Re: [aur-general] Where to put your name when you adopt an AUR package (was: TU Application)
On Thu, Oct 9, 2008 at 8:41 AM, Evangelos Foutras <foutrelis@gmail.com> wrote:
stefan-husmann@t-online.de wrote:
What should I add myself as, when adopting a package, if not using the Maintainer tag? "Contributor" maybe?
Yes!
Actually, I just read a wiki entry[1] on this which clears things up.
# Maintainer: the maintainer responsible for this pkg in the official repositories # Contributor: the person who wrote the first PKGBUILD for this package
So I shouldn't add myself at all to packages I adopt. :o /me goes to update his adopted packages.
---------------- [1] http://wiki.archlinux.org/index.php/ABS_PKGBUILD_Explained
It does make sense, at least to me, to have your name and email address in a PKGBUILD you currently maintain in AUR, even if you didn't create it. However, I don't think there is any policy about this. I have seen some "custom" tags, which I can't remember exactly and of course much misuse of the maintainer tag - I am guilty of that one myself. What one could do to prevent further misuse of the Maintainer tag would be introducing a new tag like AUR-Maintainer and put it in the wiki. Notice that this is just an example, I wouldn't be against calling this different. It wouldn't stop the "Maintainer Tag problem" all at once, but it might help reducing it.
On Thu, Oct 9, 2008 at 6:13 PM, Henning Garus <henning.garus@googlemail.com> wrote:
It does make sense, at least to me, to have your name and email address in a PKGBUILD you currently maintain in AUR, even if you didn't create it. However, I don't think there is any policy about this. I have seen some "custom" tags, which I can't remember exactly and of course much misuse of the maintainer tag - I am guilty of that one myself.
What one could do to prevent further misuse of the Maintainer tag would be introducing a new tag like AUR-Maintainer and put it in the wiki. Notice that this is just an example, I wouldn't be against calling this different. It wouldn't stop the "Maintainer Tag problem" all at once, but it might help reducing it.
I've never understood this rule, especially when I was enforcing it. Why aren't AUR maintainers allowed to use the maintainer tag? -- Callan Barrett
Callan Barrett wrote:
On Thu, Oct 9, 2008 at 6:13 PM, Henning Garus <henning.garus@googlemail.com> wrote:
It does make sense, at least to me, to have your name and email address in a PKGBUILD you currently maintain in AUR, even if you didn't create it. However, I don't think there is any policy about this. I have seen some "custom" tags, which I can't remember exactly and of course much misuse of the maintainer tag - I am guilty of that one myself.
What one could do to prevent further misuse of the Maintainer tag would be introducing a new tag like AUR-Maintainer and put it in the wiki. Notice that this is just an example, I wouldn't be against calling this different. It wouldn't stop the "Maintainer Tag problem" all at once, but it might help reducing it.
I've never understood this rule, especially when I was enforcing it. Why aren't AUR maintainers allowed to use the maintainer tag?
I agree. It is not as if people get confused and think a package is official just because of that tag. Or am I giving people too much credit... :P
On Thu, Oct 9, 2008 at 9:26 PM, Allan McRae <allan@archlinux.org> wrote:
I agree. It is not as if people get confused and think a package is official just because of that tag. Or am I giving people too much credit... :P
I don't think it has any practical reason. It's just such a damn old rule nobody questions it. -- Callan Barrett
Den Thu, 9 Oct 2008 21:57:03 +0800 "Callan Barrett" <wizzomafizzo@gmail.com> skrev:
I don't think it has any practical reason. It's just such a damn old rule nobody questions it.
It can't be that old. This is the first time I've heard of it. The Contributor tag has a long history, from way before the AUR. It's from the time when people just uploaded packages to an ftp server to contribute, and then the devs would look at them when they had time, and include it in the repo if they thought it made sense. Then the user who first uploaded the package would be named as Contributor in the PKGBUILD. If you look at the jack-audio-connection-kit and rosegarden packages, you'll still see my name as Contributor, even though the package has been in extra since 2004, and probably don't contain one line of code from my origial PKGBUILDs. I think it is wrong to put your own name in the Contributor field just because you have adopted a package. I think the Maintainer field should be used for that, since the one who has adopted a package, normal user or TU, is in fact the Maintainer, and the one who should be contacted if there's a problem with the package. Robert
On Thu, Oct 09, 2008 at 04:43:54PM +0200, Robert Emil Berge wrote:
Den Thu, 9 Oct 2008 21:57:03 +0800 "Callan Barrett" <wizzomafizzo@gmail.com> skrev:
I don't think it has any practical reason. It's just such a damn old rule nobody questions it.
