Re: [arch-general] What happened to Powerpill?
On Tue, Apr 5, 2011 at 8:46 AM, Madhurya Kakati mkakati2...@gmail.com wrote: Please, someone restart powerpill development. It was awesome. I had to remove it in order to upgrade to pacman 3.5 but now I can't find it anywhere. The developer's post on the forums indicates that he's been working on a far better replacement, but doesn't have the free time right now to complete it. Perhaps his priorities could be changed if some cash was on the table, perhaps not.
Re: [arch-general] base stuff
Am 07.04.2011 04:36, schrieb Thomas S Hatch: I like to hear that Tom! Unfortunately many people think that having SELinux compiled in means that it is running, having SELinux compiled into the core utils and the kernel but leaving it turned off has 0 negative effect on the system. If that just were true. I think we compiled SELinux into our kernel once or twice, just to remove it again. It caused random stuff to fail even though it was not enabled - at least this is what I remember (I'd have to look up the ML threads related to this). signature.asc Description: OpenPGP digital signature
Re: [arch-general] base stuff (was: Change Arch's default crond)
On Wed, 6 Apr 2011 21:22:14 -0600 Thomas S Hatch thatc...@gmail.com wrote: I think that Arch would benefit from inducing SELinux as an option because it expands the venues available for Arch Linux systems, I also think that inclusion in base of SELinux requires a minimal amount of maintenance and SELinux is completely non-intrusive if it is disabled. It's not because some folks might benefit from SELinux, that all of a sudden it should go into base. That's really not what base is for. Dieter
Re: [arch-general] Change Arch's default crond
Am 07.04.2011 04:30, schrieb Thomas S Hatch: Right, both are viable choices, btw I will be migrating my datacenters away from dcron in the near future and doing a series of tests on cronie and fcron, I will post my findings to the list. I think that will be more valuable than any continuation of this discussion, and it would be very much appreciated. Right now, cronie looks like a better candidate, although I am still a fan of fcron. signature.asc Description: OpenPGP digital signature
Re: [arch-general] base stuff (was: Change Arch's default crond)
On Thursday 07 April 2011 00:25:42 Thomas S Hatch wrote: As for adding SELinux support in base but keeping it turned off by default, +1 +1 -- Divan Santana
Re: [arch-general] base stuff (was: Change Arch's default crond)
On Thursday 07 April 2011 00:25:42 Thomas S Hatch wrote: As for adding SELinux support in base but keeping it turned off by default, +1 +1 -- Divan Santana
Re: [arch-general] Change Arch's default crond
Am Wed, 6 Apr 2011 22:03:23 -0600 schrieb Thomas S Hatch thatc...@gmail.com: I would say that we should consider compatibility with vixie cron syntax - this is and has been the expected syntax for the default cron daemon for a LONG time and avoids hindering Arch Linux adoption. Why do you need vixie cron syntax? Can't you migrate once to a new syntax? Btw., most of fcron's syntax is the same as the syntax of every cron daemon. You can easily take your previous crontabs. You probably have only to do some changes which could most likely be done by sed. In first place you need stability, reliability and useful features. And this is given by fcron. And don't tell me anything about compatibility. I would consider this argument if fcron was a new cron daemon and if it was totally incompatible. Fcron is known since years and it's known to be stable and reliable, while cronie seems to be pretty new. There are absolutely no informations about cronie's features, no documentations, no feature comparisons to other cron daemons, etc. on upstream's website. And it's still in AUR and has only 3 votes there. I spent quite some time as a trainer for Red Hat and taught classes on SELinux. Is this why you want to push cronie so heavily? Heiko
Re: [arch-general] Change Arch's default crond
Am Wed, 6 Apr 2011 20:30:46 -0600 schrieb Thomas S Hatch thatc...@gmail.com: Right, both are viable choices, btw I will be migrating my datacenters away from dcron in the near future and doing a series of tests on cronie and fcron, I will post my findings to the list. Data center? So the systems are running 24/7? But you take into consideration that there are home users whose computers are not running 24/7 and that they also need to have their cron tasks run reliably? So I hope you test this, too. See bootrun. Heiko
Re: [arch-general] Change Arch's default crond
On Thu, Apr 7, 2011 at 2:16 AM, Thomas Bächler tho...@archlinux.org wrote: Am 07.04.2011 04:30, schrieb Thomas S Hatch: Right, both are viable choices, btw I will be migrating my datacenters away from dcron in the near future and doing a series of tests on cronie and fcron, I will post my findings to the list. I think that will be more valuable than any continuation of this discussion, and it would be very much appreciated. Right now, cronie looks like a better candidate, although I am still a fan of fcron. Thanks Thomas, as you know I am very busy, but I will post my findings when I am done. As always, I hope I can help Arch get better!
