[p4] p4 uri

[Nick Barnes]

At 2007-08-05 02:37:53+0000, "Jeff A. Bowles" writes:
> Richard Brooksby, in his Perforce 1999 User Conference paper,
> talked about having depots named
>     "machinename.companyname.com"
> This made his Perforce pathnames look like:
>      //machinename.companyname.com/dir1/dir2/file.txt
> And he could easily prepend "http:" to it and feed it to
> a web  browser. The P4DB / P4Web app on that machine would
> take it from there.

As it happens we actually use Apache on the server
(info.ravenbrook.com), and have a fairly simply clientspec which puts
(more-or-less) the whole repository in the DocumentRoot, with a cron
job to keep it synced. Disk being as cheap as it is, this is the easy
way to go, and it means that some parts of the repository can
automagically work as CGI.  Very useful.  The publishable parts of our
repository are also synced to the machine www.ravenbrook.com, with
(more-or-less) the same directory structure.  So, for instance:

      //info.ravenbrook.com/project/p4dti/...

is served at

 http://info.ravenbrook.com/project/p4dti/

and also at 

  http://www.ravenbrook.com/project/p4dti/

Nick Barnes
Ravenbrook Limited