I’m thinking about moving over a team of Artists to 2018 from 2016, does anyone have any suggestions or issues or learning curves when they switched over? Are there enough new features for it to be a good idea for us to switch?
Thanks,
If you’re using PySide or PyQt4 you’ll need to update your code to work with PySide2 or PyQt5 respectively. We’ve got a few threads about it concerning the jump to 2017.
I’ve been fiddling with 2018 at work, and we’re pretty confident that we’ll roll the team over to it as well. So far it’s been stable, but I’ve been using it less than others.
But basically you’re looking at all the annoyance that the 2017 jump brought, and so far not much else.
We’re using QT for our plugin in C++ so no real worries there.
After the install I found out that they’re not updating (or adding) windows environments to the MAYA_LOCATION and MAYA_PATH so they are pointing to 2016 directory, so unless I change those it won’t open.
But everything else is good for the Artists you find? Any real issues for them do you think?
From a rigging perspective: For me, the big essential feature for rigging is the new Shape Editor for blendshapes (introduced in 2016.5) that includes delta smoothing, masking and eraser brushes. It makes a blendshape workflow 20x more powerful than before. Everything else (for my workflow) is just menu-shuffling, less contrast on the icons every year, and a new variation of crashes to contend with.
And it is easy enough to export the blendshapes back down to version 2016. However I found that I had to duplicate off the shapes so the shape masks don’t bake down as weight maps in older versions of Maya. That was causing weird issues when exporting to game engine.
From an animation perspective, there were a lot of weird bugs with the new graph editor and the “classic Coke” graph editor, depending on which 2017 SP we were on. Infinity wouldn’t work. Or it wouldn’t display properly. Insert keyframe and selecting curves has changed a lot (for better or worse, I don’t know, but our animators were griping).
I haven’t tested 2018 yet. But I’d do at least a few basic workflow tests before committing your team.
Maya 2017 great improvements are the move to Qt5 (PySide2 and PyQt5) and the addition of the multithreaded massive speed improvements. Probably 2018 will be much better in these areas…
Subscription… though…
Thanks lidiamartinez
Yeah we already have the subscription suit, for all the Artists. That’s one of the main reasons why we’re going move.
Warning to others: The 2018 Devkit contains a entirely different C++ library for Qt (There are some classes that are used but from what I see a lot have changed), and even the Qt directories are named differently. I will prob be spending the better part of a day converting 2016 Qt over to 2018.
Yeah, I can see why PyQt is not working with the changes to the C++ libraries I would have to guess the guys who developed PyQt have some work to do to integrate it (Lots of changes to the headers most of them coming from the Widget classes). Fortunately we’re not reliant on PyQt.
PyQt5 has been stable for quite awhile. The trick is just compiling it against Maya’s local environment.
This shows how to build PyQt5 for 2017, it is probably a fairly similar process for 2018.
Otherwise PySide2 is included and provides a nearly identical API, at least for the core widgets, though it is less battle tested and possibly less stable. (I personally haven’t had any problems with it that I know of)
The only downside of PySide seems to be stability… It’s younger. But there’s plenty of reasons why we should all move to pyside. The biggest one is integration with Maya, lighter weight, removed deprecated funcionality, and documentation included officially under Qt website.
Among others…