I guess I confused the matter when I said api... What I wanted to see was a
simple example of using the protocol through a socket...
Thanks
GaRy....
Donald Sharp wrote:
>
> I said 'there is no api' because the guy sounded like he wanted
> a api specification/header files and a library to link to. There is no
> library to link to...and no standard header file. Yes if you start poking
> under the covers and start pulling apart cvs you can get what you want....
> but that's not exactly usefull or quick....
>
> This is why i suggested the fork/exec method
>
> donald
> On Wed, Mar 29, 2000 at 08:30:41AM -0700, Tobias Weingartner wrote:
> > On Wednesday, March 29, Donald Sharp wrote:
> > > there is no api. Cvs client server works via STDIN/STDOUT redirection.
> > > This is why you need to have cvs run by inetd. Inetd provides the
> > > hookup of the pipes between STDIN/STDOUT and the sockets....
> >
> > Ok, there are at least 1 wrong statement in here. There is no "api", but
> > there is a client-server protocol. It is documented within the distribution.
> > Not the best, but it is there. You may be able to "snarf" parts of the
> > code to do that api from within the CVS code.
> >
> > Also, you don't NEED to run the CVS server via inetd. There are other
> > ways of invoking the server such that STDIN/STDOUT are connected to the
> > client in the right way.
> >
> >
> > > Figure out what you want to do and fork/exec the apropriate cvs command.
> >
> > This is usually easier, althout other CVS look/work-alikes have been
> > implemented using just the client-server protocol. Have a look at
> > jCVS and the like.
> >
> > --Toby.
> >
> >
--
Gary Pinkham
Reasoning, Inc. Voice: 781-359-3132
One New England Executive Park Fax: 781-229-2770
Burlington, MA 01803 Email: [EMAIL PROTECTED]
