On Friday 23 September 2011, Aaron J. Seigo wrote: > hi all... > > last night i committed to master in kde-runtime some major changes to the > code that handles the briding of Plasma::DataEngine into QML. there was > some complexity in there that was unnecessary and some performance > improvements to be found as well. > > there is still an issue with performance left that i'm not 100% sure yet > how to deal with: DataSource keeps a local copy of all > Plasma::DataEngine::Data ... which means that when the Data changes, it > will cause a bunch of allocations as data members get detached, which then > gets all thrown away instantly when dataUpdated gets called.
yeah, that was always my major gripe with datamodel.. i don't see much other ways with plasma1 tough.. a thing that could be done is try to share 100% the data copies of DataSource and DataModel. > in any case, i'm not comfortable backporting this to the KDE/4.7 branch > right now. it's functionally equivalent to the old code, so no majorly > disruptive changes (Marco: i ended up putting the source filter back in), reason? ;) > but it needs more testing. > > if you are using kde-runtime from master with QML, please try to keep an > eye out for any anomolies related to dataengines. > > everything is working perfectly here, but i may have missed something. yeah, i'll give a try and will take a look at the overall diff -- Marco Martin _______________________________________________ Plasma-devel mailing list Plasma-devel@kde.org https://mail.kde.org/mailman/listinfo/plasma-devel