[arch-general] Possibility of including xorg input patch
Can this patch https://bugs.freedesktop.org/attachment.cgi?id=33142 be applied to Arch's xorg? It solves famous problem https://bugs.freedesktop.org/show_bug.cgi?id=865 with broken keyboard input for users who use multiple keyboard layouts without any noticeable side-effects. It's known this patch has been applied in Ubuntu. -- Kirill Churin Jabber: reflex...@reflexing.ru, ICQ: 8163230, Skype on demand.
Re: [arch-general] Possibility of including xorg input patch
On 07.04.2011 17:05, Kirill Churin wrote: Can this patch https://bugs.freedesktop.org/attachment.cgi?id=33142 be applied to Arch's xorg? Has upstream accepted that patch? -- Florian Pritz -- {flo,bluewind}@server-speed.net signature.asc Description: OpenPGP digital signature
Re: [arch-general] Possibility of including xorg input patch
On Thu, Apr 7, 2011 at 5:13 PM, Florian Pritz bluew...@server-speed.netwrote: Has upstream accepted that patch? The corresponding bug report is: https://bugs.freedesktop.org/show_bug.cgi?id=865 It seems there is still some work to be done before the patch can be accepted upstream. -- Cédric Girard
Re: [arch-general] Change Arch's default crond
On Wednesday, April 06, 2011 15:27:27 Thomas Bächler wrote: Am 05.04.2011 09:19, schrieb Thomas S Hatch: I can think of three considerations for a cron daemon: 1 . Minimal - its a cron daemon, it does not need to be complex 2. Active development 3. Anacron functionality As far as I can see this leaves us with fcron, dcron and cronie. Cronie probably has the highest assurance for upstream development because it is backed by redhat. But I think that having a cron daemon as default that has issues executing jobs on time and as they are defined is highly questionable. Before the current maintainer took over dcron, we had that same discussion. Aaron even contacted the fcron maintainer (he posted the reply to arch-general or arch-dev-public, if anyone could find the link in the archives, please post it). The author responded that he considered fcron feature-complete, so didn't develop it anymore. However, he would fix bugs when they are reported, and I think there are no known bugs right now. That said, fcron lacks /etc/cron.d/ functionality which was the most important argument against it. I personally don't need that and I like fcron a lot. As for your conditions: 1) It is very small software, 1.2MB installed, and it has lots of features. It is by no means minimal though. 2) I commented on that above. 3) dcron has @daily, @hourly and so on. In fcron, you can use standard crontab entries and add bootrun to the beginning of the line to repeat missed cronjobs. I don't know cronie, so maybe you can elaborate more. Losing /etc/cron.d support is a bit of a dealbreaker for me. I think that's a rather huge feature to leave out of a crond.
