Re: [gentoo-user] [WORKAROUNDED]Nasty bugs in portage?
On Sat, 2005-09-10 at 23:03 +0200, Volker Armin Hemmann wrote: On Saturday 10 September 2005 20:09, Frank Schafer wrote: ... or which distribution to install during less than 4 days? Hi list, as I wrote yesterday I planned to complete installation after work (started ``emerge --emptytree system'' in the morning). where did you get the idea that --emptytree is needed or even a wise decision? --emptytree is almost NEVER needed and since it is a troublesom procedure, it should not be made, until you are totally sure, that you need it. emerge system is all you need to do, to get the base system. After that, emerge what you like to have, but NEVER use --emptytree, except when you are able to deal with the consequences. Obviously you are not, so do not do it. NO was that clear enough? Yes, that WAS clear enough. So we have an even more nasty documentation bug in the gentoo installation handbook (it tells to ``emerge --emptytree system'') and a nasty portage (or was it bootstrap?) bug (it tells me to be aware, that we need to add -e to emerge system now after finishing. For your gcc-problem, there is the fix script, others mentioned - but a lot of times all that is needed is to run gcc-config to set the correct gcc. gcc-config ran after the installation. (see in one fo my further posts, there was the Native compiler changed to ... amongst the messages. I've found (with some help of b.g.o.) that these ARE bugs. Here's the whole success story. scripts/bootstrap.sh #to work around the autoconf dependency bug emerge --oneshot --nodeps autoconf emerge --oneshot --nodeps aotoconf-wrapper #to work around the python-fchksum dependency bug emerge --oneshot --nodeps python emerge --emptytree system No fix_libtool_files.sh necessary. Why? There never was an older version of gcc. Some words to the autoconf ... will start a new thread. So much to the ability. Frank PS: You're a German? Probably this is the reason for your arrogance. -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] ntpd timezone
On Sun, 2005-09-11 at 23:36 -0700, Drew Tomlinson wrote: On 9/11/2005 9:34 PM Mark Shields wrote: What does ls -ln /etc/localtime return? [EMAIL PROTECTED] ~ $ ls -ln /etc/localtime lrwxrwxrwx 1 0 0 27 May 8 11:07 /etc/localtime - /usr/share/zoneinfo/PST8PDT Thanks for your reply! and what is the result of the date command? -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Build Problem
On Sun, 2005-09-11 at 15:37 +0200, Greg Armer wrote: Greetings list, Could anyone explain why 40% of all my emerges fail with this error message ? gcc-config error: Could not run/locate i386-pc-linux-gnu-gcc Upon executing this file directly, it returns the exact same message: fyre ~ # /usr/bin/i386-pc-linux-gnu-gcc gcc-config error: Could not run/locate i386-pc-linux-gnu-gcc fyre ~ # Here is my make.conf file incase this helps: fyre ~ # cat /etc/make.conf CFLAGS=-O2 -march=pentium4 -fomit-frame-pointer CHOST=i686-pc-linux-gnu CXXFLAGS=${CFLAGS} USE=gtk gtk2 gnome -qt -kde dvd alsa cdr I have a feeling it has something to do with my CHOST flag, I wonder if deleting /usr/bin/i386-pc-linux-gnu-gcc and symlinking /usr/bin/i686-pc-linux-gnu-gcc to that filename would work ? Thanks for the assistance. WEll, THIS could be solved using ``fix_libtool_files.sh''. Give it a shot! Regards Frank -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] after emerge -Du world date can't keep time
On Sun, 2005-09-11 at 20:41 -0700, maxim wexler wrote: Also, the BIOS clk is correct. How do you know that? what does hwclock tell you? -mw hum, run rc-status boot is clock started? check out /etc/conf.d/clock yup __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com -- Nick Rout [EMAIL PROTECTED] -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] ntpd timezone
On 9/11/2005 11:41 PM Nick Rout wrote: On Sun, 2005-09-11 at 23:36 -0700, Drew Tomlinson wrote: On 9/11/2005 9:34 PM Mark Shields wrote: What does ls -ln /etc/localtime return? [EMAIL PROTECTED] ~ $ ls -ln /etc/localtime lrwxrwxrwx 1 0 0 27 May 8 11:07 /etc/localtime - /usr/share/zoneinfo/PST8PDT Thanks for your reply! and what is the result of the date command? Right now it's correct: tv mythtv # date Sun Sep 11 23:52:49 PDT 2005 However after some time it will return to being 8 hours earlier. I also saw in another thread that you could show your hardware clock by this command: tv mythtv # hwclock Sun Sep 11 23:54:54 2005 -0.081182 seconds Don't know if that helps or not. Thanks for your reply! Drew -- Visit The Alchemist's Warehouse Magic Tricks, DVDs, Videos, Books, More! http://www.alchemistswarehouse.com -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Nasty bugs in portage?
On Sun, 2005-09-11 at 18:38 +0200, Volker Armin Hemmann wrote: On Sunday 11 September 2005 08:33, Nick Rout wrote: On Sat, 2005-09-10 at 23:03 +0200, Volker Armin Hemmann wrote: On Saturday 10 September 2005 20:09, Frank Schafer wrote: ... or which distribution to install during less than 4 days? Hi list, as I wrote yesterday I planned to complete installation after work (started ``emerge --emptytree system'' in the morning). where did you get the idea that --emptytree is needed or even a wise decision? --emptytree is almost NEVER needed and since it is a troublesom procedure, it should not be made, until you are totally sure, that you need it. I imagine he read the install instructions, which are pretty clear about doing an emerge --emptytree system What makes you think this is wrong? http://www.gentoo.org/doc/en/handbook/handbook-x86.xml?part=1chap=6#doc_ch ap2 (bottom of the page) WRONG it is there to show you, what emerge system wants to install. there is nothing about doing it! (Check again, see the -p) And in the 2005.1 handbook is no (!) --emptytree. LOL, LOL, LOL!! this is cut'n-pasted from the Gentoo Handbook 6d JUST NOW snip--- Building the System To start building the system, execute emerge --emptytree system. Then go do something to keep your mind busy, because this step takes a long time to complete. Code Listing 22: Building the System # emerge --emptytree system -snap LOL, LOL, LOL!!! You MUST be a German! Not able to read and arrogant (talking about ability of others)! 0,02$ Frank PS: and yes, I DID build from stage1 and DID change CFLAGS. -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Nasty bugs in portage?
On Sun, 2005-09-11 at 19:12 +0200, Volker Armin Hemmann wrote: On Sunday 11 September 2005 18:58, Mark Shields wrote: From http://www.gentoo.org/doc/en/handbook/handbook-x86.xml?part=1chap=6: // start quote Building the System To start building the system, execute emerge --emptytree system. Then go do something to keep your mind busy, because this step takes a long time to complete. Code Listing 22: Building the System # emerge --emptytree system Again, if you haven't touched the default CFLAGS and CXXFLAGS setting, using --newuse is sufficient. // end quote So you see, it does tell you to do an emerge --emptytree system, unless you haven't changed the defalt CFLAGS/CXXFLAGS, in which case you can just use the --newuse in place of --emptytree. and this one: http://www.gentoo.org/doc/en/handbook/2005.1/handbook-x86.xml?full=1 does not talk about --emptytree at all, so which one is correct? (btw, when I installed my gentoo --emptytree was totally not needed.) Then ``emerge system'' didn't replace the packages installed by ``sbin/bootstrap.sh'' with new ones (which are probably built with different compiler settings). Frank -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Nasty bugs in portage?
On Sun, 2005-09-11 at 23:01 +0100, Neil Bothwick wrote: On Sun, 11 Sep 2005 21:16:23 +0200, Holly Bostick wrote: No, Neil, this thread (or the original issue, at least), is occurring during the initial install process: Whoops, my mistake. This comes up so often it's easy to get the threads muddled up :( So possibly we might consider using our expertise to actually help the guy, in case anyone might happen to know why he's getting this breakage during his inital installation, rather than arguing about whether he should be using --emptytree or not, Hasn't this already been covered early in the thread? Run fix_libtool.sh to fix the error then do emerge --resume to carry on. If I get it right ``fix_libtool_files.sh'' corrects the settings for libtool according to the native compiler if gcc has changed. During a native install there isn't an older version of gcc. So this should be (and was) the wrong answer. Regards and thanks for all the tries to help me Frank PS: see the [WORKAROUNDED] post -- gentoo-user@gentoo.org mailing list
[gentoo-user] setting mythtv
I'm trying to set mythtv in my machine. I installed mysql and I used all default parameters. When I launch mythsetup I recieve this message: 2005-09-12 09:09:52.350 Writing settings file /home/pinna/.mythtv/mysql.txt 2005-09-12 09:09:52.547 Unable to connect to database! 2005-09-12 09:09:52.548 Driver error was [1/1045]: QMYSQL3: Unable to connect Database error was: Access denied for user: '[EMAIL PROTECTED]' (Using password: YES) 2005-09-12 09:09:52.548 Failed to init MythContext, exiting. If I try to see the database: mysql -u mythtv mythconverg -p Enter password: ERROR 1045: Access denied for user: '[EMAIL PROTECTED]' (Using password: YES) How must I set mysql? In the Gentoo howtv that I have there is no part where mysql is configured. Has someone mythtv installed? Thanks, Luigi -- Public key GPG(0x073A0960) on http://keyserver.linux.it/ pgpbx7g0EcBf4.pgp Description: PGP signature
Re: [gentoo-user] Nasty bugs in portage?
On Mon, 12 Sep 2005 11:49:52 +1200, Nick Rout wrote: No, but they are on the GRP package CDs that accompany each release. yes, but there's no requirement to use grp packages with a stage-3. There is if you're using stage 3 as quoted above. what have you been smoking Neil? A stage 3 install just means you have a biggish tarball at the start and it installs a very basic working system ready to comile on. You still do your own kernel and you can EITHER use GRP packages OR compile them yourself. I know, but the quote you removed was about using stage 3 to save time to have a running system asap. If that's your objective, you're hardly going to follow it up by compiling KDE and X from scratch, are you? -- Neil Bothwick A wok is what you throw at a wabbit. pgp21DyNneNbI.pgp Description: PGP signature
Re: [gentoo-user] Nasty bugs in portage?
On Sun, 11 Sep 2005 19:59:14 -0500, John Jolet wrote: yes, but there's no requirement to use grp packages with a stage-3. There is if you're using stage 3 as quoted above. No, there isn't. This laptop was built with a stage 3 tarball, everything else was compiled from source. I know because I watched kde build for 16 hours. Read the quote you removed IMHO, stage 3 is for those that don't want to take the lengthy build time for some of the larger packages, i.e. X and kde/gnome, in order to have a basic working gentoo system in a short timeframe. You clearly were prepared to take the lengthy build time for KDE etc. But others don't have that luxury, like when I built my laptop, so Stage 3 + GRP does the whole job in less than 10% of the time you spent on KDE alone... not that this has anything to do with the original topic :) -- Neil Bothwick Megabyte: (n.) more than you can comprehend and less than you'll need. pgpup8zTXyRxR.pgp Description: PGP signature
Re: [gentoo-user] Nasty bugs in portage?
On Mon, 12 Sep 2005 09:11:21 +0200, Frank Schafer wrote: Hasn't this already been covered early in the thread? Run fix_libtool.sh to fix the error then do emerge --resume to carry on. If I get it right ``fix_libtool_files.sh'' corrects the settings for libtool according to the native compiler if gcc has changed. During a native install there isn't an older version of gcc. So this should be (and was) the wrong answer. Fair comment, but emerge --resume is still important. your original post seemed to assume that after fixing the problem, you'd have to run emerge --emptytree again, which is not the case. It is annoying to have to stop the process to fix something, but not so nearly annoying as if you have to start over. -- Neil Bothwick Sir! Romulan warbird decloaki»®õ÷üÁ NO CARRIER pgpza3YHQTcg4.pgp Description: PGP signature
Re: [gentoo-user] Nasty bugs in portage?
