even scream at 
| 
| /Variable Data/
| 
| simply because it encourages slow and RSI-inducing click and drag
| behaviour, because such path names are impossible to type in (and this
| one even requires escaping the space to distinguish between the argument
| separator). 

Tab completion or using /Va* is about as fast as /var.

| In short, such path names are definitely no improvement from a UI
| standpoint, even though it may seem that way from a casual observer.
| Even shorter: UIs are rarely obvious. And there must be a reason why
| geeks like working with Unix.

UIs have two costs: an initial learning cost and a «running cost»
(which also includes the retaining the learnt level) which is per
operation (a little bit simplified, but you get the idea).  If you are
going to use a system a lot and over a long period of time, the
initial cost matters a lot less than if you are only going to use the
system occasionally and for short periods of time.  Also, catering for
the different conditions of users is important; I'd hate to have to
use a command line interface with obscure unix-like commands for
operating an ATM, since I fairly seldom use those and therefore the
running cost is less important than the initial cost.

Geeks use systems a lot and for long periods of time, so using more or
less obscure interfaces is good, not bad for us.

:)

-- 
Tollef Fog Heen,''`.
UNIX is user friendly, it's just picky about who its friends are  : :' :
  `. `' 
`-  

Re: How to install X-Chat in five hours (or more)

[Andreas Metzler]

Ian Hickson <[EMAIL PROTECTED]> wrote:
> On Tue, 5 Aug 2003, Steve Lamb wrote:
>>>
>>> Note that, if for some reason the user knew about the command
>>> "apropos", even that wouldn't help him -- none of dselect, aptitude,
>>> and apt-get come up for "apropos install" or "apropos setup".
>>
>> I do believe they are mentioned several times in the manual.  Y'know,
>> that thing collecting dust over yonder.

> What manual?

> I receieved the machine with Debian preinstalled and no offline
> documentation except a post it note with the root username and password.
> On other systems (Mac OS X, Windows XP, etc) I am clearly shown where to
> look for more information (on Windows, in fact, the OS goes to the other
> extreme and tries to ram help down your throat), but on Debian, there was
> no clear path to the documentation.
[...]

* In my experience the windows-help system is frequently next to
  useless, it looks pretty, but is not helpful.
* $total_computer_newbie is completely at loss with windows, too (just
  tried it two months ago).

If somebody set up your machine totally b0rked (mixed sid/testing)
without installing the docs it is not Debian's fault. If you want to
start with Debian, Suse, RedHat, or whatever you have always have to
find a decent manual or at least a person who sets up the system and
helps you.

Yes, you probably can install RedHat by just pressing enter but you'll
get stuck on the third day without manual and learning to help
yourself, Debian will just bite you earlier.
cu andreas
-- 
Hey, da ist ein Ballonautomat auf der Toilette!
Unofficial _Debian-packages_ of latest unstable _tin_
http://www.logic.univie.ac.at/~ametzler/debian/tin-snapshot/

Re: Should MUA only Recommend mail-transfer-agent?

[Andreas Metzler]

Bernd Eckenfels <[EMAIL PROTECTED]> wrote:
> On Tue, Aug 05, 2003 at 07:21:00PM +0200, Artur R. Czechowski wrote:
>> OTOH this case concerns not only mutt but also other MUA's. Feel free
>> to discuss it on debian-devel mailing list or propose a changes
>> to Debian Packaging Policy. I will leave this wishitem open until
>> an agreement is reached.

> There are enough SMTP/POP3 MUAs which do not need any MTA infrastructure on
> the local host, whatsoever.

Yes, probably almost all the others with whitespread userbase, for
example pine, mozilla, (afaik) sylpheed, kmail do support SMTP.[1]

> Mutt can fetch by pop-3, but I think it has no
> smtp support build in, or?

Yes, but iirc there is unofficial patch for linking mutt against
libesmtp on the web.

Imho mutt's dependency on mail-transfer-agent is strong enough for a
Depends. - There are usage szenarios which do not require an
mail-transfer-agent (reading mbox-archives of mailing-lists) but mutt
without MTA is seriously crippled.
  cu andreas
[1] I won't list Gnus but would be really surprised if it _needed_
/usr/sbin/sendmail ;-)
-- 
Hey, da ist ein Ballonautomat auf der Toilette!
Unofficial _Debian-packages_ of latest unstable _tin_
http://www.logic.univie.ac.at/~ametzler/debian/tin-snapshot/

Re: How to install X-Chat in five hours (or more)

[Steve Lamb]

Oh, look, someone else who CCs when it is obvious the person they're
responding to is participating right here.

On Tue, 5 Aug 2003 09:55:59 -0700 (PDT)
Ian Hickson <[EMAIL PROTECTED]> wrote:
> On Tue, 5 Aug 2003, Steve Lamb wrote:
> What manual?

I rest my case.

> I receieved the machine with Debian preinstalled and no offline
> documentation except a post it note with the root username and password.

And this is the problem of Debian... how?  I would ask why you weren't
given better help from the person who installed it or why you didn't ask them
for help.  The facilities are there.

> On other systems (Mac OS X, Windows XP, etc) I am clearly shown where to
> look for more information (on Windows, in fact, the OS goes to the other
> extreme and tries to ram help down your throat), but on Debian, there was
> no clear path to the documentation.

