[nycbug-talk] upcoming meeting

Charles Sprickman spork at bway.net
Sat Jan 6 23:58:12 EST 2007

On Sat, 6 Jan 2007, George R. wrote:

> As someone noted earlier, it would be useful to entertain pre-meeting
> questions and comments.  So here goes. . .
> Ivan approached me at NYCBSDCon about doing this meeting, so here's the
> blurb from the www site:
> Ivan Ivanov on The Version Control System Subversion

Sounds like a good one!

I don't use cvs or svn much for day-to-day stuff beyond storing some 
configs in there (unix hosts and routers).  However, I did recently do 
some work moving a client from cvs to subversion, and it's still somewhat 
fresh in my mind.  Two things that I think are essential if you've got a 
background in cvs:



That section of the (excellent) free, online svn book is a really good 
read.  While svn and cvs try to present the user with similar/compatible 
tools, the way things work behind the scenes are radically different. 
That section helps explain how and things didn't click for me until I read 
that.  The folks actually using svn were not quite understanding why a 
change in one file bumped the revision number on every file in the 

One also needs to understand that branches and tags are much more 
loose/arbitrary than in cvs and are more of a "virtual" grouping of files 
than they are in cvs.



Since many folks are not starting from scratch, but moving from cvs to 
svn, that tool is very cool.  It gets quite hairy if your cvs repo is kind 
of a mess, but it tries really hard to get things right.

I'll also add a third, which is more of a question...  With cvs, it's easy 
to set an environment variable in your shell to tell cvs where the repo 
lives.  There is no such thing in svn.  What kind of hackery is out there 
for those that don't like typing svn urls with every command?


More information about the talk mailing list