devel March 2004

devel@lists.fedoraproject.org

238 participants
275 discussions

RE: fedora.us QA process
by Erik LaBianca 04 Mar '04

04 Mar '04

> > fedora-rpmdevtools package contains the tool: > fedora-buildrpmtree > > http://fedoranews.org/tchung/rpmbuild/ > > -jef > Link and note added. Thanks. I didn't realize fedora-buildrpmtree automatically created ~/.rpmmacro's as well. Nice script :) http://www.ilsw.com/~erik/fedora-qa-quickstart.html --erik

1 0

ISDN and redhat-control-network
by Luca 04 Mar '04

04 Mar '04

Hi all, I have to report some problems with my FC1 install. I have an HFC-PCI nameless ISDN card, that got found and configured on install. When accessing redhat-control-network, the card can be found, and the connection comes up without any glitches. Problem is, the ippp0 interface does not get detected as being up, so I can connect, but not disconnect! It's not a major problem, but since I don't have a flat rate phone contract, I usually only connect when I need it. This e-mail is mainly to report the bug, but I'd also appreciate if somebody could tell me how to bring my ISDN connection down (and up, but that's less important) by command line, to avoid going broke on phone bills. Thanks all, Luca

3 3

RE: fedora.us QA process
by Jef Spaleta 04 Mar '04

04 Mar '04

Erik LaBianca wrote: > This was another assumption I made up above. I think we'll need an > accompanying document (maybe there is a good one we can just link > in?) that will explain how to setup and use rpm out of one's home > directory. I agree we DO NOT want people building out of > /usr/src/redhat. fedora-rpmdevtools package contains the tool: fedora-buildrpmtree http://fedoranews.org/tchung/rpmbuild/ -jef

1 0

Re: Logrotate RFE (2)
by Keith Lofstrom 04 Mar '04

04 Mar '04

On thursday, March 4, Konstantin Ryabitsev writes: > I'm also a little skeptical if this is a good feature. And not just > because I would have to hack my beautiful rotation-handling routines in > epylog. :) Fortunately, you would not have to change a thing. The SUSE logrotate patch is just another option that can be optionally added to the logrotate config file. If you do not explicitly turn on the dateext extension, the log files are numbered in the old way. On the other hand, if the logrotate configuration file is optionally configured to use dateext, many other programs that are sensitive to filename changes do not have to be enhanced to accomodate the special case behavior of the files in the /var/log directory. So, you have nothing to lose, and much capability to gain. The additional capability may be attractive enough for you to enhance your beautiful code. If it is not, you can ignore the new capability and leave the configuration files and the behavior the old way, forever. Still at issue, perhaps, is how the base distro should be configured. The old way, with numbers, of course! Those of us that want to use the new capability can build our own version of the /etc/logrotate.d directory, to fold into the distro version when we do upgrades. Keith -- Keith Lofstrom keithl(a)ieee.org Voice (503)-520-1993 KLIC --- Keith Lofstrom Integrated Circuits --- "Your Ideas in Silicon" Design Contracting in Bipolar and CMOS - Analog, Digital, and Scan ICs