On Mon, 2005-09-12 at 08:33 +0100, Neil Bothwick wrote: If I get it right ``fix_libtool_files.sh'' corrects the settings for libtool according to the native compiler if gcc has changed. During a native install there isn't an older version of gcc. So this should be (and was) the wrong answer. Fair comment, but emerge --resume is still important. your original post seemed to assume that after fixing the problem, you'd have to run emerge --emptytree again, which is not the case. It is annoying to have to stop the process to fix something, but not so nearly annoying as if you have to start over. That's very right and we should be glad to have this switch :) I did a lot of experiments on saturday and yesterday. So I decided to make the whole procedure (starting with making the filesystems) what took about 10 minutes and then: ``scripts/nootstrap.sh touch /var/log/portage/bootstrap emerge --oneshot --nodeps autoconf emerge --oneshot --nodeps autoconf-wrapper emerge --oneshot --nodeps python touch /var/log/portage/bugworkaround emerge --emptytree system'' After firing up this command I went to sleep. If the comp works hard meanwhile I'm having sweet dreams about my pets ... this don't hurt so much ;) In the morning the system was up and ``ls -ltr /var/log/portage'' showed me that it took something about 5 hours. Thanks for your assistence Frank -- gentoo-user@gentoo.org mailing list
[gentoo-user] how to display one X app to more than one X server simultaneously
Hi, I know how to display X one app to a remote X server, but what if I want to display one app to two remote X servers simultaneously? e.g. I want to diplay oowrite on my machine(192.168.18.2) to two remote X servers (192.168.18.8 and 192.168.18.9) and my own X server. -- gentoo-user@gentoo.org mailing list
RE: [gentoo-user] how to display one X app to more than one X server simultaneously
-Original Message- From: Bobber Cheng [mailto:[EMAIL PROTECTED] I know how to display X one app to a remote X server, but what if I want to display one app to two remote X servers simultaneously? e.g. I want to diplay oowrite on my machine(192.168.18.2) to two remote X servers (192.168.18.8 and 192.168.18.9) and my own X server. Is it possible to do this? W/O resorting to usage of VNC or some other remote view app? -- gentoo-user@gentoo.org mailing list
[gentoo-user] minicom as a user but I am getting a cant create lockfile failure
I want to use minicom as a user but I am getting a cant create lockfile failure. It works fine as root. I dont wish to change the perms on /var/lock unless I have to - so whats the gentoo way to get minicom working as a user? Also what group(s) do I need for serial port access (/dev/ttyS0 so I presume tty, or is it uucp or both/either) once I get past the lockfile problems. BillK -- gentoo-user@gentoo.org mailing list
RE: [gentoo-user] Portage logs
On Mon, 2005-09-12 at 16:18 +0800, Ow Mun Heng wrote: -Original Message- From: Frank Schafer [mailto:[EMAIL PROTECTED] I'm wondering why there are these 4 digit numbers in front of the name of each log file. This makes it hard to find the log for a single package by name. I could be wrong but it could be the PID of the emerge process. Maybe ;) but I didn't want to ask where these numbers come from but to suggest to not have them in the name of the logfile. -- gentoo-user@gentoo.org mailing list
[gentoo-user] Brother Printer
Anyone here got a Brother printer? And if so what model? Email Disclaimer http://www.aplitec.co.za/emaildisclaimer.htm -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Portage logs
On Mon, 12 Sep 2005 10:02:21 +0200, Frank Schafer wrote: I'm wondering why there are these 4 digit numbers in front of the name of each log file. This makes it hard to find the log for a single package by name. Maybe some internal feature of portage? What about changing the logfile names to: package-version.log package-version.msg? I like the idea of giving the two logs different names, it would make parsing this information with a script much easier. You'd still need some sort of unique identifier in the names, because you could merge the same package version more than once. But naming the files .log and .msg would at least allow the same identifier for both files for the same merge. -- Neil Bothwick This virus requires Microsoft Windows XP pgpselxAyAMy2.pgp Description: PGP signature
Re: [gentoo-user] Portage logs
Neil Bothwick wrote: I like the idea of giving the two logs different names, it would make parsing this information with a script much easier. You'd still need some sort of unique identifier in the names, because you could merge the same package version more than once. But naming the files .log and .msg would at least allow the same identifier for both files for the same merge. try searching for enotice on gentooexperimental. It's a nice script that stores all messages for later use. -- Naga -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Portage logs
On Mon, 2005-09-12 at 09:37 +0100, Neil Bothwick wrote: On Mon, 12 Sep 2005 10:02:21 +0200, Frank Schafer wrote: I'm wondering why there are these 4 digit numbers in front of the name of each log file. This makes it hard to find the log for a single package by name. Maybe some internal feature of portage? What about changing the logfile names to: package-version.log package-version.msg? I like the idea of giving the two logs different names, it would make parsing this information with a script much easier. You'd still need some sort of unique identifier in the names, because you could merge the same package version more than once. But naming the files .log and .msg would at least allow the same identifier for both files for the same merge. I wondered about this myself meanwhile. If the rebuild is caused by an ``emerge --sync emerge -u world``, we have a new version number. If the rebuild is caused by a USE flag change, I'd go the way I do if I change configuration files: mv package.version.log package.version.log.OLD mv package.version.msg package.version.msg.OLD emerge package if [ emerge fails ]; then luckily we have the build log of the still installed package else rm *.OLD We don't need install logs of packages which aren't installed, do we? fi -- gentoo-user@gentoo.org mailing list
[gentoo-user] USB TV card
Anyone tried a USB TV card with a laptop. I saw a few quite cheap at my nearest computer shop - but naturally they are all targeted towards Windows. Anyone having experience getting such a beast up and running under Linux?Regards,Martin S
Re: [gentoo-user] Nasty bugs in portage?
Am Montag, 12. September 2005 08:56 schrieb Frank Schafer: On Sun, 2005-09-11 at 18:38 +0200, Volker Armin Hemmann wrote: And in the 2005.1 handbook is no (!) --emptytree. LOL, LOL, LOL!! this is cut'n-pasted from the Gentoo Handbook 6d JUST NOW The 2005.1 Handbook, which Volker referes to, has no Section 6d. http://www.gentoo.org/doc/en/handbook/2005.1/handbook-x86.xml?part=1chap=6 snip--- Building the System To start building the system, execute emerge --emptytree system. Then go do something to keep your mind busy, because this step takes a long time to complete. Code Listing 22: Building the System # emerge --emptytree system -snap You are reading the Gentoo Linux x86 Handbook (http://www.gentoo.org/doc/en/handbook/handbook-x86.xml?part=1chap=6), which is a different one. LOL, LOL, LOL!!! You MUST be a German! Not able to read and arrogant (talking about ability of others)! In fact YOU did not read Volkers post carefully. Now call me arrogant, cause I am german too. Even worse, I am bavarian :) 0,02$ Frank PS: and yes, I DID build from stage1 and DID change CFLAGS. Did you change the CFLAGS after bootstrapping? If not, there is no need for --emptytree. The CFLAGS you set for bootstrap ARE your default CFLAGS. I always do stage1 installs. I even do the bootstrap ~x86. I never ever used --emptytree when emerging the system. Regards, Michael -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Nasty bugs in portage?
On Mon, 2005-09-12 at 11:12 +0200, Michael Schreckenbauer wrote: Am Montag, 12. September 2005 08:56 schrieb Frank Schafer: On Sun, 2005-09-11 at 18:38 +0200, Volker Armin Hemmann wrote: And in the 2005.1 handbook is no (!) --emptytree. LOL, LOL, LOL!! this is cut'n-pasted from the Gentoo Handbook 6d JUST NOW The 2005.1 Handbook, which Volker referes to, has no Section 6d. http://www.gentoo.org/doc/en/handbook/2005.1/handbook-x86.xml?part=1chap=6 snip--- Building the System To start building the system, execute emerge --emptytree system. Then go do something to keep your mind busy, because this step takes a long time to complete. Code Listing 22: Building the System # emerge --emptytree system -snap You are reading the Gentoo Linux x86 Handbook (http://www.gentoo.org/doc/en/handbook/handbook-x86.xml?part=1chap=6), which is a different one. LOL, LOL, LOL!!! You MUST be a German! Not able to read and arrogant (talking about ability of others)! In fact YOU did not read Volkers post carefully. Now call me arrogant, cause I am german too. Even worse, I am bavarian :) 0,02$ Frank PS: and yes, I DID build from stage1 and DID change CFLAGS. Did you change the CFLAGS after bootstrapping? If not, there is no need for --emptytree. The CFLAGS you set for bootstrap ARE your default CFLAGS. I always do stage1 installs. I even do the bootstrap ~x86. I never ever used --emptytree when emerging the system. Regards, Michael B. bavarian ;) BTW: I'm a hmmm, living in ... see the email As I mentioned, I've changed nearly everything experimenting to get the install and then ... see my [WORKAROUNDED] post. Regards Frank PS: Volker simply made me a bit angry calling me unable. I'm a UNIX programmer / system administrator since 15 years. Xcuse me. -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Portage logs
On Mon, September 12, 2005 8:02 pm, Frank Schafer said: Hi list, I'm just wondering about the manner portage manages it log files. It's a very good idea to have 2 logs for each package (one with each end every line of the make output and one with the messages of the package for the installer). I'm wondering why there are these 4 digit numbers in front of the name of each log file. This makes it hard to find the log for a single package by name. it keeps them in time order try find or ls|grep What about changing the logfile names to: package-version.log package-version.msg? A simple ``ls'' then will show the logs in alphabetical order making it easier to find a single one. A ``ls -tr'' will show the logs in chronological order. To see in which order the packages were built, rebuilt ..., how long did it take to build a package and so on. What you about this? Regards Frank -- gentoo-user@gentoo.org mailing list -- gentoo-user@gentoo.org mailing list
[gentoo-user] Building Xorg - internal compiler error: segmentation fault
Hi all, so I hit this bug too. :( Well, the log itself says it's a bug and (even worse) that nobody knows where it comes from. Somebody on b.g.o. suggests to try building with a vanilly kernel. Did someone here make the same experience and if yes, will it run then on a gentoo-kernel? Thanks for any hint Frank -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] USB TV card
On Mon, September 12, 2005 9:10 pm, Martin S said: Anyone tried a USB TV card with a laptop. I saw a few quite cheap at my nearest computer shop - but naturally they are all targeted towards Windows. Anyone having experience getting such a beast up and running under Linux? go to the mythtv site. There is mention there of one that does mpeg2 and mpeg4 compression in hardware and supports linux (obviously given that it works on mythtv) Regards, Martin S -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Portage logs
On Mon, 2005-09-12 at 21:25 +1200, Nick Rout wrote: On Mon, September 12, 2005 8:02 pm, Frank Schafer said: Hi list, I'm just wondering about the manner portage manages it log files. It's a very good idea to have 2 logs for each package (one with each end every line of the make output and one with the messages of the package for the installer). I'm wondering why there are these 4 digit numbers in front of the name of each log file. This makes it hard to find the log for a single package by name. it keeps them in time order My idea is that the time is in order with ``ls -tr'' and the alphabet is in order with ``ls''. try find or ls|grep That's fine if I know the exact package name and spelling. Reading an ordered list makes this unnecessary. What about changing the logfile names to: package-version.log package-version.msg? A simple ``ls'' then will show the logs in alphabetical order making it easier to find a single one. A ``ls -tr'' will show the logs in chronological order. To see in which order the packages were built, rebuilt ..., how long did it take to build a package and so on. What you about this? Regards Frank -- gentoo-user@gentoo.org mailing list -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Building Xorg - internal compiler error: segmentation fault
Frank Schafer wrote: so I hit this bug too. :( Well, the log itself says it's a bug and (even worse) that nobody knows where it comes from. Somebody on b.g.o. suggests to try building with a vanilly kernel. Did someone here make the same experience and if yes, will it run then on a gentoo-kernel? Reproducible? Always at the same location? If it segfaults randomly, try to check for bad memory. Christoph -- echo mailto: NOSPAM !#$.'*'|sed 's. ..'|tr * !#:2 [EMAIL PROTECTED] -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Portage logs
On Mon, September 12, 2005 9:35 pm, Frank Schafer said: On Mon, 2005-09-12 at 21:25 +1200, Nick Rout wrote: On Mon, September 12, 2005 8:02 pm, Frank Schafer said: I'm wondering why there are these 4 digit numbers in front of the name of each log file. This makes it hard to find the log for a single package by name. it keeps them in time order My idea is that the time is in order with ``ls -tr'' and the alphabet is in order with ``ls''. this is not the forum for seeking to change the way gentoo works, the devs don't really inhabit this list. Try gentoo-dev :-) try find or ls|grep That's fine if I know the exact package name and spelling. Reading an ordered list makes this unnecessary. What about changing the logfile names to: package-version.log package-version.msg? I have emerged (for example) the same version of mplayer with different use flags until I got it how I wanted it, your system does not allow for this. A simple ``ls'' then will show the logs in alphabetical order making it easier to find a single one. A ``ls -tr'' will show the logs in chronological order. To see in which order the packages were built, rebuilt ..., how long did it take to build a package and so on. genlop -t tells you how long the ebuild took. as in genlop -t mplayer Just think about this: every other log file in the system is ordered by date time :-) -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Portage logs
On Mon, 2005-09-12 at 21:55 +1200, Nick Rout wrote: On Mon, September 12, 2005 9:35 pm, Frank Schafer said: On Mon, 2005-09-12 at 21:25 +1200, Nick Rout wrote: On Mon, September 12, 2005 8:02 pm, Frank Schafer said: I'm wondering why there are these 4 digit numbers in front of the name of each log file. This makes it hard to find the log for a single package by name. it keeps them in time order My idea is that the time is in order with ``ls -tr'' and the alphabet is in order with ``ls''. this is not the forum for seeking to change the way gentoo works, the devs don't really inhabit this list. Try gentoo-dev :-) thanks :) try find or ls|grep That's fine if I know the exact package name and spelling. Reading an ordered list makes this unnecessary. What about changing the logfile names to: package-version.log package-version.msg? I have emerged (for example) the same version of mplayer with different use flags until I got it how I wanted it, your system does not allow for this. See another of my posts in this thread. A simple ``ls'' then will show the logs in alphabetical order making it easier to find a single one. A ``ls -tr'' will show the logs in chronological order. To see in which order the packages were built, rebuilt ..., how long did it take to build a package and so on. genlop -t tells you how long the ebuild took. as in genlop -t mplayer I decline to install a bunch of gen* tools which simply duplicate the functionality of UNIX commands (like ls in this case) ;p Just think about this: every other log file in the system is ordered by date time :-) Right in the looking in the syslog I'm seeng events in the order they occured in one file. Here we have a couple of logs which I'd like to find quickly by name/time (due to the question I ask and UNIX has a tool for this already) which content is again chronological. -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Building Xorg - internal compiler error: segmentation fault
