Re: [boost] Using CVS to get Boost work-in-progress

From: Phil Richards (news_at_[hidden])
Date: 2006-01-12 14:37:31

On 2006-01-12, Paul A Bristow <pbristow_at_[hidden]> wrote:
>  After downloading WinCVS (and CVSNT) I started it up,
>  but quickly decided something else was much more urgent ;-)

Can't say I find WinCVS particularly hard to use, though as has
been pointed out, TortoiseCVS is rather nice.  And I can cope with the
command line, so maybe I'm just odd...

>  It would be very helpful if someone could produce a "CVS/Boost for Dummies"
>  guide to configuring/using so that one can get the latest (and hopefully
>  greatest) for a **single project** without too much brain hurt.

Well, that might be interesting - the obvious thing to do is
something like
    cvs -d <whatever> -z3 co boost/boost/regex

But then, of course, you need boost/boost/regex.h and friends, and
also boost/libs/regex.  Then you need to scan all the files
you've just got and make sure the all the #include's from boost/boost
that are used are pulled down...  Oh, and you'll probably need the
Boost.Build stuff, too, just in case the jam files in the project use
new/fixed stuff.

>  (I don't want the full Boost update - I think?).

I'm not so sure... :-)

And subversion isn't going to help, either; not without quite a lot of
"meta" information being stored for each project (or something that will
recursively do the pull based on what is needed).

Just my thoughts...

phil
-- 
change name before "@" to "phil" for email