#391327
JDHill wrote:If it does occur on the new machine, then the questions would be: which other programs have been installed, and how similar are the graphics adapters & drivers. Because of that, I'd recommend making sure to install Maxwell first, in order to check it before other programs are installed.
I have to say that i forgot to test it on a clean machine :oops:
But it's the same problem on the new machine.
#391333
Hey JD,
You know, we've actually got a unique opportunity here. This is actually a good thing that we have two separate users and systems with the same problem. I like where your headed with this, so I'm proposing that Numerobis and myself compare the applications we have installed. Aside from the obvious (3ds Max & Maxwell), there could be a few intersection points on that list that will help us troubleshoot this. Hopefully, the uninstall (and possible a run of cCleaner) for any such overlaps cleanly remove themselves when uninstalled. Granted, there could be some large overlaps like MS Office, but most likely, we'll find some more obvious candidates and we can focus on those.

In order to ensure that a program's system wide hooks would be removed, how would I truly validate that? Are the hooks generated in a "boot strap" process upon the application's executable files loading or system startup? Can they be injected and stored into the Windows registry? Are there any other places where they may reside?

Hey Numerobis,
I'll PM a full list of my applications and their versions to you and JD (and anyone else that would need to know). I'll include the list of all installed updates as well, but hopefully we don't have to get that specific on this "bug hunt". If you have any reservations about re-configuring your system or removing applications (I totally understand that time troubleshooting does not equal time getting getting paid), then I can do the system manipulation and testing on my end. I'm do for an OS reload anyway, so if I box myself into a corner, then I just know what I'll be doing this weekend :D

Thanks guys!
#391334
It's really difficult to predict which software will need to use hooks; I used to install one simply in order to support tabbing in a modeless window in a plugin's material editor (see here). The point being, in that particular case, you are looking at a plugin in another application possibly being a culprit, and this, when you don't actually even know if that's the type of problem you're looking for.

So personally, I think you are looking at expending a large effort to try and fix a small problem, and probably with low likelihood of succeeding. That said, reviewing this issue once again, I find that I don't really like how the function works, and will suggest changing it, at least on the Windows side, since I have a simple and concrete example of an alternate implementation, which I prefer (the one I mentioned previously, from the Rhino & SW plugins).
#391336
Honestly, I saw this going one of two ways. Either we would find the culprit within the first 5 or so easily testable candidates or it would turn into a rabbit hole too big for anyone (myself included) to justify the effort. There are other system and program instabilities that I've seen on my end that would be nice to resolve with their respective vendors if found. If we do find something, it might just improve the user experience for a lot more people than just Maxwell users. If not, then at least we can say we tried, right? I think its worth an hour or two of testing.

Regarding core components of Maxwell, I'm not sure how many other cross platform issues that the development team has had to solve by refactoring code in more native API's to the respective platform, but from what I saw about Qt, it's supposed to be "one source code to rule them all". It would be nice to preserve that methodology, even if it's just coded a different way within Qt. Perhaps I'm too much of a purist :D

I thinks its at least worth a look and I'm willing to try a few things if I can compare my system setup to Numerobis. If it gets to the rabbit hole scenario, I'll just chalk it up to a "ghost in the shell" and move on. Sound like a plan?
#391337
It can't hurt to give it a shot, if that's what you feel like doing. :)

And Qt is great, but it has its own idiosyncrasies, as any developer whose done non-trivial work with it will quickly tell you. But that's the case with everything; in the end, every framework is an abstraction of base OS functionality, and even when they are not trying to do it cross-platform, where you have to deal with completely-different conceptions of what windows are and how they ought to work, various odd behaviors percolate up from the real world and manifest themselves as problems to be worked around. So while on one hand it is good to have a principled view to give you overall direction, on the other, a little pragmatism is also often in order.
#391338
So while on one hand it is good to have a principled view to give you overall direction, on the other, a little pragmatism is also often in order.
I can definitely agree with that! :D

I'm also approaching this having no experience developing a Qt application at all, so you guys have a vastly superior understanding of what you're up against than I do. The closest thing I can compare it to personally is deciding whether to use a combination of core & contributed modules for Drupal or deciding to custom code the feature that I need (thus introducing another item requiring custom maintenance). Best practices and real world deployments often diverge for very appropriate and insurmountable reasons.
where you have to deal with completely-different conceptions of what windows are and how they ought to work, various odd behaviors percolate up from the real world and manifest themselves as problems to be worked around
That's precisely why I'd rather be a smell tester in a sewage treatment plant than a front end web developer. I'll stick to the backend / server side where data is unbiased and still makes sense :D

CSS + "cross browser support" = Insanity!

Hopefully we come up with something interesting and useful in this analysis! If not, then that's just the way it goes. Thanks JD!
#391386
DISCLAIMER: I haven't tested this with Sketchup. A very similar situation was happening using 3ds Max and can be read about in http://www.maxwellrender.com/forum/view ... 35&t=44367

Hey JD,
I just started a full OS reload and posted my findings in the other forum post:

http://www.maxwellrender.com/forum/view ... 85#p391385

So far, so good! Thansk JD!
Texture/finish lost in render.

Nothing?

I answered this question myself in Rhino 6 documen[…]

New developments for Maxwell 5.1

- First one the random color don't work with hdr[…]

Material preview

Yes, I have set Maxwell for Rhino as Current Rende[…]