On Fri, 15 Jun 2012 19:13:54 +0100, Clare Macrae wrote: > We've seen some really strange behaviour with settings for disabling the eol > extension in TortoiseHg today, and I'd like to see if anyone has seen > anything similar (we can't find a bug for it). > > The oddity is: > - our user config has eol turned off, with eol = ! > - in a DOS command prompt, typing "hg help extensions" showed that eol was > disabled > - In TortoiseHg 2.4 Log window, typing "hg help extensions" showed that eol > was ENabled > - Quiting TortoiseHg, removing the "eol = !", and restarting TortoiseHg > showed that eol was DISabled > > We saw this same behaviour on another PC with TortoiseHg 2.4.1... > > So I was about to report a bug saying that TortoiseHg 2.4.1 appeared to > interpret "eol = !" incorrectly. > > However, we did a bit more testing, and found that on a TortoiseHg 2.3.2 > machine, with two open TortoiseHg processes, one showed eol on, and the other > off - but the settings had not been changed for the lifetime of these > processes. We then quit TortoiseHg and re-loaded it, and could not get it to > show eol as on. > > Finally, having upgraded the 2.4 machine to 2.4.1, I put back the "eol = !", > fully expecting to see eol as on again - and it wasn't.
If you enable eol extension on a per-repository basis, it may lead inconsistent behavior. TortoiseHg loads more than one repositories in a single process but Mercurial extension affects globally. https://bitbucket.org/tortoisehg/thg/issue/984/ It also happens when pulling from / pushing to a local repository. http://thread.gmane.org/gmane.comp.version-control.mercurial.devel/50212 Still you want to use per-repository extension, try to disable "Show Subrepos on Registry" (View -> Repository Registry Options), which currently loads all repositories listed, and avoid to open more than one repositories in the same window. Regards, > So, with several witnesses (honest!), we've been through several cycles of > seeing inconsistent and non-reproducible behaviour of the eol extension. > > We have checked to ensure that the repo's hgrc doesn't enable eol. > > This has caused several users to find a number of files to be reported > incorrectly as modified, and I am wondering whether it is also the cause of > the really slow status which I described in the thread 'What events trigger a > "Refresh status..."' > > > http://sourceforge.net/mailarchive/forum.php?thread_name=20120616001222.3b233302ff425f5c7e89a4d5%40tcha.org&forum_name=tortoisehg-discuss > > In case it's relevant, I've given more details below of our how our settings > are set up. > > We're stumped! Does this ring a bell, or can anyone offer any suggestions > from tracking this down please? ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ _______________________________________________ Tortoisehg-discuss mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/tortoisehg-discuss