And on my install the same is true.  KDE menu, Help. 
 
> >> It hurts because it scares users.

> > And?
 
> From a usability point of view, scaring users is a bad thing.

Ever hear the saying "Unix is user friendly, it's just picky about its
friends."  You missed my entire point.  Debian isn't scaring off users.  It is
scaring off *neophyte* users.  However, that segment of the population is
covered by *other distributions*.  This is a *good thing* as Debian often
attracts users with some experience who want exactly what it has to offer.


> And I'm a geek, one who has been using GNU-based distributions on multiple
> machines on a daily basis for at least 3 years, and Sun for 6 years before
> that, and I _still_ have difficulty. This should be ringing usability
> alarm bells.

No, it shouldn't.  I should be ringing alarms about the veracity of your
claims and your capabilities. 
 
> Being usable does not mean catering for the lowest common denominator; I
> fully agree that other distributions are more adequately positioned to
> target computer illiterate users.

Debian is quite usable.  *NO* computer system, however, is going to be
completely intuative to someone who is handed it with nothing more than "Here,
here's the username and password.  g'luck!"  Quick, tell me, how do you get
more software for Windows?  Help doesn't point you to TUCOWS or the like.

> Without meaning offense, that is a very selfish attitude.

No, it is selfish for people to come into a culture and flat out state,
"No, I don't want to learn so things should change to cater to ME!"

> The number of future debian users is *significantly* larger than the number
> of existing users, unless something drastic happens to either humanity or
> debian itself. Why should everyone who will use debian in future be forced
> to learn archaic commands, paths, and deal with other historical holdbacks,
> instead of the few who already use it being taught easier conventions?

Because they're not that hard to learn and the easier conventions are not
"easier" in any quantifiable sense of the word.  Furthermore any system you,
or anyone else, would propose would cause no end of upheaval to the software
that ran on it.  It would take years to make any meaningful shift and in the
end guess what would happen.  New users would still have to learn where
everything goes.

> Right. That's poor UI. However, it's not _that_ much better on Debian:

Quite the contrary.

> By "Incomprehensible status message" I mean things like:
> 
>bootlogd.
>Activating swap.
>fsck 1.35-WIP (01-Aug-2003)
>Running 0dns-down to make sure resolv.conf is ok...done.
>Please contribute if you find this software useful.
>DHCPDISCOVER on eth0 to 255.255.255.255 port 67 interval 5
>Starting Xprint servers: Xprt.

If the pause were after fsck and it was showing that it was checking the
disk that would let me know the machine is doing something.  If the pause is
after DHCPDISCOVER I can surmise that if the network isn't hooked up then it
is waiting for a timeout there.

Furthermore if I am calling up to my dad who is 400 miles away and doesn't
know all that much about Linux he can read me the lines and I can tell him
likely causes of problems and pauses.  Try that with "See, about 12 seconds
into bootup the little bar stops rotating for about 30 seconds."  

Like I said, those messages aren't meant for the neophytes.  The neophytes
aren't going to get it either way.  They're there for the people that have to
fix it when it breaks which may be through the interface of the neophyte.  Do
you REALLY think I want to walk my father through turning on boot-up logging
on his Windows box, have it boot, then talk him, through booting into safe
mode and *then* have him drill down to where the log would be?  Yeah, that's
easy alright.

> Why can't we instead have nice friendly messages? e.g.:
 
>Startup logging has begun. Log will be stored in '/var/log/boot'.
 
> ...instead of "bootlogd".

Because when it breaks what do you fix?  What mechanism logs?  Oh, the
user's going t

Re: How to install X-Chat in five hours (or more)

[Steve Greenland]

First off, error messages can always be improved, and I bet the program
maintainers would be happy to accept patches, so long as those patches
don't *decrease* the amount of information available.

But in one area you're dead wrong:

On 05-Aug-03, 11:55 (CDT), Ian Hickson <[EMAIL PROTECTED]> wrote: 
> Without meaning offense, that is a very selfish attitude. The number of
> future debian users is *significantly* larger than the number of existing
> users, unless something drastic happens to either humanity or debian
> itself. Why should everyone who will use debian in future be forced to
> learn archaic commands, paths, and deal with other historical holdbacks,
> instead of the few who already use it being taught easier conventions?

Without meaning offense, that's a very selfish attitude. You want to
ruin a system that provide incredible productivity and ease-of-use for
the experienced user, for the sake of some hypothetical users who might
be better served by another OS?

You see, a great many of us (not just Debian users, but Unix users in
general) have learned what those "archaic" commands and paths are, and
their very shortness increases usability -- not "learnability" perhaps,
but actual usuability, for those who use it every day. Combine that
with the fact that they are the same[1] from system to system, and
any changes are a complete detriment to usability. To you, 'list' may
look better than 'ls', but for someone who types it several hundred
times a week, and is *far* more likely to mis-type 'list', 'ls' is
completely superior. Compare 'ls -ltr' with the VMS equivalent[2]
'directory/sort=date/order=reverse'. Yes, the latter is easier to read
*IF* you don't know either. But which would you rather type?