On Mon, 2005-09-12 at 11:44 +0200, Christoph Gysin wrote: Frank Schafer wrote: so I hit this bug too. :( Well, the log itself says it's a bug and (even worse) that nobody knows where it comes from. Somebody on b.g.o. suggests to try building with a vanilly kernel. Did someone here make the same experience and if yes, will it run then on a gentoo-kernel? Reproducible? Always at the same location? If it segfaults randomly, try to check for bad memory. Christoph -- echo mailto: NOSPAM !#$.'*'|sed 's. ..'|tr * !#:2 [EMAIL PROTECTED] This is a well known bug. The log itself says, that it isn't reproducible (means the crash is for everyone in some file and then stable at the same point). b.g.o. isn't accessible right now so I can't send you the output. For me it crashes at sis301_init.c. Well for me it is reproducible. I only need to emerge Xorg :( BTW: The comp is brand new. I built a minimal system (91 packages) My further plans: Build X without all the dri, glx, ... and if this succeeds add features until it crashes again to see which component causes this. If even building plain X fails I'll try a kernel from kernel.org, then maybe other versions of GCC. We'll see Frank -- gentoo-user@gentoo.org mailing list
[gentoo-user] glib with nptl
I added nptl to my USE variables and tried to emerge glib. checking whether to compile timeloop... yes checking if building for some Win32 platform... no checking for thread implementation... posix checking thread related cflags... -D_REENTRANT checking for pthread_create/pthread_join... no checking for pthread_create/pthread_join in -lpthread... no checking for pthread_create/pthread_join in -lpthread32... no checking for pthread_create/pthread_join in -lpthreads... no checking for pthread_create/pthread_join in -lthread... no checking for pthread_create/pthread_join in -ldce... no configure: error: I can't find the libraries for the thread implementation posix. Please choose another thread implementation or provide information on your thread implementation. You can also run 'configure --disable-threads' to compile without thread support. !!! Please attach the config.log to your bug report: !!! /var/tmp/portage/glib-2.6.5/work/glib-2.6.5/config.log !!! ERROR: dev-libs/glib-2.6.5 failed. !!! Function econf, Line 485, Exitcode 0 !!! econf failed !!! If you need support, post the topmost build error, NOT this status message. The relevant line of /var/tmp/portage/glib-2.6.5/work/glib-2.6.5/config.log: $ ./configure --prefix=/usr --host=i686-pc-linux-gnu --mandir=/usr/share/man --infodir=/usr/share/info --datadir=/usr/share --sysconfdir=/etc --localstatedir=/var/lib --build=i686-pc-linux-gnu --with-threads=posix --disable-static --enable-gtk-doc It seems that nptl is not being taken into account. Is this normal for glib? If so, what can I do to provide the missing libraries? (I know my system is somewhat broken, I'm just trying to fix it be emerging some packages I suppose might be relevant...) -- Jorge Almeida -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Portage logs
On Mon, 2005-09-12 at 22:59 +1200, Nick Rout wrote: On Mon, September 12, 2005 10:24 pm, Frank Schafer said: On Mon, 2005-09-12 at 21:55 +1200, Nick Rout wrote: On Mon, September 12, 2005 9:35 pm, Frank Schafer said: On Mon, 2005-09-12 at 21:25 +1200, Nick Rout wrote: On Mon, September 12, 2005 8:02 pm, Frank Schafer said: I have emerged (for example) the same version of mplayer with different use flags until I got it how I wanted it, your system does not allow for this. See another of my posts in this thread. A simple ``ls'' then will show the logs in alphabetical order making it easier to find a single one. A ``ls -tr'' will show the logs in chronological order. To see in which order the packages were built, rebuilt ..., how long did it take to build a package and so on. genlop -t tells you how long the ebuild took. as in genlop -t mplayer I decline to install a bunch of gen* tools which simply duplicate the functionality of UNIX commands (like ls in this case) ;p perhaps you ought to learn how to use grep or find. People write tools for you, you are free to use them or futz around doing your own thing. I see you didn't read the whole thread. The only thing I had to learn (my own futz) is ``ls'' in this case. (see 7 lines above) The system works well for many people. But this is linux, change portage on your system to log differently and then submit a patch to bugzilla. Or just submit the idea to bugzilla. Don't forget to include a way to fix the 3 or more years of log files in /var/log/portage. Don't forget to make out a really good justification because otherwise the devs will think you are just a moaner. There shouldn't be more logs than we have actually packages installed. The log for the installation of the version of (say) python I had installed 3 years ago isn't worth anything. (Me to say) A successful installation of a new version of a package can safely overwrite the old log because IMHO nobody needs the installation logs of not installed software. Just think about this: every other log file in the system is ordered by date time :-) Right in the looking in the syslog I'm seeng events in the order they occured in one file. Here we have a couple of logs which I'd like to find quickly by name/time (due to the question I ask and UNIX has a tool for this already) which content is again chronological. what the hell are you trying to find? If it is the log for package foo, simply find the file with foo in the name in /var/log/portage. ... Which is much easier in an alphabetical ordered list. Xcuse me that I'm trying to involve. F. -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] glib with nptl
On Mon, 2005-09-12 at 11:48 +0100, Jorge Almeida wrote: I added nptl to my USE variables and tried to emerge glib. checking whether to compile timeloop... yes checking if building for some Win32 platform... no checking for thread implementation... posix checking thread related cflags... -D_REENTRANT checking for pthread_create/pthread_join... no checking for pthread_create/pthread_join in -lpthread... no checking for pthread_create/pthread_join in -lpthread32... no checking for pthread_create/pthread_join in -lpthreads... no checking for pthread_create/pthread_join in -lthread... no checking for pthread_create/pthread_join in -ldce... no configure: error: I can't find the libraries for the thread implementation posix. Please choose another thread implementation or provide information on your thread implementation. You can also run 'configure --disable-threads' to compile without thread support. !!! Please attach the config.log to your bug report: !!! /var/tmp/portage/glib-2.6.5/work/glib-2.6.5/config.log !!! ERROR: dev-libs/glib-2.6.5 failed. !!! Function econf, Line 485, Exitcode 0 !!! econf failed !!! If you need support, post the topmost build error, NOT this status message. The relevant line of /var/tmp/portage/glib-2.6.5/work/glib-2.6.5/config.log: $ ./configure --prefix=/usr --host=i686-pc-linux-gnu --mandir=/usr/share/man --infodir=/usr/share/info --datadir=/usr/share --sysconfdir=/etc --localstatedir=/var/lib --build=i686-pc-linux-gnu --with-threads=posix --disable-static --enable-gtk-doc It seems that nptl is not being taken into account. Is this normal for glib? If so, what can I do to provide the missing libraries? (I know my system is somewhat broken, I'm just trying to fix it be emerging some packages I suppose might be relevant...) -- Jorge Almeida Did you an ``emerge --newuse world'' before you tried to install further packages? Just a thought. I think I've seen nptl mentioned in the installation log of glibc. What if glibc don't know about this? Frank PS: I could be wrong. Don't have a gentoo installation accessible right now. -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] glib with nptl
Am Montag, 12. September 2005 12:48 schrieb Jorge Almeida: I added nptl to my USE variables and tried to emerge glib. you must re-emerge glibc, when adding nptl to your USE. glib does not use the nptl USE-variable ;) Regards, Michael -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] glib with nptl
On Mon, 12 Sep 2005, Frank Schafer wrote: Did you an ``emerge --newuse world'' before you tried to install further packages? Just a thought. I think I've seen nptl mentioned in the installation log of glibc. What if glibc don't know about this? Frank PS: I could be wrong. Don't have a gentoo installation accessible right now. I can't do ``emerge --newuse world' because it will certainly fail (as there is something wrong with the system). But I did emerge glibc before trying to emerge glib. I think glib doesn't use nptl (judging by emerge -pv glib), but the problem is that it can't use posix threads either. Thanks. Jorge -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] glib with nptl
On Mon, 12 Sep 2005, Michael Schreckenbauer wrote: Am Montag, 12. September 2005 12:48 schrieb Jorge Almeida: I added nptl to my USE variables and tried to emerge glib. you must re-emerge glibc, when adding nptl to your USE. glib does not use the nptl USE-variable ;) Yes, I did that before trying to emerge glib. I found out after posting (by checking a sane system) that the missing libraries are provided by glibc. And /usr/lib/libpthread.so exists in my system...(as it should, since the nptlonly USE variable is _not_ set). I think this happens because my system is broken, God knows where and why. Is there some way to fix the toolchain, short of re-installing gentoo? Thanks. Jorge Almeida -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] glib with nptl
On 12 September 2005 13:59, Michael Schreckenbauer wrote: Am Montag, 12. September 2005 12:48 schrieb Jorge Almeida: I added nptl to my USE variables and tried to emerge glib. you must re-emerge glibc, when adding nptl to your USE. glib does not use the nptl USE-variable ;) Did you see that the error occurred in glibc? Uwe -- 95% of all programmers rate themselves among the top 5% of all software developers. - Linus Torvalds http://www.uwix.iway.na (last updated: 20.06.2004) -- gentoo-user@gentoo.org mailing list
[gentoo-user] out of sync audio
Folks I have just created a vcd from one of my dvd's using dvd::rip. The problem is that the audio is out of sync ( with the speakers lips). Is there something special I need to do, or is there a filter that I need to enable? TIA, -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Audio and permissions.
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Walter Dnes wrote: On Sat, Sep 10, 2005 at 11:40:50AM -0700, gentuxx wrote OK, I tried that and I have/can verify that I am in the audio group. Logged out, then back in, even did a reboot. No joy. Now here's the weird thing. mplayer works fine, for both video (sound with movies) and audio (mp3s). But none of the GUI apps (XMMS, JuK, etc.) seem to output any sound. As a matter of fact, XMMS errors out that it failed to open the audio output: ALSA 1.2.10 output plugin. And I don't get any of the blips and whirrs from the desktop interaction (minimizing windows, etc.). Any more ideas? Two questions... 1) Run the command ls -al /dev/sound/. What's the output ? [EMAIL PROTECTED] ~ $ ls -al /dev/sound total 0 drwxr-xr-x 2 root root 220 Sep 10 11:27 . drwxr-xr-x 15 root root 29720 Sep 11 03:10 .. crw-rw 1 root audio 14, 12 Sep 10 11:27 adsp crw-rw 1 root audio 14, 4 Sep 10 11:27 audio crw--- 1 root audio 14, 20 Apr 10 15:18 audio1 crw-rw 1 root audio 14, 3 Sep 10 11:27 dsp crw--- 1 root audio 14, 19 Apr 10 15:18 dsp1 crw-rw 1 root audio 14, 0 Sep 10 11:27 mixer crw--- 1 root audio 14, 16 Apr 10 15:18 mixer1 crw-rw 1 root audio 14, 1 Sep 10 11:27 sequencer crw-rw 1 root audio 14, 8 Sep 10 11:27 sequencer2 2) Are you familiar with PAM? (That will determine how we proceed to solve your problem, if it's what I think it is). Well, it depends on how familiar you mean. ;-) I know what it is. I'm relatively comfortable with configuring it, but I haven't changed anything from the default, with the exception of pam_cracklib. - -- gentux echo hfouvyAdpy/ofu | perl -pe 's/(.)/chr(ord($1)-1)/ge' gentux's gpg fingerprint == 34CE 2E97 40C7 EF6E EC40 9795 2D81 924A 6996 0993 -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.1 (GNU/Linux) iD8DBQFDJbFKLYGSSmmWCZMRAmxFAKCEe6373lN764Y8LgNu0o0Mfn0n8wCaAokA OsULKGs9qsxEWymIc2ziysA= =cad6 -END PGP SIGNATURE- -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] openLDAP with mysql backend (problems with libmyodbc.so = undefined symbol: lt_dlclose )
Claudinei Matos wrote: [...] Looking around I've found that libmyodbc3.so is provided by myodbc package which one I've emerged and setup in both odbc.ini and odbcinst.ini like the tutorial says, but now, if I try to run iodbctest that's what I get when I type DSN=ldap: 1: [iODBC][Driver Manager]/usr/lib/libmyodbc3.so: undefined symbol: lt_dlclose (0), SQLSTATE=0 2: [iODBC][Driver Manager]Specified driver could not be loaded (0), SQLSTATE=IM003 Since I got this error, I've thinked that may I had to reinstall unixODBC, iodbc, libiodbc but even after reinstall all these packages I still getting the same error. Do somebody knows how to fix that? Is there another way to use mysql as a backend to LDAP? Tks in advice, Claudinei Matos related to bugs http://bugs.gentoo.org/show_bug.cgi?id=102026 http://bugs.gentoo.org/show_bug.cgi?id=103607 that look like sys-devel/libtool related, look like the mistery still need to be solved. -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Problem with udpate
Sébastien MORAND schreef: pyrenees:~ # emerge -uNDvp world These are the packages that I would merge, in order: Calculating world dependencies ...done! [blocks B ] dev-php/php (is blocking dev-lang/php-5.0.5) [blocks B ] dev-php/php (is blocking dev-php/PEAR-PEAR-1.3.6) [blocks B ] media-libs/giflib (is blocking media-libs/libungif-4.1.3) [ebuild N] media-libs/giflib-4.1.3 +X +gif 0 kB [ebuild N] dev-php/php-4.4.0 +X +berkdb +crypt +curl -debug +doc -fdftk +firebird -flash +freetds +gd -gd-external +gdbm -gmp -hardenedphp -imap -informix +ipv6 +java +jpeg +kerberos +ldap -mcal -memlimit -mssql +mysql +ncurses +nls -oci8 +odbc +pam +png +postgres +readline -snmp +spell +ssl +tiff +truetype +xml2 -yaz 0 kB [ebuild UD] sys-devel/gcc-3.3.5.20050130-r1 [3.3.6] (-altivec) -bootstrap -boundschecking -build +fortran +gcj +gtk -hardened -ip28 (-multilib) -multislot (-n32) (-n64) +nls -nocxx -nopie -nossp +objc -static -vanilla 0 kB Total size of downloads: 0 kB pyrenees:~ # emerge -C dev-php/php media-libs/giflib --- Couldn't find dev-php/php to unmerge. --- Couldn't find media-libs/giflib to unmerge. unmerge: No packages selected for removal. You clearly know how to clear a block, but your syntax for unmerging the blocking packages is wrong, which is why it's not working. If you want to unmerge the specific version of the package (for this example, let us assume php-5.0.5), the correct syntax is emerge -C =dev-php/php-5.0.5 If you want to unmerge any or all versions of the package currently installed, letting Portage decide which to unmerge, the correct syntax is emerge -C php So what you typed makes no sense to Portage. Hope this helps, Holly -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Copying between hard drives potential newbie question ----- Disk image post
