Mike Sullivan <Mike.Sullivan at Sun.COM> writes: > Richard Lowe wrote: >> Vladimir Marek <Vladimir.Marek at Sun.COM> writes: > > >>>> - What else? Please send me questions/issues you want to discuss >>> Documentation. Best practices (should I create branch in workspace or >>> rather create new clone; and why ?). What to do if Mercurial screws >>> repository (happens to me ~ once a week). Differences between mercurial >>> and wx/teamware. >> >> I think these are largely separate from the gate/gk issues. > > hmm, don't think I agree. Documentation is certainly one of > my big issues before the switch, particularly since as the > sfw gatekeeper I will likely be the one getting all the questions, > but without any answers.
I didn't state that it wasn't important to do before we switch. Of *course* it's important to do before we switch. I'm saying it's a separate series of tasks from making the automation of the gate function with mercurial. > I know of the hg book but haven't read it yet so I don't know > if it is all that we need, but it's probably too much to tell > someone to read who just wants to know how to bringover like > they used to, so I'd indeed like a teamware->hg > cheat sheet (which would hopefully include info about wx > operations you're used to and how they relate now in hg). I > think I've seen one of those floating around so that might > be done (but I don't know, and it was a long time ago). There's the SCMVolunteers page on the wiki, and I recall at least one person at Sun having written up something with more brevity (though I forget who). > As to how to repair a repository.. we'll, I'm not sure if we > need that though I'm alarmed if it's happening once a week. I, also, was surprised by that. I would expect to see bug reports, either to us or mercurial, when this happens. I haven't seen either. Can we get appropriate amounts of detail, please? > I know the ON teamware->bridge can have its bi-weekly troubles but > I hadn't seen hg corruption there (well, unless I interrupt > an 'hg pull -u -v' but it's not corrupted just not updated > completely). Corruption there, in approximate order of frequency: - TeamWare badness it tripped over due to its fragility. - (TeamWare) User error requiring content to be flushed for other reasons. To my knowledge, nothing having to do with mercurial has gone wrong with the bridging. -- Rich