Which should keep you (or whoever wants to do it) from building a shell
that provides more english like commands. But please don't ruin the
system for the rest of us.

> [Re: UI designers] Unfortunately we don't seem to have many of those
> in the free software community.

Now *that* is a true statement.

Steve

[1] The existing variations are painful enough. Let's not add to the problem

[2] Yes, I know that's probably not the actual VMS command. It's been a
while, and I don't have manual handy. But it's close enough to get the
flavor of it.

-- 
Steve Greenland
The irony is that Bill Gates claims to be making a stable operating
system and Linus Torvalds claims to be trying to take over the
world.   -- seen on the net

Re: How to install X-Chat in five hours (or more)

[Richard Braakman]

On Tue, Aug 05, 2003 at 11:06:53PM +0200, Tollef Fog Heen wrote:
> Tab completion or using /Va* is about as fast as /var.

Heh, teach yourself to type /Va* and you're going to get BURNED one day.
(Your co-sysadmin finds a rootkit on another machine and stores it
in /Various Dangerous Programs/ for later examination...)

Tab completion is fine in contexts where it works.

Richard Braakman

Re: How to install X-Chat in five hours (or more)

[Steve Lamb]

On Tue, 5 Aug 2003 21:38:19 +0200
Emile van Bergen <[EMAIL PROTECTED]> wrote:
> Apple has a great way of doing that. They don't dumb down, they don't
> belittle you, they assume an intelligent being who can grasp reasonably
> complex English sentences, but who has less knowledge of computer
> idiom.

*blink, blink*  Funny, I consider Apple on of the worst when it comes to
dumbing down the interface.  Let's not forget they only took about 10 years to
get *2* mouse buttons because it was too confusing.

-- 
 Steve C. Lamb | I'm your priest, I'm your shrink, I'm your
   PGP Key: 8B6E99C5   | main connection to the switchboard of souls.
   |-- Lenny Nero - Strange Days
---+-


pgpNO5gaWvjcw.pgp
Description: PGP signature

Re: How to install X-Chat in five hours (or more)

[Steve Lamb]

On Tue, 5 Aug 2003 22:16:37 +0200
Emile van Bergen <[EMAIL PROTECTED]> wrote:
> I would even scream at 
> 
> /Variable Data/
> 
> simply because it encourages slow and RSI-inducing click and drag
> behaviour, because such path names are impossible to type in (and this
> one even requires escaping the space to distinguish between the argument
> separator). 

Come now, there is always tab.  Of course there is another reason why such
descriptive names aren't all that great.  Here's me sitting in a directory
mounted off my Windows box.  Can you tell where the unix system leaves off,
where the Microsoft system catches on and why the Microsoft system was clearly
never meant to be viewed from the command line?  :)

[EMAIL PROTECTED]:/mnt/morpheus_D/Program Files/Microsoft Games/Asheron's Call}

If it were an all unix world we might have this:

[EMAIL PROTECTED]:/mnt/morpheus_D/Games/AC}

No less descriptive, just as readable but, by gum, I have more than 3
spaces before my command wraps!  And heaven forbid I go one directory deeper!


-- 
 Steve C. Lamb | I'm your priest, I'm your shrink, I'm your
   PGP Key: 8B6E99C5   | main connection to the switchboard of souls.
   |-- Lenny Nero - Strange Days
---+-


pgpf68OSEt0xK.pgp
Description: PGP signature

Re: Should MUA only Recommend mail-transfer-agent?

[Steve Lamb]

On Tue, 5 Aug 2003 21:42:43 +0200
"Artur R. Czechowski" <[EMAIL PROTECTED]> wrote:
> I would like to know Md's opinion, but for me there are no reasons to relax
> dependencies for mutt (and other MUA). I would not like to do it without
> policy requirements because it concerns also other MUA's.

But it doesn't really concern other MUAs.  Neither KMail or Sylpheed-claws
have a depends on an MTA.


-- 
 Steve C. Lamb | I'm your priest, I'm your shrink, I'm your
   PGP Key: 8B6E99C5   | main connection to the switchboard of souls.
   |-- Lenny Nero - Strange Days
---+-


pgp7SwlzTQAZU.pgp
Description: PGP signature

Re: Should this be filed as grave? Gcc-2.95

[Steve Lamb]

On Tue, 5 Aug 2003 12:16:43 -0400
"H. S. Teoh" <[EMAIL PROTECTED]> wrote:
> Downgrading sounds like overkill in this situation. I only had to edit
> /usr/src/linux/Makefile to change HOSTCC to gcc-2.95 and export
> CC=gcc-2.95 in the environment, and it worked fine for me. This is on
> 2.4.21, of course, but I suspect the same holds for 2.4.20.

[ SNIP ]
 
> ln -s /usr/bin/gcc-2.95 /usr/bin/gcc
> 
> ln -s /usr/bin/gcc-3.3 /usr/bin/gcc

Both of which I'd have to remember to do again later on if I compiled the
kernel again.  No, when 2.95 didn't work for whatever reason when I told make
to use it I preferred to reduce the variables on the system.  If 3.3 isn't
present it can't be invoked, period.  I don't have to remember to edit
Makefiles later on on a different revision.  I don't have to remember to reset
symlinks, change alternatives or equivs or anything.  I'm miffed mainly
because I asked for 2.95 installed and got 3.3 as part of the deal.  I don't
take kindly to software installing other software without a clear need and
there simply was no clear need.

