gcc-3.2: what happened to libgcc1, libg2c0, binutils
After five months of struggling ( I had connectivity very briefly before an "upgrade" make my whole installation useless and sent me back to the CD's ), it appears that I need to be able to build the nVidia kernel-patch with _the_same_version_ of gcc that was used to compile my kernel. I'm using kernel-image-2.4.26, and the nVidia installer reports this was built with gcc-3.2. OK, I go to install gcc-3.2. Including the Fortran, which I probably don't really need, I get these complaints from dpkg: "gcc-3.2 depends on libgcc1 (>= 1:3.2.3-9)" "gcc-3.2 depends on binutils (>= 2.13.90.0.10)" "g77-3.2 depends on libg2c0;" "lib.depends on libstdc++5 (>= 1:3.2.3-9)" Argh, those are not with gcc-3.2. In fact, when I check the download site (tonight) I find libgcc1_3.0.4-7_i386.deb from gcc-3.0 libgcc1_3.3.4-13_hppa.deb -- and NO i386 package from gcc-3.3 libgcc1_3.3.5-2_hppa.deb-- and NO i386 package from gcc-3.3 libgcc1_3.4.2-2_i386.deb from gcc-3.4 libgcc1_4.0-0pre0_i386.deb from gcc-4.0 libg2c0_3.3.4-13_i386.deb from gcc-3.3 libg2c0_3.3.5-2_i386.deb from gcc-3.3 binutils_2.12.90.0.1-4_i386.deb - which I already have and is too low binutils_2.15-4_i386.deb It seems I am stuck dead in the water unless I can come up with a working gcc-3.2 compiler to do the nVidia install. The alternative looks like trying to recompile my kernel with I compiler I can get. I actually tried that with gcc-3.4 but got errored out immediately. That is hardly my first choice, so I didn't pursue that further. I have gcc-2.95 and gcc-3.0. I had gcc-3.2 but had to remove it because I couldn't get 3.3 or 3.4 to coexist. I don't really want to give away space to five or six different generations of compiler. One, maybe two, but not this fiasco. Where can I go to get the missing libs?? [ backports is also no help in this ]. Anyway, why is gcc-3.2 on the download site so incomplete? I also couldn't get 3.3 up and flying for very similar reasons - key libs missing - but what I need is 3.2! -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Life is too short to tolerate crappy software! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: gcc-3.2: what happened to libgcc1, libg2c0, binutils
Rui Silva wrote: why not do apt-get install --reinstall gcc3.2 it would download all the dependecies First reason: this damn nVidia box won't give me internet connectivity from Linux until I can apply the patches. Second reason: if it isn't on the download site, it isn't gonna be downloaded anyway. My only connection, until I can get past this problem, is to download .deb files using Windoze, boot into Debian, locate the file and do dpkg --install ... . Then I find out what's missing and I have to boot back to Win to locate pieces to fix it. Serious PITA. If Windows weren't such a piece of crap I would have about run out of patience and chucked the Linux install altogether. On Tuesday 23 November 2004 00:19, David A. Cobb wrote: After five months of struggling ( I had connectivity very briefly before an "upgrade" make my whole installation useless and sent me back to the CD's ), it appears that I need to be able to build the nVidia kernel-patch with _the_same_version_ of gcc that was used to compile my kernel. I'm using kernel-image-2.4.26, and the nVidia installer reports this was built with gcc-3.2. OK, I go to install gcc-3.2. Including the Fortran, which I probably don't really need, I get these complaints from dpkg: "gcc-3.2 depends on libgcc1 (>= 1:3.2.3-9)" "gcc-3.2 depends on binutils (>= 2.13.90.0.10)" "g77-3.2 depends on libg2c0;" "lib.depends on libstdc++5 (>= 1:3.2.3-9)" Argh, those are not with gcc-3.2. In fact, when I check the download site (tonight) I find libgcc1_3.0.4-7_i386.deb from gcc-3.0 libgcc1_3.3.4-13_hppa.deb -- and NO i386 package from gcc-3.3 libgcc1_3.3.5-2_hppa.deb-- and NO i386 package from gcc-3.3 libgcc1_3.4.2-2_i386.deb from gcc-3.4 libgcc1_4.0-0pre0_i386.deb from gcc-4.0 libg2c0_3.3.4-13_i386.deb from gcc-3.3 libg2c0_3.3.5-2_i386.deb from gcc-3.3 binutils_2.12.90.0.1-4_i386.deb - which I already have and is too low binutils_2.15-4_i386.deb It seems I am stuck dead in the water unless I can come up with a working gcc-3.2 compiler to do the nVidia install. The alternative looks like trying to recompile my kernel with I compiler I can get. I actually tried that with gcc-3.4 but got errored out immediately. That is hardly my first choice, so I didn't pursue that further. I have gcc-2.95 and gcc-3.0. I had gcc-3.2 but had to remove it because I couldn't get 3.3 or 3.4 to coexist. I don't really want to give away space to five or six different generations of compiler. One, maybe two, but not this fiasco. Where can I go to get the missing libs?? [ backports is also no help in this ]. Anyway, why is gcc-3.2 on the download site so incomplete? I also couldn't get 3.3 up and flying for very similar reasons - key libs missing - but what I need is 3.2! -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Life is too short to tolerate crappy software! -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Life is too short to tolerate crappy software! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: gcc-3.2: what happened to libgcc1, libg2c0, binutils
Alexander Sack wrote: David A. Cobb wrote: On Tuesday 23 November 2004 00:19, David A. Cobb wrote: After five months of struggling ( I had connectivity very briefly before an "upgrade" make my whole installation useless and sent me back to the CD's ), it appears that I need to be able to build the nVidia kernel-patch with _the_same_version_ of gcc that was used to compile my kernel. I'm using kernel-image-2.4.26, and the nVidia installer reports this was built with gcc-3.2. You surely refer to the official nvidia driver. Why not just try to install the precompiled nvidia module provided in non-free. I am runing that driver without any problems so far. I hope you mean the NFORCE one - my problem is getting onto the 'net; graphic fanciness is way down on my list. Last time I did that it didn't work, but it is worth trying again. Another option IMHO would be simply to build your kernel with a new gcc. Urm. Yes, in fact I was trying that without much success. It still leaves me with the question of just WHAT 'gcc' I can put up that doesn't have a piece missing. Can't someone point out what is broken that critical module(s) are unavailable from at least two "current" compiler package-sets? And, OBTW, Alexander, please -- what version of the kernel are you running? Thanks, -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Life is too short to tolerate crappy software! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Available printers
