Mail Archives: cygwin/2003/05/06/21:09:39
Christopher Faylor wrote:
> Since we no longer seem to have a cygwin-daemon maintainer, I'm wondering
> if it is time to start including cygipc in the main distribution. It
> seems like the lack of this functionality is a real handicap to people.
>
> FWIW, I'm willing to take over maintainership if that is a barrier.
If folks think that it would not be counterproductive to include cygipc
in the cygwin distro, then I've no objection.
But there's no need at this time for you to take over maintaining the
package, Chris -- I'll do it. Besides, it'd be nice to keep you clear
of the taint, just in case somebody (Elfyn?) suddenly starts submitting
cygdaemon patches again.
There's one big question tho. cygwin-daemon *really* wants a 64 bit
key_t (e.g. long long). Currently key_t is a typedef long (e.g.
32bits). Eventually, if cygdaemon is EVER going to have a chance,
cygipc needs to adapt to the 64bit regime.
It would probably be a good idea to make this (backwards incompatible)
change simultaneous with the introduction of cygipc into the main distro
-- but that requires coordination with a new cygwin1.dll release and
newlib. Plus buy-in from the current users of cygipc (which is the
Xshm-enabled Xserver that Ralf used to distribute with his kde port, and
postgres, AFAIK)
I already have patches ready to adapt cygipc to the 64bit key_t, and to
allow both cygipc and cygipc(64bit) daemons and libraries to coexist on
the same system. So that's ready to go.
Comments? Corrections? (it's been awhile; I may be forgetting stuff
concerning the 64bit key_t switchover...)
--Chuck
--
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Problem reports: http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ: http://cygwin.com/faq/
- Raw text -