It can't be that old. This is the first time I've heard of it. The Contributor tag has a long history, from way before the AUR. It's from the time when people just uploaded packages to an ftp server to contribute, and then the devs would look at them when they had time, and include it in the repo if they thought it made sense. Then the user who first uploaded the package would be named as Contributor in the PKGBUILD. If you look at the jack-audio-connection-kit and rosegarden packages, you'll still see my name as Contributor, even though the package has been in extra since 2004, and probably don't contain one line of code from my origial PKGBUILDs. I think it is wrong to put your own name in the Contributor field just because you have adopted a package. I think the Maintainer field should be used for that, since the one who has adopted a package, normal user or TU, is in fact the Maintainer, and the one who should be contacted if there's a problem with the package.
Robert
I think you should be able to put whatever the hell you want in the comments of an unsupported PKGBUILD. I think this line would be cool: # Ruler of the Universe: Your Name <you@email.com>
If I understand the question correctly, and I may not..... The answer is that you are a "contributor" of a PKGBUILD (et al) unless you are a TU, in which case you are "maintaining" the *binary* package. Or at least that was what it was in at first. Very best regards; Bob Finch On Thu, Oct 9, 2008 at 7:08 AM, Callan Barrett <wizzomafizzo@gmail.com>wrote:
On Thu, Oct 9, 2008 at 6:13 PM, Henning Garus <henning.garus@googlemail.com> wrote:
It does make sense, at least to me, to have your name and email address in a PKGBUILD you currently maintain in AUR, even if you didn't create it. However, I don't think there is any policy about this. I have seen some "custom" tags, which I can't remember exactly and of course much misuse of the maintainer tag - I am guilty of that one myself.
What one could do to prevent further misuse of the Maintainer tag would be introducing a new tag like AUR-Maintainer and put it in the wiki. Notice that this is just an example, I wouldn't be against calling this different. It wouldn't stop the "Maintainer Tag problem" all at once, but it might help reducing it.
I've never understood this rule, especially when I was enforcing it. Why aren't AUR maintainers allowed to use the maintainer tag?
-- Callan Barrett
On Thu, Oct 9, 2008 at 9:49 PM, Bob Finch <w9ya@qrparci.net> wrote:
If I understand the question correctly, and I may not.....
The answer is that you are a "contributor" of a PKGBUILD (et al) unless you are a TU, in which case you are "maintaining" the *binary* package.
Or at least that was what it was in at first.
Very best regards;
Bob Finch
Oh, we understand what they're for now but not *why* they're used that way. Why aren't AUR maintainers allowed to use a maintainer tag when in most cases the word "maintainer" makes more sense and it would stop a lot of this nonsense where people get told off for using the special TU-only comment. -- Callan Barrett
-----Original Message----- Date: Thu, 09 Oct 2008 16:14:37 +0200 Subject: Re: [aur-general] Where to put your name when you adopt an AUR package (was: TU Application) From: "Callan Barrett" <wizzomafizzo@gmail.com> To: "Discussion about the Arch User Repository (AUR)" <aur-general@archlinux.org> On Thu, Oct 9, 2008 at 9:49 PM, Bob Finch <w9ya@qrparci.net> wrote:
If I understand the question correctly, and I may not.....
The answer is that you are a "contributor" of a PKGBUILD (et al) unless you are a TU, in which case you are "maintaining" the *binary* package.
Or at least that was what it was in at first.
Very best regards;
Bob Finch
Oh, we understand what they're for now but not *why* they're used that way. Why aren't AUR maintainers allowed to use a maintainer tag when in most cases the word "maintainer" makes more sense and it would stop a lot of this nonsense where people get told off for using the special TU-only comment. -- Callan Barrett 1+
Regardless of what it once was, I think the current method is silly and needlessly confusing. Contributor should be there to credit the previous handlers for a package. Maintainer should just be the person who is currently in charge of keeping it working, whether binary or otherwise. Pacman's "Packager" data is kept for the binary files, making the current use of Maintainer redundant. I think it should be changed to the more logical way. Anyone else have an opinion on that?
On Thu, Oct 9, 2008 at 10:25 PM, Daenyth Blank <daenyth+arch@gmail.com> wrote:
Regardless of what it once was, I think the current method is silly and needlessly confusing. Contributor should be there to credit the previous handlers for a package. Maintainer should just be the person who is currently in charge of keeping it working, whether binary or otherwise. Pacman's "Packager" data is kept for the binary files, making the current use of Maintainer redundant.