Re: [arch-general] base stuff (was: Change Arch's default crond)
On Wednesday, April 06, 2011 18:13:04 Grigorios Bouzakis wrote: Thomas S Hatch thatc...@gmail.com wrote: On Wed, Apr 6, 2011 at 4:16 PM, Grigorios Bouzakis grb...@xsmail.comwrote: Thomas S Hatch wrote: I am saving the include SELINUX support in base for a latter date my understanding though is that the stated position of Arch was no systemd s/was/is/g That is also my understanding in regards to selinux. Although i am not familiar with stated positions about either. PS. Ntp is fine application that will keep your clock synchronised. It seems to be 5 days off. :) Yes the systemd topic keeps popping up, right now we don't know if certain upstream changes are going to force Arch into using systemd or not. I dont think such a topic keeps popping up. In fact I dont remember reading a discussion between Arch developers about it, ever. I could probably go on ranting about stuff thats been shoved down users mouths the last years for months but its futile and a waste of time. It was a discussion that popped up here, a debate between users who felt replacing sysvinit was completely unneeded to those who seemed to want to use systemd for some useless, unneeded feature maybe less than 1% of Arch users were going to actually use. As for adding SELinux support in base but keeping it turned off by default, +1 Although this isnt a vote, mine was for no selinux at all, so its just 1. :) Selinux is another unneeded thing, but even worse is that it practically requires a doctorate in computer science to manipulate. Can't deny its security, though. +1 to leaving it out of Arch, not that anyone's asking Arch to.
Re: [arch-general] Fwd: Re: [hylafax-users] Error building 6.0.5 on Arch Linux /sys/stat.h:299:47: error: declaration of ‘int fchmod(int, __mode_t) throw ()’ throws different exceptions
On 04/01/2011 09:13 AM, Jelle van der Waa wrote: It's probably going to be build with gcc4.6 ;) Jelle, When the next version is built, we need to take a look at the file ownership for the files under /var/spool/hylafax. Currently the files are owned by root or uucp. However, we have a 'fax' user. There have been a few issues on Arch with trouble getting the faxrcvd and notify executables to work with avantfax, etc. Looking at other distros, the ownership of the files under /var/spool/hylafax is fax:uucp. Seeing that the files are world executable -- it shouldn't make a difference, but I can't help but wondering if something is checking ownership expecting to find the files owned by the 'fax' user and then not executing the faxrcvd or notify commands as a result. Dunno -- just food for thought on the next rebuild :p -- David C. Rankin, J.D.,P.E.
Re: [arch-general] Fwd: Re: [hylafax-users] Error building 6.0.5 on Arch Linux /sys/stat.h:299:47: error: declaration of ‘int fchmod(int, __mode_t) throw ()’ throws different exceptions
On 04/07/2011 11:41 AM, David C. Rankin wrote: but I can't help but wondering if something is checking ownership expecting to find the files owned by the 'fax' user and then not executing the faxrcvd or notify commands as a result. It's not just ownership of the executables, but also the files in /var/spool/hylafax/recvq, /var/spool/hylafax/doneq, It may present a read permission problem if owned by uucp:uucp, especially where ../hylafax/etc/config.ttyS1 has RecvFileMode: 0640 or 0600. Again, just food for thought... I'll try and learn more about what user hylafx is running under for the various tasks. Right now, I just don't know. -- David C. Rankin, J.D.,P.E.
Re: [arch-general] base stuff (was: Change Arch's default crond)
On Thursday 07 April 2011 18:08:39 Yaro Kasear wrote: Selinux is another unneeded thing, but even worse is that it practically requires a doctorate in computer science to manipulate. Can't deny its security, though. +1 to leaving it out of Arch, not that anyone's asking Arch to. I don't see the harm in have proper SELinux support. It would make it easier to learn SELinux if one could have it officially supported and working well on everyone's Arch laptop. Plus added security is a good practice. Not sure why having official SELinux should stir some to vote against it when one can easily disable it. -- Divan Santana
Re: [arch-general] base stuff
