devel December 2007

devel@lists.fedoraproject.org

265 participants
294 discussions

Re: f8: NFS mount problem in rescue mode
by Dario Lesca 13 Dec '07

13 Dec '07

Il giorno mer, 12/12/2007 alle 10.27 -0800, Rick Stevens ha scritto: > Can you please specify what "not work" means? What error messages are > you seeing? Start PC with Fedora-8-i386-rescuecd start network = yes IpV4 yes, ipv6 = no, Use DHCP Mount existing filesystem = yes then shell is open ... sh-3.2# ip a s dev eth0 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast qlen 100 link/ether 00:16:36:db:84:63 brd ff:ff:ff:ff:ff:ff inet 10.39.66.82/20 brd 10.39.79.255 scope global eth0 inet6 fe80::216:36ff:fedb:8463/64 scope link valid_lft forever preferred_lft forever sh-3.2# ip r 10.39.64.0/20 dev eth0 proto kernel scope link src 10.39.66.82 default via 10.39.64.10 dev eth0 sh-3.2# ping 10.39.66.100 PING 10.39.66.100 (10.39.66.100) 56(84) bytes of data. 64 bytes from 10.39.66.100: icmp_seq=1 ttl=64 time=0.074 ms 64 bytes from 10.39.66.100: icmp_seq=2 ttl=64 time=0.066 ms 64 bytes from 10.39.66.100: icmp_seq=3 ttl=64 time=0.068 ms 64 bytes from 10.39.66.100: icmp_seq=4 ttl=64 time=0.062 ms --- 10.39.66.100 ping statistics --- 4 packets transmitted, 4 received, 0% packet loss, time 3000ms rtt min/avg/max/mdev = 0.062/0.067/0.074/0.009 ms sh-3.2# ssh 10.39.66.100 The authenticity of host '10.39.64.100 (10.39.64.100)' can't be established. RSA key fingerprint is 3x:4x:2x:x7:dx:x1:4d:c8:16:b4:x7:45:97:d4:8d:a6. Are you sure you want to continue connecting (yes/no)? no sh-3.2# mkdir /tmp/f8 sh-3.2# ls -ld /tmp/f8 drwxr-xr-x 2 root root 4096 13 dic 10:33 /tmp/f8 sh-3.2# mount -tnfs 10.39.66.100:/u/f8 /tmp/f8 (wait 20/30 seconds.....) mount: mounting 10.39.66.100:/u/f8 on /tmp/f8 failed If I "tcpdump -i eth0 host 10.39.66.87" on host 10.39.66.100 I see nothing. This is the error: mount in rescue mode not work. (note which mount is a s-link to busybox) For resolve the problem I must run this command: sh-3.2# /mnt/sysimage/sbin/mount.nfs 10.39.66.100:/u/f8 /tmp/f8 -onolock sh-3.2# df /tmp/f8 Filesystem blocchi di 1K Usati Disponib. Uso% Montato su 10.39.66.100:/u/f8 30726144 28232448 907776 97% /tmp/f8 sh-3.2# sed 1q /tmp/f8/RPM-GPG-KEY The following public key can be used to verify RPM packages built and In this way, my nfs is mounted, but is this the right way to do that? > a) the network isn't started Network is started > b) your IP address isn't set properly IP is set properly > or c) the DNS system isn't responding so you're getting "unknown host" errors. I not use DNS in this case, it's not necessary. > - Rick Stevens Thanks Rick! Now I work around my problem (PXE+kickstart setup) to copy the mount.nfs into f8 structure, then use this command for mount my needed fs. Some other suggest? Thanks to all! -- Dario Lesca <d.lesca(a)solinos.it>

1 0

f8: NFS mount problem in rescue mode
by Dario Lesca 13 Dec '07

13 Dec '07

If I boot Fedora 8 with DVD in rescue mode I can't mount a NFS share. The problem also happened if I use Fedora 7 Respin, but NON if I use Fedora 7 std. or Fedora Core [6-1]. If I boot with options rescue + askmetod then do an NFS install, Anaconda mount /mnt/source correctly, but from shell if I try to mount another NFS share the mount not work All these problem happened also if I work on VMware-server. Some suggest? -- Dario Lesca <d.lesca(a)solinos.it>

