[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: outsider's perspective

From: Steve deRosier
Subject: Re: outsider's perspective
Date: Tue, 27 May 2003 15:32:49 -0700
User-agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.3) Gecko/20030313

Wait a sec...I was was not trying to suggest a solution (though I've got an idea, more in a sec), and if anything just a very specific solution to help with a very specific problem that one person was having.

I'm not that familiar with Meta-CVS, having never used it. Though perhaps I should take a look soon. The most basic CVS functionality has always been sufficient for my projects and frankly has more functionality than most people at my current company are willing to try (why I can't convince some people around here that CVS will always properly merge stuff OR complain I don't know; we've got people here that when they need to merge: grab a print out of the local file, overwrite their local with a fresh checkout, hand edit and then check in; they don't trust it to merge automatically even if it can and will; will the insanity never stop?).

A suggestion:
If Meta-CVS is essentially CVS II (or CVS III as someone pointed out as I was writing this), then why not try to merge the two projects (or parts) together. Make what is currently CVS into the "CVS engine" and make MetaCVS the "new" commandline portion of CVS. Bundle them together and call it CVS 2.0. Old clients can still connect to the CVS engine via pserver, ssh or whatever and it can handle old calls, while newer commands are handled by the CVS 2.0 client. Perhaps this requires some change in thought and maybe some people will need to upgrade, but as the usual response to peoples problems is "upgrade to the current version" I don't see how this is an issue.

Okay, so if this is not legitimate, let's hear a concrete plan about
how CVS can be extended to make a ``CVS II'' which is completely
backward compatible with CVS I clients, and works as well as Meta-CVS.
Better yet, let's see some code. It's not enough to propose alternative
ideas when the existing idea is already coded. The CVS mailing list has
seen more than a *decade* of idle discussions about this subject already.
As Don already pointed out, this is very unfair. I don't have any interest working with a project that is held tightly by a very small group of CVS gurus that feel that their way is the only way to do it. As I pointed out already, I was just trying to help an individual with a specific problem, not suggest that what CVS did or does is right or wrong (and as stated, I don't have an opinion about it, it works for me as is). But somehow despite the small and narrow scope of my help, I get pulled into this morass. You all should examine your intentions here; do you actually _want_ people to help? If so, you might want to make this place more open to new or other ideas. Why would I spend time trying to learn the CVS codebase and then make changes if I feel that whatever I do is simply going to be summarily rejected anyway? I've submitted patches on other open source projects (incl. Binutils and GCC) and frankly I'd rather spend my precious time helping out projects that want the help.

- Steve

reply via email to

[Prev in Thread] Current Thread [Next in Thread]