On Sunday 16 September 2012, Daniel Nicoletti wrote: > Hmm, well I thought you had a dataengine for the applications model, > so uninstall, favorite, sort would all be exposed as services, but it > seems it's not the case, sure it makes sence since from my experience > dataengines are slow with a good amount of data. > Anyway I just tought having a dataengine "packagekit" didn't worth > the trouble, I'd embed it in your model since it's just a DBus call.. > so maybe it makes sence for you to add a sort(), uninstall(index), > favorite(index)... up to you anyway, if you think this is best I'll just > take a look to see it the calls are fine :)
for big models, like the one in kickoff or an hypotetical one in packagekit, i really think they should be indeed, models. the performance difference in terms of speed and memory is just too ridiculously big (and is fine, they are two very different use cases) a thing i would like in plasma2 is the possibility for dataengines to request a model (that would be, an actual qaim) for a certain source, like is now for requesting services. -- Marco Martin _______________________________________________ Plasma-devel mailing list Plasma-devel@kde.org https://mail.kde.org/mailman/listinfo/plasma-devel