I think it should be changed to the more logical way. Anyone else have an opinion on that?
+1 -- Callan Barrett
Daenyth Blank wrote:
Regardless of what it once was, I think the current method is silly and needlessly confusing. Contributor should be there to credit the previous handlers for a package. Maintainer should just be the person who is currently in charge of keeping it working, whether binary or otherwise. Pacman's "Packager" data is kept for the binary files, making the current use of Maintainer redundant.
I think it should be changed to the more logical way. Anyone else have an opinion on that?
I agree and, from what I've read, many others also believe that the "Maintainer" tag should denote a package's current maintainer, be it binary or otherwise. Now, regarding contributors, the list could get unnecessarily big. Moreover, when someone adopts a package he/she won't be able to just substitute the previous maintainer but instead will need to update the contributors list first, which is complicating things (taking into account that no information is available about the work performed by the past maintainer). Does anyone else see this as an issue and share my viewpoint?
On Thu, Oct 9, 2008 at 4:25 PM, Daenyth Blank <daenyth+arch@gmail.com> wrote:
Regardless of what it once was, I think the current method is silly and needlessly confusing. Contributor should be there to credit the previous handlers for a package. Maintainer should just be the person who is currently in charge of keeping it working, whether binary or otherwise. Pacman's "Packager" data is kept for the binary files, making the current use of Maintainer redundant.
I think it should be changed to the more logical way. Anyone else have an opinion on that?
+1
On Thu, Oct 9, 2008 at 11:33 AM, Xavier <shiningxc@gmail.com> wrote:
On Thu, Oct 9, 2008 at 4:25 PM, Daenyth Blank <daenyth+arch@gmail.com> wrote:
Regardless of what it once was, I think the current method is silly and needlessly confusing. Contributor should be there to credit the previous handlers for a package. Maintainer should just be the person who is currently in charge of keeping it working, whether binary or otherwise. Pacman's "Packager" data is kept for the binary files, making the current use of Maintainer redundant.
I think it should be changed to the more logical way. Anyone else have an opinion on that?
+1
+2
On Thu, Oct 09, 2008 at 10:25:23AM -0400, Daenyth Blank wrote:
Regardless of what it once was, I think the current method is silly and needlessly confusing. Contributor should be there to credit the previous handlers for a package. Maintainer should just be the person who is currently in charge of keeping it working, whether binary or otherwise. Pacman's "Packager" data is kept for the binary files, making the current use of Maintainer redundant.
I think it should be changed to the more logical way. Anyone else have an opinion on that?
While I agree with you, there are some cases where Maintainer and Contributor overlap. Consider the following scenario: 1. X has contributed the package foo 1.0 to AUR 2. X orphans it after some time 3. Y picks it up, and adds Maintainer: tag. 4. Y updates the version to 1.1 after some time. 5. Y orphans the package. 6. Z picks it up. Now should Z a) Replace the current Maintainer tag? b) Replace and move Y to Contributor: list? I think 6b) is better because it preserves history and gives credit to everyone who worked on the PKGBUILD at some point. While the Contributor: list can get lengthy, I see no other way out. -- Abhishek Dasgupta <http://abhidg.mine.nu> GPG 67972DOF pgp.mit.edu
On Thu, Oct 09, 2008 at 11:26:36PM +0530, Abhishek Dasgupta wrote:
On Thu, Oct 09, 2008 at 10:25:23AM -0400, Daenyth Blank wrote:
Regardless of what it once was, I think the current method is silly and needlessly confusing. Contributor should be there to credit the previous handlers for a package. Maintainer should just be the person who is currently in charge of keeping it working, whether binary or otherwise. Pacman's "Packager" data is kept for the binary files, making the current use of Maintainer redundant.
I think it should be changed to the more logical way. Anyone else have an opinion on that?
While I agree with you, there are some cases where Maintainer and Contributor overlap. Consider the following scenario:
1. X has contributed the package foo 1.0 to AUR 2. X orphans it after some time 3. Y picks it up, and adds Maintainer: tag. 4. Y updates the version to 1.1 after some time. 5. Y orphans the package. 6. Z picks it up. Now should Z a) Replace the current Maintainer tag? b) Replace and move Y to Contributor: list?
I think 6b) is better because it preserves history and gives credit to everyone who worked on the PKGBUILD at some point. While the Contributor: list can get lengthy, I see no other way out.
