Thanks for fixing this mess up!

Looks good, thanks.

And Phillipp: The idea of using pipe booleans has occurred to me, and I may eventually try that, but at the moment our thoughts are to follow the "straight skeleton" method, with can deal with the intersection/merging of advancing edges when there are concave angles.

Thanks, Vitalii, Hans, and Philip for your comments.

I agree with deadpin, looks like line 152 of the file should be

Thanks, this looks good to commit.

Erik's change looks right and pretty efficient, so I will commit that change later today unless Erik wants to do it himself.

Face beveling is pretty much just Inset, right? Is the reason you want this because you want it in a modifier? Or are there edge-bevel options that you would like that aren't in Inset?
I could consider adding Face beveling, but feel we would then have to do something about the big overlap in functionality between Inset and Bevel.

There is not really a bug -- there is no obvious place to put the required new vertex in the area spanned by the double-ended blue arrow here:

When I claimed that this bug was fixed, I was referring to the problems of making zero-area polygons in UV space. I knew there were remaining cases where the UV layout, if done by hand, could be better, but also that Blender's bevel has never so far been able to do that well, so it wasn't exactly a bug or a regression.

Yes, you can combine them.

Those speedups are definitely worth this change.

I made a couple of comments which you can answer/address, but then you can commit this.

Thanks.

looks good

Makes sense. Thanks.

Thanks for the fix.

I agree, this is better.

I think we want to make this a drop-in replacement for the python importer to start, and resist the temptation to fix deficiencies of the python importer right away, because that makes it harder to field bug reports from people saying "it doesn't work like the old one". There are a lot of other requests and patches even for improvements to how materials are handled in the python OBJ importer, but I think we should hold off on all of those for now, for the reason I just stated.

This is fine to submit after you address the comment I made as you see fit.

Yes, it is in master and thus will be in 3.2

The parsing reads much more straightforwardly now, thanks. And the speedups are great. On my Linux machine (AMD Ryzen Threadripper 3960X 24-core) I got these speedups:

Thanks for fixing this. Looks good to submit.

This looks good, thanks.

This is the same bug as in T56625 , which I am working on fixing now.

I believe all of Sybren's blocking requests have been fulfilled now, so am marking this as approved in order to let this get into 3.2 before bcon2 starts. Especially as it will be marked experimental.

Thanks for the fixes. Looks good.

Sybren, since Aras has answered many of your comments with "a later patch will address this", trying to keep this first commit close to Ankit's original code, which of the comments you have made do you feel necessary to address before committing this patch? Do you think this is possible before bcon2 on Monday?

Yes, please commit this.

Agreed. I just committed a change from Pratik Borhade to put the soft limit back to 100, while leaving the possibility of typing larger number (up to 1000) in the box.

Thanks, I was wondering how to add soft limits because I felt it needed them.

Do you have an example of where vertex order is not preserved with the new exporter? We think it always is, unless you export by material group, which kind of necessitates changing the vertex order (at least, according to a recent bug report).

This code looks good to me. You can commit it yourself now. Though if you do, the backports of fixes for 3.1 might be harder to do (but they might be easier to do from the original patches anyway)

It wasn't clear that "keep vertex order" was useful. If it is, open a bug for that and we will try to add it.

Just for the record, I accepted this offline with Aras, but forgot to mark that in the tool before he committed.

I think this is likely fine but I need to spend some time checking it. The recent bug fixes were a higher priority.

Applied with commit rBb9123b806fc4d

Thanks you, looks good.

Applied with commit rB5bfdaaa80082be1