Risto Kotalampi Reporter
	Description • 24 years ago

Subject says it all... this would be really useful for browsing cvs servers.

	Axel Hecht Assignee
	Comment 1 • 24 years ago

Hi,
I've been able to checkout a file via a perl script, so basically I understood
the protocol.
The protocol is specified in docs/cvsclient.ps in the cvs source.
One important trouble maker is the fact, that I need the CVSROOT to
authenticate, so there are a few variables that I need set in the protocol.
If we do this halfway decent, this protocol could be extended to check out
specific versions or diffs, too.

I read ftp://ftp.isi.edu/in-notes/rfc2616.txt and 3.2.2 states that URLs look
like 
http_URL = "http:" "//" host [ ":" port ] [ abs_path [ "?" query ]]

So setting CVSROOT via ?CVSROOT=/cvsroot looks ok to me.

Username and password should go with pserver://user:passwd@some.host.org/ and 
default to anonymous:anonymous.
CVSROOT defaults to /cvsroot?

oh, should we call it cvs: or pserver: ? It's pserver after all, so I vote for
that.

Axel

	Axel Hecht Assignee
	Comment 2 • 24 years ago

hey, forgot mime type.
text/plain for normal files, and application/octet-stream for -kb?

Axel

	Darin Fisher
	Comment 3 • 24 years ago

Take a look at what sourceforge.net does for cvsweb.cgi

	Dan Mosedale (:dmosedale, :dmose)
	Comment 4 • 24 years ago

You might consider using protozilla <http://protozilla.mozdev.org/> for this.

	benc
	Comment 5 • 24 years ago

mass move, v2.
qa to me.

	Bradley Baetz (:bbaetz)
	Comment 6 • 23 years ago

*** Bug 135823 has been marked as a duplicate of this bug. ***

	(not currently active) Ted Mielczarek
	Comment 7 • 15 years ago

Why was this ever anything but WONTFIX? :) (an extension at best, I would think)