-- 
 Steve C. Lamb | I'm your priest, I'm your shrink, I'm your
   PGP Key: 8B6E99C5   | main connection to the switchboard of souls.
   |-- Lenny Nero - Strange Days
---+-


pgpqD18CMCOyb.pgp
Description: PGP signature

Re: Bug#202869: Should MUA only Recommend mail-transfer-agent?

[Andreas Jellinghaus]

> I can imagine a workstation without those packages but it is, IMO,
> mutilated box.

please keep your opinion outside the control file.
cron, at & friends __need__ an MTA (or to be exact:
a /usr/sbin/sendmail app), they will not work without.

mutt can do many nice things without /usr/sbin/sendmail.
a dependency is set if something is always required,
a recommends if is required for the common use, and
a suggestion is used if it improved the functionality.
so depending on mail-transport-agent is wrong,
the recommendation is fine.

or fix the policy to make a clear statement.
in that case maybe you want to reassign the bug.

> BTW, there is no need for exim4-daemon-heavy. There are other lightweight
> MTA's.

I know. but still the dependency dialog is confusing and allmost
all MTA even if not configure add poisen to a clean system
like useless cron jobs, logfile rotation etc.

an unused library only takes up a few inodes and kilobytes
of disk ram and thus is easy to bear. an unused MTA however
is quite a heavy thing compared to that. take a look
at all those silly debconf questions some packages have,
and you know why it is a good thing not to install one
on a system where you don't need it.

> Another solution is to prepare a dummy-mta package, which only
> provides mail-transfer-agent and required by policy /usr/sbin/sendmail
> and /usr/bin/newaliases binaries to do nothing[1].

sounds like shooting in ones foot to me.

If debian has one major policy, it is not to have a policy.
debian does not decide what architecture or window manager
or mail transport agent your want - you can choose. debian
does not decide whether you want a small systme or a big
fat installation - you can choose.

why should debian insist on installing a mail transport
agent where none is needed? and the easy solution is
to relax the dependency to a recommendation. The policy
supports this. Not that the wording in the policy is perfekt,
it could be improved (see my suggestion above) to support
the recommendation or to deny this bug report and put
an explicit dependency in the policy.

sure, this bug report is bigger than mutt, it will affect
many other mail readers and apps everywhere as well.

closing the bug or marking it as whishlist would be wrong.
debian claims not to hide problem. here is one. please 
accept it and handle it. 

Regards, Andreas

Re: Fw: Re: Ruby 1.8 transition plan; debian-ruby

[Joe Wreschnig]

On Tue, 2003-08-05 at 12:12, Fumitoshi UKAI wrote:
> Since ruby 1.8.0 was released recently, ruby developers will go to 
> ruby 1.8.x, so that we, ruby maintenance team (akira, tagoh, ukai), 
> are discussing about how to deal with Ruby 1.8 transition and trying to make
> debian ruby policy soon.
> 
> For now, ruby package provides /usr/bin/ruby of ruby 1.6.x, and
> ruby1.8 package provides /usr/bin/ruby1.8 of ruby 1.8.x.  Someone
> want to use /usr/bin/ruby of ruby 1.8.x, so we're considering to use 
> alternatives for /usr/bin/ruby to choice either ruby1.6, ruby1.8 (or any
> other version of ruby in future).

There was a bug about this at some point, which I can no longer find,
where I suggested doing for Ruby what Python does. Which is essentially:
- 'ruby' is a metapackage depending the current default version of Ruby
for Debian.
- 'rubyX.Y' is a specific version of Ruby. 'ruby' depends on one of
these.

- 'libfoo-bar-ruby' is a metapacakge depending on libfoo-bar-rubyX.Y,
where X.Y is the default version of Ruby (the same as the one 'ruby'
depends on).
- 'libfoo-bar-rubyX.Y' is foo-bar for a specific version of Ruby. The
above depends on one of these. (These packages may be unncessary if the
directory tree I outlined below is used, and the package is
version-independent.)

As for module include paths, they're less important since most Ruby
modules query Ruby for the right information at build-time anyway, but
I'd like to see version-independent directories, and also preferrably a
tree under /usr/share, too. Say,

These are arch-independent:
- /usr/share/ruby for Ruby modules that work with any version.
- /usr/share/ruby/X.Y for Ruby modules that depend on version X.Y.

These are arch-dependent:
- /usr/lib/ruby/version/X.Y/#{arch}-#{os} for all arch-dependent
modules. I believe most architecture-dependent modules need to be
recompiled for each version of Ruby anyway, and so a version-independent
architecture-dependent directory makes no sense.
-- 
Joe Wreschnig <[EMAIL PROTECTED]>


signature.asc
Description: This is a digitally signed message part

Re: How to install X-Chat in five hours (or more)

[Colin Watson]

On Tue, Aug 05, 2003 at 03:03:54PM -0700, Steve Lamb wrote:
> On Tue, 5 Aug 2003 09:55:59 -0700 (PDT)
> Ian Hickson <[EMAIL PROTECTED]> wrote:
> > And I'm a geek, one who has been using GNU-based distributions on multiple
> > machines on a daily basis for at least 3 years, and Sun for 6 years before
> > that, and I _still_ have difficulty. This should be ringing usability
> > alarm bells.
> 
> No, it shouldn't.  I should be ringing alarms about the veracity of your
> claims and your capabilities. 

