Re: concept of package ownership
On Tue, Jul 06, 2010 at 09:35:32AM -0700, Jesse Keating wrote: The system is fairly open with regard to just about everything except attitude. Currently it's mostly attitude that prevents openness. The ACL system restrict changes to other people packages to provenpackagers. And then the policies restrict a lot what provenpackages can do. So, the system is not very open, at least very unlike what Adam described. -- Pat -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 7/13/10 2:43 AM, Patrice Dumas wrote: On Tue, Jul 06, 2010 at 09:35:32AM -0700, Jesse Keating wrote: The system is fairly open with regard to just about everything except attitude. Currently it's mostly attitude that prevents openness. The ACL system restrict changes to other people packages to provenpackagers. And then the policies restrict a lot what provenpackages can do. So, the system is not very open, at least very unlike what Adam described. Right, the attitude has led to policy of what provenpackagers can do, but the underlying system is open. All it takes is an attitude change as nothing else is preventing provenpackagers from doing more. - -- Jesse Keating Fedora -- Freedom² is a feature! identi.ca: http://identi.ca/jkeating -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.9 (Darwin) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAkw8jdUACgkQ4v2HLvE71NUyggCgrJOtrWOEry+OUh6h6tRn5y1w RaEAn37hTdMBziF/Zb8+3DQJI4XQNLFy =hhxu -END PGP SIGNATURE- -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On Mon, Jul 05, 2010 at 01:30:37PM -0700, Adam Williamson wrote: This generally works out pretty well, and helps out with the problem of having quite a small set of maintainers for an extremely large set of packages. I was often in the situation where I happened to notice a small issue with 'someone else's' package and could just go ahead and fix it, instead of having to go through the bureaucracy of filing a bug report and waiting for them to do it. It's rarely the case that someone makes a really stupid change and causes friction. I'd say the system works more often than it doesn't, and it'd probably be good for Fedora too to - as Dave proposes - explicitly _not_ have a concept of ownership, and be more liberal about non-maintainers touching packages. If I recall well, historically, both in fedora extra and fedora core commit rights were pretty liberal. After the merge there was the provenpackagers set up and some packages are more or less protected. But I don't think that what matters is the ACL system, more interesting are the policies and how things are done and why. As you say above the open system fits well 'having quite a small set of maintainers for an extremely large set of packages'. But this is not the case for all of fedora. More precisely, with a bit exageration, there seems to be 4 sets of people. 1) The dedicated community packager works benevolently (though he may be paid for that work and, for example, be a redhat emplyee, it wouldn't matter if he wasn't) and really takes care of his packages. It corresponds, in my opinion to most of former fedora extra packagers and most people at redhat that are hired from the community. For that packager it is important to have people avoid touching his package, but he won't mind if things that are obviously broken are taken care of, especially when he is not available, and he generally have co-maintainers to do the job anyway. (You can guess that I am biased, and find that I am myself in that category). 2) The forced packager has to maintain packages as part of his job. This holds for some people @rh, but not only. For the packages maintained by those packagers, it is better if they are open since they tend not to be very dedicated. Some of them may be picky, however, since they have to make as if they were maintaining their packages, even if they are doing a poor job. Being forced to own a package doesn't mean that, necessarily, the packager will do a poor job, but this is a possibility. 3) The diletante packager is a community packager who steps to do some packaging but after some short time stops without any reason or explanation. I still haven't understood where those come from, what their motivations are, but there are quite a bit in fedora. For the corresponding packages, obviously, having an open system is good. (But having a sponsoring model that avoids them is also good...). 4) The dedicated non community packager is a specie disappearing from fedora, corresponding more or less with packagers, in general redhat old timers who are dedicated but prefer doing things their way. They obviously prefer closed packages. Historically there was a majority of packagers of type 1) in fedora extras and a mix, with, in my opinion, a majority of type 2) and 4) in fedora core. So something closed was more logical, and it was certainly quite different from mandriva. This has moved over the years, with people changing in categories, with some moves in the right direction, mainly from packagers in 2) and 4) going to 1), but, more importantly, there has been people leaving and other entering. Overall, it seems to me that the 2) and 3) are winning over in numbers (category 4) is almost extinct), especially in number of packages. Maybe Fedora should do a transition to a more open system, since the dedicated packager is less present nowadays. But it should be done carefully, in order not to piss off the remaining dedicated packagers, who are those who deliver, in my opinion, the packages with highest quality. -- Pat -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On Sat, Jul 03, 2010 at 03:28:31PM +0200, Kevin Kofler wrote: I thought rawhide should be more useful and less broken if i recall the latest threads right. Anyways, exactly that's why i do *not* want anybody can do anything with any package. That's just insane, sorry. This is Fedora. Debian is that http://www.debian.org/ way. Please don't destroy what Fedora is all about. If we don't focus on packaging the latest software anymore, we will just be another Debian or Ubuntu. There _is_ a middle ground between bleeding edge and extremely stable. A Fedora release should have a locked version of key shared packages, such as Python, Rails, etc., should be kept at a specific version (with upgrades only for bug fixes). Packages within a release can be upgraded so long as they don't require the next version of those shared packages. So, if F13 is Rails 2.3.5 and Rake 0.8.7, then an app that requires newer versions of either should wait until F14 to push _that_ update out, and not go to F13 at all. Especially since even a minor upgrade of such a shared component can break a lot of apps. -- Darryl L. Pierce, Sr. Software Engineer @ Red Hat, Inc. Delivering value year after year. Red Hat ranks #1 in value among software vendors. http://www.redhat.com/promo/vendor/ pgpyq7gr7Q6u5.pgp Description: PGP signature -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 7/6/10 2:16 AM, Patrice Dumas wrote: Maybe Fedora should do a transition to a more open system, since the dedicated packager is less present nowadays. But it should be done carefully, in order not to piss off the remaining dedicated packagers, who are those who deliver, in my opinion, the packages with highest quality. The system is fairly open with regard to just about everything except attitude. Currently it's mostly attitude that prevents openness. - -- Jesse Keating Fedora -- Freedom² is a feature! identi.ca: http://identi.ca/jkeating -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.9 (Darwin) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAkwzW1IACgkQ4v2HLvE71NUGigCgkVZUzkxqfhhcS95PbwEpjLxz 9noAn2c7LTeD2CWs6QNhxezm+rXaC07L =RNzg -END PGP SIGNATURE- -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