clip Can/has anyone used a 12V lead acid car battery as a UPS for their box? I've got a spare one in the garage and the thought has crossed my mind . . -- Regards, Mick Mick, It can be done but is slightly more involved than just plugging in teh car bat. I have 3 deepcycle batts, set up for my FD base radio and 1 server using a 12V inverter and a trickle charger and just running the power through the inverter and batt all the time, that way I did not have to set up s a switching mechanism that would switch it over. It works well as long as you get a good inverter and one that is sized correctly for your power requirements. Mike -- Michael W. Holdeman Powered by Gentoo Linux www.gentoo.org | Kernel 2.6.11-ck8 | Win4Lin 5-1-20 netraverse.com | Win4LinPro 6.1.1-03 win4lin.com | | -- gentoo-user@gentoo.org mailing list
[gentoo-user] Eclipse vs. Unifont
Does anyone have an idea what the Eclipse ebuild doesn't like about Unifont? huxley ~ # emerge -DNupt dev-util/eclipse-sdk These are the packages that I would merge, in reverse order: Calculating dependencies ...done! [blocks B ] media-fonts/unifont (is blocking dev-util/eclipse-sdk-3.0.1-r2) [ebuild N] dev-util/eclipse-sdk-3.0.1-r2 The RDEPEND line in the ebuild is not commented, and at work I installed Eclipse on the server and it runs just fine on my Gentoo box that's virtually identical (including unifont) to this laptop's setup... regards Matthias -- I prefer encrypted and signed messages. KeyID: 90CF8389 Fingerprint: 8E 1F 10 81 A4 66 29 46 B9 8A B9 E2 09 9F 3B 91 pgpb8CdLXudvM.pgp Description: PGP signature
[gentoo-user] public_html on apache2
Does someone know how to allow everyone to access public_html on home folder of my users ? I had inserted the following lines on my commonapache2.conf and it still doesn't work Directory /home/*/public_html AllowOverride None Order allow,deny Allow from all /Directory anyone had any idea about it ? -- gentoo-user@gentoo.org mailing list
[gentoo-user] emerge -e world / --resume
If I do an emerge -e world , and interrupt it, will I be able to --resume it? Including after a reboot? -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] after emerge -Du world date can't keep time
--- Willie Wong [EMAIL PROTECTED] wrote: On Sun, Sep 11, 2005 at 08:41:31PM -0700, maxim wexler wrote: --- Willie Wong [EMAIL PROTECTED] wrote: On Sun, Sep 11, 2005 at 11:59:58AM -0700, maxim wexler wrote: Hello everybody, After emerge --deep --update world I compiled and installed the new kernel, 2.6.12-gentoo-r10, then I noticed that no matter when I boot the system, date always starts marking time at 6:00 AM, although it gets the date and zone right. No matter what time string I give to date(as root), after a re-boot it alway replies with 6AM plus whatever time has elapsed since the original boot. FWIW rtc support is compiled as a module and is *not* currently loaded. Also, the BIOS clk is correct. -mw hum, run rc-status boot is clock started? yup Let me check my understanding of the problem: 1) If you cold boot, the time starts at 6AM of the right day. A closer look: It appears 6AM was a one-off. Last night before shutdown I set the correct time. This morning at 9:17AM date gave 3:17 w/ correct day, date, time zone. So it looks like it's picking an hour at random but gets everything else correct. 2) If you reboot, the time counts from 6AM + the elapsed time of all previous reboots up until the last cold boot. whatever time has elapsed since first cold boot is added *correctly* to the incorrect original time every time I reboot. For instance, the clock on the wall says 12:40PM. I've just rebooted for the second time and date gives 6:40 which is what it *would* be if the original time, 3:17 had been correct, since approx two and a half hrs have elapsed. 3) Behaviours 1 and 2 doesn't change if you set the time yourself. i.e., on the next boot/reboot the same thing will happen. Does that summary seem correct? Yes. Whatever time string I give to date after rebooting, the clock resets to the original, wrong time after another reboot. BTW this applies to xclock as well. __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] after emerge -Du world date can't keep time
--- Willie Wong [EMAIL PROTECTED] wrote: On Sun, Sep 11, 2005 at 08:41:31PM -0700, maxim wexler wrote: On Sun, Sep 11, 2005 at 11:59:58AM -0700, maxim wexler wrote: Hello everybody, After emerge --deep --update world I compiled and installed the new kernel, 2.6.12-gentoo-r10, then I noticed that no matter when I boot the system, date always starts marking time at 6:00 AM, although it gets the date and zone right. No matter what time string I give to date(as root), after a re-boot it alway replies with 6AM plus whatever time has elapsed since the original boot. FWIW rtc support is compiled as a module and is *not* currently loaded. Also, the BIOS clk is correct. -mw hum, run rc-status boot is clock started? yup what about the output from hwclock? run it as root, does it give the same time/date as date? the same. In fact when I update the time w/ date -s hwclock continues to give the incorrect one. Until a reboot when both will give the incorrect original time plus whatever time has elapsed since the first boot of the day. __ Yahoo! Mail - PC Magazine Editors' Choice 2005 http://mail.yahoo.com -- gentoo-user@gentoo.org mailing list
[gentoo-user] reiserfs file size issue
Hi people This question follows on from a post I made regarding a hard drive death and backup problems, I made a little script that tar and gzip's each of the root directories in order. At the point of /usr which is a very large one I get the following error /root/fullsysbackup.sh: line 30: 18492 File size limit exceededcp /tmp/fullsysbackupusr.tgz When checking the file size I see that it is approx 2.1Gb and to be exact = 2^31 -1 After a fair bit of checking around into possible causes of the error [samba, cp, backup box runs XP, and finally reiserfs] I discovered from the reiser site that it has a file size limit of ... 2^31 -1. However this seemed strange but I accepted it. I asked a friend who is of the beleif that reiserfs 3.5 had the limit but 3.6 does not. Now for the silly question of the day, how do I know what version of reiserfs I have installed? OK well it is possible that I have got the wrong end of the stick at this point and indeed the file size limit is 2.1Gb, the solution would be to split the tgz on backup but I also have some rather large movie type files and Iwould rather have a fs that works for me rather than the other way around. I dont want to turn this thread into everyones favorite which fs for me? but any simplistic advice on the status of reiserfs file size limits and such would be helpful. stu ps. Comon England and up the Aussies :) we won the ashes! pps. Its a cricket thing ^^ ;) -- There are 10 types of people in this world: those who understand binary, those who don't --Unknown -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Enabling DMA on SE7520BD2S
--- JarosÅaw Kapica [EMAIL PROTECTED] wrote: Hi, I'm using Intel Server Board SE7520BD2S with 3 disks, 2xSATA WDC WD1200JD-00H (sda sdb) and 1xATA133 Seagate ST3120026A (hda). SATA disks seem to work fine (though they show some warnings during tests), but I can't enable DMA on /dev/hda. I've had this problem on two systems so far. The solution, for me, was to compile *into* the kernel support for the respective mobo chipsets: nvidia for one, ali for the other. -mw __ Yahoo! for Good Donate to the Hurricane Katrina relief effort. http://store.yahoo.com/redcross-donate3/ -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Proliant 3000
Guess I could, just wanted to keep the SCSI's clean for data only... The MBR doesn't use any space, and it won't alter your partitions in anyway. So to me it seems like the best route. --- Chris Covington IT Plus One Health Management 75 Maiden Lane Suite 801 NY, NY 10038 646-312-6269 http://www.plusoneactive.com -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] emerge -e world / --resume
On Mon, 12 Sep 2005 14:43:31 -0400, fire-eyes wrote: If I do an emerge -e world , and interrupt it, will I be able to --resume it? Including after a reboot? Yes. -- Neil Bothwick ABORT: Drivel filter is compromised! pgp4hmfp5UyID.pgp Description: PGP signature
Re: [gentoo-user] The new apache2
Grant wrote: Hello, I've been on the new apache2 layout for some time now via ~x86. It looks like it's now marked stable so I'd like to remove the ~x86 keyword, but after that I get this: [blocks B ] dev-libs/apr-util (is blocking net-www/apache-2.0.54-r15) [blocks B ] dev-libs/apr (is blocking net-www/apache-2.0.54-r15) Should I unmerge those packages? I thought they were important for apache? As posted on g-dev: These changes will stabilized on Sunday, September 18th. Are you sure that the package is stable now? -- Naga I'm running 2.0.54-r14 now and when I remove the ~x86 keyword, emerge world wants to upgrade to 2.0.54-r15. - Grant -- gentoo-user@gentoo.org mailing list
[gentoo-user] hardware advice for dual-processor set-up on gentoo
Folks, I am currently running Gentoo on a dual-processor machine that has 2 Intel Xeons with hyperthreading on a Tyan board, with 2GB of RDRAM. Some 3 years ago when I bought this machine, it was a marvel, and it still is a solid computational workhorse for me, doing things like Monte Carlo and Finite Element calculations. Because of hyperthreading, there are 4 logical processors, which gives me a good multi-tasking ability. I can run 3 or 4 Monte Carlo tasks at the same time and not get any noticeable slow-down compared to running only one Monte Carlo task. That has been indispensable in my productivity. Now, I am about to purchase a new machine, but I'd like to solicit some advice first, so I can get the most bang for the buck. The main thing is, I still need to have those 4 logical processors to maintain my productivity. So that gives me two choices: - 2 64-bit Irwindale Xeons (EM64T) with Hyperthreading, 2MB L2 Cache, 800 MHZ FSB, somewhere in 3.4-3.8 GHz range - 2 64-bit AMD Dual-Core Opterons somewhere in the 265-275 range With my need for parallel execution of several single-thread Monte Carlo and Finite Element tasks, what would be the pros and cons in considering the Xeons and Opterons? I looked at several benchmarks, and the Opterons seem to be leading the way significantly in pretty much every category except things like Matrix multiplications (because of vectoring?)... Is that the general impression of folks with experience on these systems running Linux? What would be some issues and software limitations for me in trying to get the AMD Opterons running on Gentoo? And would I use an x86 platform or an AMD64 platform? Thanks! Denis -- gentoo-user@gentoo.org mailing list
RE: [gentoo-user] mailman stopped working after upgrade
Someone on the 'mailman' list pointed me at, but that didn't seem to be very helpful: http://www.python.org/cgi-bin/faqw-mm.py?req=showfile=faq06.016.htp I don't know why this would have broken/changed. I've been running mailman, so this was really just an upgrade. Same user/groups, lists, dirs, etc. The only 'USE' flag I have availabe to me is +apache2 [ebuild R ] net-mail/mailman-2.1.5-r4 +apache2 5,611 kB Furthermore, it appears that the .ebuild has those --with flags mentioned in the FAQ. # more /usr/portage/net-mail/mailman/mailman-2.1.5-r4.ebuild ... src_compile() { econf \ --prefix=${INSTALLDIR} \ --with-mail-gid=${MAILGID} \ --with-cgi-gid=${APACHEGID} \ || die configure failed make || die make failed } ... -Original Message- From: Daevid Vincent [mailto:[EMAIL PROTECTED] Sent: Monday, September 12, 2005 12:35 PM To: gentoo-user@lists.gentoo.org Subject: [gentoo-user] mailman stopped working after upgrade Mailman was working, then a (Gentoo) 'emerge -Davu system' update happened the other day and looks like it broke mailman. daevid portage-logs # ll *mailman* -rw-r--r-- 1 root root 62662 Sep 7 11:34 3485-mailman-2.1.5-r4.log -rw-r--r-- 1 root root 1105 Sep 7 11:34 3486-mailman-2.1.5-r4.log Sep 12 12:16:19 [Mailman mail-wrapper] Group mismatch error. Mailman expected the mail_wrapper script to be executed as group mailman, but_the system's mail server executed the mail script as_group mail. Try tweaking the mail server to run the_script as group mailman, or re-run configure, _providing the command line option `--with-mail-gid=mail'._ Sep 12 12:16:19 [exim] 2005-09-12 12:16:19 1EEtms-00021M-60 ** |/usr/local/mailman/mail/mailman post rbc [EMAIL PROTECTED] R=system_aliases T=address_pipe: Child process of address_pipe transport returned 2 from command: /usr/local/mailman/mail/mailman daevid bin # locate mail_wrapper daevid bin # locate as_group daevid portage-logs # ll /usr/local/mailman/mail/mailman -rwxr-sr-x 1 mailman mailman 7768 Sep 7 11:34 /usr/local/mailman/mail/mailman daevid bin # ll total 372 drwxrwsr-x 2 mailman mailman 4096 Sep 7 11:34 . drwxrwsr-x 20 mailman mailman 4096 Apr 4 13:52 .. -rwxr-xr-x 1 mailman mailman 7612 Sep 7 11:34 add_members -rwxr-xr-x 1 mailman mailman 5406 Sep 7 11:34 arch -rw-r--r-- 1 mailman mailman 3748 Sep 7 11:34 auto -rwxr-xr-x 1 mailman mailman 2559 Sep 7 11:34 b4b5-archfix -rwxr-xr-x 1 mailman mailman 6130 Sep 7 11:34 change_pw -rwxr-xr-x 1 mailman mailman 4264 Sep 7 11:34 check_db -rwxr-xr-x 1 mailman mailman 12223 Sep 7 11:34 check_perms -rwxr-xr-x 1 mailman mailman 6632 Sep 7 11:34 check_perms_grsecurity.py -rwxr-xr-x 1 mailman mailman 5492 Sep 7 11:34 cleanarch -rwxr-xr-x 1 mailman mailman 6409 Sep 7 11:34 clone_member -rwxr-xr-x 1 mailman mailman 12326 Sep 7 11:34 config_list -rwxr-xr-x 1 mailman mailman 1499 Sep 7 11:34 convert.py -rwxr-xr-x 1 mailman mailman 3238 Sep 7 11:34 discard -rwxr-xr-x 1 mailman mailman 4399 Sep 7 11:34 dumpdb -rwxr-xr-x 1 mailman mailman 4951 Sep 7 11:34 find_member -rwxr-xr-x 1 mailman mailman 2633 Sep 7 11:34 fix_url.py -rw-r--r-- 1 mailman mailman 2497 Apr 4 13:52 fix_url.pyc -rwxr-xr-x 1 mailman mailman 3097 Sep 7 11:34 genaliases -rwxr-xr-x 1 mailman mailman 2815 Sep 7 11:34 inject -rwxr-xr-x 1 mailman mailman 2561 Sep 7 11:34 list_admins -rwxr-xr-x 1 mailman mailman 3329 Sep 7 11:34 list_lists -rwxr-xr-x 1 mailman mailman 8316 Sep 7 11:34 list_members -rwxr-xr-x 1 mailman mailman 3165 Sep 7 11:34 list_owners -rwxr-xr-x 1 mailman mailman 21124 Sep 7 11:34 mailmanctl -rw-r--r-- 1 mailman mailman 18995 Sep 7 11:34 majordomo2mailman.pl -rw-r--r-- 1 mailman mailman 5951 Sep 7 11:34 mm-handler -rw-r--r-- 1 mailman mailman 7048 Sep 7 11:34 mm-handler.readme -rwxr-xr-x 1 mailman mailman 2989 Sep 7 11:34 mmsitepass -rwxr-xr-x 1 mailman mailman 5530 Sep 7 11:34 msgfmt.py -rwxr-xr-x 1 mailman mailman 7127 Sep 7 11:34 newlist -rw-r--r-- 1 mailman mailman 2412 Sep 7 11:34 paths.py -rw-r--r-- 1 rootmailman 652 Sep 7 11:34 paths.pyc -rwxr-xr-x 1 mailman mailman 4076 Sep 7 11:34 qmail-to-mailman.py -rwxr-xr-x 1 mailman mailman 9196 Sep 7 11:34 qrunner -rwxr-xr-x 1 mailman mailman 2957 Sep 7 11:34 rb-archfix -rwxr-xr-x 1 mailman mailman 4699 Sep 7 11:34 remove_members -rwxr-xr-x 1 mailman mailman 4256 Sep 7 11:34 rmlist -rwxr-xr-x 1 mailman mailman 3463 Sep 7 11:34 rotatelogs.py -rwxr-xr-x 1 mailman mailman 1346 Sep 7 11:34 show_qfiles -rwxr-xr-x 1 mailman mailman 9664 Sep 7 11:34 sync_members -rwxr-xr-x 1 mailman mailman 11999 Sep 7 11:34 transcheck -rwxr-xr-x 1 mailman mailman 2286 Sep 7 11:34 unshunt -rwxr-xr-x 1 mailman mailman 29648 Sep 7 11:34 update
Re: [gentoo-user] Problem with udpate
You clearly know how to clear a block, but your syntax for unmerging the blocking packages is wrong, which is why it's not working. If you want to unmerge the specific version of the package (for this example, let us assume php-5.0.5), the correct syntax is emerge -C =dev-php/php-5.0.5 If you want to unmerge any or all versions of the package currently installed, letting Portage decide which to unmerge, the correct syntax is emerge -C php Ok, but this will remove php-5, and that's not what I want. I want to remove dev-php/php any version, but I do that (I think the right syntax to do it is emerge -C dev-php/php), I got the previous message, obvisoualy there is an inconsistence on my system: - To update a package that is blocked by another package is required - The required package is not installed. It looks like the point is php has migrated. The version 4.x and below are located in dev-php and the version 5.x and above are locate in dev-lang, so is there an inconsistency in the portage tree when unmasking php-5.x and above? This message was sent using IMP, the Internet Messaging Program. -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Problem with udpate