Hixie's pretty well-known in certain other free software circles. What
I've seen of him elsewhere implies to me that he isn't incompetent in
the least, and frankly I think you're going way overboard in the
hostility of your responses. I'm sure it's possible for us to recognize
that things aren't perfect even if we disagree on how to fix them.

Cheers,

-- 
Colin Watson  [EMAIL PROTECTED]

Re: How to install X-Chat in five hours (or more)

[Alan Shutko]

Steve Lamb <[EMAIL PROTECTED]> writes:

> Oh, look, someone else who CCs when it is obvious the person they're
> responding to is participating right here.

Maybe you should stop whining and just set the Mail-Copies-To header,
which is generally respected by posters on Debian lists?

-- 
Alan Shutko <[EMAIL PROTECTED]> - I am the rocks.
What's all this I hear about "endangered feces"?

Re: How to install X-Chat in five hours (or more)

[Steve Lamb]

On Tue, 5 Aug 2003 23:30:11 +0100
Colin Watson <[EMAIL PROTECTED]> wrote:
> Hixie's pretty well-known in certain other free software circles. What
> I've seen of him elsewhere implies to me that he isn't incompetent in
> the least, and frankly I think you're going way overboard in the
> hostility of your responses. I'm sure it's possible for us to recognize
> that things aren't perfect even if we disagree on how to fix them.
 
You're most likely right, I'll bow out.

-- 
 Steve C. Lamb | I'm your priest, I'm your shrink, I'm your
   PGP Key: 8B6E99C5   | main connection to the switchboard of souls.
   |-- Lenny Nero - Strange Days
---+-


pgp1k6Er4gb9A.pgp
Description: PGP signature

Re: How to install X-Chat in five hours (or more)

[Alan Shutko]

Richard Braakman <[EMAIL PROTECTED]> writes:

> Heh, teach yourself to type /Va* and you're going to get BURNED one day.
> (Your co-sysadmin finds a rootkit on another machine and stores it
> in /Various Dangerous Programs/ for later examination...)

And gee, your shell beeps, completes up to /Various\ , and doesn't go
further.  How is this going to get you burned?

-- 
Alan Shutko <[EMAIL PROTECTED]> - I am the rocks.
Your rifle won't leave a wet spot on the bed after you use it.

Re: Bug#202869: Should MUA only Recommend mail-transfer-agent?

[Hans Fugal]

* Andreas Jellinghaus [Wed,  6 Aug 2003 at 00:27 +0200]
> mutt can do many nice things without /usr/sbin/sendmail.
> a dependency is set if something is always required,
> a recommends if is required for the common use, and
> a suggestion is used if it improved the functionality.
> so depending on mail-transport-agent is wrong,
> the recommendation is fine.
Mutt can read mail without an MTA, but cannot send mail without one.
Whether sending mail is considered simply "common use" or is a core
functionality has been and certainly will be nit-picked. Your argument
rests on the former. 

-- 
 Hans Fugal | De gustibus non disputandum est.
 http://hans.fugal.net/ | Debian, vim, mutt, ruby, text, gpg
 http://gdmxml.fugal.net/   | WindowMaker, gaim, UTF-8, RISC, JS Bach
-
GnuPG Fingerprint: 6940 87C5 6610 567F 1E95  CB5E FC98 E8CD E0AA D460


pgpVhuUMAFDkO.pgp
Description: PGP signature

Re: Bug#202869: Should MUA only Recommend mail-transfer-agent?

[Rene Engelhard]

Hi,

[ note that I atm have the tendency to say that the Depends should
  remain... ]

Hans Fugal wrote:
> * Andreas Jellinghaus [Wed,  6 Aug 2003 at 00:27 +0200]
> > mutt can do many nice things without /usr/sbin/sendmail.
> > a dependency is set if something is always required,
> > a recommends if is required for the common use, and

Right,  although I normally, I decide from pkg to pkg with this
and if the feature needing a package is one of main ones
of the software or not. And the MUAs main features are
reading and _sending_ mails

> > a suggestion is used if it improved the functionality.
> > so depending on mail-transport-agent is wrong,
> > the recommendation is fine.
> Mutt can read mail without an MTA, but cannot send mail without one.

... but this is bullshit.

I don't need a MTA to send my mail with mutt.

I use offlineimap, courier's outbox feature, direct the mail to
the local Outbox[1], synchronize it and the _mail server_ sends it out.

Or if someone accesses IMAP directly it could

set sendmail=/dev/null

and set Fcc to the Outbox which has the same effect.

There are so many possibilities to send mails from mutt without an MTA..

Yes, sure, this is not the "common" configuration, but possible
and the argument that people using mutt _need_ a MTA is evidently
wrong...

