-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Am 19.03.2011 10:30, schrieb Mark de Wever: > On Fri, Mar 18, 2011 at 12:08:38AM +0200, Iurii Chernyi wrote: >> Hi, >> >> There were some patches committed (1.9) that make 'side=1' no longer a >> default value for some of WML direct action tags (i.e. >> https://gna.org/patch/?1883 - change for [gold] tag) > > I had a look at the patch and I've a problem with it. We normally give a > two releases transition when we change behaviour. This change silently > changes behaviour and can break UMC. IMO we should warn two releases > that an empty value is not allowed, before changing what an empty value > does.
Yeah, mordante mentioned this one to me some days ago as possible blocker for 1.9.5. Would be nice to know the status of this one and when it will be fixed. I'd love to release 1.9.5 in some hours but this sounds like a rather bad case of breakage. >> 1. ignore it, let the mess with 'some tags having side=1 default and >> some tags not having it' continue. >> 2. revert the patches, returing to 'side=1 is default for all direct >> action tags' status quo. >> 3. modify all direct action tags to have "side=1 should be explicitly >> set, if it's not set it's 'work for all sides'" behavior. >> 4. modify all direct action tags to have "side=1 should be explicitly >> set, if it's not set it's an error" behavior. >> 5. modify all direct action tags to have "side=1 should be explicitly >> set, if it's not set it's an error" behavior, plus add support for >> [filter_side] tag (a standard side filter, a standard way to filter for >> sides like 'all enemies of side 1') for all the direct action WML tags. >> >> I personally think that (5) is the way to go, but I'm only strongly >> opposed to (1). > > To me 4 sounds as the best solution. However since I don't do much > content development I can't judge whether 5 better as 4. So if you > prefer 5 I think we should go for 5. > > When we go to sides as a comma separated list, would it make sense to > also allow > > side = all > > which sounds user friendlier as > > side = 0 In a query Espreon (who applied the patch) mentioned that Crab_ will probably handle the "real" fixing, but I have no idea what the progress of this is. Would be nice to know how long this should block 1.9.5... Cheers, Nils Kneuper aka Ivanovic -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.17 (GNU/Linux) iEYEARECAAYFAk2Og6kACgkQfFda9thizwUa3gCgm7UH4Tj6Zkd6ysgt+BuIOa84 9ngAmwWRLjs7bgs16xO3eT1/xPjLxy+F =yiNj -----END PGP SIGNATURE----- _______________________________________________ Wesnoth-dev mailing list [email protected] https://mail.gna.org/listinfo/wesnoth-dev
