This is the mail archive of the cygwin@cygwin.com mailing list for the Cygwin project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: An easy (but possibly time-consuming) test case for Make hanging


>  Date: Wed, 7 Mar 2001 11:37:26 -0500
>  From: Christopher Faylor <cgf@redhat.com>
>  
>  Dunno.  Frankly, I think that if you want to get this working you should
>  debug it yourself.  That's just a pragmatic suggestion.

I tried.  Cygwin's signal-handling code is extraordinarily complex.
It would take me at least a couple of weeks dedicated to this just to
begin to understand how it all works.  And I suspect that just
beginning to understand how it all works would not be enough to be
able to figure out this bug.

Is the there an overview of Cygwin's signal-handling architecture
somewhere that I can start with?  I could not find one.  Trying to
understand code this complex with only the source itself as a guide
is sort of like the story of the blind men trying to describe what an
elephant looks like.

>  I sincerely doubt that anyone else is going to be interested in fixing
>  your problem for you.

Yes, it is "our problem," in that it is affecting our work, but it is
also a bug in Cygwin.  Most people who maintain software, free or
otherwise, appreciate it when people report bugs.  Most software
maintainers understand that they know much more about their code than
the people who use it, and thus if a user reports a bug and gives
instructions for reproducing it, the software maintainer is more
qualified at that point to debug it than the user is.  Certainly, in
all the years that I have been maintaining free software, any time
someone has sent me a bug report and explained how to reproduce the
bug, I have taken the time to attempt to reproduce it, and to fix it
if I could indeed reproduce it.

But hey, if you're not interested in fixing bugs in the software you
maintain, that's your right.  Whatever.

>  The other alternative is to purchase a support contract with Red Hat and
>  pay for our professional services.

It's so nice to see a free software maintainer demanding money from a
user before being willing to fix a bug.  That's so in the spirit of
free software.

In any case, I have already suggested this idea to our management.  If
I can convince them to consider it, we will be contacting RedHat
shortly for a quote.

>  If someone submits a patch or has a specific question beyond "Here's an
>  strace log.  What's going on???" I'll respond to it with my usual
>  alacrity.

But Chris, you've already responded about this "with your usual
alacrity."

  jik

--
Want to unsubscribe from this list?
Check out: http://cygwin.com/ml/#unsubscribe-simple


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]