Grüße/Regards,

René

[1] set sendmail=~/bin/imap/mailout which contains
safecat /home/rene/Mail/INBOX.Outbox/tmp /home/rene/Mail/INBOX.Outbox/new
hi Omnic :)
-- 
 .''`.  René Engelhard -- Debian GNU/Linux Developer
 : :' : http://www.debian.org | http://people.debian.org/~rene/
 `. `'  [EMAIL PROTECTED] | GnuPG-Key ID: 248AEB73
   `-   Fingerprint: 41FA F208 28D4 7CA5 19BB  7AD9 F859 90B0 248A EB73


pgpbY9uc6wnP3.pgp
Description: PGP signature

Re: Have Linux boot with eye-candy

[Erich Schubert]

Hi,
i have built packages for the bootsplash tools (no package for the patch
itself though. just download and apply the diff).
They are available on http://people.debian.org/~erich/boot/bootsplash/
and work fine on my notebook as well as my sisters.
I didn't get the bootsplash support of swsusp working though; but maybe
i just forgot applying that "connector" patch... ;)

No AA Text rendering yet (what text to use? that would require
modifications to the init scripts... i didn't build this fbtruetype tool
yet, but i probably is as easy as the rest) and no animations (packages
are built, but i don't have nice animations, i'm no artist. ;) )
Progress bar is working fine if you use the supplied rc.splash
(just change it in /etc/inittab)

You'll find a theme i built for Debian there, too.
It uses a really cool background made by  Alexis Younes (ayo),
http://www.73lab.com/  -- unfortunately i don't know if it is DFSG-free.
using the debian logos it probably isn't. ;)

Greetings,
Erich Schubert
-- 
erich@(vitavonni.de|debian.org)--GPG Key ID: 4B3A135C(o_
There was never a good war or a bad peace. - Benjamin Franklin   //\
   Die kürzeste Verbindung zwischen zwei Menschen ist ein Lächeln.   V_/_

Re: How to install X-Chat in five hours (or more)

[Gustavo Noronha Silva]

Em Tue, 05 Aug 2003 17:39:06 -0500, Alan Shutko <[EMAIL PROTECTED]> escreveu:

> Richard Braakman <[EMAIL PROTECTED]> writes:
> 
> > Heh, teach yourself to type /Va* and you're going to get BURNED one day.
> > (Your co-sysadmin finds a rootkit on another machine and stores it
> > in /Various Dangerous Programs/ for later examination...)
> 
> And gee, your shell beeps, completes up to /Various\ , and doesn't go
> further.  How is this going to get you burned?

I guess he's not talking about tab completion in this case, but about
'*'... I actually use this stuff when I cannot use tabs...

[~]
[EMAIL PROTECTED] $ cd /ho*/k*
[/home/khaotikuz]
[EMAIL PROTECTED] $

This should make the problem clearer... I actually wanted to go to 
'/home/kov', but my dumb sysadmin (me =P) created a khaotikuz user...

[]s!

-- 
[EMAIL PROTECTED]: Gustavo Noronha 
Debian:   *  
Dúvidas sobre o Debian? Visite o Rau-Tu: http://rautu.cipsga.org.br

Re: How to install X-Chat in five hours (or more)

[Gustavo Noronha Silva]

Em Tue, 5 Aug 2003 10:19:53 -0700 (PDT), Ian Hickson <[EMAIL PROTECTED]> 
escreveu:

> > You can't get there from here, I think. Unix admins coming to Debian
> > will scream blue murder if it starts being "/My Variable Data/Logs", and
> > that group is important to us.
> 
> Note that there is at least one project which is looking at doing exactly
> that while retaining backwards compatability (GoboLinux). It may be worth,
> on the long term, looking at how it may be possible to migrate from
> obscure paths like "/opt", "/bin", "/sbin", "/usr/bin", etc, to more
> sensible names, in that way.

They had a presentation on the IV Free Software International Forum in
Porto Alegre (Brazil) this year and I can assure you I didn't hear a
lot of good things about this (I didn't actually attend to the thing,
but other DD's did).

It seems like a mess of symlinks and they say it doesn't need 'package
management' when it actually has a bunch of scripts to handle the symlink
'farm' they grow on /usr/bin and such.

Doesn't seem like a clean solution to me, no no no...

It also seems like they want to have each package installed in its
own directory, which I think sucks tremendously... I hated that when
I lived with windows... I always wanted to have all my binaries in
the PATH, etc. And the real end user does not even care about that.

I don't really care about /. I think path abstraction should be achieved
by graphical file management utilities like Nautilus, not by messing
with something that's actually working, to then cause more problems.

[]s!

-- 
[EMAIL PROTECTED]: Gustavo Noronha 
Debian:   *  
Dúvidas sobre o Debian? Visite o Rau-Tu: http://rautu.cipsga.org.br

Re: How to install X-Chat in five hours (or more)

[Gunnar Wolf]

Gustavo Noronha Silva dijo [Tue, Aug 05, 2003 at 09:26:20PM -0300]:
> > Note that there is at least one project which is looking at doing exactly
> > that while retaining backwards compatability (GoboLinux). It may be worth,
> > on the long term, looking at how it may be possible to migrate from
> > obscure paths like "/opt", "/bin", "/sbin", "/usr/bin", etc, to more
> > sensible names, in that way.
> 
> They had a presentation on the IV Free Software International Forum in
> Porto Alegre (Brazil) this year and I can assure you I didn't hear a
> lot of good things about this (I didn't actually attend to the thing,
> but other DD's did).
> (...)
> I don't really care about /. I think path abstraction should be achieved
> by graphical file management utilities like Nautilus, not by messing
> with something that's actually working, to then cause more problems.