Debian kernel-2.4.29-f3, mostly sarge, some sid. Gnome Desktop 2.8.3 Cups: 1.0.23-10 mozilla-firefox browser, 2 versions, debian 1.0.2 for everybody but myself Epson Color Stylus CX5400 printer When I print from my own desktop, the choices of printer offered by the browser's print dialog are: * CUPS/scx5300_5400 * Default Postscript And things seem to print correctly When my wife tries to print from her desktop, the choices of printer are * scx5400_5400@:64 * xp_ps_spooldir_HOME_xprint ... @:64 * xp_pdf_spooldir_HOME_xprint ...@:64 * Postscript/scx5300_5400 And everything she tries to print comes out with very large fonts and only about half on the page, as though the dots-per-inch setting is way off. My SWAG is that the application is picking up the wrong default printer information. But I have no idea where the printer info comes from. It's worth noting that everything was working fine until one of my frequent aptitude--upgrade runs included the cupsys packages. Meanwhile, this has convinced my wife that I only installed Linux to annoy her. -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Free at last! Free at last! Using Linux, I'm FREE at last! Life is too short to tolerate crappy software! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
cron, or something, chatters too much
Trying slowly to recover from what I think was a "dist-upgrade" that downgraded everything until damn little works. Ah, but I did finally get nVidia stuff working, so at least I'm on-line after a fashion. PROB: While I'm working (TTY1 as root) I get interrupted every minute or so by about 10 lines of: PAM_unix[12345] (cron) session started for user [root, mail, amvis-stats] . . . and PAM_unix[23456] (cron) session ended for I guess the information is useful ( except that none of the components involved are configured to do anything useful yet ). But I don't want it on my TTY ! I pops into the middle of an info screen or my editor or whatever else I'm trying to do. And it's not always easy to get rid of. How can I tell PAM or cron or whatever to log its complaints somewhere else; or to block any daemon logging messages from my TTY? TIA -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Life is too short to tolerate crappy software! begin:vcard fn:David A. Cobb n:Cobb;David A. adr:;;7 Lenox Av #1;West Warwick;RI;02893-3918;USA email;internet:[EMAIL PROTECTED] title:Independent Software Consultant note:PGP Key ID#0x4C293929 effective 01/28/2004 x-mozilla-html:TRUE version:2.1 end:vcard
dist-upgrade . . . to what?
I can't find anything in the docs to give me a clue. If I do apt-get dist-upgrade, how do I direct it which of [stable, testing, unstable] to target? Or am I misunderstanding the whole thing? Dist-upgrade seems to have been the cause of my present brokenness -- a lot of packages got removed or downgraded. Right now I've fiddled sources.list so only 'testing' stuff is visible, since that is where I want to go. Isn't there a "better" way? It looked as though the "-o=whatever" was the trick, but I can't put anything there that doesn't get a complaint. -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Life is too short to tolerate crappy software! begin:vcard fn:David A. Cobb n:Cobb;David A. adr:;;7 Lenox Av #1;West Warwick;RI;02893-3918;USA email;internet:[EMAIL PROTECTED] title:Independent Software Consultant note:PGP Key ID#0x4C293929 effective 01/28/2004 x-mozilla-html:TRUE version:2.1 end:vcard
dpkg-query --showformat
Trying to use --showformat to get information from the package
database. I'm fine with everything EXCEPT the Filename.
All the other items I've tried, the syntax "${Package} . . . ${Version}
etc" the "variable" matches the tag on the lines of the available file --
Package: ...
Version:
However, ${Filename} does not output anything -- always blanks.
So, what's the magic? I'm running short on goats here ( and the smell
of burnt goat hair is really annoying ).
--
David A. Cobb, Software Engineer, Public Access Advocate
"By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a
Pilgrim: R.French, Tr.
Life is too short to tolerate crappy software!
begin:vcard
fn:David A. Cobb
n:Cobb;David A.
adr:;;7 Lenox Av #1;West Warwick;RI;02893-3918;USA
email;internet:[EMAIL PROTECTED]
title:Independent Software Consultant
note:PGP Key ID#0x4C293929 effective 01/28/2004
x-mozilla-html:TRUE
version:2.1
end:vcard
Newbie problems galore
Hi! I purchased a CD set of "Debian 3.0 "Woody" Official" and started to install it. First woe: my hardware is by nVidia; the Official kernel on the CD is 2.4.18, nVidia provides a driver for 2.4.20/21 and the Debian archive has nVidia patches for 2.4.26. My plan is to go to .26, but in the meanwhile I can't use X Windows and I can't access my NIC, so all that stuff requires rebooting into Windows. So, my first question: given the kernel images on $MIRROR/Debian/main/k/kernel-image-2.4.26-i386, how do I change my kernel? I need to pass things back-and-forth between Linux and Windoze. I see references to VFAT FS on the web site, but for the life of me, I can't find a trace of the software. It's really bad to have to play games with tar at both sides of the route in order not to munge up the "magic" pathnames. PLEASE don't tell me that the evil beast of Redmond has buried VFAT under a patent claim!! If not, please, where can I find it? Right now, I've managed to hork up my package data so dpkg gets hung up trying to fix things. My best bet seems to be to restart from scratch. How do I get dpkg / apt / aptitude to clean my machine totally, or what files should I remove to make all this stuff go away? Or, would it really be quicker to re-init my partitions and start again from the CD? Thanks, all, for your patience. I hope I see some answers before I do something more silly than what I've already done here. -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Life is too short to tolerate crappy software! begin:vcard fn:David A. Cobb n:Cobb;David A. adr:;;7 Lenox Av #1;West Warwick;RI;02893-3918;USA email;internet:[EMAIL PROTECTED] title:Independent Software Consultant note:PGP Key ID#0x4C293929 effective 01/28/2004 x-mozilla-html:TRUE version:2.1 end:vcard
Re: Newbie problems galore
inline-> Kent West wrote: David A. Cobb wrote: First woe: my hardware is by nVidia; the Official kernel on the CD is 2.4.18, nVidia provides a driver for 2.4.20/21 and the Debian archive has nVidia patches for 2.4.26. My plan is to go to .26, but in the meanwhile I can't use X Windows and I can't access my NIC, so all that stuff requires rebooting into Windows. You might be able to get your network working, in which case you won't need to reboot into Windows. Run "lspci" to see what type of Ethernet device you have, then run "modprobe" to install the correct driver for it. You might then need to modify /etc/network/interfaces (see "man interfaces" for examples) and restart networking (/etc/init.d/networking stop && /etc/init.d/networking start). PASTING ALL THE PROBABLY-RELEVANT 00:02.2 USB Controller: nVidia Corporation: Unknown device 0068 (rev a3) (prog-if 20) Subsystem: Micro-star International Co Ltd: Unknown device 5700 Flags: bus master, 66Mhz, fast devsel, latency 0, IRQ 10 Memory at dd00 (32-bit, non-prefetchable) [size=256] Capabilities: [44] #0a [2080] Capabilities: [80] Power Management version 2 00:04.0 Ethernet controller: nVidia Corporation: Unknown device 0066 (rev a1) Subsystem: Micro-star International Co Ltd: Unknown device 570c Flags: bus master, 66Mhz, fast devsel, latency 0, IRQ 11 Memory at dd001000 (32-bit, non-prefetchable) [size=4K] I/O ports at d000 [size=8] Capabilities: [44] Power Management version 2 00:06.0 Multimedia audio controller: nVidia Corporation: Unknown device 006a (rev a1) Subsystem: Micro-star International Co Ltd: Unknown device 5700 Flags: bus master, 66Mhz, fast devsel, latency 0, IRQ 5 I/O ports at d400 [size=256] I/O ports at d800 [size=128] Memory at dd002000 (32-bit, non-prefetchable) [size=4K] Capabilities: [44] Power Management version 2 ... 