3 4

Outage Notification (Buildsystem) - 2008-12-13 03:45 UTC
by Mike McGrath 13 Dec '07

13 Dec '07

There will be an outage starting at 2008-12-13 04:00 UTC, which will last approximately 1 hour. To convert UTC to your local time, take a look at http://fedoraproject.org/wiki/Infrastructure/UTCHowto or run: date -d '2008-12-13 03:45 UTC' Affected Services: Buildsystem Unaffected Services: Websites CVS / Source Control Database DNS Mail Torrent Ticket Link: https://hosted.fedoraproject.org/projects/fedora-infrastructure/ticket/277 Reason for Outage: koji.fedoraproject.org is on a machine that is no longer under warranty and recently had a drive fail. We will be koji to new hardware. NOTE TO PACKAGERS: If you have to build a package that typically takes over an hour to build, please make sure to start your job so it is not still running during the outage. We will be disabling the builders one by one and will try to let builds finish but won't wait forever. Contact Information: Please join #fedora-admin in irc.freenode.net or respond to this email to track the status of this outage. _______________________________________________ Fedora-devel-announce mailing list Fedora-devel-announce(a)redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-announce

2 1

doxygen updates
by Braden McDaniel 13 Dec '07

13 Dec '07

F-8/doxygen.spec is at 1.5.3-1; though it seems no update corresponding to this was built/pushed. How come? Meanwhile, devel/doxygen.spec is at 1.5.4; any reason that can't be pushed to F-8? -- Braden McDaniel e-mail: <braden(a)endoframe.com> <http://endoframe.com> Jabber: <braden(a)jabber.org>

1 0

Re: flash-plugin-9.0.115.0 v. nspluginwrapper
by Louis Garcia 12 Dec '07

12 Dec '07

Just to be clear, nspluginwrapper is useless on i386 systems? This is for x86_64. -Louis

2 1

yum, x86_64 and i386: the never ending story
by Christoph Wickert 12 Dec '07

12 Dec '07

sorry that I bring up this topic once again: why does yum install i386 packages on a x86_64 system during groupinstall? # yum groupinstall XFCE ... Dependencies Resolved ============================================================================= Package Arch Version Repository Size ============================================================================= Installing: Thunar i386 0.8.0-3.fc8 fedora 5.2 M xfce-mcs-plugins x86_64 4.4.1-3.fc8 fedora 593 k xfce-utils x86_64 4.4.1-3.fc8 fedora 346 k xfce4-appfinder x86_64 4.4.1-2.fc8 fedora 264 k xfce4-mailwatch-plugin x86_64 1.0.1-7.fc8 fedora 365 k xfce4-mixer x86_64 4.4.1-3.fc8 fedora 216 k xfce4-session-engines x86_64 4.4.1-2.fc8 fedora 325 k xfprint i386 4.4.1-2.fc8 fedora 584 k xfprint x86_64 4.4.1-2.fc8 fedora 587 k Installing for dependencies: Terminal x86_64 0.2.6-3.fc8 fedora 1.2 M Thunar x86_64 0.8.0-3.fc8 fedora 5.2 M exo i386 0.3.2-3.fc8 fedora 679 k exo x86_64 0.3.2-3.fc8 fedora 681 k libxfce4mcs i386 4.4.1-3.fc8 fedora 46 k libxfce4mcs x86_64 4.4.1-3.fc8 fedora 48 k libxfce4util x86_64 4.4.1-3.fc8 fedora 74 k libxfce4util i386 4.4.1-3.fc8 fedora 73 k libxfcegui4 x86_64 4.4.1-3.fc8 fedora 278 k libxfcegui4 i386 4.4.1-3.fc8 fedora 277 k mousepad x86_64 0.2.12-3.fc8 fedora 89 k xfce-mcs-manager x86_64 4.4.1-3.fc8 fedora 382 k xfce4-panel i386 4.4.1-4.fc8 fedora 530 k xfce4-panel x86_64 4.4.1-4.fc8 fedora 535 k xfce4-session i386 4.4.1-2.fc8 fedora 509 k xfce4-session x86_64 4.4.1-2.fc8 fedora 510 k xfdesktop x86_64 4.4.1-3.fc8 fedora 2.7 M xfwm4 x86_64 4.4.1-3.fc8 fedora 1.3 M To me this looks like xfprint.i386 and Thunar.i386 are installed by accident and then pull in the all the rest. Can anybody explain this to me? I don't think that this behavior is desired by 99 % of the users. Christoph