Nils Philippsen wrote: AIUI, a SIG are more people than those who actually work on related packages as maintainers, or are competent and responsible enough to not break things in the process of updating packages with which they're not familiar (otherwise they'd be (co-)maintainers, wouldn't they?). If we ever get group ACLs, I think we should have $SIG-packager groups, consisting of SIG members who fulfill the above, who get that kind of access. That's true, not all SIG members are packagers, only those that actually are should get packaging access. (As for what packagers to trust with such access, that's really the individual SIG's business. I'm personally a defender of the give everyone the benefit of the doubt, but yell loudly and threaten revoking access if they screw up approach. :-) ) Kevin Kofler -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
Darryl L. Pierce wrote: There _is_ a middle ground between bleeding edge and extremely stable. A Fedora release should have a locked version of key shared packages, such as Python, Rails, etc., should be kept at a specific version (with upgrades only for bug fixes). Well, I don't know how Rails works, but for Python, yes, of course we don't want Python upgraded to an ABI-incompatible version (say, from 2.6 to 2.7) in updates! If you thought I was asking for that kind of updates (which require rebuilding half of the distro!), you misunderstood me! On the other hand, point releases (as in 2.6.n to 2.6.n+1) ARE only bug fixes and as such SHOULD get pushed. Kevin Kofler -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On Sat, 2010-07-03 at 03:34 +0200, Kevin Kofler wrote: Ralf Corsepius wrote: We need groups, with grouped privileges/acls etc. It's essentially what e.g. the perl-sig originally was meant to be. Yes, group ACLs are definitely needed, but in addition to that technical feature, we also need to make sure that the SIG actually gets commit access to ALL packages related to the SIG. ALL perl-* packages should be committable to by the Perl SIG. That's what a SIG is for. And it'd have prevented this whole why were X and Y added as maintainers to my perl-* packages fiasco, it would just have been the SIG's decision to add the people to the SIG and this should automatically give commit access to all perl-* packages. Similarly, all packages using Qt should be committable to by the KDE SIG etc. AIUI, a SIG are more people than those who actually work on related packages as maintainers, or are competent and responsible enough to not break things in the process of updating packages with which they're not familiar (otherwise they'd be (co-)maintainers, wouldn't they?). If we ever get group ACLs, I think we should have $SIG-packager groups, consisting of SIG members who fulfill the above, who get that kind of access. Nils -- Nils Philippsen Those who would give up Essential Liberty to purchase Red Hat a little Temporary Safety, deserve neither Liberty n...@redhat.com nor Safety. -- Benjamin Franklin, 1759 PGP fingerprint: C4A8 9474 5C4C ADE3 2B8F 656D 47D8 9B65 6951 3011 -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On Fri, 2010-07-02 at 03:18 +0200, Kevin Kofler wrote: I think we need to get rid of the concept of ownership entirely, that'd also make orphaned or de-facto orphaned packages less of a problem. You see a problem, you fix it. Who cares whether the package has an active maintainer or not? I think there's a reasonable middle ground here. Mandriva has a very open ACL policy. Anyone with commit access to a repository has commit access to every package in that repository, so any MDV maintainer can change any package in contrib, and anyone who maintains a package in main - which, practically speaking, is a lot of people including many non-staff - can change any package in main. There's a very small list of restricted packages to which this doesn't apply; I think that's pretty much just kernel and glibc or something very minimal like that. In practice, packages still have maintainers who are recognized for practical reasons and generally you would check with the listed maintainer of a package before making a change to it. (But, hey, if they don't reply in a day or two, it's fair game!) This generally works out pretty well, and helps out with the problem of having quite a small set of maintainers for an extremely large set of packages. I was often in the situation where I happened to notice a small issue with 'someone else's' package and could just go ahead and fix it, instead of having to go through the bureaucracy of filing a bug report and waiting for them to do it. It's rarely the case that someone makes a really stupid change and causes friction. I'd say the system works more often than it doesn't, and it'd probably be good for Fedora too to - as Dave proposes - explicitly _not_ have a concept of ownership, and be more liberal about non-maintainers touching packages. -- Adam Williamson Fedora QA Community Monkey IRC: adamw | Fedora Talk: adamwill AT fedoraproject DOT org http://www.happyassassin.net -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On Mon, 2010-07-05 at 13:30 -0700, Adam Williamson wrote: In practice, packages still have maintainers who are recognized for practical reasons and generally you would check with the listed maintainer of a package before making a change to it. (But, hey, if they don't reply in a day or two, it's fair game!) This generally works out pretty well, and helps out with the problem of having quite a small set of maintainers for an extremely large set of packages. I was often in the situation where I happened to notice a small issue with 'someone else's' package and could just go ahead and fix it, instead of having to go through the bureaucracy of filing a bug report and waiting for them to do it. I should add that obviously it works best if you exercise common sense about what to change. I'd never go into someone else's package and change their whitespace preferences or variable naming scheme or whatever to match my own preferences just so I'd have an easier time looking at their .spec file. Obviously it works best if you follow a minimal approach of changing just what's necessary to address the practical issue you're fixing. -- Adam Williamson Fedora QA Community Monkey IRC: adamw | Fedora Talk: adamwill AT fedoraproject DOT org http://www.happyassassin.net -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On Sat, Jul 3, 2010 at 3:40 AM, Kevin Kofler kevin.kof...