01:06.0 VGA compatible controller: nVidia Corporation Vanta [NV6] (rev 15) (prog-if 00 [VGA]) Flags: bus master, 66Mhz, medium devsel, latency 32, IRQ 10 Memory at da00 (32-bit, non-prefetchable) [size=16M] Memory at d800 (32-bit, prefetchable) [size=32M] Expansion ROM at [disabled] [size=64K] Capabilities: [60] Power Management version 1 01:07.0 Communication controller: Lucent Microelectronics LT WinModem Subsystem: Risq Modular Systems, Inc.: Unknown device 044e Flags: bus master, medium devsel, latency 32, IRQ 11 Memory at dc00 (32-bit, non-prefetchable) [size=256] I/O ports at c000 [size=8] I/O ports at c400 [size=256] Capabilities: [f8] Power Management version 2 I can't find anything at debian.org that looks likely -- the "nforce" driver from nVidia is the right one but I can't seem to match my distro. However, I found that "TUN" is generic enough to allow the installation to see the network. Vidio . . . well, that's another matter, and I'll worry about it later. I'll try to insmod vesa later on. So, my first question: given the kernel images on $MIRROR/Debian/main/k/kernel-image-2.4.26-i386, how do I change my kernel? Once you have networking working, just "apt-get install kernel-image-2.4.26-i386" (although I suspect you really want the i686 variant instead of i386). K6, actually. I need to pass things back-and-forth between Linux and Windoze. If your Windows system is on FAT32 (not NTFS), no problem. If it is on NTFS, you'll need to create a "shared" FAT32 partition. To mount a FAT32 partition, you can do something similar to: mount /dev/hda1 /home/superbskt/MyWinFiles (assuming you've created a "MyWinFiles" directory in your home directory, and assuming Windows is on /dev/hda1). You can modify /etc/fstab to do this for you automatically if you want. Yeah. I cannot get my NTFS partitions visible at this point. I reconfigured several things, used Partition Magic to convert a big chunk of disk real estate. That, at least, is temporarily satisfactory. I swear, the first time I tried using VFAT, if failed so I fell back to MSDOS and lost long filenames. Anyway, it works now. I'll think about "why's" after I have a working system. Right now, I've managed to hork up my package data so dpkg gets hung up trying to fix things. My best bet seems to be to restart from scratch. How do I get dpkg / apt / aptitude to clean my machine totally, or what files should I remove to make all this stuff go away? Or, would it really be quicker to re-init my partitions and start again from the CD? Without knowing what you mean when you say "hork up", it's hard to say. I will say that it's almost never necessary to wipe/rebuild a Debian box (although sometimes it's easier to do that, depending on circumstances / skill level / etc). Well, when I get all the "Broken"
Re: Newbie problems galore
Jacob S. wrote: On Thu, 15 Jul 2004 12:08:14 -0500 Kent West <[EMAIL PROTECTED]> wrote: David A. Cobb wrote: First woe: my hardware is by nVidia; the Official kernel on the CD is 2.4.18, nVidia provides a driver for 2.4.20/21 and the Debian archive has nVidia patches for 2.4.26. My plan is to go to .26, but in the meanwhile I can't use X Windows and I can't access my NIC, so all that stuff requires rebooting into Windows. You might be able to get your network working, in which case you won't need to reboot into Windows. Actually, he mentions nVidia hardware above. By this I assume he means an nVidia chipset on his motherboard. In which case, his network card (built into the motherboard) is not recognized unless you have a kernel newer than 2.4.20 (not sure the exact version where they added it) using the forcedeth module or by applying nVidia's patches to a 2.4.20 kernel and using nvnet. URGGH. Actually, as I said in answer to Kent, "TUN" is able to load and recognize my card. Video matters -- later. I've looked around here at Debian.org enough to know that nVidia presents a major PITA until I can upgrade my kernel. And, until I can reach the net and let apt-get worry about what's missing, that is a thought to horrid to be borne. Unfortunately, either way would require him to go into Windows to download stuff, for his nVidia hardware to work in Woody. David, if your Windows partition really is vfat format, and not running something like XP which uses NTFS, you probably just need to modprobe vfat before trying to mount it. Look at Monique's e-mail for a good fstab example, if you need it. HTH, Jacob -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Life is too short to tolerate crappy software! begin:vcard fn:David A. Cobb n:Cobb;David A. adr:;;7 Lenox Av #1;West Warwick;RI;02893-3918;USA email;internet:[EMAIL PROTECTED] title:Independent Software Consultant note:PGP Key ID#0x4C293929 effective 01/28/2004 x-mozilla-html:TRUE version:2.1 end:vcard
Re: Newbie problems galore
Monique Y. Mudama wrote: On 2004-07-15, David A. Cobb penned: [snip] I don't have the brain power right now to answer your question about kernels ... so moving on ... I need to pass things back-and-forth between Linux and Windoze. I see references to VFAT FS on the web site, but for the life of me, I can't find a trace of the software. It's really bad to have to play games with tar at both sides of the route in order not to munge up the "magic" pathnames. PLEASE don't tell me that the evil beast of Redmond has buried VFAT under a patent claim!! If not, please, where can I find it? Vfat support isn't provided as an application; it's a capability you need to have enabled in the kernel. I would think the kernel image you have would include that support by default. What are the relevant /etc/fstab lines and 'mount' commands you've tried to use? For example, I have this in my /etc/fstab : /dev/sda1 /cigar vfatdefaults,user,noauto0 0 It allows me to mount my usb keychain drive, formatted as vfat, onto my linux system. Then I just type mount /cigar and I have access to the drive. Yeah. I swear it didn't work the first time but it does now. I converted a lot of disk space ( backwards ) to FAT32 to create my space. Actually, I'm making my whole "/var" file system shared between the two because I'm planning to do some cross-platform stuff and my installation looks very *nix-y already. Right now, I've managed to hork up my package data so dpkg gets hung up trying to fix things. My best bet seems to be to restart from scratch. How do I get dpkg / apt / aptitude to clean my machine totally, or what files should I remove to make all this stuff go away? Or, would it really be quicker to re-init my partitions and start again from the CD? Hard to answer without you pasting some example of your horkage. -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Life is too short to tolerate crappy software! begin:vcard fn:David A. Cobb n:Cobb;David A. adr:;;7 Lenox Av #1;West Warwick;RI;02893-3918;USA email;internet:[EMAIL PROTECTED] title:Independent Software Consultant note:PGP Key ID#0x4C293929 effective 01/28/2004 x-mozilla-html:TRUE version:2.1 end:vcard
