Mail Archives: djgpp/1997/10/08/11:51:39
Eli Zaretskii wrote:
> > How about a list of bugs in the downloaded version? That wouldn't take much
> > to update, just a list of all 'open' bugs in te current system, updated say
> > once a month.
>
> You mean, in a separate file? What are the chances that people don't
> forget to download another non-essential file?
No, I meant in the zip files you need to start (specifically in this case
the library binary zip file - anyone who starts by downloading the library
source before the binary isn't going to have much luck).
> Anyway, it still calls for a volunteer to maintain such a list. For
> starters, people will need to know who is that person, and make a point
> of reporting the bugs to him/her. Right now, it's a mess: some people
> post to c.o.m.d., others write to DJ or to me, still others submit to the
> bug-tracking system and never tell anybody else, etc.
Do all bugs go into the tracking system eventually? If not, then there
is a real problem (i.e. if bugs are being fixed without being tracked).
I don't know how the BTR is organised, but surely it must be possible to
extract the bug reports from it in text format (or something convertable
to text format, like TeX or HTML)?
> DJ has decided to separate the core DJGPP from the rest, so now DJGPP
> releases are not necessarily synchronized with the compiler and the rest
> of the GNU tools.
I suspect there's some confusion about what exactly is 'DJGPP', then.
I've always taken it to mean the entire system, comprising the compiler,
libraries, and binutils, and that seems to be the sense of the recent
"what does DJGPP stand for?" cpompetition, the words "programming
platform" being used a lot.
> > Indeed. But isn't this partly because of trying to do all the patches
> > at once? I would have thought that incremental releases, with a couple
> > of easily-tested patches in each one, would be better. That's the way I'm
> > used to releasing code in a multi-user project in my job.
>
> I suspect that in your job (as me in mine) you don't have to ask your
> wife whether you could spend several hours making an update...
I don't at home either - no wife <g>...
But it shouldn't be just one person who has to do it, surely? Isn't it a
distributed system, with several developers doing the work? Or does DJ
have to do all the 'releasable' builds himself?
> The point is that it takes time anyway. Obviously, a lot of small patches
> done separately will incur a lot of overhead. Whether it is easier to
> install the patches one by one or all of them together, is a function of
> your arrangements with your Powers That Be at home. It depends on the
> personal circumstances of each one of us.
Not just that. My point is that it's a lot easier to fix a couple of
unrelated bugs than to have to suddenly have to get them all in at once,
with undoubtedly some modules having more than one patch.
> It is doable, no doubt. If you are telling that you are willing to try
> to make it happen, then it is up to you to define the procedures for
> reporting library bugs. Just coordinate with DJ and make the announcement.
> Of course, if you want to discuss what is the best format and procedure,
> before you decide, that is OK too.
It seems he's the one who knows the bug tracking system, anyway. At very
least, any bug which is entered into the tracking system should generate an
email to the maintainers, preferably to a special list - from what I've seen
of it djgpp-workers might be usable as a list for that purpose, or another
list might be better. Perhaps this should be transferred to djgpp-workers
anyway - thoughts?
Chris C
- Raw text -