[darcs-users] newbie question about 'darcs send'

Andrei A. Voropaev av at simcon-mt.com
Fri Apr 29 11:40:46 UTC 2005


On Fri, Apr 29, 2005 at 01:59:23PM +0300, Adrian Maier wrote:
> On 4/29/05, Andrei A. Voropaev <av at simcon-mt.com> wrote:
> > 
> > I guess the failure here is the fact that send is trying to figure out
> > what shall be sent based on the difference between your current
> > repository and the "remote" repository that you shall specify on the
> > command line. In your case you can't access your remote repository
> > because it is in the office while you are at home.
> > 
> > Unfortunately I don't know the way, how to tell darcs to trust the user
> > input and create bundle using patches that user specified.
> > 
> > Silly idea. You may have 2 repositories at home and 2 at work. Then you
> > can do your changes, use send to create the patchfile comparing against
> > second directory, apply the changes to the second directory, take the
> > file to work and apply it to both repositories there. Hm. I feel like
> > simple carrying around the whole repository is more justified in your
> > case :)
> 
> Hm... so :   i need to tell 'darcs send'  that it should compare the
> current repository to an older version of the same repository ...
> 
> It's quite messy to have a total of 4 repositories . Very-very ugly ... 
> 
> 
> Since darcs has all the changes stored locally,  it guess someday 
> it could have the ability to create patch files without a direct connection
> to the other repository ... 
> Perhaps none of the developers needed this yet, and therefore 
> it's not implemented. 
> Do you have any idea if other versioning software is able to do it ?

I've used gnuarch and CVS before. But both of those (and most others)
require presence of the central archive. Which is essentially that
"second repository" on your machine (or accessible somehow). So you
don't win anything with those. So, I guess, darcs is no better than others
in this area.

-- 
Minds, like parachutes, function best when open




More information about the darcs-users mailing list