[NTLUG:Discuss] any CVS experts out there?

[kbrannen@gte.net]

Paul Elliott wrote:
> 
> Any CVS experts out there?
> 
> I want to import a project that has both text and binary
> files:
> 
> 51.jpg         banner.html   crset.jpg     decadence.html
> help.html      opinion.html  socrel.html   why.html
> adherent.html  card.html     crypto.html
> diegeda2.mid   index.html   regcom.cgi    toler.html
> art.html       comment.html  culture.html  gk1.gif
> maj16.jpg      set.jpg       tromba.mid
> 
> The .jpg, .gif, and .mid files are binary. The .html files are not.
> 
> Is the following the correct command to do the import?
> 
> import -m "transfer files from OS/2 " -W "*.mid -k 'b'" -W "*.jpg -k 'b'" -W "*.gif -k 'b'"  cfi pelliott start
> 
> If cfi is the project name. Pelliott is the vendor and start is the tag?

That looks about right (minus the missing "cvs" at the front. :-)

To simplify your life a bit, there is a file in the CVSROOT dir that allows
you to specify extensions that will be treated as binary, so you don't have to
do all the "-k" stuff.  I don't remember the filename just now, and I'm not at
work.  Grep for "jpg" in CVSROOT and it should be obvious (as that's an
out-of-the-box one treated as binary).  You can also check the docs for the
filename too.

HTH,
Kevin