This is Gentoo's daft syntax again! What you need here is not # emerge -C giflib, as you would reasonably expect, but # emerge -C libungif I can't do that, libungif is required on my system. I'm pretty sure some programs won't run anymore if I remove it, but I can test anyway This message was sent using IMP, the Internet Messaging Program. -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] hardware advice for dual-processor set-up on gentoo
On Monday 12 September 2005 21:40, Denis wrote: With my need for parallel execution of several single-thread Monte Carlo and Finite Element tasks, what would be the pros and cons in considering the Xeons and Opterons? Opterons everytime. Under certain circumstances HT can make the CPU *slower*, and at best only adds around a 20% performance boost. A dual core CPU *is* two CPUs, with single digit % performance drop over 2 full CPUs. What would be some issues and software limitations for me in trying to get the AMD Opterons running on Gentoo? And would I use an x86 platform or an AMD64 platform? Issues? Almost exactly nil. AMD64 support is excellent, if a little smaller and perhaps the tiniest bit behind, the x86 profile. -- Mike Williams -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] reiserfs file size issue
On Monday 12 September 2005 20:09, Stuart Howard wrote: Now for the silly question of the day, how do I know what version of reiserfs I have installed? debugreiserfs /dev/blah I've got a 2.3G file on a 3.6 formatted filesystem, but I can't help any further I'm afraid. -- Mike Williams -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] ntpd timezone
Right now it's correct: tv mythtv # date Sun Sep 11 23:52:49 PDT 2005 However after some time it will return to being 8 hours earlier. I also saw in another thread that you could show your hardware clock by this command: tv mythtv # hwclock Sun Sep 11 23:54:54 2005 -0.081182 seconds Don't know if that helps or not. Thanks for your reply! Drew Possibly /var/lib/ntp/ntp.drift has a skewed value. what does it say? -- -- Nick Rout [EMAIL PROTECTED] -- gentoo-user@gentoo.org mailing list
[gentoo-user] Fwd: reiserfs file size issue
Further reading further problem, I have been through the reiserfs FAQ and discovered that with 2.4 - 2.6 kernels that the file size limit I have posted below should not be an issue ie. I should be version 3.6 and therefore file size is well ... biG So I now need to discover where the fault lies, the error message was File size limit exceededcp no google, dogpile or yahoo search admits to ever knowing this error however I suppose the clue is in -cp ? /me reads GNU docs for tar report no issue's with file sizes below 8Gb and then only when using specific formats that I am not so far as I am aware. To recap the commands I am calling are tar -czf /tmp/fullsysbackupusr.tgz /usr cp /tmp/fullsysbackupusr.tgz /mnt/NEWSTU/genstubackup/fullsysbackup The source is Gentoo box 2.6.11.5 on a reiserfs partition, the file is created but will not exceed 2Gb or 2,147,483,647. Once created [though the file should be larger] it is dutifully transfered to the XP box via cp. any suggestions on this ? Beginning to suspect I am not seeing wood for tree's but if anyone has an axe please throw it my way. stu ps. For my North American cousins, cricket is a game where 13 men stand in a field for 5 days and hurl a small leather ball at 3 sticks in the ground with the aim of disturbing 2 smaller pieces of wood from the top of the aforesaid 3 sticks :) -- Forwarded message -- From: Stuart Howard [EMAIL PROTECTED] Date: Sep 12, 2005 9:09 PM Subject: reiserfs file size issue To: gentoo-user@lists.gentoo.org Hi people This question follows on from a post I made regarding a hard drive death and backup problems, I made a little script that tar and gzip's each of the root directories in order. At the point of /usr which is a very large one I get the following error /root/fullsysbackup.sh: line 30: 18492 File size limit exceededcp /tmp/fullsysbackupusr.tgz When checking the file size I see that it is approx 2.1Gb and to be exact = 2^31 -1 After a fair bit of checking around into possible causes of the error [samba, cp, backup box runs XP, and finally reiserfs] I discovered from the reiser site that it has a file size limit of ... 2^31 -1. However this seemed strange but I accepted it. I asked a friend who is of the beleif that reiserfs 3.5 had the limit but 3.6 does not. Now for the silly question of the day, how do I know what version of reiserfs I have installed? OK well it is possible that I have got the wrong end of the stick at this point and indeed the file size limit is 2.1Gb, the solution would be to split the tgz on backup but I also have some rather large movie type files and Iwould rather have a fs that works for me rather than the other way around. I dont want to turn this thread into everyones favorite which fs for me? but any simplistic advice on the status of reiserfs file size limits and such would be helpful. stu ps. Comon England and up the Aussies :) we won the ashes! pps. Its a cricket thing ^^ ;) -- There are 10 types of people in this world: those who understand binary, those who don't --Unknown -- There are 10 types of people in this world: those who understand binary, those who don't --Unknown -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] hardware advice for dual-processor set-up on gentoo
On Monday 12 September 2005 15:40, Denis wrote: Folks, I am currently running Gentoo on a dual-processor machine that has 2 Intel Xeons with hyperthreading on a Tyan board, with 2GB of RDRAM. Some 3 years ago when I bought this machine, it was a marvel, and it still is a solid computational workhorse for me, doing things like Monte Carlo and Finite Element calculations. Because of hyperthreading, there are 4 logical processors, which gives me a good multi-tasking ability. I can run 3 or 4 Monte Carlo tasks at the same time and not get any noticeable slow-down compared to running only one Monte Carlo task. That has been indispensable in my productivity. Now, I am about to purchase a new machine, but I'd like to solicit some advice first, so I can get the most bang for the buck. The main thing is, I still need to have those 4 logical processors to maintain my productivity. So that gives me two choices: - 2 64-bit Irwindale Xeons (EM64T) with Hyperthreading, 2MB L2 Cache, 800 MHZ FSB, somewhere in 3.4-3.8 GHz range - 2 64-bit AMD Dual-Core Opterons somewhere in the 265-275 range With my need for parallel execution of several single-thread Monte Carlo and Finite Element tasks, what would be the pros and cons in considering the Xeons and Opterons? I looked at several benchmarks, and the Opterons seem to be leading the way significantly in pretty much every category except things like Matrix multiplications (because of vectoring?)... Is that the general impression of folks with experience on these systems running Linux? What would be some issues and software limitations for me in trying to get the AMD Opterons running on Gentoo? And would I use an x86 platform or an AMD64 platform? Thanks! Denis I don't know the benchmarks, but I DO know that I got a 64-bit gentoo system running on a dual-core athlon 4400. It's a single-dual-core cpu, using a Tyan Tomkat k8e board (only one cpu slot on the board). There were no problems or gotchas I could tell. It's got 4 gigs of ram and screams (at least on kernel compiles and suchhaven't benchmarked mysql on it yet). -- John Jolet Your On-Demand IT Department 512-762-0729 www.jolet.net [EMAIL PROTECTED] -- gentoo-user@gentoo.org mailing list
[gentoo-user] Re: reiserfs file size issue
thanks mike using debugreiserfs -J /dev/hda3 answers my earlier question of Now for the silly question of the day, how do I know what version of reiserfs I have installed? =. 3.6 which means from reiserfs FAQ that :- 2^60 - bytes = 1 Ei, but page cache limits this to 8 Ti on architectures with 32 bit int ie. biG or at least I think so, I am afraid that this is somewhat out of my league. stu On 9/12/05, Stuart Howard [EMAIL PROTECTED] wrote: Further reading further problem, I have been through the reiserfs FAQ and discovered that with 2.4 - 2.6 kernels that the file size limit I have posted below should not be an issue ie. I should be version 3.6 and therefore file size is well ... biG So I now need to discover where the fault lies, the error message was File size limit exceededcp no google, dogpile or yahoo search admits to ever knowing this error however I suppose the clue is in -cp ? /me reads GNU docs for tar report no issue's with file sizes below 8Gb and then only when using specific formats that I am not so far as I am aware. To recap the commands I am calling are tar -czf /tmp/fullsysbackupusr.tgz /usr cp /tmp/fullsysbackupusr.tgz /mnt/NEWSTU/genstubackup/fullsysbackup The source is Gentoo box 2.6.11.5 on a reiserfs partition, the file is created but will not exceed 2Gb or 2,147,483,647. Once created [though the file should be larger] it is dutifully transfered to the XP box via cp. any suggestions on this ? Beginning to suspect I am not seeing wood for tree's but if anyone has an axe please throw it my way. stu ps. For my North American cousins, cricket is a game where 13 men stand in a field for 5 days and hurl a small leather ball at 3 sticks in the ground with the aim of disturbing 2 smaller pieces of wood from the top of the aforesaid 3 sticks :) -- Forwarded message -- From: Stuart Howard [EMAIL PROTECTED] Date: Sep 12, 2005 9:09 PM Subject: reiserfs file size issue To: gentoo-user@lists.gentoo.org Hi people This question follows on from a post I made regarding a hard drive death and backup problems, I made a little script that tar and gzip's each of the root directories in order. At the point of /usr which is a very large one I get the following error /root/fullsysbackup.sh: line 30: 18492 File size limit exceededcp /tmp/fullsysbackupusr.tgz When checking the file size I see that it is approx 2.1Gb and to be exact = 2^31 -1 After a fair bit of checking around into possible causes of the error [samba, cp, backup box runs XP, and finally reiserfs] I discovered from the reiser site that it has a file size limit of ... 2^31 -1. However this seemed strange but I accepted it. I asked a friend who is of the beleif that reiserfs 3.5 had the limit but 3.6 does not. Now for the silly question of the day, how do I know what version of reiserfs I have installed? OK well it is possible that I have got the wrong end of the stick at this point and indeed the file size limit is 2.1Gb, the solution would be to split the tgz on backup but I also have some rather large movie type files and Iwould rather have a fs that works for me rather than the other way around. I dont want to turn this thread into everyones favorite which fs for me? but any simplistic advice on the status of reiserfs file size limits and such would be helpful. stu ps. Comon England and up the Aussies :) we won the ashes! pps. Its a cricket thing ^^ ;) -- There are 10 types of people in this world: those who understand binary, those who don't --Unknown -- There are 10 types of people in this world: those who understand binary, those who don't --Unknown -- There are 10 types of people in this world: those who understand binary, those who don't --Unknown -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] reiserfs file size issue
I would gamble your shell doesn't have support for files larger than 2GB. You'll run into this with tcsh. :( The pipes created by the shell for child processes will inherit this problem. On Mon, 12 Sep 2005, Stuart Howard wrote: Hi people This question follows on from a post I made regarding a hard drive death and backup problems, I made a little script that tar and gzip's each of the root directories in order. At the point of /usr which is a very large one I get the following error /root/fullsysbackup.sh: line 30: 18492 File size limit exceededcp /tmp/fullsysbackupusr.tgz When checking the file size I see that it is approx 2.1Gb and to be exact = 2^31 -1 After a fair bit of checking around into possible causes of the error [samba, cp, backup box runs XP, and finally reiserfs] I discovered from the reiser site that it has a file size limit of ... 2^31 -1. However this seemed strange but I accepted it. I asked a friend who is of the beleif that reiserfs 3.5 had the limit but 3.6 does not. Now for the silly question of the day, how do I know what version of reiserfs I have installed? OK well it is possible that I have got the wrong end of the stick at this point and indeed the file size limit is 2.1Gb, the solution would be to split the tgz on backup but I also have some rather large movie type files and Iwould rather have a fs that works for me rather than the other way around. I dont want to turn this thread into everyones favorite which fs for me? but any simplistic advice on the status of reiserfs file size limits and such would be helpful. stu ps. Comon England and up the Aussies :) we won the ashes! pps. Its a cricket thing ^^ ;) -- Bryan Whitehead Email:[EMAIL PROTECTED] -- gentoo-user@gentoo.org mailing list
[gentoo-user] ebuild - .src.rpm
Does anyone know how to do this? There are some simple pachakes (like netselect) that I'd love to use at work - but don't want to go thru the hastle of making a rpm .spec file. ebuild file rpm only poops out a binary RPM... and using an updated gentoo system doesn't move over well with old versions of redhat... -- Bryan Whitehead Email:[EMAIL PROTECTED] -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] after emerge -Du world date can't keep time
On Mon, Sep 12, 2005 at 12:03:27PM -0700, maxim wexler wrote: what about the output from hwclock? run it as root, does it give the same time/date as date? the same. In fact when I update the time w/ date -s hwclock continues to give the incorrect one. Until a reboot when both will give the incorrect original time plus whatever time has elapsed since the first boot of the day. hum... fishy! On Mon, Sep 12, 2005 at 12:20:55PM -0700, maxim wexler wrote: check out /etc/conf.d/clock # /etc/conf.d/clock # Set CLOCK to UTC if your system clock is set to UTC (also known as # Greenwich Mean Time). If your clock is set to the local time, then # set CLOCK to local. CLOCK=UTC ^ | Problem #1. # If you wish to pass any other arguments to hwclock during bootup, # you may do so here. CLOCK_OPTS= # If you want to set the Hardware Clock to the current System Time # during shutdown, then say yes here. CLOCK_SYSTOHC=no ^ --- | | THIS is your problem #2. Date only manipulates the kernel clock. If you don't sync it with your hardware clock, how do you expect the computer to remember the change? During a power cycle, the kernel is not running, you know q= Also, I'd wager that your clock is consistently off by the same number of hours after each boot. Why? If the BIOS clock is correctly set to the time in your time-zone, you need to set CLOCK=local (see problem #1 above). Right now even though the BIOS clock is correct, the software thinks that the time it is keeping is Greenwich Mean Time, so it adds/subtracts the suitable number of hours according to your timezone. W -- This is obviously not how one does science, but in retrospect you get the right answer. ~DeathMech, S. Sondhi. P-town PHY 205 Sortir en Pantoufles: up 32 days, 1:07 -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Proliant 3000
On Monday 12 September 2005 03:34 pm, Covington, Chris wrote: Guess I could, just wanted to keep the SCSI's clean for data only... The MBR doesn't use any space, and it won't alter your partitions in anyway. So to me it seems like the best route. Trouble is when the SCSI is replaced and it is not thought of because the OS is on the IDE HD. For me it would be cleaner to boot with a floppy or CD and (even keep a couple of spare floppies). I have trouble remembering certian things, the location of the MBR on SCSI 0 would be one of those things. Mike -- Michael W. Holdeman Powered by Gentoo Linux www.gentoo.org | Kernel 2.6.11-ck8 | Win4Lin 5-1-20 netraverse.com | Win4LinPro 6.1.1-03 win4lin.com | | -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Why autoconf?