@chello.at wrote: Thomas Janssen wrote: You have to accept the maintainers decision to not update it yet? What do you think will happen if everyone builds the wishes he has and breaks a lot of stuff with it? Anarchy? We have processes for that in Fedora: https://fedoraproject.org/wiki/MikeKnox/AWOL_Maintainers It is part of the Fedora Objectives: https://fedoraproject.org/wiki/Objectives to be on the leading edge of free and open source technology. Given that, it is completely unacceptable to not upgrade software to the current release in Rawhide (within a reasonable timeframe, of course we're all not available 24/7) unless there's a really good reason to (in which case that reason ought to be given in the bug report asking for the upgrade!), especially when upstream is asking for their software to be upgraded. snip.. We should really be more aggressive about allowing to upgrade other people's packages in Rawhide if the maintainers don't do it within a reasonable timeframe and don't document any good reason not to do the upgrade. I'm sorry, i can't agree with you here. Being more aggressive, putting pressure on whatever just to have the latest versions of all the software around in rawhide, sounds to me like we would go and break rawhide a lot. I thought rawhide should be more useful and less broken if i recall the latest threads right. Anyways, exactly that's why i do *not* want anybody can do anything with any package. That's just insane, sorry. -- LG Thomas Dubium sapientiae initium -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On Sat, 03 Jul 2010 03:40:57 +0200, Kevin wrote: Thomas Janssen wrote: You have to accept the maintainers decision to not update it yet? What do you think will happen if everyone builds the wishes he has and breaks a lot of stuff with it? Anarchy? We have processes for that in Fedora: https://fedoraproject.org/wiki/MikeKnox/AWOL_Maintainers It is part of the Fedora Objectives: https://fedoraproject.org/wiki/Objectives to be on the leading edge of free and open source technology. Given that, it is completely unacceptable to not upgrade software to the current release in Rawhide (within a reasonable timeframe, of course we're all not available 24/7) unless there's a really good reason to (in which case that reason ought to be given in the bug report asking for the upgrade!), especially when upstream is asking for their software to be upgraded. So the maintainer's decision (assuming there even WAS a decision rather than just lack of time or worse) goes against Fedora's Objectives and so it's not OK to say that it should just get accepted. We should really be more aggressive about allowing to upgrade other people's packages in Rawhide if the maintainers don't do it within a reasonable timeframe and don't document any good reason not to do the upgrade. Ridiculous. :( The way you've phrased it doesn't meet the be excellent guidelines IMO. There is nothing completely unacceptable or against Fedora's objectives with skipping certain upstream releases. And I hope that nobody will become more aggressive or try to force me (or other packagers) to upgrade packages. I don't want anyone among the Fedora contributors to be aggressive in any way when talking to me or when trying to make me do something. As a user or fellow packager [or upstream developer], you are free to suggest upgrades in a bugzilla ticket. And hopefully you evaluate the new release to examine it for changes compared with the previous release in Fedora, so you can give a rationale for your upgrade request. If you meet resistance, you'll have to live with that or return with a competent mediator. If you demonstrate interest in the packaged software (and I truly hope more people will do that), become a maintainer of the packages you want to upgrade and collaborate with existing maintainers. It *can* be a great deal of extra fun to see multiple people work on the same packages, push fixes and updates, communicate with upstream, and experience progress. -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
2010/7/3 Michael Schwendt mschwe...@gmail.com: On Sat, 03 Jul 2010 03:40:57 +0200, Kevin wrote: It is part of the Fedora Objectives: https://fedoraproject.org/wiki/Objectives to be on the leading edge of free and open source technology. Given that, it is completely unacceptable to not upgrade software to the current release in Rawhide (within a reasonable timeframe, of course we're all not available 24/7) unless there's a really good reason to (in which case that reason ought to be given in the bug report asking for the upgrade!), especially when upstream is asking for their software to be upgraded. So the maintainer's decision (assuming there even WAS a decision rather than just lack of time or worse) goes against Fedora's Objectives and so it's not OK to say that it should just get accepted. We should really be more aggressive about allowing to upgrade other people's packages in Rawhide if the maintainers don't do it within a reasonable timeframe and don't document any good reason not to do the upgrade. Ridiculous. :( The way you've phrased it doesn't meet the be excellent guidelines IMO. There is nothing completely unacceptable or against Fedora's objectives with skipping certain upstream releases. And I hope that nobody will become more aggressive or try to force me (or other packagers) to upgrade packages. I don't want anyone among the Fedora contributors to be aggressive in any way when talking to me or when trying to make me do something. As a user or fellow packager [or upstream developer], you are free to suggest upgrades in a bugzilla ticket. And hopefully you evaluate the new release to examine it for changes compared with the previous release in Fedora, so you can give a rationale for your upgrade request. If you meet resistance, you'll have to live with that or return with a competent mediator. I'm fully agree with you, but there are some maintainers who don't respond on bugzilla at all or for a very long time. They may be still active on koji, but they don't respond even when you attach a patch/spec to solve known issues or request for co-maintainership. Obviously, they cannot be defined as nonresponsive package maintainers, so we have no process/policy to treat those packages. I filled dozens of reports in bugzilla to request for updating long unmaintained packages(more than 3 years) several months ago, no packager respond yet. Regards. Chen Lei -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On Fri, Jul 02, 2010 at 07:43:26PM +0530, Rahul Sundaram wrote: On 07/02/2010 07:37 PM, Patrice Dumas wrote: Ok, this policy was for the other case, a case when the maintainer does not respond. I am not saying that it happens a lot, but it happened in the past, and the syslog-ng case exposed in the thread is another recent case. Maybe a policy is not needed and a case by case handling by escalation to FESCo is enough, though. In my days as a Fedora contributor, however, this issue was annoying enough that I proposed the policy, maybe things have changed now. A global view of package versions in rawhide vs the latest upstream similar to http://wiki.debian.org/DEHS would be useful to know how we stand. Rakesh Pandit was looking into this earlier. Not sure of the status on that now. Most of the packages listed here are not up to date: https://bugzilla.redhat.com/buglist.cgi?emailreporter1=1emailtype1=exactquery_format=advancedbug_status=ASSIGNEDemail1=upstream-release-monitoring%40fedoraproject.orgproduct=Fedora Regards Till pgpIMNkrEmufg.pgp Description: PGP signature -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On Sat, 3 Jul 2010 18:08:03 +0800, Chen wrote: I'm fully agree with you, but there are some maintainers who don't respond on bugzilla at all or for a very long time. They may be still active on koji, but they don't respond even when you attach a patch/spec to solve known issues or request for co-maintainership. Obviously, they cannot be defined as nonresponsive package maintainers, so we have no process/policy to treat those packages. Not responding = non-responsive. For non-upgrade patches, the provenpackagers' guidelines make it possible to apply those fixes if the maintainer doesn't do that. The maintainer _should_ see the commit, and you are even permitted to submit a build request after a few days. It just doesn't work, if some maintainers hate Fedora