Re: Newbie problems galore
Ryan Waye wrote: On Thu, 15 Jul 2004 12:19:03 -0400, David A. Cobb <[EMAIL PROTECTED]> wrote: Hi! I purchased a CD set of "Debian 3.0 "Woody" Official" and started to install it. First woe: my hardware is by nVidia; the Official kernel on the CD is 2.4.18, nVidia provides a driver for 2.4.20/21 and the Debian archive has nVidia patches for 2.4.26. My plan is to go to .26, but in the meanwhile I can't use X Windows and I can't access my NIC, so all that stuff requires rebooting into Windows. So, my first question: given the kernel images on $MIRROR/Debian/main/k/kernel-image-2.4.26-i386, how do I change my kernel? There are literally tons of guides on this. Just google it and you will get all the info you need More than I need. That's the problem. And Googling Debian.org for "nVidia" what I find is a thousand reasons to wish I had either a newer kernel or a different card-set. I need to pass things back-and-forth between Linux and Windoze. I see references to VFAT FS on the web site, but for the life of me, I can't find a trace of the software. It's really bad to have to play games with tar at both sides of the route in order not to munge up the "magic" pathnames. PLEASE don't tell me that the evil beast of Redmond has buried VFAT under a patent claim!! If not, please, where can I find it? I am pretty sure linux can read the default FAT32 and NTFS partitions without any trouble, but I may be wrong. Half and half. VFAT is built into the kernel and accesses FAT32; NTFS is there but is unhappy with my partitions. Perhaps because of a version difference (NTFS 2.1 in WinXP - mine, vs NTFS 2.0 in WinNT 4, Win98). Anyway, what I read on line makes me nervous about allowing my NTFS partitions to be mounted "R/W" from Linux. Right now, I've managed to hork up my package data so dpkg gets hung up trying to fix things. My best bet seems to be to restart from scratch. How do I get dpkg / apt / aptitude to clean my machine totally, or what files should I remove to make all this stuff go away? Or, would it really be quicker to re-init my partitions and start again from the CD? No! Don't start over, anything can be fixed. Try these commands in this order:(As superuser) apt-get clean apt-get check apt-get dist-upgrade post the results of these commands here. Ryan Waye The result of those was several screens full sailing by. However, when I used aptitude to look over the situation, the things that were iffy before are still marked and aptitude still hangs after printing "Preconfiguring modules . . .". Because of the total of these things, I'm going back to the CD's this time only. -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Life is too short to tolerate crappy software! begin:vcard fn:David A. Cobb n:Cobb;David A. adr:;;7 Lenox Av #1;West Warwick;RI;02893-3918;USA email;internet:[EMAIL PROTECTED] title:Independent Software Consultant note:PGP Key ID#0x4C293929 effective 01/28/2004 x-mozilla-html:TRUE version:2.1 end:vcard
Re: Newbie problems galore
Kent West wrote: [EMAIL PROTECTED] wrote: I still haven't figured out how to make the Linux data visible from within Windoze, other than scribbling files from Linux onto one of the VFAT-mounted drives. And you won't, unless you resort to third-party products (I've heard of such utilities, but have never used them); Microsoft doesn't seem to think there's any reason to be able to read any other type of partition other than Microsoft-branded flavors. I thought I saw someone in the free world working on an ext3 driver for windows (installable fs, I assume). But my email bounced. Anyway, it's not very mature - obviously. -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Life is too short to tolerate crappy software! begin:vcard fn:David A. Cobb n:Cobb;David A. adr:;;7 Lenox Av #1;West Warwick;RI;02893-3918;USA email;internet:[EMAIL PROTECTED] title:Independent Software Consultant note:PGP Key ID#0x4C293929 effective 01/28/2004 x-mozilla-html:TRUE version:2.1 end:vcard
Re: Newbie problems galore
Not a very successful day!!! Trying to go with the fact that TUN driver sees my network controller. Two occurrences of "Warning" Failure trying to run chroot /target dpkg --force-depends \ /var/cache/apt/archives/base-files_3.0.2_i386.deb \ (same basepath)/base-passwd_3.4.1_i386.deb and several more where all I get is "debootstrap exited with an error (return code 1)" I was getting the debootstrap errors before, seeming to be related to the way I mounted my partitions. Also, eliminating drivers may have helped in the past. When there is no error info, it's hard to guess what correction was a good idea and what wasn't. Doesn't the Base-System create process leave a log behind anywhere??? David A. Cobb wrote: Hi! I purchased a CD set of "Debian 3.0 "Woody" Official" and started to install it. First woe: my hardware is by nVidia; the Official kernel on the CD is 2.4.18, nVidia provides a driver for 2.4.20/21 and the Debian archive has nVidia patches for 2.4.26. My plan is to go to .26, but in the meanwhile I can't use X Windows and I can't access my NIC, so all that stuff requires rebooting into Windows. So, my first question: given the kernel images on $MIRROR/Debian/main/k/kernel-image-2.4.26-i386, how do I change my kernel? I need to pass things back-and-forth between Linux and Windoze. I see references to VFAT FS on the web site, but for the life of me, I can't find a trace of the software. It's really bad to have to play games with tar at both sides of the route in order not to munge up the "magic" pathnames. PLEASE don't tell me that the evil beast of Redmond has buried VFAT under a patent claim!! If not, please, where can I find it? Right now, I've managed to hork up my package data so dpkg gets hung up trying to fix things. My best bet seems to be to restart from scratch. How do I get dpkg / apt / aptitude to clean my machine totally, or what files should I remove to make all this stuff go away? Or, would it really be quicker to re-init my partitions and start again from the CD? Thanks, all, for your patience. I hope I see some answers before I do something more silly than what I've already done here. -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Life is too short to tolerate crappy software! begin:vcard fn:David A. Cobb n:Cobb;David A. adr:;;7 Lenox Av #1;West Warwick;RI;02893-3918;USA email;internet:[EMAIL PROTECTED] title:Independent Software Consultant note:PGP Key ID#0x4C293929 effective 01/28/2004 x-mozilla-html:TRUE version:2.1 end:vcard
"Warning" failure in chroot . . .
