dist-hg proof-of-concept ready for use

David Woodhouse dwmw2 at infradead.org
Wed Nov 15 05:09:42 UTC 2006


On Tue, 2006-11-14 at 10:50 -0500, Alan Cox wrote:
> On Tue, Nov 14, 2006 at 07:31:51AM -0800, Carl Worth wrote:
> > Alan, you've mentioned this trashing twice in this thread. Do you have
> > anything more than anecdotal evidence about this? If we are to take
> > your recommendation seriously, how do you recommend we measure the
> > robustness of git?
> 
> In my case I was doing a rebase and the box crashed. End of repository
> had to haul diffs out of log files by hand.

Do you have a tarball of this repository? I'd like to do a post-mortem.
I don't actually believe it was trashed, unless the crash of the box
_also_ scribbled on the filesystem, in which case all bets are off.

And a rebase on the _server_ side isn't a use case we're interested in
anyway -- server repositories we only ever push to; never merge or
commit locally. Not that it makes your alleged problem any less serious
in the general case, of course.

-- 
dwmw2




More information about the Fedora-maintainers mailing list