Closed MaelRL closed 5 years ago
Shouldn't be able to call functions if the input mesh does not meet preconditions, e.g. calling "does_self_intersect" on a quad mesh will return that everything is fine because the mesh is not triangulated, regardless of the geometry.
The architecture used by Qt for the "applicable" system would make the demo freeze for a century anytime you open the menu "Operation" if you start calling mesh function to enable/disable action. The only viable option I think is to return a warning when the preconditions are not met, and interrupt the action, which is supposed to be done already. The real bug is that the self-intersection plugin does not perform this test before calling the algorithm.
What is a Filter item ?
By "filter item", I meant the different kinds of selection, sorry. Apparently, you can't have multiple selection files (e.g. intersecting simplicies + selection) anymore, so this bug goes away too.
Failing to reload an item crashes the demo (take an .off, load it, change the number of a vertices to 1 fewer in the file, reload, kaputt).
Actually if you lie aboutthe number of vertices in an OFF file, reloading OR loading will lead to a crash, because the loader doesn't know how to detect that a call to an invalid index is out of bounds. The loaders expect the files to be coherent with there own standards, at least for the OFF. I think fixing that peculiar bug would alter performance. @sloriot what do you think ?
Can't load polygon soups
I don't reproduce. Is it possible you loaded an off with the lcc_io_plugin and then tried to load a soup, which was created empty ? In that case, it is because the lcc_plugin became default after the first load and it failed to create the soup, but the off_plugin works well.
That bug is from @sloriot, so I don't know.
That bug is from @sloriot, so I don't know
Can't load polygon soups
I don't reproduce. Is it possible you loaded an off with the lcc_io_plugin and then tried to load a soup, which was created empty ? In that case, it is because the lcc_plugin became default after the first load and it failed to create the soup, but the off_plugin works well.
I guess so.
For the colors, I usually use the HSV model to play with generated colors, where VSA uses a cheat-sheet filled with randomish RGB values. Is the problem specific to VSA or should I think more about my color generation in general ?
Adding "Under Testing" to an issue is useless. You should label the PR #3879 instead.
I know, I must have mistaken it for a PR
Bugs
Feature Requests
Environment