1 0

RE: fedora.us QA process
by Erik LaBianca 04 Mar '04

04 Mar '04

Toshio, Thanks for taking the time to look it over. > > Architectural Comments: > * I like that you've split out blocker criteria from non-blocker > criteria. I would split security out too so the new reviewer can > realize this list is _really_ important. I purposefully didn't examine > what was in your lists of critical vs non-critical because I think we > need to discuss what standards we as a group want our packages to be > held up to. I've been going on the assumption that everything on the > QAChecklist is mandatory. If that's not so, it has to be discussed (and > perhaps voted) as it's our collective name as a quality repository > that's at stake. [We might be seeking to make a bigger repository, but > we still have to balance that against quality.] Agreed. My initial impression was that the QACheckList was mandatory too. However, I quickly found out that several of those items are subject to a LOT of contention, and are generally NOT considered showstoppers. I think unless there is some sort of unanimity about a given item, it has to be considered a non-showstopper. The alternative is to not accept packages or input from those who disagree, which is partially why we have as many "non-standard" repositories as we do. This is a subject for much discussion, I agree. I've tried to put something down for people to argue over and be proactive instead of whining. > > * I don't like the fact that the QAChecklist contains most of the things > in this document but is a separate document. I think the QAChecklist > needs to be tuned up with categories of blocker, non-blocker. And > perhaps in terms of order of doing things. And then it can be > referenced from this page. Having another page with overlapping > Checklists adds to the confusion of "which list do I follow?" and means > policy changes have to be updated in two places. > Agreed here as well. I'd like to see the QACheckList be the definitive list of blockers and non-blockers. Right now it's a hodge podge of how-to's, critical checks, and opinions. My guide is meant as a one-stop shop to understanding the QA checklist and its accompanying documents. I don't see it being a definitive list of things to check, or necessarily of fedora.us QA policy. Personally, I'd like to see my document become the official "high level overview" of the policy, and the qa checklist become the definitive list of policy, at a detailed level. > Editorial comments: > [3.1] - I remember having to add fedora.us to my yum.conf manually but > I'm not sure if that was before or after I installed FC1. If FC1 > doesn't include it, it needs to be mentioned. > - mach needs apt installed as well. Yes. However, I stated that I assumed the QA volunteer had set up their system to use fedora.us already. Mach installs apt as a dependency, and I'm working with Thomas to get my changes merged into the upstream mach. They should be available via fedora.us by the time this document goes live. I've changed the document to assume that the user can install mach from fedora.us. > [3.2] - I like gpasswd more than vigr because `gpasswd -a [USERNAME] > mach` is a more complete explanation than edit groups with vigr. Good suggestion. I forgot about gpasswd. Change made. > [3.x] - There needs to be something in here about modifying ~/.rpmmacros > to point to another build directory as /usr/src/redhat is FHS and > root-user problematic. This was another assumption I made up above. I think we'll need an accompanying document (maybe there is a good one we can just link in?) that will explain how to setup and use rpm out of one's home directory. I agree we DO NOT want people building out of /usr/src/redhat. I don't really know how many people know how to do this. As I see it, if they don't know how, they've got a lot to learn before they can do QA effectively. > [4] - I'd add some note to pay attention to see what good and bad things > to look for. How things are structured into Critical and non-critical > comments. Reviews should include an SRPM MD5sum _at_minimum_. My belief is that reviews should contain ALL the things I have in the template. That includes the SRPM filename and MD5Sum, and the MD5Sums for the source packages. The non-critical comments should go in the notes section of my template. I don't see much point in QA without some accountability that says "Yes, I checked the basics, and the package seems ok". I realize not everyone agrees with me here. A standard format also has the benefit of being able to be automated. For instance, one could write a script to troll bugzilla comments for reviews and validate them for completeness, check md5sums, or whatever. > [6] - Not all packages have MD5sums. And the improtant part of this > step is to correlate a package to a gpg key. Either MD5sums are > provided and signed with a gpg key (which needs to be gpg verified) or > the package itself needs to be rpm --checksig verified. The Fedora.us > policy says that all packages need to be rpm --addsign'ed so you can > leave out the MD5Sum verification if you're feeling lazy. My explanations of verifying package integrity obviously need some work. As I see it there are two aspects. 1. Are you checking the same package the author uploaded? Clearsigned md5sums or gpg sigs on the SRPM should do this. This is part of step 5, but should be elaborated on. How do you easily get somebody's gpg key on your keyring? One of the things that slowed me down a lot was attempting to track down peoples public keys. I've changed the document to require checking with rpm --checksig in addition to checking md5sums. 2. Is the author using legimitate sources. Verifying md5sums or gpg sigs on the tarballs and patches in the SRPM is the only way to do this. This part is the non-obvious one, because it's tough to find the sources the author used sometimes. I've attempted to make this more clear... Please review > [8] - Note to google. Note to _not_trust_the_spec_ completely. It's > important that the reviewer verify that the URL's and etc are not > leading to false download sources. Good point. I've added a note. Please review. > [14] - A use your judgement about additional things should be added. > (And where to ask QA/Packaging questions -- irc #fedora-devel, mailing > lists, asking on the bugzilla, etc) > - Other keywords need mentioning. PUBLISH, NEEDSWORK, QA come to > mind. REVIEWED only applies if no one else has already REVIEWED it. > > [Example Review Template] > - There needs to be an example of a negative review as well. Hmm. I've added one. See what you think. > - GPG signing is _important_. It's not even mentioned. Yes it is, in brief. For a newbie I don't believe it's critical since their reviews will be somewhat suspect anyway. > - I like MD5Sums of all files in the SRPM. > - It's better to have the output of md5sum verbatim in the review. > Easier to automate verification than having it split over multiple > lines. Good idea. See changes in document. > It's a good start on a helpful document. Thanks, and thanks for the input. --erik

