Join GitHub today
GitHub is home to over 28 million developers working together to host and review code, manage projects, and build software together.Sign up
mrview Connectome tool: Streamtubes not displaying #1330
A continuation of #1311; re-listing as the two bugs originally reported appear to in fact be unrelated, and so need a "clean sheet" to investigate the second.
Attempting to visualise edges as streamtubes within the
If any devs can reproduce this it would be greatly appreciated.
referenced this issue
May 9, 2018
Thanks for the input @LucSam.
Okay, that's ... awkward. The problematic line is here: specifically statement "
Since this all looks as it should, I can only suggest that this must be some sort of GL update timing problem with Macs: Something is invoking
@LucSam: One thing you could try is replacing file
Then, see if after some period of time the display starts working correctly.
Hmmm. I'd expected that bypassing that draw call would have been sufficient: I don't see anywhere else in the code where the streamtube data are accessed without verification. There might be something even more exotic going on, which unfortunately I have no hope of figuring out without some more debugger action...
It looks like you've retained the original line of code in addition to adding the new one, rather than replacing the line. There should only be one line of code that begins with "
OK, I have to admit I'm not familiar with what's going on here - it's not my code...
But looking through the code you point to, it strikes me that given this line, we should see a line with
None of these occur anywhere near the constructor...
@LucSam, I think you need to add an explicit initialisation at line 85 in the file src/gui/mrview/tool/connectome/connectome.cpp, so it looks like this:
That sounds like the most likely explanation to me...
Also of note: this is an AMD/ATI system, and their drivers are always more picky. You often get issues with the OpenGL handling not being exactly to standard, and it's rarely all that obvious when it isn't... But it doesn't look like this is the issue here (at least the outright segfault isn't...).
No need to apologise - it's hard to second-guess how our brains work...
So that's good news, in that at least that's one issue sorted. Can I check that this now runs without crashing or assert failures within the
added a commit
May 21, 2018
Had to be done...
On the plus side: this probably means the issue occurs on any ATI systems, macOS or not. We've got plenty of those in my department. Is there any test data for me to have a play with?