bugzilla or private email and prefer IRC. Or if others don't seem to see cvs commit notifications or pkgdb requests. I filled dozens of reports in bugzilla to request for updating long unmaintained packages(more than 3 years) several months ago, no packager respond yet. IMO you've waited too long to see a response from those people. You could have started the non-responsive maintainer procedure much sooner. And if it results in a quick reply and NOTABUG/WONTFIX closing of your tickets, contact FESCo and ask for a mediator. -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On 07/03/2010 04:05 PM, Till Maas wrote: Most of the packages listed here are not up to date: https://bugzilla.redhat.com/buglist.cgi?emailreporter1=1emailtype1=exactquery_format=advancedbug_status=ASSIGNEDemail1=upstream-release-monitoring%40fedoraproject.orgproduct=Fedora Yeah but this is a partial view only since very few of the package maintainers sign up for upstream monitoring. I forget to do it for new packages myself. If we were monitoring ALL upstreams for packages in Fedora and creating a web page to note differences like the DEHS, we would understand how far we are behind in a glance. Rahul -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
Thomas Janssen wrote: I'm sorry, i can't agree with you here. Being more aggressive, putting pressure on whatever just to have the latest versions of all the software around in rawhide, sounds to me like we would go and break rawhide a lot. I thought rawhide should be more useful and less broken if i recall the latest threads right. Anyways, exactly that's why i do *not* want anybody can do anything with any package. That's just insane, sorry. This is Fedora. Debian is that http://www.debian.org/ way. Please don't destroy what Fedora is all about. If we don't focus on packaging the latest software anymore, we will just be another Debian or Ubuntu. Kevin Kofler -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
Michael Schwendt wrote: Ridiculous. :( The way you've phrased it doesn't meet the be excellent guidelines IMO. There is nothing completely unacceptable or against Fedora's objectives with skipping certain upstream releases. And I hope that nobody will become more aggressive or try to force me (or other packagers) to upgrade packages. I don't want anyone among the Fedora contributors to be aggressive in any way when talking to me or when trying to make me do something. I didn't mean to say that we should be aggressive towards some person, just that we should aggressively, i.e. proactively, quickly, routinely, systematically, frequently etc., update packages to new upstream versions because that's part of the Fedora Objectives. Rawhide should always have the latest upstream release unless there's a strong reason why a particular release needs to be skipped (i.e. it's broken, it contains illegal stuff or something like that). Kevin Kofler -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On Sat, 03 Jul 2010 15:33 +0200, Kevin wrote: Rawhide should always have the latest upstream release unless there's a strong reason why a particular release needs to be skipped (i.e. it's broken, it contains illegal stuff or something like that). How would you find out whether that's the case? - You would need to talk to the package maintainer(s). Having arbitrary provenpackagers perform random upgrades won't do it. -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On Sat, Jul 3, 2010 at 3:28 PM, Kevin Kofler kevin.kof...@chello.at wrote: Thomas Janssen wrote: I'm sorry, i can't agree with you here. Being more aggressive, putting pressure on whatever just to have the latest versions of all the software around in rawhide, sounds to me like we would go and break rawhide a lot. I thought rawhide should be more useful and less broken if i recall the latest threads right. Anyways, exactly that's why i do *not* want anybody can do anything with any package. That's just insane, sorry. This is Fedora. Debian is that http://www.debian.org/ way. Please don't destroy what Fedora is all about. If we don't focus on packaging the latest software anymore, we will just be another Debian or Ubuntu. I'm with Fedora because we have the latest and greatest. Though blindly, aggressively bomb in the latest releases if they're compatible or not is the wrong way. Doing that breaks even koji from time to time (thanks for our real quick rel-eng helping out there). BTW is there a big difference between being a stable Debian or doing the latest and greatest updates the smart way. I know that you know that as well. Just recall from time to time that putting too much pressure and being too aggressive is often counter productive. I generally trust in our maintainers that they know what they do. So if they hold an update back will be for something good. We have processes for everything else. -- LG Thomas Dubium sapientiae initium -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
Michael Schwendt wrote: How would you find out whether that's the case? - You would need to talk to the package maintainer(s). Having arbitrary provenpackagers perform random upgrades won't do it. We need to get packagers to document the reason why they're not upgrading some package in a standard place. That'll also save them repetitive Why is XYZ not up to date? questions. Kevin Kofler -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On Thu, 2010-07-01 at 23:28 -0500, Adam Miller wrote: I don't think it really matters what we call it, I just think that package maintainers are starting to get a sense of entitlement and I feel that's counter productive to the open environment we're used to and are trying to help continue to grow. Absolutely. I've often commented on this problem -- we really don't want package maintainers to throw their toys out of the pram when someone touches their package; this is a collaborative effort. In the old days of RHL and beehive, I think we had it about right... with the obvious exception that it was Red Hat only, but the attitude to packaging was right, IMHO. There _was_ someone who knew most about a package and was expected to deal with it most of the time, but it was also perfectly reasonable for other people to work on the packages too. Fedora seems to have regressed a lot in that respect, although it did improve after we started approving ProvenPackagers. -- David WoodhouseOpen Source Technology Centre david.woodho...@intel.com Intel Corporation -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
Dne 2.7.2010 06:28, Adam Miller napsal(a): I don't think it really matters what we call it, I just think that package maintainers are starting to get a sense of entitlement and I feel that's counter productive to the open environment we're used to and are trying to help continue to grow. I don't think it is that much matter of words, but matter of culture. Wandering provenpackagers (although they were not called that then) were the most striking difference for me when I came from the Debian world, where the possesion of the package is much stronger concept. Probably with increasing number of Ubuntu converts coming to Fedora world, we need to stress this more? Matěj -- Q: Is vi an easy editor to learn, is it intuitive? A: Yes, some of us think so. But most people think that we are crazy. -- vi FAQ -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