I completely agree with you... I was arguing with a friend of mine, a
Ximian developer. He insisted me that they were bringing Unix to the
desktop of people, just like what Apple did. I insst that is *not* what
they are doing. Even though under every Ximian and MacOS X system there
is a Unix box, the user only uses a strange abstraction that -maybe
without the user's knowledge- uses Unix itself.

Debian is a Unix system (ok, Unix-like for purism). It should stay a
Unix system. We already have a very important user base, we are obliged
not to give them such a headache.

Besides, I really doubt that even one tenth of the Debian developers
would be happy to switch ;-)

Greetings,

-- 
Gunnar Wolf - [EMAIL PROTECTED] - (+52-55)5630-9700 ext. 1366
PGP key 1024D/8BB527AF 2001-10-23
Fingerprint: 0C79 D2D1 2C4E 9CE4 5973  F800 D80E F35A 8BB5 27AF

Re: How to install X-Chat in five hours (or more)

[Brian Nelson]

Alan Shutko <[EMAIL PROTECTED]> writes:

> Steve Lamb <[EMAIL PROTECTED]> writes:
>
>> Oh, look, someone else who CCs when it is obvious the person they're
>> responding to is participating right here.
>
> Maybe you should stop whining and just set the Mail-Copies-To header,
> which is generally respected by posters on Debian lists?

Har!  The people that use a decent MUA already know how to participate
in list discussions, and those that don't won't pay a bit of attention
to that header.

-- 
I'm sick of being the guy who eats insects and gets the funny syphilis.


pgpSznaaJXdBF.pgp
Description: PGP signature

Request for maintainer

[jmd]

Hi!

Subject: RFP: GRubik -- A 3D Rubik cube game
Package: wnpp
Version: N/A; reported 2003-08-06
Severity: wishlist

* Package name: GRubik
  Version : 1.16
  Upstream Author : John Darrington <[EMAIL PROTECTED]>
* URL : http://www.freesoftware.fsf.org/rubik/grubik.html
* License : GPL-2
  Description : A 3D Rubik cube game

This is an OpenGL / GTK+ package which I have written.  I've had
positive feedback from the people who've looked at it so far. It's fully
internationalised, has a number (5?) localisations.

I'm not a DD, and am not currently able to commit to being one.
However if a DD wants to become a maintainer of this package I will
co-operate with him/her (eg upstream patches where needed). 

I have created an unofficial deb for this software, and you can get it
from my website http://darrington.wattle.id.au/deb if you want to use
that as a starting point.  


-- System Information
Debian Release: 3.0
Architecture: i386
Kernel: Linux freyja.cellform.com 2.2.17 #1 Sun Jun 25 09:24:41 EST 2000 i?86
Locale: LANG=en_AU, LC_CTYPE=en_AU

-- 
PGP Public key ID: 1024D/2DE827B3 
fingerprint = 8797 A26D 0854 2EAB 0285  A290 8A67 719C 2DE8 27B3
See http://wwwkeys.pgp.net or any PGP keyserver for public key.








pgpdTirubV8Nf.pgp
Description: PGP signature

OT (Re: Excessive wait for DAM - something needs to be done)

[Goswin Brederlow]

Martin Michlmayr - Debian Project Leader <[EMAIL PROTECTED]> writes:

> * Jamin W. Collins <[EMAIL PROTECTED]> [2003-07-21 18:52]:
> > Perhaps that is because only the DPL can appoint them (as far as I can
> > tell) and we haven't seen a request from you for them.
> 
> Request for help are usually very ineffective; examples: apt's
> maintainer asked for help and didn't get any (with the exception that
> mdz has started more apt work, but he worked on apt before so
> effectively there are no new volunteers), Bdale is looking for a
> co-maintainer of ntp as is md for mutt.

Mails with suggestions and offers to help (and I'm esspecially mean
apt here :( ) also go unanswered or patches for bugs or improvements
get just overlooked and never included.

Shit happens. Don't stop trying.

MfG
Goswin

Re: Excessive wait for DAM - something needs to be done

[Goswin Brederlow]

"Dwayne C. Litzenberger" <[EMAIL PROTECTED]> writes:

> On Sun, Jul 13, 2003 at 01:09:47PM -0600, Jamin W. Collins wrote:
> > 2001-01-24 - Dwayne Litzenberger <[EMAIL PROTECTED]>
> >http://nm.debian.org/nmstatus.php?email=dlitz%40dlitz.net
> 
> For the record, I'm still interested in becoming a DD.
> 
> Nice to see this finally being addressed!  Thanks!

http://nm.debian.org/nmstatus.php?email=brederlo%40informatik.uni-tuebingen.de

Received application2000-08-25

Still waiting too.

