[BusyBox] Problem with mkdir of busybox-0.60.5

[Allan (pls reply to list only)]

Hey Glenn;

First, let me say "cool head"; very sane cool-headed response.  Further
inline.

Glenn McGrath wrote:
> On Tue, 26 Nov 2002 17:22:38 -0500
> "Allan (pls reply to list only)" <discard at chickenandporn.com> wrote:
> > Hey Glenn;
> >
> > > Im not sure if you realise, but the stable branch is on its last legs.
> > >
> > > The unstable branch has been in development for over a year now, Erik
> > > has said he intends make a stable release from the unstable code and
> > > remove the current fork soonish.
> >
> > The problem with that logic is that the "stable" busybox is actually
> > "the busybox that appears to have been tested"
> >
> 
> You do have a point here.
> 
> > The problem you state -- people reporting bugs in a 31-day-old release
> > (BB-0.60.5 on 20021026) -- can be avoided by simply cutting a new
> > release and have some organized, coherent testing... maybe even
> > self-tests in the source base.  Wouldn't that be scary?
> >
> 
> Its not a problem reporting bugs in any timeframe, i value bug reports
> very much, its time managment.

Yeah, that's a common thing with Open Source.  :(

> The bug could be solved by cutting a new release from unstable as well,
>  is it better to spend a little bit of time patching an old branch, or a
> bit more time and releasing the unstable branch.

Oh, if your point is that you prefer to release the unstable, I
wholeheartedly agree.


> I have spent a fair bit of time in the last few days helping to get
> unstable upto scratch, and personally think it is very close.
> 
> I dont co-ordinate releases though, Erik is pretty busy right now, if we
> can all exercise a little more patients untill he has some time then we
> can get unstable out the door, removing the branch in cvs should make it
> easier to manage in the long run.
> 
> There are tests in ./testsuite it could do with some more tests added to
> it if you have time :)

Yeah, since I'm not that savvy inside the busybox hood, I think that's
where my input would go.  I do support that approach, too; self-tests
make bugs disappear for good :)

Allan