2010-07-02 03:18 keltezéssel, Kevin Kofler írta: Dave Airlie wrote: So I've noticed maintainers of packages in Fedora seem to have a concept of ownership, and I'm wondering if we could remove that word from usage about maintainership. +1 IMHO any sponsored packager should be free to do changes which benefit the Fedora Project to any package, no matter who officially maintains the package. And such changes include things like upgrading the package to the current upstream release in Rawhide, especially when that release is needed for other packages. Even a provenpackager can't always make such changes without getting yelled at. I think we need to get rid of the concept of ownership entirely, that'd also make orphaned or de-facto orphaned packages less of a problem. You see a problem, you fix it. Who cares whether the package has an active maintainer or not? +1 I'd like to get syslog-ng updated to the latest version in Rawhide (I work part time for the upstream developer and I'm also an occasional Fedora user). I contacted the package owner, no response. Created a bugreport to get it updated ( https://bugzilla.redhat.com/show_bug.cgi?id=598961 ), and also provided an updated package, which compiles and works fine on Fedora 12, 13 and Rawhide. After waiting for weeks, I started a maintainer time out. It was closed within an hour. I got some comments on bugzilla, but nothing happened ever since. The updated package was never downloaded from my website. What can I do in this situation? Obviously I'm not a proven packager to update the package myself, as I'm not a Fedora developer. I worked a lot to update and test the package, but still I'm stuck. And as you can see, the maintainer timeout procedure does not help either... Bye, CzP -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On Fri, Jul 2, 2010 at 2:23 PM, Peter Czanik pcza...@fang.fa.gau.hu wrote: 2010-07-02 03:18 keltezéssel, Kevin Kofler írta: Dave Airlie wrote: So I've noticed maintainers of packages in Fedora seem to have a concept of ownership, and I'm wondering if we could remove that word from usage about maintainership. +1 IMHO any sponsored packager should be free to do changes which benefit the Fedora Project to any package, no matter who officially maintains the package. And such changes include things like upgrading the package to the current upstream release in Rawhide, especially when that release is needed for other packages. Even a provenpackager can't always make such changes without getting yelled at. I think we need to get rid of the concept of ownership entirely, that'd also make orphaned or de-facto orphaned packages less of a problem. You see a problem, you fix it. Who cares whether the package has an active maintainer or not? Well, who's the one in charge for bugreports within that system? Everyone? Nobody? I see a lot of who cares and frustration coming up with that everyone builds what he think he have to build system. +1 I'd like to get syslog-ng updated to the latest version in Rawhide (I work part time for the upstream developer and I'm also an occasional Fedora user). I contacted the package owner, no response. Created a bugreport to get it updated ( https://bugzilla.redhat.com/show_bug.cgi?id=598961 ), and also provided an updated package, which compiles and works fine on Fedora 12, 13 and Rawhide. After waiting for weeks, I started a maintainer time out. It was closed within an hour. I got some comments on bugzilla, but nothing happened ever since. The updated package was never downloaded from my website. What can I do in this situation? Obviously I'm not a proven packager to update the package myself, as I'm not a Fedora developer. I worked a lot to update and test the package, but still I'm stuck. And as you can see, the maintainer timeout procedure does not help either... You have to accept the maintainers decision to not update it yet? What do you think will happen if everyone builds the wishes he has and breaks a lot of stuff with it? Anarchy? We have processes for that in Fedora: https://fedoraproject.org/wiki/MikeKnox/AWOL_Maintainers I'm really trying very hard to see it positive. -- LG Thomas Dubium sapientiae initium -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On Fri, Jul 02, 2010 at 02:23:43PM +0200, Peter Czanik wrote: 2010-07-02 03:18 keltezéssel, Kevin Kofler írta: I think we need to get rid of the concept of ownership entirely, that'd also make orphaned or de-facto orphaned packages less of a problem. You see a problem, you fix it. Who cares whether the package has an active maintainer or not? That's very much against the fedora policies. I'd like to get syslog-ng updated to the latest version in Rawhide (I work part time for the upstream developer and I'm also an occasional Fedora user). I contacted the package owner, no response. Created a bugreport to get it updated ( https://bugzilla.redhat.com/show_bug.cgi?id=598961 ), and also provided an updated package, which compiles and works fine on Fedora 12, 13 and Rawhide. After waiting for weeks, I started a maintainer time out. It was closed within an hour. Indeed. Maintainer time out are for completly missing maintainers, not to force them to apply a change. Obviously I'm not a proven packager to update the package myself, as I'm not a Fedora developer. Even if you were a provenpackager you would be forbidden from doing that. Provenpackagers right to modify other people packages are far from being that large. Have a look at the relevant policy if you want more information. I worked a lot to update and test the package, but still I'm stuck. And as you can see, the maintainer timeout procedure does not help either... And the provenpackager policy wouldn't help either. In the past we proposed a policy for that kind of issues with Rahul, but it was never approved (nor really considered). http://fedoraproject.org/wiki/RahulSundaram/CollectiveMaintenance The only thing that can be done, right now for such issues is the traditional escalation procedure. I don't know if it is documented anywhere, but it is along * make yourself clear in a bugreport (which is already done) * explain the issue on the devel list (guess you already did that) * escalate to FESCo -- Pat -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
2010/7/2 Patrice Dumas pertu...@free.fr: On Fri, Jul 02, 2010 at 02:23:43PM +0200, Peter Czanik wrote: 2010-07-02 03:18 keltezéssel, Kevin Kofler írta: I think we need to get rid of the concept of ownership entirely, that'd also make orphaned or de-facto orphaned packages less of a problem. You see a problem, you fix it. Who cares whether the package has an active maintainer or not? That's very much against the fedora policies. I'd like to get syslog-ng updated to the latest version in Rawhide (I work part time for the upstream developer and I'm also an occasional Fedora user). I contacted the package owner, no response. Created a bugreport to get it updated ( https://bugzilla.redhat.com/show_bug.cgi?id=598961 ), and also provided an updated package, which compiles and works fine on Fedora 12, 13 and Rawhide. After waiting for weeks, I started a maintainer time out. It was closed within an hour. Indeed. Maintainer time out are for completly missing maintainers, not to force them to apply a change. Obviously I'm not a proven packager to update the package myself, as I'm not a Fedora developer. Even if you were a provenpackager you would be forbidden from doing that. Provenpackagers right to modify other people packages are far from being that large. Have a look at the relevant policy if you want more information. I worked a lot to update and test the package, but still I'm stuck. And as you can see, the maintainer timeout procedure does not help either... And the provenpackager policy wouldn't help either. In the past we proposed a policy for that kind of issues with Rahul, but it was never approved (nor really considered). http://fedoraproject.org/wiki/RahulSundaram/CollectiveMaintenance The only thing that can be done, right now for such issues is the traditional escalation procedure. I don't know if it is documented anywhere, but it is along * make yourself clear in a bugreport (which is already done) * explain the issue on the devel list (guess you already did that) * escalate to FESCo -- I think escalating to FESCo is only suitable for changes which are controversial between different people, we should have another policy to treat those non-responsive issues, maintainers should respond on bugzilla report in time. Chen Lei -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On 07/02/2010 06:46 PM, Patrice Dumas wrote: In the past we proposed a policy for that kind of issues with Rahul, but it was never approved (nor really considered). http://fedoraproject.org/wiki/RahulSundaram/CollectiveMaintenance I had forgotten about this but since becoming provenpackager I have helped out in simple rebuilds or even version bumps on occasions and have gotten positive feedback. The written policy is rather rigid but people tend to be more flexible about others helping out in effect. We should generally assume goodwill and not try to mandate common sense via minute policies. Rahul -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
Hello, 2010-07-02 14:48 keltezéssel, Thomas Janssen írta: +1 I'd like to get syslog-ng updated to the latest version in Rawhide (I work part time for the upstream developer and I'm also an occasional Fedora user). I contacted the package owner, no response. Created a bugreport to get it updated ( https://bugzilla.redhat.com/show_bug.cgi?id=598961 ), and also provided an updated package, which compiles and works fine on Fedora 12, 13 and Rawhide. After waiting for weeks, I started a maintainer time out. It was closed within an hour. I got some comments on bugzilla, but nothing happened ever since. The updated package was never downloaded from my website. What can I do in this situation? Obviously I'm not a proven packager to update the package myself, as I'm not a Fedora developer. I worked a lot to update and test the package, but still I'm stuck. And as you can see, the maintainer timeout procedure does not help either... You have to accept the maintainers decision to not update it yet? What do you think will happen if everyone builds the wishes he has and breaks a lot of stuff with it? Anarchy? We have processes for that in Fedora: https://fedoraproject.org/wiki/MikeKnox/AWOL_Maintainers Well, syslog-ng is a leaf package. AFAIK, there are no others depending on it, so it can't really break a lot of stuff. Also, I tested the package heavily, not just updated, and no problems came up during testing. Version 2.X of syslog-ng, which is currently included in Fedora, is now obsolate. The current release is 3.1.1, this is the version I prepared for Fedora. And it is not just a random wish: I did the update for openSUSE and FreeBSD, where they are accepted already and in use. Gentoo, Arch, Mandriva, Debian, Ubuntu, NetBSD, Solaris, etc. all did the switch at least in their devel versions for the above reasons. Only Fedora is missing from the list... Bye, CzP -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On Fri, Jul 02, 2010 at 09:36:34PM +0800, Chen Lei wrote: I think escalating to FESCo is only suitable for changes which are controversial between different people, we should have another policy to treat those non-responsive issues, maintainers should respond on bugzilla report in time. I agree with you, and that's why I proposed a policy with Rahul. This may certainly be enhanced. If I recall well this was rejected. http://fedoraproject.org/wiki/RahulSundaram/CollectiveMaintenance I am no more involved in Fedora (except for unfrequent whining on the devel list) so I won't be able to follow on this issue, but I still agree that it is a problematic issue. Until such a policy is available, however, ther is no other possibility than escalation to FESCo. -- Pat -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On Fri, Jul 02, 2010 at 07:15:54PM +0530, Rahul Sundaram wrote: I had forgotten about this but since becoming provenpackager I have helped out in simple rebuilds or even version bumps on occasions and have gotten positive feedback. You mean that you didn't only send a patch but you did commit version bumps without prior communication with the maintainer or with a maintainer not responding to your offer to do a version bump? The written policy is rather rigid but people tend to be more flexible about others helping out in effect. We should generally assume goodwill and not try to mandate common sense via minute policies. I am not that positive. Maybe the fedora project changed since these days, but there were quite a bit of cases where bugs with fixes were not considered for a long time. -- Pat -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On 07/02/2010 07:27 PM, Patrice Dumas wrote: On Fri, Jul 02, 2010 at 07:15:54PM +0530, Rahul Sundaram wrote: I had forgotten about this but since becoming provenpackager I have helped out in simple rebuilds or even version bumps on occasions and have gotten positive feedback. You mean that you didn't only send a patch but you did commit version bumps without prior communication with the maintainer or with a maintainer not responding to your offer to do a version bump? I offered to do versions bumps and uniformly got a positive response. In some cases, I filed a patch in bugzilla. Rahul -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On 07/02/2010 07:37 PM, Patrice Dumas wrote: Ok, this policy was for the other case, a case when the maintainer does not respond. I am not saying that it happens a lot, but it happened in the past, and the syslog-ng case exposed in the thread is another recent case. Maybe a policy is not needed and a case by case handling by escalation to FESCo is enough, though. In my days as a Fedora contributor, however, this issue was annoying enough that I proposed the policy, maybe things have changed now. A global view of package versions in rawhide vs the latest upstream similar to http://wiki.debian.org/DEHS would be useful to know how we stand. Rakesh Pandit was looking into this earlier. Not sure of the status on that now. Rahul -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On Thu, Jul 01, 2010 at 11:28:09PM -0500, Adam Miller wrote: I don't think it really matters what we call it, I just think that package maintainers are starting to get a sense of entitlement and I feel that's counter productive to the open environment we're used to and are trying to help continue to grow. The package owner gets emails about cvs commits, so they are always aware of what's going on and can review the changes to packages they maintain. In the event of a discrepancy then the person receiving the email obviously has an email account and can easily email the person who made the edit in order to extend a friendly inquiry as to the change. It doesn't need to be any more complicated than that. I for one welcome co-maintainers because I'm a big fan of collaboration and a sense of community, and if I can't trust my fellow community member and contributor to help in the maintenance of packages that I just so happen to have a bit flipped for in the pkgdb, then I'm in the wrong place. Good points all, Adam. My personal experience with a couple of my packages, where for example Matthias Clasen found and stomped a bug, or Jesse Keating took care of a rebuild when I wasn't around[1], gave me confidence that fellow contributors have my back, as opposed to sneaking around behind it. I prefer to presume goodwill. At worst I've caused them to grumble for taking up my slack -- in which case they know where to find my inbox to complain. :-) * * * [1] These happened both when I was a volunteer, and when I was a Red Hat employee, FWIW. -- Paul W. Frieldshttp://paul.frields.org/ gpg fingerprint: 3DA6 A0AC 6D58 FEC4 0233 5906 ACDB C937 BD11 3717 http://redhat.com/ - - - - http://pfrields.fedorapeople.org/ Where open source multiplies: http://opensource.com -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
