GSoC 2019: Plugin Manager (Week 12)

Posted 5 years ago

This week I still managed to add more comments and do more code cleaning. I must admit these were not done completely in previous days, partially because some functions are being added and renamed.

I did some fixes, too. One small but interesting one among them is, the space between the following QLabel and QTextEdit, was not constant when we switched between plugin items and package items. I fixed this by setting a minimum width value to the QLabel on the left.

What to do next

I'm now fixing various artifacts(including typos, format improvements, ) with the help from Jojo, and indeed I didn't notice so many flaws before his examination... These should definitely be fixed even if the codes work properly already.

And I'd also like to improve the contents:

in https://musescore.org/en/handbook/3/plugins, for users, mainly on how to use the newly added plugin store;
in https://musescore.org/en/handbook/developers-handbook/plugins-3x for developers, mainly on how to publish an acceptable distribution in the plugin repository.

Of course, the above documentation should be done when the work product is approved or merged. (But I guess the latter could be changed in advance?)

I'll also probably write a work product blog post within this week, and summarize all features this project brings to MuseScore via GIF animations.

Hope everything goes well, and see you all next week!

Comments

[DELETED] 1831606

• Aug 19, 2019 - 11:29

The stuff @DLLarson and I have been working on is going to augment the property repertoire significantly, adding play-events (note events) ties, and a feint at the object-uniquization problem. I hope you have seen my new plugins for which they were created at https://musescore.org/en/project/articulation-and-ornamentation-control .

Both Dale and I would very much like to hear from whoever made the significant internal change that uses disposable wrappers as opposed to stable Qt objects as to the rationale and tradeoffs made, although the desideratum of removing Qt overhead from objects not asked for is a conceivable one.