REPLY INLINE Mark Pictor wrote: Well I guess I'm gonna reply to two people at once... --- "David A. Cobb" <[EMAIL PROTECTED]> wrote: <>Not a very successful day!!! Trying to go with the fact that TUN driver sees my network controller. Two occurrences of "Warning" Failure trying to run <>Warning *doesn't* mean failure... and having to use <><>--force is a bad sign. Are you sure you're doing it <>the right way? (I can't recommend anything, not sure <>what you are doing.) <> Which makes two of us. Anyway, I'm just quoting what's in the message that pops up. The whole thing is from the installer. And it is a Failure because immediately DEBOOTSTRAP exits with Error Return-Code 1. I see I wasn't very clear: the DEBOOTSTRAP error is constant. Twice it was preceded by the "Warning" Failure . . . message. chroot /target dpkg --force-depends \ /var/cache/apt/archives/base-files_3.0.2_i386.deb \ (same basepath)/base-passwd_3.4.1_i386.deb and several more where all I get is "debootstrap exited with an error (return code 1)" I was getting the debootstrap errors before, seeming to be related to the way I mounted my partitions. ??? Once you've got things set up, you should use /etc/fstab - try 'nano /etc/fstab' as root, to edit it. for one-time mounts, use 'mount -t vfat /dev/hda1 /mnt/windows' where hda1 is a partition [1], and /mnt/windows is an empty folder which you have permissions for. You must be root to use mount this way. Ur, Yeah. I'm doing the install from CD's. I start out by Initializing and mounting partitions into which the new system is to be installed. In my first attempt, it seemed some bad selections on what to mount where led to problems -- at least the problems went away when I stopped trying to be so damn clever. The only thing I'm doing that isn't a simple agreement with the script suggestions is mounting a 2Gb partition for /tmp, instead of leaving it in the rather undersized primary (/) partition. Undersized is relative -- about 1Gb. I have 12Gb for /var and 8Gb for /usr. See comments above, and [1] below. To make sure your kernel can use it, (-bf24 almost certainly can) type 'cat /proc/filesystems' and read the output... one of the lines should be vfat, which is fat32/long file names. A-Ha! Yes, they mount fine. As I say somewhere above, some early failures led me to the erroneous conclusion that VFAT wasn't functional. <>PLEASE don't tell me that the evil beast of Redmond has buried VFAT under a patent claim!! They are trying!!! Not surprised. Microsoft delenda est! It might be... if you have a spare hard disk laying around, I would do an install on it, and work out the kinks/practice... If you haven't done much personalization or got much data on the debian partition, I would just erase it and try again. Practice, that I'm getting. It's good for my soul to learn to control my frustration. Or something like that. Sure, we all have a spare HD laying around. [1] partitions - use something like cfdisk to find out which number corresponds to which partition; alternately just go through the numbers (hda1 thru hda9) until you don't get an error that it isn't vfat, then do 'ls /mnt/windows' and see if everything looks right. Actually easier than that: there's a menu choice in the installer for "VIEW PARTITION TABLE" If I could get it into a text file I could post the results here. I can probably do a df to get it, using the ash shell on the CD. -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Life is too short to tolerate crappy software! begin:vcard fn:David A. Cobb n:Cobb;David A. adr:;;7 Lenox Av #1;West Warwick;RI;02893-3918;USA email;internet:[EMAIL PROTECTED] title:Independent Software Consultant note:PGP Key ID#0x4C293929 effective 01/28/2004 x-mozilla-html:TRUE version:2.1 end:vcard
Re: "Warning" failure in chroot . . .
