Re: Policy question about handling bugs and patches
On Thu, Mar 20, 2003 at 05:52:40PM +0100, Frank Lichtenheld wrote:
> - Who is responsible for the www.debian.org pseudopackage on the BTS?
> This means: Who decides what bugs to be tagged wontfix, moreinfo,
> help, done, what to be merged?
Well, that should be rather obvious, except for wontfix... and wontfix is
decided on a case-by-case basis.
> - Who decides what patches from the BTS should be applied?
Depends on the patch. :)
> - How should someone interpret silence after suggesting changes to the
> webpage? As 1) agreement, 2) rejection, 3) "no one has time, so wait".
> There are also some bugs in the BTS with no responses at all. It is
> worthy to produce patches for this bugs or are they implictly marked
> 'wontfix'?
Of course they're not!
You should simply think about it. (I.e. don't implement whatever is
requested without thinking about it.) If it's a large change, it probably
takes time for others to read it through. However, after some time has
passed, and your change doesn't do anything that could piss anyone off, you
can assume that nobody objects. And before you say "How could I possibly
know if something could piss anyone off?!" -- do your research and you it
won't fail you. :)
> #133800: Sorting of security items hides "new" old bugs
> #156679: www.debian.org: Recent security advisory list too short.
> Provided new version of recent_list, but no feedback yet.
The procedure seems obvious to me, but after Alfie you're the second person
to ask... find all mentions of recent_list, test them after the change. Fix
everything (if anything) that gets broken, barring objections commit, update
translations if possible.
> #177531: www.debian.org:
> http://www.debian.org/devel/passwordlessssh gives a dangerous
> advice about ssh
> Provided patch, but after that I've found some errors in it
> myself, so not ready for committing.
I thought it was pretty much okay.
> #126952: Add a "repeatmerged" button
> Provided patch, but no feedback
Obviously the idea's fine, since other similar things (like mbox=yes) are
already there. And the patch is fine because it does exactly the analogous
thing as the "no ordering by status or severity" thing does.
> #175474: www.debian.org: Move a link on MailingLists/ to make it
> easier to read
> Bug submitted by me with patch, one (mostly positive) response by
> Joy but patch not comitted (at the time I filed the bug, I
> couldn't do it myself)
My response is entirely positive on the issue you mentioned.
I was commenting on the generic issue.
> no number:
> New version of /devel/join/nm-step2 proposed, no feedback.
If you want to be sure, get a native speaker or two to proofread it, and
then commit it.
> #181872: www.debian.org: Inproper handling of special HTML
> characters in package descriptions
> Provided patch, discussed it with Joy, perhaps he applies it (?)
When members of the debwww group (which can commit there) get a round tuit.
Don't worry, your work is done :)
> #162588: packages.debian.org: please add a last-modified timezone
> Provided patch, but no feedback
Ditto.
(Tue Mar 18 19:53:13 2003 -> Tue, Mar 18 2003 19:53:13 +0000 sounds fine to
me at least.)
Overall, there's not much in the way of policy decisions in these issues.
Simply fix the bugs and make sure you don't screw anything up in the process.
If you do screw something up, fix it ASAP, and we will not behead
you^W^W^W^W^W everything will be fine.
--
2. That which causes joy or happiness.
Reply to: