[elephant-devel] Re: public_html/darcs in c-l.net elephant

[Robert L. Read]

I know this sounds weird coming from a LISP developer, but in the
interest of leveraging mainstream knowledge I insist that we used either
darcs or subversion.



On Sat, 2007-03-10 at 16:23 +0000, Ian Eslick wrote:

> Well, its not as if we have that many developers.  I think the idea is that we can have individual developer branchs, a stable branch and a release branch.  Developers update stable when a set of changes are, uh, stable.  That way we can avoid branches getting too far out of date and sidestep the simultaneous change issue.
> 
> Ian
> 
> Sent via BlackBerry from T-Mobile  
> 
> -----Original Message-----
> From: "Attila Lendvai" <attila.lendvai at gmail.com>
> Date: Sat, 10 Mar 2007 17:07:30 
> To:eslick at csail.mit.edu
> Subject: Re: public_html/darcs in c-l.net elephant
> 
> > What conditions lead to performance issues?  Volume of patches or type of patches?
> > Is it due to haskell or the underlying algorithms?
> 
> in my experience the only problem was when two branches were actively
> developed and the same files at the same positions received several
> changes. then trying to pull these (nested conflicting) changes into
> the other repo ended up in some exponential algorithms instead of
> bailing out and marking the conflicts.
> 
> but other then this, those "performance issues" everyone talks about
> are non-existent.
> 
> i've asked the darcs guys, and they know about this issue and have
> solutions, too, but as far as i know it's not yet implemented.
> 
> hth,
> 
> _______________________________________________
> elephant-devel site list
> elephant-devel at common-lisp.net
> http://common-lisp.net/mailman/listinfo/elephant-devel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.common-lisp.net/pipermail/elephant-devel/attachments/20070310/77dc053c/attachment.html>