[aur-general] Mandatory comment along with OOD-flag
good day! recently i experienced some wild out-of-date flagging without any comments or even backupped reasons for codelite [1] from different users (even multiple times by the same) over the past two weeks. someone has been kind enough to help me unflagging these. however this is not only very annoying without any reasons (comments or email of the flaggers) whatsoever but also kind of lighted some concern about the carelessness of how people are flagging. they don't seem to even read the latest comments of the package or try to get a grasp of their reasoning ([1] is the stable version. they probably want the svn version which is also available). i thought about that and putting a fulltext description in the pkg-desc field is probably not what it's meant to be. so what came to my mind was a mandatory comment along with flagging to get an idea, why the person thinks it's out of date. that would make them think before flagging and also make them come back and read the response and not flag it out-of-date again. anyone else having similar issues? or am i overreacting here? regards, nem [1] https://aur.archlinux.org/packages.php?ID=18809
2011/9/13 nem <nem@ikitten.co.uk>:
good day!
recently i experienced some wild out-of-date flagging without any comments or even backupped reasons for codelite [1] from different users (even multiple times by the same) over the past two weeks. someone has been kind enough to help me unflagging these.
however this is not only very annoying without any reasons (comments or email of the flaggers) whatsoever but also kind of lighted some concern about the carelessness of how people are flagging. they don't seem to even read the latest comments of the package or try to get a grasp of their reasoning ([1] is the stable version. they probably want the svn version which is also available).
i thought about that and putting a fulltext description in the pkg-desc field is probably not what it's meant to be. so what came to my mind was a mandatory comment along with flagging to get an idea, why the person thinks it's out of date. that would make them think before flagging and also make them come back and read the response and not flag it out-of-date again.
anyone else having similar issues? or am i overreacting here?
regards, nem
IMO, shouldn't be mandatory. If you have a stable version package and user flags it because there is a new devel version or the PKGBUILD has a bug, then the problem is a misunderstanding of the OOD feature usage by the user. In this case, leave a educational comment.
On Tue, 13 Sep 2011 08:58:34 -0300 rafael ff1 <rafael.f.f1@gmail.com> wrote:
IMO, shouldn't be mandatory. If you have a stable version package and user flags it because there is a new devel version or the PKGBUILD has a bug, then the problem is a misunderstanding of the OOD feature usage by the user. In this case, leave a educational comment.
i did that twice (!) during the past two weeks and still some people kept reflagging it without any notification. it seems that the comment section doesn't get a whole lot of attention if it comes to OOD flagging and the mandatory comment would be a way to ensure that. so if some other maintainers experience the same it would be something nice to have.
so if some other maintainers experience the same it would be something nice to have.
Let the ood flag be there, it doesn't hurt anyone if that's not true. You already stated it's the latest stable release in the comments, and people will eventually check and risk installing it either way. That will avoid any trolls be fed. Then you can start thinking about a feature request for improving the AUR's functionality in a way that makes some more sense. People who would flag a package out of date would write a comment like "because it's autof date" The software in question seems to be used by ignorant people that haven't finished their first book on a programming language. Besides the constant unflagging and thinking about what's wrong with the system, common sense would also suggest you consider orphaning. Umm, a wiki paragraph that empowers maintainers to let accounts be disabled for this way of trolling, would also be nice. I just don't know how far disabling accounts and/or logging of the AUR goes already, maybe that would be a place if you really want to get serious. cheers. mar77i
On Tue, Sep 13, 2011 at 04:19:27PM +0200, Martti Kühne wrote:
so if some other maintainers experience the same it would be something nice to have.
Let the ood flag be there, it doesn't hurt anyone if that's not true. You already stated it's the latest stable release in the comments, and people will eventually check and risk installing it either way. That will avoid any trolls be fed. Then you can start thinking about a feature request for improving the AUR's functionality in a way that makes some more sense. People who would flag a package out of date would write a comment like "because it's autof date"
I kind of like the idea of adding a comment field. I wondered why some package has been flagged out-of-date a couple of times and I think having an additional comment might make things easier. -1 to making it mandatory, though. Martti already gave a reason. Anyway, could you please file a feature request?
On Wed, 14 Sep 2011 12:23:54 +0200 Lukas Fleischer <archlinux@cryptocrack.de> wrote:
I kind of like the idea of adding a comment field. I wondered why some package has been flagged out-of-date a couple of times and I think having an additional comment might make things easier. -1 to making it mandatory, though. Martti already gave a reason.
Anyway, could you please file a feature request?
i'm not sure what you have in mind by 'additional comment' - but if someone _wants_ to make a comment, he probably will anyway in the normal comments. another commentfield just for OOD would be redundant, or am i missing your point?
On Fri, Sep 16, 2011 at 11:26:35AM +0200, nem wrote:
On Wed, 14 Sep 2011 12:23:54 +0200 Lukas Fleischer <archlinux@cryptocrack.de> wrote:
I kind of like the idea of adding a comment field. I wondered why some package has been flagged out-of-date a couple of times and I think having an additional comment might make things easier. -1 to making it mandatory, though. Martti already gave a reason.