MfG
Goswin

Re: NM non-process

[Goswin Brederlow]

Steve Langasek <[EMAIL PROTECTED]> writes:

> On Mon, Jul 21, 2003 at 10:17:24AM -0400, Nathanael Nerode wrote:
> 
> > Martin Schulze is listed as the other DAM member.  He's also the Press 
> > Contact, so I certainly hope he has good communication skills!
> 
> And the Stable Release Manager, and a member of the Security Team, and
> a member of debian-admin.  What makes you think he would give a higher
> priority to DAM work than James currently does?
> 
> Actually, given that Joey is already listed as part of DAM, and isn't
> actively involved, doesn't this suggest he already gives a lower
> priority to this work?

As far as I heard Matrin is only there in case the DAM dies. He won't
create or delete account on his own while the DAM is still breathing
(or thought to be).

MfG
Goswin

Re: NM non-process

[Goswin Brederlow]

[EMAIL PROTECTED] (Nathanael Nerode) writes:

> Steve Langasek said:
> >I don't think it irrelevant that those clamouring loudest for the DPL
> >to do something to fix the situation are people who don't actually have
> >a say in the outcome of DPL elections.  While I'm not happy to see such
> >long DAM wait times, I'm also not volunteering to take on the thankless
> >job myself.
> 
> No, it's not irrelevant.  It means precisely that Debian is in danger of 
> becoming an unresponsive, closed group which does not admit new people.  
> If this continues for, say, 2 more years, I would expect a new Project 
> to be formed, replicating what Debian is doing, but admitting new 
> people.  I'd probably be right there starting it.
> 
> That would be a stupid waste of effort, so I hope it turns out to be 
> unnecessary.

I know of several DDs and non-DDs thinking about creating a Debian2 (or
whatever named) project due to this and other lack of responce
problems and the group is growing. The danger is already there and
should not be ignored.

MfG
Goswin

Re: NM non-process

[Goswin Brederlow]

Kalle Kivimaa <[EMAIL PROTECTED]> writes:

> Roland Mas <[EMAIL PROTECTED]> writes:
> > with.  The MIA problem is significant enough that NM might be the only
> > way to tackle with it seriously.  That means taking time to examine
> > applications.
> 
> BTW, has anybody done any research into what types of package
> maintainers tend to go MIA? I would be especially interested in a
> percentage of "old" style DD's, DD's who have gone through the NM
> process, people going MIA while in the NM queue, and people going MIA
> without ever even entering the NM queue. I'll try to do the statistics
> myself if nobody has done it before.

And how many NMs go MIA because they still stuck in the NM queue after
years? Should we ask them? :)

MfG
Goswin

Re: Request for maintainer

[Goswin Brederlow]

[EMAIL PROTECTED] writes:

> Hi!
> 
> Subject: RFP: GRubik -- A 3D Rubik cube game
> Package: wnpp
> Version: N/A; reported 2003-08-06
> Severity: wishlist
> 
> * Package name: GRubik
>   Version : 1.16
>   Upstream Author : John Darrington <[EMAIL PROTECTED]>
> * URL : http://www.freesoftware.fsf.org/rubik/grubik.html
> * License : GPL-2
>   Description : A 3D Rubik cube game
> 
> This is an OpenGL / GTK+ package which I have written.  I've had
> positive feedback from the people who've looked at it so far. It's fully
> internationalised, has a number (5?) localisations.
> 
> I'm not a DD, and am not currently able to commit to being one.
> However if a DD wants to become a maintainer of this package I will
> co-operate with him/her (eg upstream patches where needed). 
> 
> I have created an unofficial deb for this software, and you can get it
> from my website http://darrington.wattle.id.au/deb if you want to use
> that as a starting point.  

Maybe you should ask on the new maintainer list for some new
maintainer thats intrested and needs a package to maintain. Just an
idea.

MfG
Goswin

Re: Bug#202869: Should MUA only Recommend mail-transfer-agent?

[Erik Steffl]

Hans Fugal wrote:
* Andreas Jellinghaus [Wed,  6 Aug 2003 at 00:27 +0200]
mutt can do many nice things without /usr/sbin/sendmail.
a dependency is set if something is always required,
a recommends if is required for the common use, and
a suggestion is used if it improved the functionality.
so depending on mail-transport-agent is wrong,
the recommendation is fine.
Mutt can read mail without an MTA, but cannot send mail without one.
  it does not have to be on the same machine
erik

The results of your email commands

[Cc-bounces]

The results of your email command are provided below. Attached is your
original message.

- Results:
Invalid confirmation string.  Note that confirmation strings expire
approximately 3 days after the initial subscription request.  If your
confirmation has expired, please try to re-submit your original request or
message.

- Unprocessed:
-- 
Reasumuj±c i upraszczaj±c, wysy³anie czegokolwiek do mnie przez internet
nie jest niczyim prawem, jest przywilejem.
/Marcin Frankowski, pl.news.mordplik, dyskusja ze spamerem/

- Done.

--- Begin Message ---

-- 
Reasumując i upraszczając, wysyłanie czegokolwiek do mnie przez internet
nie jest niczyim prawem, jest przywilejem.
/Marcin Frankowski, pl.news.mordplik, dyskusja ze spamerem/


--- End Message ---