Kevin Fenzi wrote: On Fri, 02 Jul 2010 00:44:29 -0400 Tom Lane t...@redhat.com wrote: I think it's counterproductive to downgrade that responsibility, or even worse pretend that it doesn't matter --- and Kevin's lead statement in this thread is damn close to pretending that. Sorry Kevin, we are not interchangeable parts. Which Kevin? I never intended to say any such thing... :) I did, so I guess he's talking about me. :-) Kevin Kofler -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
Ralf Corsepius wrote: We need groups, with grouped privileges/acls etc. It's essentially what e.g. the perl-sig originally was meant to be. Yes, group ACLs are definitely needed, but in addition to that technical feature, we also need to make sure that the SIG actually gets commit access to ALL packages related to the SIG. ALL perl-* packages should be committable to by the Perl SIG. That's what a SIG is for. And it'd have prevented this whole why were X and Y added as maintainers to my perl-* packages fiasco, it would just have been the SIG's decision to add the people to the SIG and this should automatically give commit access to all perl-* packages. Similarly, all packages using Qt should be committable to by the KDE SIG etc. Kevin Kofler -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
Thomas Janssen wrote: You have to accept the maintainers decision to not update it yet? What do you think will happen if everyone builds the wishes he has and breaks a lot of stuff with it? Anarchy? We have processes for that in Fedora: https://fedoraproject.org/wiki/MikeKnox/AWOL_Maintainers It is part of the Fedora Objectives: https://fedoraproject.org/wiki/Objectives to be on the leading edge of free and open source technology. Given that, it is completely unacceptable to not upgrade software to the current release in Rawhide (within a reasonable timeframe, of course we're all not available 24/7) unless there's a really good reason to (in which case that reason ought to be given in the bug report asking for the upgrade!), especially when upstream is asking for their software to be upgraded. So the maintainer's decision (assuming there even WAS a decision rather than just lack of time or worse) goes against Fedora's Objectives and so it's not OK to say that it should just get accepted. We should really be more aggressive about allowing to upgrade other people's packages in Rawhide if the maintainers don't do it within a reasonable timeframe and don't document any good reason not to do the upgrade. Kevin Kofler -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
David Woodhouse wrote: In the old days of RHL and beehive, I think we had it about right... with the obvious exception that it was Red Hat only, but the attitude to packaging was right, IMHO. There _was_ someone who knew most about a package and was expected to deal with it most of the time, but it was also perfectly reasonable for other people to work on the packages too. Fedora seems to have regressed a lot in that respect, although it did improve after we started approving ProvenPackagers. But the official provenpackager policy is way too restrictive on allowed changes, so I'm often left wondering Will I get away with doing that change?, and several times, I have to err on the side of caution, wasting both my and the maintainer's time by filing bugs etc. when I could just fix the issue. Kevin Kofler -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On 07/03/2010 03:49 AM, Kevin Kofler wrote: David Woodhouse wrote: In the old days of RHL and beehive, I think we had it about right... with the obvious exception that it was Red Hat only, but the attitude to packaging was right, IMHO. There _was_ someone who knew most about a package and was expected to deal with it most of the time, but it was also perfectly reasonable for other people to work on the packages too. Fedora seems to have regressed a lot in that respect, although it did improve after we started approving ProvenPackagers. But the official provenpackager policy is way too restrictive on allowed changes, so I'm often left wondering Will I get away with doing that change? Yep. On such occasions, I usually resort to being ultra conservative, i.e. to only apply changes when I am sure about them. and several times, I have to err on the side of caution, wasting both my and the maintainer's time by filing bugs etc. when I could just fix the issue. So do I. Ralf -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
concept of package ownership
So I've noticed maintainers of packages in Fedora seem to have a concept of ownership, and I'm wondering if we could remove that word from usage about maintainership. I'm come from working as a maintainer in the kernel, and its long been said that kernel maintainers don't *own* the code, they are merely the stewards of the code and the code belongs to everyone. Linus reminds you of this by routinely doing stuff to code behind your back and when you give out he reminds you that maintainership doesn't imply ownership. I see Fedora maintainers as doing things on behalf of the Fedora project, and not merely providing the Fedora project with stuff they own, but I see a lot of others see me owns this package as the priority not me enjoys maintaining things on behalf of Fedora. Dave. -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
I agree. The relevant concept is not owner, but sucker, or victim. When businessspeak people say someone owns a piece of work, what they mean is to identify the person as the recipient of problems, complaints, pleas for help, and perhaps even, rarely, praise, regarding the state of the work. We don't own the code, the code owns us. It knows where we live, and it keeps bringing people over and expecting us to feed them. When a robot sends me email via an alias that ends with -owner, I never, ever, get the feeling that I am the one in charge. Thanks, Roland -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
