[arch-general] Upstream bugs, patches

Nagy Gabor ngaba at bibl.u-szeged.hu
Thu May 1 16:43:29 EDT 2008


Hi!

I'm new to this list, and my English is far from perfect, please
tolerate it ;-)

I am/was a big fan of ArchLinux, I'm using it from 0.5 version, I also
made some little contributions to pacman, but now I noticed some
tendencies which I cannot accept. So I would like to hear the official
"standpoint" about a certain question, before I may drop my good old
distro. I also think that this question is important to all end-users.

(Aaron:) I know that I'm not popular here, but I hope that you won't
answer "against me" ;-)

Some foreword (sorry, maybe off, I would like to manipulate
thoughts :-D):
Basically I think that principles are _not_ rules. So things like KISS
and "vanilla packages" are good for defining ourselves in one sentence
(in wikipedia for example), but I don't really understand reasoning
like "foo violates KISS" (IMHO the reasoning should say, _why_ it worth
applying KISS here.). When I _describe_ myself as liberal, I won't
deduce my acts from this "rule"... (theoretical example).

So the question: Do/Should Archlinux packagers apply unofficial or
merged-but-not-yet-released patches to fix an existing _bug_ of a
released package?
An example: http://bugs.archlinux.org/task/5861
This is quite an old bug, and we are just waiting and waiting...

If the answer no, do packagers forward the bug to the official developer
or the end-user should forward his discovered bug?

An other small example, which is much less important; but I think this
belongs to the same "category", the reasoning is much more mysterious to
me, since the "patch" clearly cannot break anything here:
http://bugs.archlinux.org/task/10307 This was closed 4 minutes after
opening, so I couldnot discuss it (I wanted to revert my last sentence
there, this is not true now, I didn't want to lie). Evince developers
like that option, and they don't want to change it. (But I don't agree
with them here, of course). I simply cannot imagine any reason for
"won't fix" (apart from "lazyness" :-P). I'm pretty sure that
"implementing" it has _no drawback_, and at least 6 users (from the
number of requests) would like it. Again, it is a marginal issue (I put
my evince.desktop to NoUpgrade of pacman), but I would like to
understand the reason of close.

Bye

I hope, that I didn't hurt anybody, overall I think that AL is still one
of best distros around, and I must say a big "thank you" for your work.




More information about the arch-general mailing list