Just a general question I guess. I’m interested to see how long you all spend debugging issues inside Max/Maya as we seem to be coding round more and more bugs recently, spendoing more time on support and testing new releases. Gone are the days when we could rely on a release, I can’t see us taking 2012 till at least ServicePack1, or even later due to continued issues.
So how long do you spend debugging the apps themselves and not fixing your own code?
For Max, not that much anymore as I think I’ve learnt most gotchas by now… But it’s annoying to say the least…
Yeah, ServicePack1 before seriously thinking of taking anything into use these days. I wish I had more time to test OLD FUNCTIONALITY in the beta programs, but I just don’t… And often I get a feeling that old functionality breaking down is pretty much overlooked over new shiny bulletpoint features. 
SamiV.
Maya 2011 has a couple of brokah features. (I’m looking at you draggerContext)
My main gripe with the packages now is that I remember when the new features were massive and game changing. Now, besides things that are ported over from other packages (Animation Layers), the only features that even look exciting are what used to be plugins (Maintain UVs)
So you take the risk of breaking your pipeline so you can get a couple of minor features…
Second that, I was kind of looking forward to getting VP2 into some more testing over a few users here, but with the characterSet bug we just can’t push it out at all (still not fixed in Hotfix2 by the way, or at least not forour setups).
I think 2011 was a big one to take purely as Pymel and QT were managed for us. That said we still over-write the default pymel build so we can get latest from the GIT repro if needed.
I was a bit disappointed with the qt integration. The loadUI comand is pretty useless, so you still need to bring in PyQT.
I haven’t messed with the latest version that has the .uic stuff built in so you don’t have to run bat files on the background, that might be closer that what the loadUI should’ve been doing.
We also manage our own pymel, so all we really got out of the Qt upgrade was the ability to dock custom windows. =/
Well, dock at least a few of them, in a few places… if you’re really lucky 
There were some big reference fixes for us in 2011 which have helped.
I’ve been scarred for life with references and avoid it like the plague. It might’ve gotten better, but I never saw a rig get updated and not blow most of the animation data away. Or even the skinning update and not break stuff.
I haven’t looked at the asset stuff, tuned out when they said it was like references, but if they give you the option of flag a reference as a version and then run a custom script to move between versions than it might be worth it.
So if the rig goes from version 1 to 2, I can run a little retargetting script to deal with the changes.
How do you guys use references?
I use referenced rigs in our pipeline in maya 2011. The only problem I’ve really come across when updating the rig is that attributes that are not keyed (and have never been set by the user in that scene) will occasionally get garbage values written into them when the animation scene is re-opened. To fix this I just have a script that batches all the anim files searching for that rig type and then sets unkeyed and unset attributes back to their default value (sounds like a bit of a dirty fix but in practice it has always worked fine). This is also the way I tend to deal with rig updates if there’s something significant changed e.g. a control has been removed or replaced by one that works differently. Just using a script that batches all the anim scenes and swaps the rig (doing any retargetting or fixes as necessary).
A system to do this based on rig version on file load might be quite nice though.
Regarding the point about lack of significant updates to the software, I’d say that HIK has proved really nice for retargetting.
I’ve spend quite a few hours getting plugins and scripts to run in 3dsMax 2012.
I would put a whole rant here to complain about Autodesk, bla bla bla…but whats the point really?! 
It is what it is!
Definitely some of my tools are large ‘master-pieces’ of band-aids around max bugs.
Referencing, hhhmm, well we’ve been running a fully referenced animation pipeline for about 5years now and I think most of the bugs in the reference handling have gone away. It’s pretty rock solid. That said we’ve wrapped most of the calls here so that the animators don’t use the reference editor, they use our AssetBrowser that displays all available rigs in the team project and just hit the swap button, the rest is done behind the scenes. The only real management being done is the namespace handling to ensure that rigs are always consistant, things like the character Tag being switched so that they always sit in a namespace that reflects the character, top group node is renamed, refNode renamed etc.
But on the whole, I’d say referencing for rigs in anim scenes has been crucial for us, and works well. I should add that we run characterSets on all rigs which changes the way referencing links and manages the connections on file open… I think I put something in my blog about it a while ago
[QUOTE=Mark-J;10829]
So how long do you spend debugging the apps themselves and not fixing your own code?[/QUOTE]
40+ hrs a week Mark
And that’s just for me!
You and Seth mostly!
Well just think, if every thing was perfect and we didn’t find you bugs what would you do with your time! ( Fix Trax maybe!
I’d probably sit around trying to bend spoons with my mind. Or maybe start knocking things off my list of things I want to re-write.
I enjoy the bugs as it gives me problems to solve (I like bugs that I didn’t introduce more than ones I did of course) and I get to reminisce on the days of old when Maya would do dumb things on me in production. Difference is before I’d hack my way through fixing it and now I get to hack my way and then hand it to a developer to do a ‘proper’ fix.
If we do some really dumb $#!+ just ping me in an email and I’ll look at it… offers open to anyone really.
New one in 2012: If you use the browse button in UV Snapshot (UV Editor, Polygons menu) and don’t specify an extension, it returns a name with "." (dot-star) on the end of it. Windows does not allow '’ in a filename, so the save fails.
Workaround is to delete those two characters first. Easy enough, but this should be an embarrassing bug (as in how did we miss that).
<* Wes *>
386646 UV Map Save generates bad filename from browse
uvSnapshot -n "K:\12-3DWork\ estbadname..tga" -xr 1024 -yr 1024 -r 255 -g 0 -b 0 -o -ff tga;
// // Saved file: K:\12-3DWork estbadname..tga
// Error: Failed to open output file: K:\12-3DWork estbadname.*.tga //
Shawn,
I will send you your choice of sweet (hint: My wife makes handmade Gelato) if you fix the 2011 bug with the holdCommand arguments for MPxContexts not working. Either the MPxContext on the API or draggerContext in script.
Just send me a patch and an address!
Luiz: is this related to MPxContext::doHold? There’s a few drag contexts so I just want to make sure I go chasing after the right one. Let me know what’s busted and where and I’ll go bang on some developers doors.
Yes it is, it’s the MPcContext::doHold (doesn’t seem to get called) and by proxy the draggerContext flag -hc doesn’t get called
holdCommand(-hc) Command called when mouse dragger is held.
I started off with the draggerContext mel command, and when that didn’t work, I tried making a custom one with the API which didn’t work either.
Thanks very very much!
Hmmmm yeah that’s a bit frustrating… wonder if it’s something to do with QT Signal stuff. dragCommand and pressCommand appear to be fine but holdCommand is just doing nothing.