Roland McGrath <[EMAIL PROTECTED]> writes: > So, um, wtf? I've read http://www.venge.net/mtn-wiki/DatabaseLocking. > > Is it really the expected situation that mtn ssh:// users sharing a server > will get random failures because someone else is connecting at the same time? > > <snip> > c) no three people actually collaborate > by using ssh:// to the same place (when e.g. sysadmin paranoia or firewalls > prevent using a netsync server).
I'm planning on doing exactly that (once I get 'sync file:' working reliably on Win32, and the Emacs DVC frontend spiffed up). Just be clear, I think by "random failure" above you mean "failed to obtain write lock, try again later" sorts of failures, not "corrupted database" sorts of failures. Is that correct? That's my situation now with CVS via ssh, so I can live with it. Hmm. CVS does file/directory locking, but mtn locks the whole database, so there will be more collisions with mtn. We'll see how bad it gets. -- -- Stephe _______________________________________________ Monotone-devel mailing list [email protected] http://lists.nongnu.org/mailman/listinfo/monotone-devel
