Inetd.c sigpause() --> __xpg_sigpause() not in uClibc

[Rob Landley]

On Friday 11 November 2005 03:20, Vladimir N. Oleynik wrote:
> John Z. Bohach,
>
> > Well said.  So for busybox, this means one of two choices:  revert to old
> > inetd.c which uses POSIX signals, then add IPV6 support, or, leave as is
> > (has IPV6 support) and convert BSD signals to POSIX signals.  Both paths
> > will lead to the same end result, I don't know which is easier, though
> > I'm suspecting the BSD --> POSIX signal conversion would be easier,
> > keeping in mind that I'm more familiar with signals than IPV6...
>
> There are three more choices:
> 1. Before being indignant, look, at last, on current corrections
> 2. Before offending another's work to try to help in the beginning the real
> contribution.
> 3. I can and remove all my work and then you do that want.

See, the reason I _haven't_ touched the inetd.c code (and instead expressed my 
concerns to you, on the list because I can't email you directly due to your 
spam filter) is that you are the maintainer, and I respect that.

I _don't_ want you to "take your ball and go home".  (I put quotes around it 
because I'm not sure the idiom will translate.)  I value your contributions 
to busybox, I think you generally do good work, and the project would be 
poorer if we lost that.

It seems that the problem here is one of design goals.  You see inetd as a 
port of an existing applet, not an independent applet.  I apply the same 
criteria to all busybox applets: if we can do better, we should.  (I also 
realize that developers have a finite amount of time and energy to put into 
work, so potential improvements remain undone indefinitely.  But when changes 
_are_ made, I think they should clearly be improvements over what was there 
before.)

If your response is that you don't have time and you would prefer someone else 
fix it in this case, I'm happy to accept that.  If your response is that you 
don't think there are any improvements to make, I'd like to understand why.

Rob