8 10

libflashsupport - Shouldn't this be libflashsupport-pulse or something like that?
by Kelly Miller 12 Dec '07

12 Dec '07

It just seems to me that there is a good chance for conflict if there is another plugin for Flash sound ever added to the repository; technically, any sound extension is providing libflashsupport... Wouldn't it be better to call the package libflashsupport-pulse and have it "Provides: libflashsupport" or something along those lines?

2 1

Zope/Plone F7+
by Jonathan Steffan 12 Dec '07

12 Dec '07

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Wed, 12 Dec 2007 14:42:14 -0500 "Thomas J. Baker" <tjb(a)unh.edu> wrote: > > Hello, > > What have you been doing for your sites with plone/zope since FC6 is > now EOL'd? I have always had a hard time not pushing updates, I almost pushed Plone 3.0.4 to FC6, but have decided to follow policy on this one. > Do you have unofficial F8 packages anywhere that include a > python-compat or something similar? Yes. I have the full stack needed to run Zope2/Plone for both F7 and F8. I just have compat-python24-setuptools left, which will enable the installation of most any python egg/module/etc (for those with needs outside of the stack I'm willing to maintain.) I'll do this now, as it has needed to be done for some time. The goal is to stick them in RPMFusion. I suppose I should just stick them in Livna for now. They have passed review, I just need to start building. I currently have them in my personal repo, which has a poor upload and I prefer to not host them from there. > > I've had my plone site on a Xen instance of FC6 but F8 Xen is not > running that FC6 instance very well. I'm trying to figure out what to > do next. As indicated below, you have a few options. > > Thanks, > > tjb Some options for Zope/Plone: Run RHEL/Centos 5, EPEL will be updated promptly. I run this for my production sites. Run FC6, rebuilding your own updates from the srpms for EPEL5. Run F7 with the compat stack. Fun F8 with the compat stack, I do this currently. Use the Plone unified installer which includes everything needed, including it's own python. - -- Jonathan Steffan daMaestro GPG Fingerprint: 93A2 3E2F DC26 5570 3472 5B16 AD12 6CE7 0D86 AF59 -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (GNU/Linux) iD8DBQFHYGazrRJs5w2Gr1kRAv/UAJwKyefSDaqi62xnkJSsxDqRNg4oZgCdEvfL hlG7jV1XpxQOyRZSH9ZSWtk= =NbIh -----END PGP SIGNATURE-----

