This may be worse than we thought... When I fixed the storage and retrieval of the CSS properties, it still didn't work properly in release mode. Some classes are found and bound, others are not. Turns out that the Closure Compiler doesn't resolve all dependencies accurately, the classes it misses are never 'considered' during compilation :-(
I will look into the custom dependency algorithm in the Publisher next. Wish me luck ;-) Also, the fix will literally affect all JS classes, so prepare for some interesting merges. If I find a solution, I'll publish it first in a branch, so we can look at it together before we "commit". EdB On Wed, Nov 13, 2013 at 8:12 PM, Erik de Bruin <e...@ixsoftware.nl> wrote: >>>One thought is that we might store both the 'name' and the 'qName' in >>>the class metadata (where currently only the interfaces - if any - >>>live) and adopt the 'getValue' routines to search that instead of the >>>entire namespace chain. This would get rid of the need for the dreaded >>>'__proto__' as well... >> Sounds good. We need to find the superclass somehow as well. > > Alex, can you please create a JIRA issue for this and assign it to me. > I don't think I'll have time in the next few days to work on this, and > I don't want any details to get lost in the avalanche of emails on the > list. > > EdB > > > > -- > Ix Multimedia Software > > Jan Luykenstraat 27 > 3521 VB Utrecht > > T. 06-51952295 > I. www.ixsoftware.nl -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
