Re: Perforce vs CVS
[email protected] (John Tobey)
| Newsgroups | perl.bootstrap |
|---|---|
| Message-ID | <[email protected]> |
Dan Sugalski <[email protected]> wrote: > At 01:03 PM 7/26/00 +1000, [email protected] wrote: > >I'd hate to see us create a barrier to entry which discourages people > >from reading or contributing to the perl source. > > Here's a bigger issue with any source control system we choose. We *must* > make sure that everyone, on any reasonable platform, has access to at least > timely snapshots of the code, and I'd really prefer a full-blown client. > > Don't forget, p5 runs on VMS, MacOS, Windows, AS/400, RISCOS, OS/2, and > OS/390 (at least) in addition to the various flavors of Unix. Forcing > someone to port a client just to participate, or to require a Unix box, is > a nasty thing, and one we should consider long and hard before we do it. > (Both perforce and CVS clients run on most of the named OSes AFAIK, and if > we have a pure perl client that's OK too) OK, this is starting to bug me. I would contribute more to Perl 5 if it were in CVS instead of Perforce. Having to use rsync to bring bleadperl up-to-date is a royal PITA. CVS *has* 3-way merge, and the Emacs client gives it a reasonable front end. > Having said that, we ought to choose the source control tool that meets our > needs best and gets in the way least. The P in perl stands for Practical, > not Political, after all... Who is "we" in "our needs"? The developers with write access, or contributors who just want to change a half-dozen files in various directories and send in the output of cvs diff -u? Best -John -- John Tobey, late nite hacker <[email protected]> \\\ /// ]]] With enough bugs, all eyes are shallow. [[[ /// \\\