Re: The GNUstep Roadmap

[Richard Frith-Macdonald]

On Friday, October 24, 2003, at 09:32 AM, Markus Hitter wrote:

> Am Freitag, 24.10.03 um 06:01 Uhr schrieb Chris B. Vetter:
>
> > Just consider it. You spend a sunny Saturday tracking down a bug that 
> > is
> > apparently particular to your system (for whatever reason), you find 
> > it,
> > you fix it, you create a patch, you send it in -- and nothing happens.
>
> Totally different experience here. Each patch I've commited so far was 
> handled within a few days. Thanks a lot to the maintainers.

Realistically it's going to vary a lot ...

Sometimes developers are all just very busy on other stuff.
Sometimes a patch is in an area that none of the developers knows 
anything much about ... so everyone expects someone else more expert to 
deal with it.
Sometimes it's a really simple thing fixing an obvious bug and seeming 
to have no repercussions elsewhere ... so it can be fitted in in a few 
minutes.
Sometimes it's an obvious fix ... but there is a lot of non-obvious 
code which will be broken by making the fix, so it gets left until 
people can find time to update all that other code too.
Sometimes it's not an obvious fix at all .. and the developers 
misunderstand the patch or simply can't be sure it's correct.
Sometimes, for all the work that's been put in, the patch is actually 
wrong.

Irrespective of all these 'sometimes'  ... I for one am grateful for 
patches and bug reports on savannah.  The bug reporting mechanism beats 
simply sending to the mailing list for all but the most trivial 
problems ... it keeps track of things and the reporters can see the 
state of the report, so if it's been overlooked (nobody is working on 
it for a long time etc) the reporter can  nag about it, and if it's 
closed but the reporter disagrees (eg because reporter/developer 
miscommunicated, or because the developer did what they thought was a 
complete fix, and missed some aspect of the problem) they know about it 
and can raise the issue.  Things sent to mailing list all too easily 
get overlooked/forgotten.

What do people think about actively discouraging direct mailing of bug 
reports to the bug and discuss mailing lists?