[Top][All Lists]

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

Re: How to set up a complex project

From: Alexander Kamilewicz
Subject: Re: How to set up a complex project
Date: Wed, 13 Jun 2001 15:24:48 -0500

Fran Fabrizio wrote:
> Hello,
> Until now I've only used CVS to set up relatively simple projects, and
> now I'm needing to do something a little more complicated.  I have a lot
> of questions about how to properly leverage CVS in this situation.  The
> situation is as follows:
> We're working with uClinux.  We have need to customize some of the stuff
> for our purposes and would like to keep our changes in CVS.  uClinux is
> distributed as a big tarball which, from what I can tell, is mostly
> source code with maybe a few binary files (definitely saw some .gif and
> .jpg, not sure if there were any compiled libraries or 3rd-party
> software in there as well).  It's pretty big, a 32 meg gzipped tarball.
> So, I was wondering, if we were to use their latest tarball as the base
> of our CVS (first question: will there be a problem having a CVS project
> that large?) and then edit and check in our changes off of that, I think
> that would work, with a few questions.

32 Mb?  Well, my repository is currently 13 Gb large with 146,000 files,
in which are 2 modules of 8 Gb & 3 Gb, so CVS can definitely handle it.

Your issue is going to be your end users.  If they do cvs get module,
how long are they going to have to wait for 32 Mb to transfer?  If
they're on the LAN, not a problem, if on a dial-up:  EEEK!

> - What happens when they release a new tarball?  We've already made
> changes to some of their stuff so do I create a new vendor branch, put
> their new stuff on it, and then merge the branches?

That's how my 8 Gb module works.  Run cvs import and look for
conflicts.  Learn to love cvs update -j.

> - Is there any way to check in the whole project at once if there are
> binary files in it?

If you mean cvs import, remember to set your cvswrappers to -kb those

If you mean cvs commit, well, sure, it's just "cvs commit".

> - Will I have to change all of the 100+ (rough estimate) Makefiles so
> that they compile somewhere else instead of right there in the working
> directory?

Depends on your setup, I guess.

> - Is this feasible at all or am I going about this all wrong?

It sounds OK to me - as long as you're not averse to cvs diffs.  :)

> I'm sure a lot of these are just newbie questions that expose my total
> lack of understanding regarding managing large, open source projects
> like this. =)  So if someone could shine some light on this subject for
> me, it'd be greatly appreciated!  Thanks.

Not at all.  It's all groovy.

Have you read Karl Fogel's CVS book?  Very recommended.  

Good luck!

reply via email to

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