Yaro Kasear wrote: On Wednesday, April 06, 2011 18:13:04 Grigorios Bouzakis wrote: Thomas S Hatch thatc...@gmail.com wrote: Yes the systemd topic keeps popping up, right now we don't know if certain upstream changes are going to force Arch into using systemd or not. I dont think such a topic keeps popping up. In fact I dont remember reading a discussion between Arch developers about it, ever. I could probably go on ranting about stuff thats been shoved down users mouths the last years for months but its futile and a waste of time. It was a discussion that popped up here, a debate between users who felt replacing sysvinit was completely unneeded to those who seemed to want to use systemd for some useless, unneeded feature maybe less than 1% of Arch users were going to actually use. I guess you mean http://thread.gmane.org/gmane.linux.arch.general/32759 Didnt enjoy skimming through it much, except maybe this part: http://article.gmane.org/gmane.linux.arch.general/32874 And indeed, it was just user talk. The only developer who got even remotely interested in participating got flamed. As for adding SELinux support in base but keeping it turned off by default, +1 Although this isnt a vote, mine was for no selinux at all, so its just 1. :) Selinux is another unneeded thing, but even worse is that it practically requires a doctorate in computer science to manipulate. Can't deny its security, though. +1 to leaving it out of Arch, not that anyone's asking Arch to. All these seem like the natural sideffects of Arch's growth along with the (d)evolution (as in degeneration) of Linux. Greg
Re: [arch-general] base stuff
On Thu, Apr 7, 2011 at 12:51 PM, Grigorios Bouzakis grb...@xsmail.comwrote: Yaro Kasear wrote: On Wednesday, April 06, 2011 18:13:04 Grigorios Bouzakis wrote: Thomas S Hatch thatc...@gmail.com wrote: Yes the systemd topic keeps popping up, right now we don't know if certain upstream changes are going to force Arch into using systemd or not. I dont think such a topic keeps popping up. In fact I dont remember reading a discussion between Arch developers about it, ever. I could probably go on ranting about stuff thats been shoved down users mouths the last years for months but its futile and a waste of time. It was a discussion that popped up here, a debate between users who felt replacing sysvinit was completely unneeded to those who seemed to want to use systemd for some useless, unneeded feature maybe less than 1% of Arch users were going to actually use. I guess you mean http://thread.gmane.org/gmane.linux.arch.general/32759 Didnt enjoy skimming through it much, except maybe this part: http://article.gmane.org/gmane.linux.arch.general/32874 And indeed, it was just user talk. The only developer who got even remotely interested in participating got flamed. As for adding SELinux support in base but keeping it turned off by default, +1 Although this isnt a vote, mine was for no selinux at all, so its just 1. :) Selinux is another unneeded thing, but even worse is that it practically requires a doctorate in computer science to manipulate. Can't deny its security, though. +1 to leaving it out of Arch, not that anyone's asking Arch to. All these seem like the natural sideffects of Arch's growth along with the (d)evolution (as in degeneration) of Linux. Greg Thanks for the link, I did not want to bring up SELinux yet, because I will not be getting to it for a few months, but this will help a great deal! -Thomas S Hatch
Re: [arch-general] base stuff
Thomas S Hatch wrote: On Thu, Apr 7, 2011 at 12:51 PM, Grigorios Bouzakis grb...@xsmail.comwrote: I guess you mean http://thread.gmane.org/gmane.linux.arch.general/32759 Thanks for the link, I did not want to bring up SELinux yet, because I will not be getting to it for a few months, but this will help a great deal! -Thomas S Hatch Glad i could help :) Greg
Re: [arch-general] Change Arch's default crond
On Thursday 07 of April 2011 12:32:50 Heiko Baums wrote: Am Wed, 6 Apr 2011 22:03:23 -0600 I spent quite some time as a trainer for Red Hat and taught classes on SELinux. Is this why you want to push cronie so heavily? Heiko Sorry to sound rude, but Heiko, it's you who is pushing fcron so unhealthily heavily. I wouldn't have no opinion on the two crons but after reading the discussion I'd stick to cronie. Just my 2c. Cheers, mark -- Marek Otahal :o) signature.asc Description: This is a digitally signed message part.
