[asterisk-dev] Optional api and weak symbol problem

[Tzafrir Cohen]

On Tue, Nov 08, 2011 at 01:18:50PM +0200, Yaroslav Panych wrote:
> No, it is childish to introduce trash into source code(trash - is how
> I named cygwin compatibility code) when it still buggy without it. 

It was introduced by someone who, at the time, contributed large
ammounts of code. He moved along to other ventures and (sadly?) noone
else picked that up. Do you want to?

Terry explained what it would take to get a platform supported
(implicitly, by naming those tasks and stating he won't do them). It's
not that difficult, if you actually care for the code.

The basic thing is that the code actually builds. All the time. If a
specific commit breaks it, you should hopefully provide at least a
workaround. If not: a bug report.


So assuming that (a) the code actually builds and (b) basically works on
cygwin and (c) someone cares enough to maintain it: keeping it
maintained is doable.

Do you want to work on (a) and (b)? Do you want to keep wasting mailing
list bandwidth on pointless arguments? If you want to work on (a) and
(b), do you intend to commit to (c)?

> I
> don't meant Kevin or somebody else in person, but whole team. Asterisk
> already contains a lot of potential deadlocks, resource leaks,
> etc(judging of their number they have fixed each release). They do not
> spend any time on cygwin to recheck patches, so they cannot be sure
> one or another patch will no introduce any deadlock(conclusion "it
> seems it will not harm" is not acceptable).

I don't tell "them" what to do.

-- 
               Tzafrir Cohen
icq#16849755              jabber:tzafrir.cohen at xorcom.com
+972-50-7952406           mailto:tzafrir.cohen at xorcom.com
http://www.xorcom.com  iax:guest at local.xorcom.com/tzafrir