Why the hick are there so much versions of autoconf (in system)? Well, somewhere in the [Nasty bug..] thread someone (again) mentioned, that different packaged depend on different versions of autoconf. That's NOT the truth for building a package. You're making some assumptions here... Part of the emerge process applies many patches to the original package. Sometimes these patches are applied to the configure.in and makefile.in scripts. When changes like this are made, autoconf automake are called to rebuild the configure script before it is executed. And because the original package used autoconf version x.yz, x.yz is needed to rebuild the configure script appropriately for the package. Depending upon the release time, an older version of autoconf was used, and therefore the older version of autoconf is a dependency and is installed in a slot. It's really not a big deal to get upset over. The autoconf and automake packages are pretty small, so they don't take up a lot of disk nor do they really require any attention on your part. Portage will ensure the correct versions are available for building packages, and there's really nothing that you, as the gentoo user, would need to do about it. -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Portage logs
What about changing the logfile names to: package-version.log package-version.msg? Because multiple installs of the same version would overwrite the log. The added prefix is, I believe, either some sort of order number or a time-based reference, not sure which. -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Fwd: reiserfs file size issue
On Mon, Sep 12, 2005 at 11:40:05PM +0200, Stuart Howard wrote: Further reading further problem, I have been through the reiserfs FAQ and discovered that with 2.4 - 2.6 kernels that the file size limit I have posted below should not be an issue ie. I should be version 3.6 and therefore file size is well ... biG So I now need to discover where the fault lies, the error message was File size limit exceededcp no google, dogpile or yahoo search admits to ever knowing this error however I suppose the clue is in -cp ? the cp is from your command: cp /tmp/fullsysbackupusr.tgz ... The error message is NOT File size limit exceededcp. The error message should be line 30: cp /tmp/fullsysbackupusr.tgz ... File size limit exceeded I am not quite sure why the error message got jumbled up the way it did in your case. To recap the commands I am calling are tar -czf /tmp/fullsysbackupusr.tgz /usr cp /tmp/fullsysbackupusr.tgz /mnt/NEWSTU/genstubackup/fullsysbackup The source is Gentoo box 2.6.11.5 on a reiserfs partition, the file is created but will not exceed 2Gb or 2,147,483,647. Once created [though the file should be larger] it is dutifully transfered to the XP box via cp. The problem is here. The kernel code used to access smbfs (I presume that is how you mount your XP box to /mnt/NEWSTU?) has a file size limit of 2/4GB (I am not quite sure about the number). http://lists.suse.com/archive/suse-amd64/2004-May/0058.html W -- Thesaurus: ancient reptile with an excellent vocabulary. Sortir en Pantoufles: up 32 days, 1:31 -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] ebuild - .src.rpm
That would indeed be an interesting addition! I can see a few problems already. Different dependency names would be a problem - ie on gentoo a dependency might be named foo, on redhat it might be named foo2, on SuSE it might be named libfoo and on Mandrivel it might be named libfoo-2. In other words spec files may be different on different rpm distros and may use different namings to gentoo. SuSE installs kde to /opt, others don't etc etc . So you not only have to create a spec file, but you have to create one for each distro, and you have to have a lot of knowledge about how $DISTRO works. Another strategy for you is to find a src.rpm for the package and adapt its .spec file. I found an old (0.2) src rpm for netselect (from SuSE 7.3!) via google. If I cared enough I guess I could adapt it's spec file to another distro. Actually I stumbled on a part of the Mandrake site once that had just the spec files for all their packages - that might be worth a look given RH and MD's common (although now distant) origins. One of the things I like about gentoo is the ease of writing a .ebuild file as opposed to a spec file or whatever the .deb equivalent is. On Mon, 12 Sep 2005 14:54:55 -0700 (PDT) Bryan Whitehead wrote: Does anyone know how to do this? There are some simple pachakes (like netselect) that I'd love to use at work - but don't want to go thru the hastle of making a rpm .spec file. ebuild file rpm only poops out a binary RPM... and using an updated gentoo system doesn't move over well with old versions of redhat... -- Bryan Whitehead Email:[EMAIL PROTECTED] -- gentoo-user@gentoo.org mailing list -- Nick Rout [EMAIL PROTECTED] -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] howto list packages that depend on the package?
On Mon, Sep 12, 2005 at 01:32:42PM -0600, Peter Ziobrzynski wrote: Is there a way to get the list of packages that depend on a package (dependent packages) then doing grep through the /var/db/pkgs like that? cd /var/db/pkg find . -name DEPEND | xargs grep avifile Unfortunately this only lists locally installed packages. The reverse operation for finding a list of dependent packages for a package is provided with 'emerge --emptytree -p mozilla'. Is there something like that for the first question? [06:24 PM]wwong ~ $ equery belongs equery [ Searching for file(s) equery in *... ] app-portage/gentoolkit-0.2.0 (/usr/bin/equery) app-portage/gentoolkit-0.2.0 (/usr/share/doc/gentoolkit-0.2.0_pre10/equery) [06:38 PM]wwong ~ $ equery depends avifile [ Searching for packages depending on avifile... ] [06:39 PM]wwong ~ $ equery depends ffmpeg [ Searching for packages depending on ffmpeg... ] media-libs/xine-lib-1.0.1-r3 (I don't have avifile installed) so... emerge gentoolkit run equery with no arguments to see what it can do. W -- Inside some of us is a thin person struggling to get out, but she can usually be sedated with a few pieces of chocolate cake. Sortir en Pantoufles: up 32 days, 1:42 -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] howto list packages that depend on the package?
oops, I missed 'qpkg -q'. Ignore this Peter Ziobrzynski wrote: Is there a way to get the list of packages that depend on a package (dependent packages) then doing grep through the /var/db/pkgs like that? cd /var/db/pkg find . -name DEPEND | xargs grep avifile Unfortunately this only lists locally installed packages. The reverse operation for finding a list of dependent packages for a package is provided with 'emerge --emptytree -p mozilla'. Is there something like that for the first question? -- Peter Ziobrzynski, mailto:[EMAIL PROTECTED] -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Proliant 3000
Trouble is when the SCSI is replaced and it is not thought of because the OS is on the IDE HD. For me it would be cleaner to boot with a floppy or CD and (even keep a couple of spare floppies). I have trouble remembering certian things, the location of the MBR on SCSI 0 would be one of those things. Well it's just as hard to remember that the MBR is on the floppy or the cdrom as it is on the SCSI disk. If it's about remembering which SCSI disk, just run a cron job or init script which installs the Grub MBR in all of your SCSI disks so you won't have to worry about remembering. --- Chris Covington IT Plus One Health Management 75 Maiden Lane Suite 801 NY, NY 10038 646-312-6269 http://www.plusoneactive.com -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Why autoconf?
On Mon, 12 Sep 2005 18:26:45 -0400 Dave Nebinger wrote: It's really not a big deal to get upset over. No but I am glad it has been explored, as i had been wondering the same thing. The autoconf and automake packages are pretty small, so they don't take up a lot of disk nor do they really require any attention on your part. Portage will ensure the correct versions are available for building packages, and there's really nothing that you, as the gentoo user, would need to do about it. -- gentoo-user@gentoo.org mailing list -- Nick Rout [EMAIL PROTECTED] -- gentoo-user@gentoo.org mailing list
[gentoo-user] org-x11 GLSA 200509-07. Is bug #96053 fixed in -r3? (black icons)
Re the xorg-x11 GLSA 200509-07. Is bug #96053 fixed in -r3? The -r2 version is unusable for me because of this. According to the bug its resolved, fixed upstream, but there is no detail which versions of xorg it applies to. BillK -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] ebuild - .src.rpm
Of course you might like to consider this post again :-) http://www.gentoo.org/news/en/gwn/20030401-newsletter.xml Be careful to note the date of the article ! On Mon, 12 Sep 2005 14:54:55 -0700 (PDT) Bryan Whitehead wrote: Does anyone know how to do this? There are some simple pachakes (like netselect) that I'd love to use at work - but don't want to go thru the hastle of making a rpm .spec file. ebuild file rpm only poops out a binary RPM... and using an updated gentoo system doesn't move over well with old versions of redhat... -- Bryan Whitehead Email:[EMAIL PROTECTED] -- gentoo-user@gentoo.org mailing list -- Nick Rout [EMAIL PROTECTED] -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Fwd: reiserfs file size issue
On Monday 12 September 2005 22:40, Stuart Howard wrote: To recap the commands I am calling are tar -czf /tmp/fullsysbackupusr.tgz /usr cp /tmp/fullsysbackupusr.tgz /mnt/NEWSTU/genstubackup/fullsysbackup Is NEWSTU a samba mount from the XP box? The source is Gentoo box 2.6.11.5 on a reiserfs partition, the file is created but will not exceed 2Gb or 2,147,483,647. Once created [though the file should be larger] it is dutifully transfered to the XP box via cp. OK, I assume you're running this from some sort of script? If the file isn't getting created fully, I'd go for it being tar at fault, as I know reiser has no problems with files well over 2GB. Personally, I'd just bypass the whole issue, and use a better method for backing up. My favourite system is rdiff-backup (rdiff-backup /usr /mnt/NEWSTU/genstubackup/fullsysbackup). -- Mike Williams -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Portage logs
On Mon, 12 Sep 2005 13:38:41 +0200 Frank Schafer wrote: There shouldn't be more logs than we have actually packages installed. The log for the installation of the version of (say) python I had installed 3 years ago isn't worth anything. (Me to say) A successful installation of a new version of a package can safely overwrite the old log because IMHO nobody needs the installation logs of not installed software. It does provide some useful functionality IMHO, (and obviously in the NSHO of the designer of the logging system.) I quite _like_ to see when i first installed mplayer, what versions i had installed at certain times, how long it will take to emerge (averaged over all the times I have previously emerged it). I particularly like being able to pipe emerge -puD world into genlop -p and find out how long a whole lot of ebuilds is likely to take. -- Nick Rout [EMAIL PROTECTED] -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] hardware advice for dual-processor set-up on gentoo
I was leaning toward the Opterons myself. While we're on the topic of hardware, has anyone here used Tyan S2895 Thunder K8WE? I understand it's an NVIDIA-based high-end board that supports dual Opterons. I've always liked Tyan's products, as well as NVIDIA's, so I would be happy with such a choice... But I am curious to hear about people's experiences with or what they heard (good and bad) about this particular board... -- gentoo-user@gentoo.org mailing list
[gentoo-user] Samba/Names/XPPro
List, I'm using Samba to share with Windows and am using the %m macro as the share name so that each Windows computer sees only one share with it's name to use. This works great for almost every machine on my network: other Sambas, Win2K, Win2K3, WinXP home but not this one XPPro. I've tried tweaking the lmhosts file and waiting for Netbios to share the names but this hasn't worked, any ideas? /djb -- gentoo-user@gentoo.org mailing list
[gentoo-user] Samba/Names/XPPro
List, I'm using Samba to share with Windows and am using the %m macro as the share name so that each Windows computer sees only one share with it's name to use. This works great for almost every machine on my network: other Sambas, Win2K, Win2K3, WinXP home but not this one XPPro. I've tried tweaking the lmhosts file and waiting for Netbios to share the names but this hasn't worked, any ideas? /djb -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Problem with udpate