1 0

SDL pulseaudio workaround hack
by Warren Togami 12 Dec '07

12 Dec '07

* Tue Oct 30 2007 Warren Togami <wtogami(a)redhat.com> - 1.2.8-4 - SDL_AUDIODRIVER=esd temporary hack until SDL supports pulseaudio directly avoids applications from locking up. (#358341) At the last minute before F-8's release I added this ugly hack to SDL_mixer to force SDL to use esd for sound instead of the default ALSA. This change was made because it would at least allow sound on popular games like wesnoth to work, and make the Fedora Games spin not dead-on-arrival. Unfortunately, this had the side-effect that we knew at the time of disabling SDL sound if you have pulseaudio disabled or removed. I should have thought of this earlier, but the /etc/profile.d/* scripts that set the environment variable commanding SDL to use esd could have been conditional. Still not perfect, but better to tide us over until Lennart implements native pulseaudio for SDL that he promised to do before F9. The following changes should improve our workaround. I am uncertain if the .csh version needs a matching unset for nonomatch or does it happen automatically? Should we push this in a SDL_mixer update? Warren Togami wtogami(a)redhat.com cvs diff: Diffing . Index: SDL_pulseaudio_hack.csh =================================================================== RCS file: /cvs/pkgs/rpms/SDL_mixer/devel/SDL_pulseaudio_hack.csh,v retrieving revision 1.1 diff -u -p -r1.1 SDL_pulseaudio_hack.csh --- SDL_pulseaudio_hack.csh 31 Oct 2007 14:39:10 -0000 1.1 +++ SDL_pulseaudio_hack.csh 27 Nov 2007 01:33:43 -0000 @@ -1,2 +1,4 @@ # Temporary hack until SDL directly supports pulseaudio -setenv SDL_AUDIODRIVER esd +# If alsa-plugins-pulseaudio is installed, force SDL to output sound to esd +set nonomatch +if ( -e /usr/lib*/alsa-lib/libasound_module_pcm_pulse.so ) setenv SDL_AUDIODRIVER esd Index: SDL_pulseaudio_hack.sh =================================================================== RCS file: /cvs/pkgs/rpms/SDL_mixer/devel/SDL_pulseaudio_hack.sh,v retrieving revision 1.1 diff -u -p -r1.1 SDL_pulseaudio_hack.sh --- SDL_pulseaudio_hack.sh 31 Oct 2007 14:39:10 -0000 1.1 +++ SDL_pulseaudio_hack.sh 27 Nov 2007 01:33:43 -0000 @@ -1,2 +1,3 @@ # Temporary hack until SDL directly supports pulseaudio -export SDL_AUDIODRIVER=esd +# If alsa-plugins-pulseaudio is installed, force SDL to output sound to esd +[ -e /usr/lib*/alsa-lib/libasound_module_pcm_pulse.so ] && export SDL_AUDIODRIVER=esd

3 3

Argyll color management system integration
by Nicolas Mailhot 12 Dec '07

12 Dec '07

Hi all, The FLOSS color management scene (calibrating screens for pictures or video) is very small. In fact there is only one reference application every other project depends on or compares itself to: Argyll CMS. Argyll is free, supports many common (commercial and do-it-yourself) color meters, has been actively developed for more than a decade. It was on the Fedora package wishlist. Unfortunately its main author is not following best release practises and has chosen to use a non-standard unmaintained make replacement (jam), with static linking, private copies of common libraries, and all kinds of things distributions are not fond of. As a result no first-level distribution shipped Argyll. Despite the fact every color management HOWTO on the net points to it. I started untangling the mess in October but gave up after a frustrating week. Fortunately I had the good idea to put the result on fedorapeople, and lately Frédéric Crozat from Mandriva googled it and completed the work. On hindsight it seems I had stopped just short of the finish line. http://twinpeaks.dyndns.org/blog/general/2007/12/11/monitor-calibration-epi… Since it'd be a shame to let my packaging work end up in Mandriva only, I picked up his package, changed back a few bits, added some stuff I had planned but not done yet, and pushed the result in the review queue: https://bugzilla.redhat.com/show_bug.cgi?id=421921 I'd appreciate if a friendly reviewer looked at the result. The following work has been done: – link against system libs, not built-in copies – build with modular X – remove the shell wrapper that was hiding build errors from rpm – reorder build logic to fix those errors – hal/pam/udev logic so colorimeters do not need root access The result is IMHO good enough to be merged. However, since this software has suffered from a secluded life due to the aforementioned problems, it would probably be a good idea if more clueful people than me checked the following points: – go over the build logs, check if the warnings are really harmless and fix the code if needed (where is the list of people that wanted to do this kind of work a few months ago?) http://nim.fedorapeople.org/argyllcms/build.log I doubt argyllcms was ever built with a modern gcc with all the fancy options we use and for all the architectures we target – check if I got the hal/pam/udev logic right, push bits upstream if needed (I didn't find two Fedora packages that did it the same way, all I can say the way I did it works on my system) – check if upstream didn't mess up licensing as it did the code Then we'll have one key app for photographers and video buffs nailed, and there will only be the matter of pushing fixes upstream. Regards, -- Nicolas Mailhot

3 4

← Newer
1
...
14
15
16
17
18
19
20
...
30
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 December 2007