cvs is designed to leak memory like a sieve and to exit from the middle of
deeply nested function calls. To make it a persistent process would require
a rewrite from the ground up.
Tony Hoyle established a "better cvs" project on SourceForge
(http://sourceforge.net/projects/bcvs/) to attempt to do just that (rewrite
the code cleanly without redesigning the fundamentals), but the project is
dormant at the moment. Carrying that off would probably be the first step
towards being able to rewrite cvs as a daemon.
What you probably would like to do is to have a daemon running that calls
cvs when requested to by a client, but isn't that just the same as what
inetd does?
Also, why run as a daemon? Is your server really creaking under the load of
people checking code in and out?
Just my simple-minded understanding of things. If I'm wrong or just plain
ignorant, sorry for wasting time.
At 03:40 PM 6/28/01, Matthew Von-Maszewski wrote:
>Ok, I stupid. Would someone be so kind as to tell me where the information
>exists for establishing cvs as a daemon, not inetd? If it is in the html
>manual, I apologies for being blind (and stupid).
_______________________________________________
Info-cvs mailing list
[EMAIL PROTECTED]
http://mail.gnu.org/mailman/listinfo/info-cvs