Sébastien MORAND schreef: You clearly know how to clear a block, but your syntax for unmerging the blocking packages is wrong, which is why it's not working. If you want to unmerge the specific version of the package (for this example, let us assume php-5.0.5), the correct syntax is emerge -C =dev-php/php-5.0.5 If you want to unmerge any or all versions of the package currently installed, letting Portage decide which to unmerge, the correct syntax is emerge -C php Ok, but this will remove php-5, and that's not what I want. I want to remove dev-php/php any version, but I do that (I think the right syntax to do it is emerge -C dev-php/php), I got the previous message, obvisoualy there is an inconsistence on my system: - To update a package that is blocked by another package is required - The required package is not installed. No, the right syntax to do that (remove php any and all versions) is emerge -C php There is no syntax that includes the cat-egory without both a version number and a mathematical symbol, such as emerge -C =dev-php/php-5.0.5 (unmerges all versions greater than or equal to php-5.0.5) emerge -C =dev-php/php-5.0.5 (unmerges only php-5.0.5) emerge -C dev-php/php-5.0.5 (unmerges all versions lower than php-5.0.5) emerge -C dev-lang/php is not valid, because it includes the cat-egory (dev-lang) without specifying a version. If you don't want to specify a version, only use the package name, without the category. If you want to specify a version, use a mathematical symbol (for precision, using =, , and either alone or in combination with each other), aand the version number. It looks like the point is php has migrated. The version 4.x and below are located in dev-php and the version 5.x and above are locate in dev-lang, so is there an inconsistency in the portage tree when unmasking php-5.x and above? Well, yes and no. If you ultimately want php v.5 on your system, unmerge the current version, then re-emerge =dev-lang/php-5.0.5, then see if it still blocks the program you are trying to emerge. It's possible that the block exists because of the php migration to a new category, and the program itself needs to be unmerged and re-emerged with the correct category, rather than a block due to functionality clash between the two programs. If the block remains, it's because the program you're trying to emerge either will emerge php v.5 as a dependency or the program replicates its function in some way, so unmerge 5.0.5 (again), and emerge what was blocked by it, then re-(re-)emerge php-5.0.5 (if you still need to do that). HTH, Holly -- gentoo-user@gentoo.org mailing list
RE: [gentoo-user] mailman stopped working after upgrade
Well, since I've gotten no replies here, I figured I'd file a report: http://bugs.gentoo.org/show_bug.cgi?id=105744 -Original Message- From: Daevid Vincent [mailto:[EMAIL PROTECTED] Sent: Monday, September 12, 2005 1:53 PM To: gentoo-user@lists.gentoo.org Subject: RE: [gentoo-user] mailman stopped working after upgrade Someone on the 'mailman' list pointed me at, but that didn't seem to be very helpful: http://www.python.org/cgi-bin/faqw-mm.py?req=showfile=faq06.016.htp I don't know why this would have broken/changed. I've been running mailman, so this was really just an upgrade. Same user/groups, lists, dirs, etc. The only 'USE' flag I have availabe to me is +apache2 [ebuild R ] net-mail/mailman-2.1.5-r4 +apache2 5,611 kB Furthermore, it appears that the .ebuild has those --with flags mentioned in the FAQ. # more /usr/portage/net-mail/mailman/mailman-2.1.5-r4.ebuild ... src_compile() { econf \ --prefix=${INSTALLDIR} \ --with-mail-gid=${MAILGID} \ --with-cgi-gid=${APACHEGID} \ || die configure failed make || die make failed } ... -Original Message- From: Daevid Vincent [mailto:[EMAIL PROTECTED] Sent: Monday, September 12, 2005 12:35 PM To: gentoo-user@lists.gentoo.org Subject: [gentoo-user] mailman stopped working after upgrade Mailman was working, then a (Gentoo) 'emerge -Davu system' update happened the other day and looks like it broke mailman. daevid portage-logs # ll *mailman* -rw-r--r-- 1 root root 62662 Sep 7 11:34 3485-mailman-2.1.5-r4.log -rw-r--r-- 1 root root 1105 Sep 7 11:34 3486-mailman-2.1.5-r4.log Sep 12 12:16:19 [Mailman mail-wrapper] Group mismatch error. Mailman expected the mail_wrapper script to be executed as group mailman, but_the system's mail server executed the mail script as_group mail. Try tweaking the mail server to run the_script as group mailman, or re-run configure, _providing the command line option `--with-mail-gid=mail'._ Sep 12 12:16:19 [exim] 2005-09-12 12:16:19 1EEtms-00021M-60 ** |/usr/local/mailman/mail/mailman post rbc [EMAIL PROTECTED] R=system_aliases T=address_pipe: Child process of address_pipe transport returned 2 from command: /usr/local/mailman/mail/mailman daevid bin # locate mail_wrapper daevid bin # locate as_group daevid portage-logs # ll /usr/local/mailman/mail/mailman -rwxr-sr-x 1 mailman mailman 7768 Sep 7 11:34 /usr/local/mailman/mail/mailman daevid bin # ll total 372 drwxrwsr-x 2 mailman mailman 4096 Sep 7 11:34 . drwxrwsr-x 20 mailman mailman 4096 Apr 4 13:52 .. -rwxr-xr-x 1 mailman mailman 7612 Sep 7 11:34 add_members -rwxr-xr-x 1 mailman mailman 5406 Sep 7 11:34 arch -rw-r--r-- 1 mailman mailman 3748 Sep 7 11:34 auto -rwxr-xr-x 1 mailman mailman 2559 Sep 7 11:34 b4b5-archfix -rwxr-xr-x 1 mailman mailman 6130 Sep 7 11:34 change_pw -rwxr-xr-x 1 mailman mailman 4264 Sep 7 11:34 check_db -rwxr-xr-x 1 mailman mailman 12223 Sep 7 11:34 check_perms -rwxr-xr-x 1 mailman mailman 6632 Sep 7 11:34 check_perms_grsecurity.py -rwxr-xr-x 1 mailman mailman 5492 Sep 7 11:34 cleanarch -rwxr-xr-x 1 mailman mailman 6409 Sep 7 11:34 clone_member -rwxr-xr-x 1 mailman mailman 12326 Sep 7 11:34 config_list -rwxr-xr-x 1 mailman mailman 1499 Sep 7 11:34 convert.py -rwxr-xr-x 1 mailman mailman 3238 Sep 7 11:34 discard -rwxr-xr-x 1 mailman mailman 4399 Sep 7 11:34 dumpdb -rwxr-xr-x 1 mailman mailman 4951 Sep 7 11:34 find_member -rwxr-xr-x 1 mailman mailman 2633 Sep 7 11:34 fix_url.py -rw-r--r-- 1 mailman mailman 2497 Apr 4 13:52 fix_url.pyc -rwxr-xr-x 1 mailman mailman 3097 Sep 7 11:34 genaliases -rwxr-xr-x 1 mailman mailman 2815 Sep 7 11:34 inject -rwxr-xr-x 1 mailman mailman 2561 Sep 7 11:34 list_admins -rwxr-xr-x 1 mailman mailman 3329 Sep 7 11:34 list_lists -rwxr-xr-x 1 mailman mailman 8316 Sep 7 11:34 list_members -rwxr-xr-x 1 mailman mailman 3165 Sep 7 11:34 list_owners -rwxr-xr-x 1 mailman mailman 21124 Sep 7 11:34 mailmanctl -rw-r--r-- 1 mailman mailman 18995 Sep 7 11:34 majordomo2mailman.pl -rw-r--r-- 1 mailman mailman 5951 Sep 7 11:34 mm-handler -rw-r--r-- 1 mailman mailman 7048 Sep 7 11:34 mm-handler.readme -rwxr-xr-x 1 mailman mailman 2989 Sep 7 11:34 mmsitepass -rwxr-xr-x 1 mailman mailman 5530 Sep 7 11:34 msgfmt.py -rwxr-xr-x 1 mailman mailman 7127 Sep 7 11:34 newlist -rw-r--r-- 1 mailman mailman 2412 Sep 7 11:34 paths.py -rw-r--r-- 1 rootmailman 652 Sep 7 11:34 paths.pyc -rwxr-xr-x 1 mailman mailman 4076 Sep 7 11:34 qmail-to-mailman.py -rwxr-xr-x 1 mailman mailman 9196 Sep 7 11:34 qrunner -rwxr-xr-x 1 mailman mailman 2957 Sep 7 11:34 rb-archfix -rwxr-xr-x 1 mailman mailman 4699
Re: [gentoo-user] The new apache2
On Mon, Sep 12, 2005 at 12:41:29PM -0700, Grant wrote: Grant wrote: Hello, I've been on the new apache2 layout for some time now via ~x86. It looks like it's now marked stable so I'd like to remove the ~x86 keyword, but after that I get this: [blocks B ] dev-libs/apr-util (is blocking net-www/apache-2.0.54-r15) [blocks B ] dev-libs/apr (is blocking net-www/apache-2.0.54-r15) Should I unmerge those packages? I thought they were important for apache? As posted on g-dev: These changes will stabilized on Sunday, September 18th. Are you sure that the package is stable now? I'm running 2.0.54-r14 now and when I remove the ~x86 keyword, emerge world wants to upgrade to 2.0.54-r15. The latest stable seems to be -r15. I am running r14 now, emerge --update wants to upgrade to -r30, but I have no blocks from apr. Considering apr and apr-util are all masked by ~x86, it wouldn't make sense for stable apache to depend on apr. In fact, if you look at the various apache ebuilds, it seems apache2.x that is ~x86 all depends on apr, and apache2.x that is x86 all have !dev-libs/apr-util in it. W -- M: Even when I know what I am doing, I don't know what I am doing. M: Like, this first semester, I knew I was going to be an ORFE major, but look at me now... I am definitely a classics major... Sortir en Pantoufles: up 32 days, 1:47 -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Building Xorg - internal compiler error: segmentation fault
I get segment faults randomly when emerging packages... and I know it is bad ram... so I would advise you to test it... :) Regards. On 9/12/05, Frank Schafer [EMAIL PROTECTED] wrote: Hi all, so I hit this bug too. :( Well, the log itself says it's a bug and (even worse) that nobody knows where it comes from. Somebody on b.g.o. suggests to try building with a vanilly kernel. Did someone here make the same experience and if yes, will it run then on a gentoo-kernel? Thanks for any hint Frank -- gentoo-user@gentoo.org mailing list -- gentoo-user@gentoo.org mailing list
[gentoo-user] Kernel says hdd is my dvd drive, /dev/hdd does not exist
Howdy. I have my kernel telling me that it found my dvd drive at hdd : hdd: PLEXTOR DVDR PX-708A, ATAPI CD/DVD-ROM drive However, /dev/hdd doesn't even exist: # stat /dev/hdd stat: cannot stat `/dev/hdd': No such file or directory I am using udev-068. Any clues as to what's going on here? This worked up until today and I can't really think of what I've changed. -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] mailman stopped working after upgrade
On Mon, 12 Sep 2005 16:34:26 -0700 Daevid Vincent wrote: Well, since I've gotten no replies here, I figured I'd file a report: http://bugs.gentoo.org/show_bug.cgi?id=105744 -Original Message- From: Daevid Vincent [mailto:[EMAIL PROTECTED] Sent: Monday, September 12, 2005 1:53 PM To: gentoo-user@lists.gentoo.org Subject: RE: [gentoo-user] mailman stopped working after upgrade -Original Message- From: Daevid Vincent [mailto:[EMAIL PROTECTED] Sent: Monday, September 12, 2005 12:35 PM To: gentoo-user@lists.gentoo.org Subject: [gentoo-user] mailman stopped working after upgrade Mailman was working, then a (Gentoo) 'emerge -Davu system' update happened the other day and looks like it broke mailman. daevid portage-logs # ll *mailman* -rw-r--r-- 1 root root 62662 Sep 7 11:34 3485-mailman-2.1.5-r4.log -rw-r--r-- 1 root root 1105 Sep 7 11:34 3486-mailman-2.1.5-r4.log Odd that you took this long to upgrade mailman as mine was upgraded to 2.1.5-r4 in February ? Sep 12 12:16:19 [Mailman mail-wrapper] Group mismatch error. Mailman expected the mail_wrapper script to be executed as group mailman, but_the system's mail server executed the mail script as_group mail. Try tweaking the mail server to run the_script as group mailman, or re-run configure, _providing the command line option `--with-mail-gid=mail'._ what MTA are you running? What is the result of running: id mail id mailman ls -l /usr/local/mailman/mail/mailman -- Nick Rout [EMAIL PROTECTED] -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Eclipse vs. Unifont
On Mon, 12 Sep 2005 20:18:41 +0200 Matthias Bethke [EMAIL PROTECTED] wrote: Does anyone have an idea what the Eclipse ebuild doesn't like about Unifont? Look at the ebuild Changelog: quote 04 Jul 2004; Karl Trygve Kalleberg [EMAIL PROTECTED] eclipse-sdk-3.0.0.ebuild: New upstream version. Added block against media-fonts/unifont, at the request of Billy Biggs from Eclipse. /quote Cheers, Renat -- Probleme kann man niemals mit derselben Denkweise loesen, durch die sie entstanden sind. (Einstein) pgpyofo9Nfouo.pgp Description: PGP signature
[gentoo-user] Recent booting problems
Hi, i rebooted my firewall yesterday, and it stopped just after the initrd, when trying to fsck the root filesystem, because it had failed to remount the / filesystem read-only because it was busy, and thus fsck was warning about checking a live filesystem i rebooted from livecd, and fsck'ed my partitions manually, to make sure they were ok, and then rebooted and said yes to fsck the live FS, and it carried on and booted ok i suspected i had baselayout out of sync or some such, so did emerge sync and then upgraded baselayout. that didn't help, so i tried emerge -e system, and also upgrade of world, and then i also rebuilt my initrd with genkernel, in case something had changed in there. still the same, so now i'm quite stuck anyone else had this recently? my kernels and ramdisk are always built with genkernel, and as i say above i'm bang up to date now any help greatly appreciated cheers, Dunc -- gentoo-user@gentoo.org mailing list
[gentoo-user] recent booting problems
Hi, i rebooted my firewall yesterday, and it stopped just after the initrd, when trying to fsck the root filesystem, because it had failed to remount the / filesystem read-only because it was busy, and thus fsck was warning about checking a live filesystem i rebooted from livecd, and fsck'ed my partitions manually, to make sure they were ok, and then rebooted and said yes to fsck the live FS, and it carried on and booted ok i suspected i had baselayout out of sync or some such, so did emerge sync and then upgraded baselayout. that didn't help, so i tried emerge -e system, and also upgrade of world, and then i also rebuilt my initrd with genkernel, in case something had changed in there. still the same, so now i'm quite stuck anyone else had this recently? my kernels and ramdisk are always built with genkernel, and as i say above i'm bang up to date now any help greatly appreciated cheers, Dunc -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Proliant 3000