Hmm. Possible, and maybe I can experiment with the idea a bit. How much space do I need for an installation that all goes into one place? My immediate problem would be that that space is not currently available on hda - the normal boot device. I was also worried about the 1024-Cylinder problem, although I guess that's now a non-issue. If you have any knowledge of the internals of the installer process, could there also be problems because of trying to set permissions on something in /target/var/cache/... ? That's a vfat partition at present and would not fully support file permissions. I don't know to what extent the driver "fakes" it. My ability to experiment with that is rather more limited but I'll see. Mark Pictor wrote: Sorry it took so long to reply. I don't know if this has anything to do with your problems, but I have never used separate partitions. (Unless, of course, it was a windows partition!) Less install headache, and you don't run out of space on one partition while wasting space elsewhere... -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Life is too short to tolerate crappy software! begin:vcard fn:David A. Cobb n:Cobb;David A. adr:;;7 Lenox Av #1;West Warwick;RI;02893-3918;USA email;internet:[EMAIL PROTECTED] title:Independent Software Consultant note:PGP Key ID#0x4C293929 effective 01/28/2004 x-mozilla-html:TRUE version:2.1 end:vcard
Using APT with a local mirror
[With apology that this duplicates a mail to 'backports'] I see in the SOURCES.LIST document that one can use the "file:" scheme URI to designate a source. I am in a position where I need to do just that. I cannot get my nVidia cardset to work, so I cannot reach the 'net from Linux; I need to use Windoze for that. I set up a FAT32/VFAT area to cache all my downloads -- it looks like a huge but fragmented RSYNC of pieces of the net. I duplicate the directory structure of my source. I set it up in what looked correct, but APTITUDE complains that it cannot stat. the package list files. I assume that is the "Packages.gz" or "Sources.gz" file. Also, the way the path looks in the error message is seriously munged from what it is on disk: looks like "/" replaced by "." [IIRC] and a few other translations. My hypothesis is that I haven't formatted the file:/top/next/third/ . . . URI correctly, or that I have pointed it to the wrong place in the tree. Of course, a formally correct URI would need to begin "file:///top/next . . .". Can someone send me a copy of a correctly formatted file URI from SOURCES.LIST? And, as a P.S., do the debian and backport servers / mirrors support RSYNC? It would surely save me a lot of manual stuff. -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Life is too short to tolerate crappy software! begin:vcard fn:David A. Cobb n:Cobb;David A. adr:;;7 Lenox Av #1;West Warwick;RI;02893-3918;USA email;internet:[EMAIL PROTECTED] title:Independent Software Consultant note:PGP Key ID#0x4C293929 effective 01/28/2004 x-mozilla-html:TRUE version:2.1 end:vcard
Cannot get this virtual package right!
I currently am running kernel-2.6.7-1 and I've slowly brought many key components up to the bleeding edge. I have apt at 0.6.25 I keep getting (aptitude, and others) depends on "libapt-pkg-libc6.3-5-3.3; however: Package libapt-pkg-libc6.3-5-3.3 is not installed." I googled the site, and find that this is a virtual package and that it caused a batch of problems a while back; also that it should be provided by apt-utils. I did not get any errors listed when I installed from . Please, what do I need to do to get this library to show up?? There is, evidently, a shared-object library by the same name that I should have but do not. -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Life is too short to tolerate crappy software! begin:vcard fn:David A. Cobb n:Cobb;David A. adr:;;7 Lenox Av #1;West Warwick;RI;02893-3918;USA email;internet:[EMAIL PROTECTED] title:Independent Software Consultant note:PGP Key ID#0x4C293929 effective 01/28/2004 x-mozilla-html:TRUE version:2.1 end:vcard
[Linux newbie asks:] Why "dhelp_parse: no title found for directory "
On several of my recent installs, (I'm using dpkg --install -R because I still haven't got my Linux talking to my network controller ) I get a boatload of "Unpacking replacement libapt-pkg-dev ... Preparing to replace libapt-pkg-doc 0.6.25 (using .../libapt-pkg-doc_0.6.25_all.deb) ... dhelp_parse: no title found for directory Apps/Programming dhelp_parse: no title found for directory doc dhelp_parse: no title found for directory libs dhelp_parse: no title found for directory music dhelp_parse: no title found for directory windowmanagers dhelp_parse: no title found for directory Apps " So, what am I missing here? -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Life is too short to tolerate crappy software! begin:vcard fn:David A. Cobb n:Cobb;David A. adr:;;7 Lenox Av #1;West Warwick;RI;02893-3918;USA email;internet:[EMAIL PROTECTED] title:Independent Software Consultant note:PGP Key ID#0x4C293929 effective 01/28/2004 x-mozilla-html:TRUE version:2.1 end:vcard
Re: Cannot get this virtual package right!
REPLIES INLINE Travis Crump wrote: David A. Cobb wrote: I currently am running kernel-2.6.7-1 and I've slowly brought many key components up to the bleeding edge. I have apt at 0.6.25 I keep getting (aptitude, and others) depends on "libapt-pkg-libc6.3-5-3.3; however: Package libapt-pkg-libc6.3-5-3.3 is not installed." it is provided by apt in testing/unstable, if you want to use experimental's apt you will have to rebuild aptitude et al against libapt-pkg-dev from experimental. Ah. OK, for the moment, I'll drop back to apt 0.5.whatever. By rebuild, do you mean configure & make from sources? Or do I just need more of the latest packages? In any case, thanks. It does feel better when I stop banging my head against the same wall. -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Life is too short to tolerate crappy software! begin:vcard fn:David A. Cobb n:Cobb;David A. adr:;;7 Lenox Av #1;West Warwick;RI;02893-3918;USA email;internet:[EMAIL PROTECTED] title:Independent Software Consultant note:PGP Key ID#0x4C293929 effective 01/28/2004 x-mozilla-html:TRUE version:2.1 end:vcard
Problems with .deb signatures
Well ! After 3 months of tweaking this then tweaking that I suddenly
got the right combination and my internet connection is now working from
Debian. I did several updates, couldn't get things quite stable.
That's stable only in the sense that aptitude actually completes its
list of things to do. There are always things that don't get
done and so are still highlighted for update.
Commented out SOURCES.LIST references to 'sid' ('unstable') & tried
--dist-upgrade.
Now every run of APT-GET --INSTALL gets stopped by this --
apt-listchanges: Mailing superbiskit: apt-listchanges: changelogs for
Cobb028933918S1Tux
apt-listchanges: Mailing superbiskit: apt-listchanges: news for
Cobb028933918S1Tux
Preconfiguring packages ...
grep: /usr/share/iso-codes/iso_639.tab: No such file or directory
grep: /usr/share/iso-codes/iso_639.tab: No such file or directory
Fetched 405MB in 19m46s (341kB/s)
Authenticating /var/cache/apt/archives/apache-utils_1.3.31-6_i386.deb ...
debsig: Origin Signature check failed. This deb might not be signed.
dpkg: error processing
/var/cache/apt/archives/apache-utils_1.3.31-6_i386.deb (--unpack):
Verification on package
/var/cache/apt/archives/apache-utils_1.3.31-6_i386.deb failed!
