On Fri, Apr 08, 2011 at 08:45:44AM +0300, Niko Tyni wrote: > IMO, it should absolutely go in 5.12/experimental straight away to > spot any arch-specific test failures.
I see you've made a start on this - thanks! > After that, a new round of manual > rebuilds of the lib*-perl packages against 5.12 could be done to find any > "obvious" breakage. It's probably about time I did another rebuild and review of outstanding perl 5.12 related bugs - I'll probably be able to do this next weekend. > However, if the 5.12 (or 5.14?) transition is still going to take months, > it might be good to decouple the db transition in order to not waste > those months. They could just as well be used for finding and fixing > any db related bugs in testing/unstable. Okay, let's see how things go in the next couple of weeks. > I'm a bit concerned about user databases breaking on upgrades, but IIUC > the transactional mode needs to be enabled with a DB_INIT_TXN flag [1] > (among other things), and the Perl sources don't use that anywhere. > Therefore, it seems possible that transactions can't be used through > DB_File anyway and we're safe from the incompatibility. This probably > needs further investigation. Okay. Perhaps we can check whether there was any similar situations with previous libdb upgrades. > [1] http://download.oracle.com/docs/cd/E17076_02/html/gsg_txn/C/enabletxn.html -- Dominic Hargreaves | http://www.larted.org.uk/~dom/ PGP key 5178E2A5 from the.earth.li (keyserver,web,email) -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected]
