Re: [bug-gcron] gcron status question

[Ryan M. Golbeck]

I restored the CC to bug-gcron, I hope that is alright.

> > It is alive somewhat.
> 
> What does that mean?  Please be more specific.

I mean I work on things when I have time, but I haven't had a lot of
time lately.  Things are a bit stalled at a design point, see below.

> OK.  I can see (a) a drop in replacement without Vixie
> cron's holes and (b) a super cron with all the stuff I
> wrote about.
> 
> New functions should be made available only to those who
> need them and know how to turn them on.

Yes.  This is the idea.  New functions should not interfere with
normal Vixie cron compatability.  Instead, there will be an addition
of a gcron config file, or some other similar mechanism, for managing
extra features.

> > I, of course, welcome help.
> 
> I can program in C, C++, any Unix shell except the California
> shell by choice, and a whole bunch of Unix text manipulation
> tools.

gcron will be written mostly in C and will try to keep library
dependance to a minimum because it is an core system utility.

> I design, and I can write standard English.

Design is good, maybe you have comments for mine, then.

> I have a Mac OS X box, and if pressed I can boot up Linux,
> OpenBSD, BeOS, and Windog 2000 at home on the same PC --
> different discs.

You should be able to build the stuff fine on MACOSX, provided you
have the GNU autoconf tools setup there.  gcron should be portable so
developing there would help with that. (This goes for s/MACOSX/Other
POSIX system/ as well).

> Is the design complete?

Not entirely.  I have not decided how communication with the cron
daemon should be handled reliably, and without using setuid binaries.
Please see [1] for an explanation of what I mean.  Your comments are
welcome.

> Is there a list of tasks?  Milestones?

There are some documents in the doc/ subdirectory off of the cvs root
in texinfo format.  These documents attempt to explain some design
decisions and motivations for gcron, but are not wholly complete.

There is a fairly broad task list on the savannah project webpage, but
nothing concrete.  Once this one design issue is settled, it should be
more clear where to move to.

> What would you like me to do?

Help me decide on part of the design and then we can move on.

> Do you want a copy of my resume circa 1996?

No need.

thanks,
ryan

[1] http://mail.gnu.org/archive/html/bug-gcron/2003-01/msg00001.html