Why not just use multiple lines? One for each contributor? It's far easier this way. And can expand from 1 to 100 contributors (though the PKGBUILD would be annoying to look at). Cheers, Aaron PS Test sending from muttng. Woo
Why not just use multiple lines? One for each contributor? It's far easier this way. And can expand from 1 to 100 contributors (though the PKGBUILD would be annoying to look at).
+1
-----Original Message----- Date: Thu, 09 Oct 2008 16:25:23 +0200 Subject: Re: [aur-general] Where to put your name when you adopt an AUR
package (was: TU Application)
From: "Daenyth Blank" <daenyth+arch@gmail.com> To: "Discussion about the Arch User Repository (AUR)" <aur-general@archlinux.org>
Regardless of what it once was, I think the current method is silly and needlessly confusing. Contributor should be there to credit the previous handlers for a package. Maintainer should just be the person who is currently in charge of keeping it working, whether binary or otherwise. Pacman's "Packager" data is kept for the binary files, making the current use of Maintainer redundant.
I think it should be changed to the more logical way. Anyone else have an opinion on that?
I think this is about the same as Callan said, and I gave him a 1+, which mees that I agree. Regards Stefan
On Thu, Oct 9, 2008 at 4:25 PM, Daenyth Blank <daenyth+arch@gmail.com> wrote:
Regardless of what it once was, I think the current method is silly and needlessly confusing. Contributor should be there to credit the previous handlers for a package. Maintainer should just be the person who is currently in charge of keeping it working, whether binary or otherwise. Pacman's "Packager" data is kept for the binary files, making the current use of Maintainer redundant.
I think it should be changed to the more logical way. Anyone else have an opinion on that?
+1 Hm, this is quite a "me too" post :) -- Anders Bergh
Bob Finch wrote:
If I understand the question correctly, and I may not.....
The answer is that you are a "contributor" of a PKGBUILD (et al) unless you are a TU, in which case you are "maintaining" the *binary* package.
Or at least that was what it was in at first.
Very best regards;
Bob Finch
On Thu, Oct 9, 2008 at 7:08 AM, Callan Barrett <wizzomafizzo@gmail.com <mailto:wizzomafizzo@gmail.com>> wrote:
On Thu, Oct 9, 2008 at 6:13 PM, Henning Garus <henning.garus@googlemail.com <mailto:henning.garus@googlemail.com>> wrote: > It does make sense, at least to me, to have your name and email > address in a PKGBUILD you currently maintain in AUR, even if you > didn't create it. However, I don't think there is any policy about > this. I have seen some "custom" tags, which I can't remember exactly > and of course much misuse of the maintainer tag - I am guilty of that > one myself. > > What one could do to prevent further misuse of the Maintainer tag > would be introducing a new tag like AUR-Maintainer and put it in the > wiki. Notice that this is just an example, I wouldn't be against > calling this different. It wouldn't stop the "Maintainer Tag problem" > all at once, but it might help reducing it.
I've never understood this rule, especially when I was enforcing it. Why aren't AUR maintainers allowed to use the maintainer tag?
-- Callan Barrett
So, I guess we can all agree on the following two: 1) Someone who submits a *new* package puts himself as "Contributor". 2) The developers and TUs that maintain binary versions of packages (and only they) may list themselves as "Maintainers". The above is what's written in the wiki as well. An alternative would be to list all people who modify the original PKGBUILD (i.e. updating/patching/restructuring) as "Contributors". I'm not fond of this approach as, considering an extreme case where 10+ individuals are involved, having a lengthy list of "Contributor" tags would look pretty damn ugly. This could be bent a bit though, by allowing a person that heavily modifies a package to include himself/herself at his/her own discretion. PS: I don't mind not being mentioned in an orphan package I've adopted, I just wasn't aware that the "Maintainer" tag is applicable only to Devs/TUs. I think that this point should be made clear on the wiki, in a relevant to the AUR article [1]. ---------------- [1] Maybe here: http://wiki.archlinux.org/index.php/AUR_User_Guidelines#Maintaining_Packages...
participants (13)
-
Aaron Griffin
-
Abhishek Dasgupta
-
Allan McRae
-
Anders Bergh
-
Bob Finch
-
Callan Barrett
-
Daenyth Blank
-
Evangelos Foutras
-
Henning Garus
-
Loui
-
Robert Emil Berge
-
stefan-husmann＠t-online.de
-
Xavier