Dave Airlie wrote: So I've noticed maintainers of packages in Fedora seem to have a concept of ownership, and I'm wondering if we could remove that word from usage about maintainership. +1 IMHO any sponsored packager should be free to do changes which benefit the Fedora Project to any package, no matter who officially maintains the package. And such changes include things like upgrading the package to the current upstream release in Rawhide, especially when that release is needed for other packages. Even a provenpackager can't always make such changes without getting yelled at. I think we need to get rid of the concept of ownership entirely, that'd also make orphaned or de-facto orphaned packages less of a problem. You see a problem, you fix it. Who cares whether the package has an active maintainer or not? Kevin Kofler -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 7/1/10 6:18 PM, Kevin Kofler wrote: I think we need to get rid of the concept of ownership entirely, that'd also make orphaned or de-facto orphaned packages less of a problem. You see a problem, you fix it. Who cares whether the package has an active maintainer or not? While I agree that package ownership should not feel possessive, I do strongly feel that there still should be some single person (or team I suppose...) who is ultimately responsible for the package. A place for bug reports, for autoqa activity, for reviewing potential patches and changes from other people, etc... - -- Jesse Keating Fedora -- Freedom² is a feature! identi.ca: http://identi.ca/jkeating -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.9 (Darwin) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAkwtaF8ACgkQ4v2HLvE71NXCCwCgxoNtzgQ/DDpx78uI4jjodHSu GTYAnAxN9OwDW/qnXMDnZKfp4zCNG8NO =F1ef -END PGP SIGNATURE- -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On Fri, 02 Jul 2010 08:16:35 +1000 Dave Airlie airl...@redhat.com wrote: So I've noticed maintainers of packages in Fedora seem to have a concept of ownership, and I'm wondering if we could remove that word from usage about maintainership. ...snip... I agree. I think 'stewards' or 'guardian' or something might be better, but those don't quite work either. ;( The packages I steward for Fedora (with a few exceptions of things I should orphan or the like) I use and enjoy using, and wish to make sure they are in good shape for others to use and enjoy too. I think it's always been the case that something you use and enjoy and want other people to use and enjoy ends up handled much better than something you own for other reasons. kevin signature.asc Description: PGP signature -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
I don't think it really matters what we call it, I just think that package maintainers are starting to get a sense of entitlement and I feel that's counter productive to the open environment we're used to and are trying to help continue to grow. The package owner gets emails about cvs commits, so they are always aware of what's going on and can review the changes to packages they maintain. In the event of a discrepancy then the person receiving the email obviously has an email account and can easily email the person who made the edit in order to extend a friendly inquiry as to the change. It doesn't need to be any more complicated than that. I for one welcome co-maintainers because I'm a big fan of collaboration and a sense of community, and if I can't trust my fellow community member and contributor to help in the maintenance of packages that I just so happen to have a bit flipped for in the pkgdb, then I'm in the wrong place. -AdamM -- http://maxamillion.googlepages.com - () ascii ribbon campaign - against html e-mail /\ www.asciiribbon.org - against proprietary attachments -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On Thu, 01 Jul 2010 21:17:38 -0700 Jesse Keating jkeat...@j2solutions.net wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 7/1/10 6:18 PM, Kevin Kofler wrote: I think we need to get rid of the concept of ownership entirely, that'd also make orphaned or de-facto orphaned packages less of a problem. You see a problem, you fix it. Who cares whether the package has an active maintainer or not? While I agree that package ownership should not feel possessive, I do strongly feel that there still should be some single person (or team I suppose...) who is ultimately responsible for the package. A place for bug reports, for autoqa activity, for reviewing potential patches and changes from other people, etc... Agreed. While wandering provenpackagers or whoever can assist with sticky issues, there needs to be a group of people who manage bugs, build a relationship with upstream, follow upstream development, etc. So, while I think we should try and reduce the possessiveness of owning packages, we still need a group of stewards or whatever for packages. kevin signature.asc Description: PGP signature -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
Kevin Fenzi ke...@scrye.com writes: Jesse Keating jkeat...@j2solutions.net wrote: While I agree that package ownership should not feel possessive, I do strongly feel that there still should be some single person (or team I suppose...) who is ultimately responsible for the package. A place for bug reports, for autoqa activity, for reviewing potential patches and changes from other people, etc... Agreed. While wandering provenpackagers or whoever can assist with sticky issues, there needs to be a group of people who manage bugs, build a relationship with upstream, follow upstream development, etc. Yeah. There needs to be somebody in the Fedora community with a long-term commitment to each package. Perhaps the term owner is politically incorrect but nonetheless there is always going to be somebody who knows more about that package than anybody else in Fedora. I think it's counterproductive to downgrade that responsibility, or even worse pretend that it doesn't matter --- and Kevin's lead statement in this thread is damn close to pretending that. Sorry Kevin, we are not interchangeable parts. regards, tom lane -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On 07/02/2010 06:34 AM, Kevin Fenzi wrote: On Thu, 01 Jul 2010 21:17:38 -0700 Jesse Keatingjkeat...@j2solutions.net wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 7/1/10 6:18 PM, Kevin Kofler wrote: I think we need to get rid of the concept of ownership entirely, that'd also make orphaned or de-facto orphaned packages less of a problem. You see a problem, you fix it. Who cares whether the package has an active maintainer or not? While I agree that package ownership should not feel possessive, Should the name owner be an issue, why not call them by what they actually are, maintainer and co-maintainer? Agreed. While wandering provenpackagers or whoever can assist with sticky issues, there needs to be a group of people who manage bugs, build a relationship with upstream, follow upstream development, etc. Agreed. So, while I think we should try and reduce the possessiveness of owning packages, we still need a group of stewards or whatever for packages. We need groups, with grouped privileges/acls etc. It's essentially what e.g. the perl-sig originally was meant to be. Unfortunately, technical limitations of Fedora's packager infrastructure so far have prevented to take full advantage of this (c.f. Petr's mass acl-changes in recent weeks). Ralf -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel
Re: concept of package ownership
On Fri, 02 Jul 2010 00:44:29 -0400 Tom Lane t...@redhat.com wrote: Yeah. There needs to be somebody in the Fedora community with a long-term commitment to each package. Perhaps the term owner is politically incorrect but nonetheless there is always going to be somebody who knows more about that package than anybody else in Fedora. Sure. But they should also welcome others joining them, not say I own this package, its mine all mine!, IMHO. I think it's counterproductive to downgrade that responsibility, or even worse pretend that it doesn't matter --- and Kevin's lead statement in this thread is damn close to pretending that. Sorry Kevin, we are not interchangeable parts. Which Kevin? I never intended to say any such thing... :) kevin signature.asc Description: PGP signature -- devel mailing list devel@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/devel