Authenticating /var/cache/apt/archives/apache-common_1.3.31-6_i386.deb ...
debsig: Origin Signature check failed. This deb might not be signed.
dpkg: error processing
/var/cache/apt/archives/apache-common_1.3.31-6_i386.deb (--unpack):
Verification on package
/var/cache/apt/archives/apache-common_1.3.31-6_i386.deb failed!
Authenticating /var/cache/apt/archives/debwrap_0.9.2_i386.deb ...
debsig: Origin Signature check failed. This deb might not be signed.
[ more of the above until the error quota stops things ]
Has anyone an idea of what might be happening?
TIA
--
David A. Cobb, Software Engineer, Public Access Advocate
"By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a
Pilgrim: R.French, Tr.
Life is too short to tolerate crappy software!
begin:vcard
fn:David A. Cobb
n:Cobb;David A.
adr:;;7 Lenox Av #1;West Warwick;RI;02893-3918;USA
email;internet:[EMAIL PROTECTED]
title:Independent Software Consultant
note:PGP Key ID#0x4C293929 effective 01/28/2004
x-mozilla-html:TRUE
version:2.1
end:vcard
Problems with apt
After using it successfully several times a week, every run of apt-get or aptitude now ends with: E: Dynamic MMap ran out of room E: Error occurred while processing python2.3 (NewVersion2) E: Problem with MergeList /var/lib/apt/lists/mail.linuxvar.it_%7egianluca_athlon-xp_dists_sarge_main_binary-i386_Packages E: The package lists or status file could not be parsed or opened. The specific /var/lib/apt/lists file that show up as MergeList varies depending on how many active sources I hav in sources.list I'm guessing this is more likely a configuration problem than an apt bug. Can anyone give me a clue? TIA -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Free at last! Free at last! Using Linux, I'm FREE at last! Life is too short to tolerate crappy software! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Problems with apt
Hugo Vanwoerkom wrote: > David A. Cobb wrote: > >> After using it successfully several times a week, every run of apt-get >> or aptitude now ends with: >> >> E: Dynamic MMap ran out of room >> E: Error occurred while processing python2.3 (NewVersion2) >> E: Problem with MergeList >> /var/lib/apt/lists/mail.linuxvar.it_%7egianluca_athlon-xp_dists_sarge_main_binary-i386_Packages >> >> E: The package lists or status file could not be parsed or opened. >> >> The specific /var/lib/apt/lists file that show up as MergeList varies >> depending on how many active >> sources I hav in sources.list >> >> I'm guessing this is more likely a configuration problem than an apt >> bug. >> Can anyone give me a clue? >> >> TIA >> > > http://groups-beta.google.com/group/linux.debian.devel/browse_thread/thread/4ea8ee8629bee9ab/b3a5a18d4bb161ca?lnk=st&q=Dynamic+MMap&rnum=3&hl=ia#b3a5a18d4bb161ca > > > Yes, that did the trick. The particular Cache-Limit number is obviously meaningful -- I don't lack for memory so I also tried doubling the number given in the article -- that also crashed. However, to make the entire process succeed, I also needed to reduce the number of sources in my sources.list below what I like. Q: Is this an active bug? -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Free at last! Free at last! Using Linux, I'm FREE at last! Life is too short to tolerate crappy software! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Broken /var filesystem
I found a huge problem with my /var filesystem. Or, at least, fsck was going to take all day and more fixing 11 inode block numbers each pass. It's supposed to be possible to clean /var, or at least FHS suggests so. And it was messing me up horribly anyway. So, I re-initialized /var. One of the things that vanished, of course, is dpkg/apt's memory of what is installed. Right now, dpkg doesn't think dpkg is installed! Everything runs, until it needs to refer to the database. I googled around, and found some similar reports. The proposed cure involved installing 'mawk' and 'dpkg' with '--force-depends.' Then I should be able to re-install libc6, on which nearly everything depends. However, before I can get to that point, I get " unable to create updated files list file for package mawk: No such file or directory". So, it looks as though I need to manually re-create at least one more directory. I checked the "contents.gz" files, but they only mention files that are actually present in the tarball, not those created by the installation scripts. I think, if I can get past this, I can get dpkg to re-install itself and that will cure whatever other things are destroyed. Can anyone identify what directories and files dpkg requires to exist? TIA -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Free at last! Free at last! Using Linux, I'm FREE at last! Life is too short to tolerate crappy software! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Broken /var filesystem
Joey Hess wrote: David A. Cobb wrote: I found a huge problem with my /var filesystem. Or, at least, fsck was going to take all day and more fixing 11 inode block numbers each pass. It's supposed to be possible to clean /var, or at least FHS suggests so. Not really, it only says you can delete data from /var/cache. So, I re-initialized /var. Welcome to a nearly endless world of pain. Where are your backups, BTW? Ah, yes. Backups. Waiting 'til I can afford a RW+DVD or some other suitably high-capacity device to make them onto. :-! Can anyone identify what directories and files dpkg requires to exist? /var/lib/dpkg/status (empty) If that doesn't work, I figure by tonight I will go back to my Woody-CD's and build up again from there. -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Free at last! Free at last! Using Linux, I'm FREE at last! Life is too short to tolerate crappy software! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Broken /var filesystem
Joey Hess wrote: David A. Cobb wrote: I found a huge problem with my /var filesystem. Or, at least, fsck was going to take all day and more fixing 11 inode block numbers each pass. It's supposed to be possible to clean /var, or at least FHS suggests so. Not really, it only says you can delete data from /var/cache. So, I re-initialized /var. Welcome to a nearly endless world of pain. Where are your backups, BTW? Can anyone identify what directories and files dpkg requires to exist? /var/lib/dpkg/status (empty) Didn't seem to help. Scr** it! Back to the CD's -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Free at last! Free at last! Using Linux, I'm FREE at last! Life is too short to tolerate crappy software! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: setting up clamav-freshclam on debian-user mailing list
Chris Isomaki wrote: Did you ever resolve this? I'm running into the same problem, and there is no solution posted to the list. Sorry to bother you. Thanks in advance. I can't really say I resolved it. But it seems to have gone away with the most recent ClamAV version(s). I'm showing ClamAV 0.85.1-2; and I do not get any failures showing during apt install or boot up. -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Free at last! Free at last! Using Linux, I'm FREE at last! Life is too short to tolerate crappy software! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Two questions about building kernels