Anyway, could you please file a feature request?
i'm not sure what you have in mind by 'additional comment' - but if someone _wants_ to make a comment, he probably will anyway in the normal comments. another commentfield just for OOD would be redundant, or am i missing your point?
A separate comment box on a separate "flag package out-of-date" page (similar to what archweb does) to * Allow for sending that comment alongside the notification mail. * Allow for linking the comment to the out-of-date flag (allows TU to quickly check for a reason without having to read all the comments). * Show some extra information about what "out-of-date" means.
2011/9/20 Lukas Fleischer <archlinux@cryptocrack.de>:
On Fri, Sep 16, 2011 at 11:26:35AM +0200, nem wrote:
On Wed, 14 Sep 2011 12:23:54 +0200 Lukas Fleischer <archlinux@cryptocrack.de> wrote:
I kind of like the idea of adding a comment field. I wondered why some package has been flagged out-of-date a couple of times and I think having an additional comment might make things easier. -1 to making it mandatory, though. Martti already gave a reason.
Anyway, could you please file a feature request?
i'm not sure what you have in mind by 'additional comment' - but if someone _wants_ to make a comment, he probably will anyway in the normal comments. another commentfield just for OOD would be redundant, or am i missing your point?
A separate comment box on a separate "flag package out-of-date" page (similar to what archweb does) to
* Allow for sending that comment alongside the notification mail.
* Allow for linking the comment to the out-of-date flag (allows TU to quickly check for a reason without having to read all the comments).
* Show some extra information about what "out-of-date" means.
Lukas, I might be wrong, but this is sort of how flagging out-of-date works in official repos, right? It sounds good. Better than being mandatory to add a comment in the comments list after flagging ood - which I disagree.
Excerpts from rafael ff1's message of 2011-09-21 00:07:47 +0200:
2011/9/20 Lukas Fleischer <archlinux@cryptocrack.de>:
On Fri, Sep 16, 2011 at 11:26:35AM +0200, nem wrote:
On Wed, 14 Sep 2011 12:23:54 +0200 Lukas Fleischer <archlinux@cryptocrack.de> wrote:
I kind of like the idea of adding a comment field. I wondered why some package has been flagged out-of-date a couple of times and I think having an additional comment might make things easier. -1 to making it mandatory, though. Martti already gave a reason.
Anyway, could you please file a feature request?
i'm not sure what you have in mind by 'additional comment' - but if someone _wants_ to make a comment, he probably will anyway in the normal comments. another commentfield just for OOD would be redundant, or am i missing your point?
A separate comment box on a separate "flag package out-of-date" page (similar to what archweb does) to
* Allow for sending that comment alongside the notification mail.
* Allow for linking the comment to the out-of-date flag (allows TU to quickly check for a reason without having to read all the comments).
* Show some extra information about what "out-of-date" means.
Lukas, I might be wrong, but this is sort of how flagging out-of-date works in official repos, right? It sounds good. Better than being mandatory to add a comment in the comments list after flagging ood - which I disagree.
Still this assumes that the person who flags will leave a sensible comment. There's no way to guarantee that this will happen, he could leave it empty or write asdf or something equally helpful. Yes, having a sensible reason sent with the out-of-date mail would be nice, but such a field wont make sure it will help. Currently you get the reason in a second mail if the person who flags out-of-date leaves a comment, big deal. I'm not convinced we are talking about a real problem here, and I don't think the proposed solutions would really improve matters in any significant way.
Excerpts from nem's message of 2011-09-13 11:41:56 +0200:
good day!
recently i experienced some wild out-of-date flagging without any comments or even backupped reasons for codelite [1] from different users (even multiple times by the same) over the past two weeks. someone has been kind enough to help me unflagging these.
however this is not only very annoying without any reasons (comments or email of the flaggers) whatsoever but also kind of lighted some concern about the carelessness of how people are flagging. they don't seem to even read the latest comments of the package or try to get a grasp of their reasoning ([1] is the stable version. they probably want the svn version which is also available).
i thought about that and putting a fulltext description in the pkg-desc field is probably not what it's meant to be. so what came to my mind was a mandatory comment along with flagging to get an idea, why the person thinks it's out of date. that would make them think before flagging and also make them come back and read the response and not flag it out-of-date again.
anyone else having similar issues? or am i overreacting here?
regards, nem
I think you just had bad luck. There are people who make an error when flagging out of date, but at least in my experience this happens rarely. Most of the time it's legit and the reason is obvious (new release). If the reason is non-obvious then the 'flagger' can leave a normal comment. Luckily the relatively new default on AUR is that the maintainer gets notified in case of a new comment, so you'll get a mail and all is well. However, my experience is with a larger number of not so popular packages, it's not necessarily the same as with few popular ones. So again, I think this was bad luck. See whether this annoyance keeps going on or not. If it does keep up we should think about a change. Regards, Philipp
participants (5)
-
Lukas Fleischer
-
Martti Kühne
-
nem
-
Philipp Überbacher
-
rafael ff1