1 0

Re: Logrotate RFE (2)
by Keith Lofstrom 04 Mar '04

04 Mar '04

On Tuesday 02 March 2004 22:55, Keith Lofstrom wrote: > Logrotate is the program called by cron to rename and expire log > files. Ruedinger Oertel at SUSE (ro(a)suse.de) has some patches that > enhance the basic Redhat logrotate with "dateext". This allows a > dated log file extension rather than a numbered one, for example > /var/log/messages.20031029 . The old logfiles do not get renamed, > just discarded after they get too old. This is a lot easier on > rsync, and it also is easier to administer. on Thu, 4 Mar 2004 10:42:17, Doncho N. Gunchev responds: > I was seeking a such thing and would love to test it! However I see a > small problem with it - webalizer or whatever log analyzer expects to > find rotated logs as <log>.[1..x] will fail, but it's not that hard > to make symlinks to workaround this. Is this done? on Thu, 4 Mar 2004 Keith Lofstrom responds: The bugzilla description: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=108775 ... contains a fairly complete description, including a pointer to Ruediger's patch to logrotate 3.6.10: ftp.suse.com/pub/people/ro/logrotate So you can try it out yourself. Again, this is an additional capability, not a replacement - you can keep using logrotate the old way if you want, or even do some log files with numbers and some with dates. Someday, the log analysis tools can be upgraded to use dated extensions as an option. Until then, just leave the appropriate log files configured to use the old numbering scheme, and everything will keep working. This permits incremental improvements and upgrades, while permitting the use of any SUSE-oriented analysis tools that might come along. On average, I like the way Fedora does things better than I like the way SUSE does things. But in the cases where SUSE is better, why not have the best of both? Keith -- Keith Lofstrom keithl(a)ieee.org Voice (503)-520-1993 KLIC --- Keith Lofstrom Integrated Circuits --- "Your Ideas in Silicon" Design Contracting in Bipolar and CMOS - Analog, Digital, and Scan ICs

1 0

Logrotate RFE (2)
by Keith Lofstrom 04 Mar '04

04 Mar '04

Logrotate is the program called by cron to rename and expire log files. Ruedinger Oertel at SUSE (ro(a)suse.de) has some patches that enhance the basic Redhat logrotate with "dateext". This allows a dated log file extension rather than a numbered one, for example /var/log/messages.20031029 . The old logfiles do not get renamed, just discarded after they get too old. This is a lot easier on rsync, and it also is easier to administer. Ruediger's patches do not change the current logrotate behavior, just adds new features. The code patches are stable and very well tested - all SUSE users are using them - and they work fine with the old integer-suffix scheme (which I ran for some time, to test). What are the proper steps to get these patches folded into the main development path for Fedora Core? I mentioned this on this list in October with a few positive comments, and submitted an RFE to bugzilla: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=108775 ... but there have been no responses since early November. This patch can significantly increase the efficiency of rsync disk-to-disk backups, and overall is a cleaner way to do things. Hundreds of thousands of SUSE users have tested the snarf out of it. So what can I do? Organize more Redhat-style testers? Run more compatability tests? Or is this one of those not-invented-here or not-important-enough things, and I should just patch and reinstall every time I upgrade Fedora? In the larger sense, what can I do to make it easier for the Fedora organizers to incorporate small enhancements like this? Help with other, more popular projects to earn credibility? Send beer and pizza? Keith -- Keith Lofstrom keithl(a)ieee.org Voice (503)-520-1993 KLIC --- Keith Lofstrom Integrated Circuits --- "Your Ideas in Silicon" Design Contracting in Bipolar and CMOS - Analog, Digital, and Scan ICs

8 10

Archive `dev' on gmane
by Harry Putnam 04 Mar '04

04 Mar '04

I'd like to submist this least to the mail to news gateway at news.gmane.org. Its main guy. Lars, Asked me to make sure it was ok with whoever decides such things here. Any one here that can speak to this? It would be an extra place to do research. Has an excellent search engine. http://search.gmane.org/ And thousands of lists archived.

