What we did at UBZ

[James Blackwell]

On Wed, Nov 16, 2005 at 09:11:39AM +0100, Jan Hudec wrote:
> On Wed, Nov 16, 2005 at 13:53:31 +1100, Martin Pool wrote:
> > On 15 Nov 2005, Matthieu Moy <Matthieu.Moy at imag.fr> wrote:
> > 
> > > It's a relatively common naming convention to put a _ at the beginning
> > > of an identifier to mean "the user shouldn't use such identifier" (not
> > > for filenames, but typically in C++). Forbidding _ at the beginning of
> > > branch names sounds to me to be an acceptable limitation (you
> > > shouldn't forbid just "+storage" or "_storage", but "+*" or "_*" so
> > > that future archive format can use "_whatever").
> > 
> > I'd be happy with that.
> > 
> > I think darcs allows you to have either .darcs or _darcs in the working
> > directory; do people want that too.

> 
> On Unices it does not make sense. But on Windows there is some br0ken
> software that does not like files with names starting with '.' lying
> around. There even exists a modified version of subversion that uses
> '_svn' for exactly this purpose. So it might make sense to support using
> '_bzr', though it's not high priority.

Two of the uses for an exposed rcs directory is to hunt around for the
sorts of archives^wbranches that are laying about in google, or to tell by
looking at an http rendered directory listing whether or not you're
looking at a tree dump or an rcs.

Case in point: a trivial look at
http://people.ubuntu.com/~robertc/pqm/trunk doesn't expose that it's a bzr
branch.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
Url : https://lists.ubuntu.com/archives/bazaar/attachments/20051116/17abadd9/attachment.pgp