Re: [arch-general] Change Arch's default crond
Am Thu, 7 Apr 2011 21:53:58 +0200 schrieb Marek Otahal markota...@gmail.com: Sorry to sound rude, but Heiko, it's you who is pushing fcron so unhealthily heavily. I wouldn't have no opinion on the two crons but after reading the discussion I'd stick to cronie. Just my 2c. Well, on the one hand yes, on the other hand no. But let's try to get objective again. dcron: Known to be buggy. fcron: Known to work reliably since years, works for 24/7 servers as well as for home desktop computers, features are known and well documented on upstream's website. cronie: Quite new, not known well, if at all, and at least I can't find neither a feature list nor a documentation on upstream's website. Doesn't seem to work for home desktop computers, at least not as easy as fcron (separate crontab and anacrontab), from what I read in this thread and from the very few descriptions on upstream's website. A feature comparison between all the cron daemons, at least between fcron and cronie would be nice. That are my concerns. Again, it's just a question about the default cron daemon, not the one and only in the repos. I wouldn't care, btw., if cronie would go into the repos, too, even if it has only 3 votes in AUR, yet. On the other hand this issue could be solved in a different way without any further discussions. There's a need for installing one cron daemon, but no need for a default cron daemon. It's pretty the same issue as with the bootloaders. There's no default bootloader anymore and currently it doesn't make sense anymore to define one bootloader as the default, because they all have different features and it depends on the system configuration which bootloader is the best. In the development isos AIF asks the user to choose one of currently two bootloaders (grub and syslinux), more (grub2 and lilo) could or should be added. And this bootloader is automatically chosen by AIF in the package selection. The same could be made for the cron daemons. Put every cron daemon into [core] and let the user choose his preferred cron daemon during installation. But if this shouldn't be done, and if there shall be a default cron daemon it must be a daemon which fulfills every use case and not only the needs of servers which are running 24/7. Heiko
Re: [arch-general] Change Arch's default crond
Am Thu, 7 Apr 2011 23:16:46 +0200 schrieb Heiko Baums li...@baums-on-web.de: But let's try to get objective again. Btw., generally it doesn't really matter that much which cron daemon is installed by AIF. Another cron daemon can easily be installed afterwards. A cron daemon is not such an important and sensible software as a bootloader. So the question about a default cron daemon is rather a question of user-friendliness. Heiko
Re: [arch-general] Change Arch's default crond
On Thu, Apr 7, 2011 at 6:32 AM, Heiko Baums li...@baums-on-web.de wrote: Why do you need vixie cron syntax? Can't you migrate once to a new syntax? Btw., most of fcron's syntax is the same as the syntax of every cron daemon. You can easily take your previous crontabs. You probably have only to do some changes which could most likely be done by sed. In first place you need stability, reliability and useful features. And this is given by fcron. And don't tell me anything about compatibility. I would consider this argument if fcron was a new cron daemon and if it was totally incompatible. Fcron is known since years and it's known to be stable and reliable, while cronie seems to be pretty new. There are absolutely no informations about cronie's features, no documentations, no feature comparisons to other cron daemons, etc. on upstream's website. And it's still in AUR and has only 3 votes there. The thing is that cronie is forked from vixie-cron which is much older than fcron. And I would venture to say that vixie-cron or some derivative is the default crond for the vast majority of distributions out there. --Kaiting. -- Kiwis and Limes: http://kaitocracy.blogspot.com/
Re: [arch-general] Change Arch's default crond