1 0

rawhide report: 20040304 changes
by Build System 04 Mar '04

04 Mar '04

Updated Packages: a2ps-4.13b-36 ------------- * Wed Mar 03 2004 Tim Waugh <twaugh(a)redhat.com> 4.13b-36 - Oops, use system C compiler. * Tue Mar 02 2004 Elliot Lee <sopwith(a)redhat.com> - rebuilt anaconda-9.91-0.20040303144435 ------------------------------ * Wed Mar 03 2004 Anaconda team <bugzilla(a)redhat.com> - built new version from CVS * Tue Feb 24 2004 Jeremy Katz <katzj(a)redhat.com> - buildrequire libselinux-devel * Thu Nov 06 2003 Jeremy Katz <katzj(a)redhat.com> - require booty (#109272) bug-buddy-2.5.90-2 ------------------ * Wed Mar 03 2004 Alexander Larsson <alexl(a)redhat.com> 1:2.5.90-2 - add COPYING.ximian-logos to docs (#115950) cyrus-imapd-2.2.3-6 ------------------- * Wed Mar 03 2004 Karsten Hopp <karsten(a)redhat.de> 2.2.3-6 - strip binaries - fix file ownership (#116627) * Tue Mar 02 2004 Elliot Lee <sopwith(a)redhat.com> - rebuilt esound-0.2.33-1 --------------- * Wed Mar 03 2004 Alexander Larsson <alexl(a)redhat.com> 1:0.2.33-1 - update to 0.2.33, hopefully fixes alsa issues * Tue Mar 02 2004 Elliot Lee <sopwith(a)redhat.com> - rebuilt gcc34-3.4.0-0.6 --------------- * Tue Mar 02 2004 Jakub Jelinek <jakub(a)redhat.com> 3.4.0-0.6 - update from gcc-3_4-branch - PRs 14240, bootstrap/14207, bootstrap/14356, c++/13944, c++/14138, c++/14246, c++/14250, c++/14260, c++/14278, c++/14284, c++/14324, c++/14337, c++/14359, c++/14360, c++/14361, c++/14369, c/14188, debug/12103, debug/14328, java/14296, libstdc++/10246, libstdc++/14248, libstdc++/14252, middle-end/13448, optimization/7871, target/14209 - fix PR optimization/12419 - warn if VLAs might free alloca before end of function (Alexandre Oliva, PR 14236) - add --param min-pretend-dynamic-size=NNN option (Alexandre Oliva) - remove uninstalled binaries and libraries from the install tree, so that they don't end up in gcc34-debuginfo (which prevents installing both gcc-debuginfo and gcc34-debuginfo together) - include gcov34 in the package gkrellm-2.1.26-2 ---------------- * Wed Mar 03 2004 Karsten Hopp <karsten(a)redhat.de> 2.1.26-2 - fix -devel provision (#117105) - remove stringfreeze hack * Tue Mar 02 2004 Elliot Lee <sopwith(a)redhat.com> - rebuilt gnome-icon-theme-1.1.8-2 ------------------------ * Wed Mar 03 2004 Alexander Larsson <alexl(a)redhat.com> 1.1.8-2 - remove redhat-main-menu symlink (#100407) gnome-panel-2.5.91-2 -------------------- * Wed Mar 03 2004 Mark McLoughlin <markmc(a)redhat.com> 2.5.91-2 - Use the main menu icon in the menu bar (#100407) * Tue Mar 02 2004 Elliot Lee <sopwith(a)redhat.com> - rebuilt gpdf-0.123-2 ------------ * Wed Mar 03 2004 Alexander Larsson <alexl(a)redhat.com> 0.123-2 - pre-require scrollkeeper (#116628) * Tue Mar 02 2004 Elliot Lee <sopwith(a)redhat.com> - rebuilt gtk2-2.3.5-1 ------------ * Wed Mar 03 2004 Mark McLoughlin <markmc(a)redhat.com> 2.3.5-1 - Update to 2.3.5 - Bump the required glib and pango versions - Make it build on x86_64 * Tue Mar 02 2004 Elliot Lee <sopwith(a)redhat.com> - rebuilt hpoj-0.91-6 ----------- * Wed Mar 03 2004 Tim Waugh <twaugh(a)redhat.com> 0.91-6 - Updated Till Kamppeter's 2.6 kernel patch. * Tue Mar 02 2004 Elliot Lee <sopwith(a)redhat.com> - rebuilt jpilot-0.99.7-1 --------------- * Wed Mar 03 2004 Than Ngo <than(a)redhat.com> 0.99.7-1 - update to 0.99.7 * Tue Mar 02 2004 Elliot Lee <sopwith(a)redhat.com> - rebuilt kdebase-3.2.0-1.10 ------------------ * Wed Mar 03 2004 Than Ngo <than(a)redhat.com> 6:3.2.0-1.10 - respin kde-redhat-config * Tue Mar 02 2004 Elliot Lee <sopwith(a)redhat.com> - rebuilt * Tue Mar 02 2004 Than Ngo <than(a)redhat.com> 3.2.0-1.9 - don't rename to the contents of the translated Name field - create desktop files if Desktop directory is empty nfs-utils-1.0.6-8.fc2 --------------------- * Mon Mar 01 2004 <SteveD(a)RedHat.com> - Removed gssd tarball and old nfsv4 patch. - Added new nfsv4 patches that include both the gssd and idmapd daemons - Added redhat-only v4 patch that reduces the static librpc.a to only contain gss rpc related routines (I would rather have gssd use the glibc rpc routines) -Changed the gssd svcgssd init scripts to only start up if SECURE_NFS is set to 'yes' in /etc/sysconfig/nfs openssh-3.6.1p2-33 ------------------ * Wed Mar 03 2004 Daniel Walsh <dwalsh(a)redhat.com> 3.6.1p2-33 - Close file descriptors on exec pam-0.77-38 ----------- * Wed Mar 03 2004 Dan Walsh <dwalsh(a)redhat.com> 0.77-38 - Fix error handling of pam_unix * Tue Mar 02 2004 Elliot Lee <sopwith(a)redhat.com> - rebuilt policy-1.7-3 ------------ * Wed Mar 03 2004 Dan Walsh <dwalsh(a)redhat.com> 1.7-3 - Fix creation of aliases.db * Wed Mar 03 2004 Dan Walsh <dwalsh(a)redhat.com> 1.7-2 - Fix mailman, su and other problems. * Tue Mar 02 2004 Dan Walsh <dwalsh(a)redhat.com> 1.7-1 - New version from NSA. rpm-4.3-0.17 ------------ * Tue Mar 02 2004 Jeff Johnson <jbj(a)jbj.org> 4.3-0.17 - permit globs in macrofiles: directive (#117217). - fix: segfault generating transaction serialization lock path. - use /etc/security/selinux/file_contexts instead. * Wed Feb 25 2004 Jeff Johnson <jbj(a)jbj.org> 4.3-0.15 - serialize rpmtsRun() using fcntl on /var/lock/rpm/transaction. rpmdb-fedora-1.90-0.20040304 ---------------------------- squid-2.5.STABLE5-1 ------------------- * Tue Mar 02 2004 Jay Fenlason <fenlason(a)redhat.com> 7:2.5.STABLE5-1 - New upstream version, obsoletes many patches. - Fix --datadir passed to configure. Configure automatically adds /squid so we shouldn't. - Remove the problematic triggerpostun trigger, since is's broken, and FC2 never shipped with that old version. - add -j4 to make line. * Tue Mar 02 2004 Elliot Lee <sopwith(a)redhat.com> - rebuilt * Mon Feb 23 2004 Tim Waugh <twaugh(a)redhat.com> - Use ':' instead of '.' as separator for chown. system-config-services-0.8.7-2 ------------------------------ * Wed Mar 03 2004 Brent Fox <bfox(a)redhat.com> 0.8.7-2 - add a BuildRequires on automake17 * Tue Mar 02 2004 Brent Fox <bfox(a)redhat.com> 0.8.7-1 - remove dependency on gnome-python2 and gnome-python2-canvas - try to load glade file in the cwd, if not, pull from /usr/share/ - apply patch from bug #117277 * Tue Jan 06 2004 Daniel J Walsh <dwalsh(a)redhat.com> 0.8.6-3 - Fix console app so it launches properly vim-6.2.311-1 ------------- * Wed Mar 03 2004 Karsten Hopp <karsten(a)redhat.de> 6.2.311-1 - patchlevel 311

1 0

optparse/Optik
by Toshio Kuratomi 04 Mar '04

04 Mar '04

Just noticed that test1 contains both python-2.3.3 with optparse and python-Optik modules. From my reading, optparse _is_ Optik so the only thing gained from this is being able to 'import Optik' in scripts instead of 'import optparse' (Which is a forward compatibility feature of Optik 1.4.1 too.) Not a big deal, but it may confuse some people new to python if they are trying to decide what module to use to parse their command line arguments. Python and optik pages detailing that Optik was merged into python as optparse: http://www.python.org/doc/2.3.3/whatsnew/node18.html#SECTION000182000000000… http://optik.sourceforge.net/ -Toshio -- Toshio <toshio(a)tiki-lounge.com>

1 0

← Newer
1
...
19
20
21
22
23
24
25
...
28
Older →

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

devel March 2004