I'm building 2.6.12 kernels for my own machine. Two things are rather problematical: The content of .config looks like a Makefile snippet. If it is one, can I add to it the values I want for CC & CFLAGS so they are tied to a particular build configuration when I need to re-do it? and The "official" kernel images are all built with initrd. My build complains about the size (number of blocks) in the initrd fs. Where is the parameter to create a smaller number of larger "disks" in the initrd fs? Aw, heck! I might as well add the third -- Why my non-initrd kernel won't boot? It panics with Cannot open root device on (3,01), please specify a correct root. My lilo.conf indeed says root=/dev/hda1 which is correct. -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Free at last! Free at last! Using Linux, I'm FREE at last! Life is too short to tolerate crappy software! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Concerning APPEND-TO-VERSION in make-kpkg
I have been trying various configurations building LINUX-SOURCE-2.6.12 using make-kpkg. To minimize avoidable errors, I run the whole make from a bash script. If I use --append-to-version "x6+p0c40" for example, the first build is fine. However, the appended codes get written into '.config' If I then re-run the script (having corrected something), the make-kpkg apparently concatenates the value in '.config' with my value and gives me a horror "2.6.12x6+p0c40x6+p0c40" -- which is in any case too long for LILO to accomodate (it is also incorrect). If, on the other hand, I manually enter my code in '.config' and do NOT use the --append-to-version option, make-kpkg goes through the make process but then declares: usr/bin/make -f /usr/share/kernel-package/rules real_stamp_image make[1]: Entering directory `/usr/src/linux-source-2.6.12_2.6.12-6' The UTS Release version in include/linux/version.h 2.6.12x6+p0c40 does not match current version 2.6.12, reconfiguring. It apparently proceeds to REMOVE the appended "brand" from the version code. At least it isn't in the .deb file name any more. Do I have to forcably get rid of the header line in '.config' where the LocalVersion appears? I am already removing 'include/linux/version.h' (and debian/official) before building. Are there other places where I need to remove remnants of the code? -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Free at last! Free at last! Using Linux, I'm FREE at last! Life is too short to tolerate crappy software! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Concerning make-kpkg --append-to-version
I have been trying various configurations building LINUX-SOURCE-2.6.12 using make-kpkg. To minimize avoidable errors, I run the whole make from a bash script. If I use --append-to-version "x6+p0c40" for example, the first build is fine. However, the appended codes get written into '.config' If I then re-run the script (having corrected something), the make-kpkg apparently concatenates the value in '.config' with my value and gives me a horror "2.6.12x6+p0c40x6+p0c40" -- which is in any case too long for LILO to accomodate (it is also incorrect). If, on the other hand, I manually enter my code in '.config' and do NOT use the --append-to-version option, make-kpkg goes through the make process but then declares: usr/bin/make -f /usr/share/kernel-package/rules real_stamp_image make[1]: Entering directory `/usr/src/linux-source-2.6.12_2.6.12-6' The UTS Release version in include/linux/version.h 2.6.12x6+p0c40 does not match current version 2.6.12, reconfiguring. It apparently proceeds to REMOVE the appended "brand" from the version code. At least it isn't in the .deb file name any more. Do I have to forcably get rid of the header line in '.config' where the LocalVersion appears? I am already removing 'include/linux/version.h' (and debian/official) before building. Are there other places where I need to remove remnants of the code? -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Free at last! Free at last! Using Linux, I'm FREE at last! Life is too short to tolerate crappy software! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
cvspserver and frequent "inetd address already in use"
Every few minutes, my syslog gets a message such as: "Sep 15 16:07:08 Cobb028933918A-Tux inetd[8465]: cvspserver/tcp: bind: Address already in use" I googled around. Such a thing was reported back in 2000. At the time, the answer was "You have two copies of inetd running. However: [20:30:56] # ps -A -f | grep 'netd$' - root 5934 1 0 18:22 ?00:00:00 /usr/sbin/inetd I don't. I looked through /etc/services and I don't see any port# appearing more than once. I suppose, if the error message identified the address involved in the conflict it might help, but other than that, can anyone give me a clue? -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Free at last! Free at last! Using Linux, I'm FREE at last! Life is too short to tolerate crappy software! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Ubuntu's unstable vs Debian unstable
I haven't read the whole thread, so please pardon me if I repeat something. One area that causes many conflicts has Ubuntu "ahead" of Debian. Ubuntu is stabilized on Python 2.4, whereas Debian packages all demand <2.4 (Python 2.3 is "official"). The problem is exacerbated because 'apt' interprets the dependency on what seems a "broken" manner. "Depends: Python < 2.4" would seem to be satisfied by the presence of Python-2.3. However, apt is interpreting this as "Conflicts: Python >= 2.4". (I'll try to say more on that subject elsewhere and later). Anyway, it's very difficult to pick and choose on the basis of who is more or less stable (speaking, of course, only of the software). It's bound to vary from week to week. Tshepang Lekhonkhobe wrote: Hello, >From what I read on the lists it seems that Ubuntu's unstable is generally more broken than Debian's, making me feel safer using Sid. Could anyone confirm this. Thanks... -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Free at last! Free at last! Using Linux, I'm FREE at last! Life is too short to tolerate crappy software!
Understanding "Depends: Package <= Version"
I don't know whether this is the right place to ask this. Someone, perhaps, will "straighten me out." It is not usually a problem to have both Python-2.3 and Python-2.4 on the same machine. In fact, when Python gets pulled down as a dependency, it's pretty likely to happen. However, a number of Debian packages, notably OpenOffice.org Python-Uno, have a dependency "Depends: Python <= 2.4". Now, if I parse that according to the usual human language rules, it says I must have an earlier Python; and it would be satisfied because I have both 2.3 and 2.4. But 'apt' evidently interprets this as though it said: "Conflicts: Python >= 2.4" and is UNSATISFIED because of the presence of my 2.4. Along with my "English language" interpretation, it would seem strange that the presence of a newer package which does not directly conflict with the older is not likely to break an application. Is this a bug in apt? If so, is it a known bug? -- David A. Cobb, Software Engineer, Public Access Advocate "By God's Grace, I am a Christian man; by my actions a great sinner." -- The Way of a Pilgrim: R.French, Tr. Free at last! Free at last! Using Linux, I'm FREE at last! Life is too short to tolerate crappy software!