On Monday 12 September 2005 06:43 pm, Covington, Chris wrote: Trouble is when the SCSI is replaced and it is not thought of because the OS is on the IDE HD. For me it would be cleaner to boot with a floppy or CD and (even keep a couple of spare floppies). I have trouble remembering certian things, the location of the MBR on SCSI 0 would be one of those things. Well it's just as hard to remember that the MBR is on the floppy or the cdrom as it is on the SCSI disk. If it's about remembering which SCSI disk, just run a cron job or init script which installs the Grub MBR in all of your SCSI disks so you won't have to worry about remembering. Thats an idea, although it isn't that hard remembering the mbr is on a floppy if there is one in teh machine, I don't think any of my other machines have a floppy in them ever, let alone at boot time... Mike -- Michael W. Holdeman Powered by Gentoo Linux www.gentoo.org | Kernel 2.6.11-ck8 | Win4Lin 5-1-20 netraverse.com | Win4LinPro 6.1.1-03 win4lin.com | | -- gentoo-user@gentoo.org mailing list
RE: [gentoo-user] mailman stopped working after upgrade
Thank you for taking the time to reply Nick. Odd that you took this long to upgrade mailman as mine was upgraded to 2.1.5-r4 in February ? I run a fairly stable x86 server and try not to 'fix' what ain't broke. I dunno. I usually run 'emerge -Davu system' and 'world' every couple of days and see if anything stands out. I also try not to update things right a way so as to let the -rN versions stabalize. There is almost always a -r1 right away ;-) There was a GLSA 3 months back on June 9th. http://www.gentoo.org/security/en/glsa/glsa-200406-04.xml what MTA are you running? [ebuild R ] mail-mta/exim-4.50-r1 +X -dnsdb -exiscan -exiscan-acl +ipv6 -ldap -lmtp -mailwrapper -mbox +mysql -nis +pam +perl -postgres -sasl +ssl -syslog +tcpd 1,401 kB But I don't believe that's changed since: daevid portage-logs # ll *exim* -rw-r--r-- 1 root root 23369 Jul 7 00:57 3190-exim-4.50-r1.log -rw-r--r-- 1 root root 242 Jul 7 00:57 3191-exim-4.50-r1.log What is the result of running? id mail uid=8(mail) gid=12(mail) groups=12(mail),408(shadow) id mailman uid=280(mailman) gid=280(mailman) groups=280(mailman),16(cron) ls -l /usr/local/mailman/mail/mailman -rwxr-sr-x 1 mailman mailman 7768 Sep 12 15:47 /usr/local/mailman/mail/mailman I tried to emerge mailman again today just incase that helped. Same results. daevid portage-logs # ll *mailman* -rw-r--r-- 1 root root 62662 Sep 7 11:34 3485-mailman-2.1.5-r4.log -rw-r--r-- 1 root root 1105 Sep 7 11:34 3486-mailman-2.1.5-r4.log -rw-r--r-- 1 root root 62663 Sep 12 15:47 3528-mailman-2.1.5-r4.log -rw-r--r-- 1 root root 1105 Sep 12 15:48 3529-mailman-2.1.5-r4.log From the .log file: checking for --with-permcheck... yes checking for --with-username... mailman checking for user name mailman... okay checking for --with-groupname... mailman checking for group name mailman... okay checking permissions on /usr/local/mailman... okay checking for mail wrapper group; i.e. --with-mail-gid... mailman checking for CGI wrapper group; i.e. --with-cgi-gid... apache -- gentoo-user@gentoo.org mailing list
RE: [gentoo-user] mailman stopped working after upgrade
BTW, someone on the mailman list said, Yes, it has the flags, but what is the value of the ${MAILGID} parameter? The error message you received says it's 'mailman' and should be 'mail'. And looking at the log, he seems to be correct: Sep 12 12:16:19 [Mailman mail-wrapper] Group mismatch error. Mailman expected the mail_wrapper script to be executed as group mailman, but_the system's mail server executed the mail script as_group mail. Try tweaking the mail server to run the_script as group mailman, or re-run configure, _providing the command line option `--with-mail-gid=mail'._ But the .ebuild script sets it to gid 280, shouldn't it be 8? why is that? Is that the problem? How can I fix that. Re-emerging doesn't let me change those .ebuild values. MD5 errors and such. APACHEGID=81 MAILGID=280 src_compile() { econf \ --prefix=${INSTALLDIR} \ --with-mail-gid=${MAILGID} \ --with-cgi-gid=${APACHEGID} \ || die configure failed make || die make failed } # ll /etc/exim/exim.conf -rw-r--r-- 1 root root 29452 Apr 29 13:54 /etc/exim/exim.conf # User and group for Mailman, should match your --with-mail-gid # switch to Mailman's configure script. # Value is normally mailman MM_UID=mailman MM_GID=mailman So, these match what the comment says there, and what the .ebuild is doing. I'm so confused. *sigh* -Original Message- From: Daevid Vincent [mailto:[EMAIL PROTECTED] Sent: Monday, September 12, 2005 6:22 PM To: gentoo-user@lists.gentoo.org Subject: RE: [gentoo-user] mailman stopped working after upgrade Thank you for taking the time to reply Nick. Odd that you took this long to upgrade mailman as mine was upgraded to 2.1.5-r4 in February ? I run a fairly stable x86 server and try not to 'fix' what ain't broke. I dunno. I usually run 'emerge -Davu system' and 'world' every couple of days and see if anything stands out. I also try not to update things right a way so as to let the -rN versions stabalize. There is almost always a -r1 right away ;-) There was a GLSA 3 months back on June 9th. http://www.gentoo.org/security/en/glsa/glsa-200406-04.xml what MTA are you running? [ebuild R ] mail-mta/exim-4.50-r1 +X -dnsdb -exiscan -exiscan-acl +ipv6 -ldap -lmtp -mailwrapper -mbox +mysql -nis +pam +perl -postgres -sasl +ssl -syslog +tcpd 1,401 kB But I don't believe that's changed since: daevid portage-logs # ll *exim* -rw-r--r-- 1 root root 23369 Jul 7 00:57 3190-exim-4.50-r1.log -rw-r--r-- 1 root root 242 Jul 7 00:57 3191-exim-4.50-r1.log What is the result of running? id mail uid=8(mail) gid=12(mail) groups=12(mail),408(shadow) id mailman uid=280(mailman) gid=280(mailman) groups=280(mailman),16(cron) ls -l /usr/local/mailman/mail/mailman -rwxr-sr-x 1 mailman mailman 7768 Sep 12 15:47 /usr/local/mailman/mail/mailman I tried to emerge mailman again today just incase that helped. Same results. daevid portage-logs # ll *mailman* -rw-r--r-- 1 root root 62662 Sep 7 11:34 3485-mailman-2.1.5-r4.log -rw-r--r-- 1 root root 1105 Sep 7 11:34 3486-mailman-2.1.5-r4.log -rw-r--r-- 1 root root 62663 Sep 12 15:47 3528-mailman-2.1.5-r4.log -rw-r--r-- 1 root root 1105 Sep 12 15:48 3529-mailman-2.1.5-r4.log From the .log file: checking for --with-permcheck... yes checking for --with-username... mailman checking for user name mailman... okay checking for --with-groupname... mailman checking for group name mailman... okay checking permissions on /usr/local/mailman... okay checking for mail wrapper group; i.e. --with-mail-gid... mailman checking for CGI wrapper group; i.e. --with-cgi-gid... apache -- gentoo-user@gentoo.org mailing list -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] mailman stopped working after upgrade
On Mon, 12 Sep 2005 18:52:35 -0700 Daevid Vincent wrote: BTW, someone on the mailman list said, Yes, it has the flags, but what is the value of the ${MAILGID} parameter? The error message you received says it's 'mailman' and should be 'mail'. My system runs postfix, not exim, but mailman is owned and run as mailman:mailman or 280:280 And looking at the log, he seems to be correct: Sep 12 12:16:19 [Mailman mail-wrapper] Group mismatch error. Mailman expected the mail_wrapper script to be executed as group mailman, but_the system's mail server executed the mail script as_group mail. Try tweaking the mail server to run the_script as group mailman, or re-run configure, _providing the command line option `--with-mail-gid=mail'._ But the .ebuild script sets it to gid 280, shouldn't it be 8? why is that? Is that the problem? How can I fix that. Re-emerging doesn't let me change those .ebuild values. MD5 errors and such. I am not sure how mailman comes to be run as user:group mailman:mailman on my system, but it does! postfix runs as user postfix, not sure where user mail comes into it at all. I suspect you need to tweak something in order to encourage exim to pass off to the mailman program as user mailman Ahhh hold on the penny drops: -rwxr-sr-x 1 mailman mailman 7880 Feb 16 2005 /usr/local/mailman/mail/mailman The s in group permissions is set gid or something isn't it? Which means it gets run as group mailman even though someone else is starting it? Then again yours has the same permissions - back to square one. -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] ebuild - .src.rpm
Yea... I'm not that much of n00b! I remember that newsletter when it was first posted! :) On Tue, 13 Sep 2005, Nick Rout wrote: Of course you might like to consider this post again :-) http://www.gentoo.org/news/en/gwn/20030401-newsletter.xml Be careful to note the date of the article ! On Mon, 12 Sep 2005 14:54:55 -0700 (PDT) Bryan Whitehead wrote: Does anyone know how to do this? There are some simple pachakes (like netselect) that I'd love to use at work - but don't want to go thru the hastle of making a rpm .spec file. ebuild file rpm only poops out a binary RPM... and using an updated gentoo system doesn't move over well with old versions of redhat... -- Bryan Whitehead Email:[EMAIL PROTECTED] -- gentoo-user@gentoo.org mailing list -- Bryan Whitehead Email:[EMAIL PROTECTED] -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] after emerge -Du world date can't keep time
--- Willie Wong [EMAIL PROTECTED] wrote: On Mon, Sep 12, 2005 at 12:03:27PM -0700, maxim wexler wrote: what about the output from hwclock? run it as root, does it give the same time/date as date? the same. In fact when I update the time w/ date -s hwclock continues to give the incorrect one. Until a reboot when both will give the incorrect original time plus whatever time has elapsed since the first boot of the day. hum... fishy! On Mon, Sep 12, 2005 at 12:20:55PM -0700, maxim wexler wrote: check out /etc/conf.d/clock # /etc/conf.d/clock # Set CLOCK to UTC if your system clock is set to UTC (also known as # Greenwich Mean Time). If your clock is set to the local time, then # set CLOCK to local. CLOCK=UTC ^ | Problem #1. # If you wish to pass any other arguments to hwclock during bootup, # you may do so here. CLOCK_OPTS= # If you want to set the Hardware Clock to the current System Time # during shutdown, then say yes here. CLOCK_SYSTOHC=no ^ --- | | THIS is your problem #2. Thanks W, that works! Date only manipulates the kernel clock. If you don't sync it with your hardware clock, how do you expect the computer to remember the change? Dunno, but on another box(2.6.12-gentoo-r6) with the same settings date gives the correct, local time. During a power cycle, the kernel is not running, you know q= ? Also, I'd wager that your clock is consistently off by the same number of hours after each boot. Why? If the BIOS clock is correctly set to the time in your time-zone, you need to set CLOCK=local (see problem #1 above). Right now even though the BIOS clock is correct, the software thinks that the time it is keeping is Greenwich Mean Time, so it adds/subtracts the suitable number of hours according to your timezone. W -- This is obviously not how one does science, but in retrospect you get the right answer. ~DeathMech, S. Sondhi. P-town PHY 205 Sortir en Pantoufles: up 32 days, 1:07 -- gentoo-user@gentoo.org mailing list __ Yahoo! for Good Donate to the Hurricane Katrina relief effort. http://store.yahoo.com/redcross-donate3/ -- gentoo-user@gentoo.org mailing list
[gentoo-user] emerge --oneshot vs emerge --uDvp differences
emerge --ask --oneshot --verbose =dev-lang/python-2.3.5-r2, as directed in the GLSA, returns a download for xemacs, which I do not now or ever had installed. It is not in use flags. However, an emerge -uDvp python doesn't even mention xemacs. Any clue as to why the difference? Any problems if the --oneshot is not used? Thanks. -- Ed Jabbour -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Fwd: reiserfs file size issue
Stuart Howard wrote: Yes I am using samba to connect to the XP box as you suggested. The error report arrived via mail from a cron, prehaps the garbling of the message occured there? Anyway thanks again for responses, I usually clean up /tmp after the script runs so I made the assumption later on that the error was at the point of creation of the file and not during the cp, so as a way of confirming this I have set it off again to make the very large tgz and then try too cp it via samba transport to the XP box. [will run overnight now] If it works ie. if the full file is created [waiting in /tmp] then it is samba or if not then a post back to the shell limit. Should know by morning. An interesting chase regardless and if it is samba then will have to split the tgz. Mounting it as a smbfs filesystem (which does not support file larger than 2G) is almost certainly the problem. Try mounting the XP box as a cifs (mount.cifs) filesystem and you should have better luck. -Richard -- gentoo-user@gentoo.org mailing list