On 08.04.2011 00:15, Kaiting Chen wrote: On Thu, Apr 7, 2011 at 6:32 AM, Heiko Baums li...@baums-on-web.de wrote: Why do you need vixie cron syntax? Can't you migrate once to a new syntax? Btw., most of fcron's syntax is the same as the syntax of every cron daemon. You can easily take your previous crontabs. You probably have only to do some changes which could most likely be done by sed. In first place you need stability, reliability and useful features. And this is given by fcron. And don't tell me anything about compatibility. I would consider this argument if fcron was a new cron daemon and if it was totally incompatible. Fcron is known since years and it's known to be stable and reliable, while cronie seems to be pretty new. There are absolutely no informations about cronie's features, no documentations, no feature comparisons to other cron daemons, etc. on upstream's website. And it's still in AUR and has only 3 votes there. The thing is that cronie is forked from vixie-cron which is much older than fcron. And I would venture to say that vixie-cron or some derivative is the default crond for the vast majority of distributions out there. --Kaiting. cronie also appears to be the nicest migration choice for users who are not used to fcron. It seems to support anachron features, cron.d, daily/weekly/etc, is able to actually keep time and works just like expected whereas fcron has fcrontab with a slightly different syntax. We could actually make cronie replace dcron. fcron would be nice but it is not a drop in like cronie. What do you say? If you agree, I shall make (or somebody who steps up) a plan to the replacement and that's that. -- Sven-Hendrik
Re: [arch-general] Change Arch's default crond
cronie also appears to be the nicest migration choice for users who are not used to fcron. It seems to support anachron features, cron.d, daily/weekly/etc, is able to actually keep time and works just like expected whereas fcron has fcrontab with a slightly different syntax. We could actually make cronie replace dcron. fcron would be nice but it is not a drop in like cronie. What do you say? If you agree, I shall make (or somebody who steps up) a plan to the replacement and that's that. -- Sven-Hendrik I agree with Sven, the more I look into it the more I think that cronie is the right way to go for a default cron daemon. With that said, the more I have learned about fcron the more I have liked it! +1
Re: [arch-general] Change Arch's default crond
Kaiting Chen wrote: The thing is that cronie is forked from vixie-cron which is much older than fcron. And I would venture to say that vixie-cron or some derivative is the default crond for the vast majority of distributions out there. --Kaiting. Why do you have --disable-anacron in the build in the AUR? That probably makes it as useless as vixie-cron or the old dcron (3.x versions). Do you use a seperate anacron? Is there something wrong with cronie's anacron? Greg
Re: [arch-general] [arch-announce] GNOME3 in [testing]
On 04/07/2011 03:01 PM, Arch Linux: Recent news updates: Ionuț Mircea Bîru wrote: Ionuț Mircea Bîru wrote: GNOME 3.0.0 packages are now available in the [testing] repository. These bring with it an update to gtk2, as well as the new gtk3. This is a major update and you should take note of a couple of things: * GNOME3 will replace GNOME2 once it gets moved to [extra]. * GNOME3 has two modes, standard mode (gnome-shell) and fallback mode (gnome-panel + metacity). * Panel applets using Bonobo aren't supported anymore and packages depending on it will be dropped. * pulseaudio is now required to run the GNOME desktop. * Some packages exist in separate versions for gtk2 and gtk3. These typically have a name like packagename3. Examples are vte3, libwebkit3, gtkhtml4. * pygobject is now available for Python 3 in the package py3gobject. Have fun testing these packages! Update and installation instructions are available at [https://wiki.archlinux.org/index.php/GNOME_3][1]. Bugs related to packaging should be reported to [http://bugs.archlinux.org/][2] . Crashes and feature requests should be reported to [https://bugzilla.gnome.org/][3] . [1]: https://wiki.archlinux.org/index.php/GNOME_3 (gnome 3 wiki) [2]: http://bugs.archlinux.org/ (arch linux bugtracker) [3]: http://bugs.archlinux.org (gnome bugtracker) URL: http://www.archlinux.org/news/gnome3-in-testing/ ___ arch-announce mailing list arch-annou...@archlinux.org http://mailman.archlinux.org/mailman/listinfo/arch-announce Is pulseaudio mandatory no matter what? Can it be removed if some of the packages are built custom? I like having volume control for HDMI out. And I like that this mysterious starving audio, playing silence, glitch has disappeared. But I don't like that pulse uses so much CPU and i think it's a major reason why my computer is runn 5-8 degrees C hotter with gnome3
Re: [arch-general] Change Arch's default crond
On Fri, 08 Apr 2011 00:34:42 +0200 Sven-Hendrik Haase wrote: cronie also appears to be the nicest migration choice for users who are not used to fcron. It seems to support anachron features, cron.d, daily/weekly/etc, is able to actually keep time and works just like expected whereas fcron has fcrontab with a slightly different syntax. We could actually make cronie replace dcron. I agree with this instead i'm a fcron user. But because it is say so much often, is there really anyone who have something in /etc/cron.d? See you, Attila