Brainapp/bugs and feature requests: Difference between revisions
< Brainapp
No edit summary |
Added cutting plane rendering bug →Bugs |
||
| Line 3: | Line 3: | ||
* Brainapp often segfaults on startup. NVIDIA direct rendering fails, which causes an assertion failure in [[GLG3D]]. Click for [[/startup segfault|the full debug/error output]]. | * Brainapp often segfaults on startup. NVIDIA direct rendering fails, which causes an assertion failure in [[GLG3D]]. Click for [[/startup segfault|the full debug/error output]]. | ||
* A non-fatal [[G3D]] assertion failure can happen while adding tubes to the model even when all the NVIDIA stuff is okay. Click for [[/tube add assertion failure|the full debug/error output]]. | * A non-fatal [[G3D]] assertion failure can happen while adding tubes to the model even when all the NVIDIA stuff is okay. Click for [[/tube add assertion failure|the full debug/error output]]. | ||
* The cutting plane uses the wrong raster scan width on certain datasets, so in some directions the image on the plane is just translated sideways, while in others it looks totally scrambled. My bet is that the program is using hard-coded image dimensions rather than reading the image dimensions from the <tt>parameters</tt> file on disk. Use <code>$G/src/brainapp/obj/brainapp-gcc3-d -f $G/data/diffusion/edinburgh/tumor_fullstudy.2009.11.23/tumor1/tubes.2009.11.25/brainapp/settings.cfg -f $G/src/brainapp/settings-desktop.cfg</code> to reproduce. | |||
== Feature Requests == | == Feature Requests == | ||
Latest revision as of 15:13, 3 December 2009
Bugs
- Selecting ROIs is really flaky; it seems to work only about a quarter of the time. This needs to be more reliable.
- Brainapp often segfaults on startup. NVIDIA direct rendering fails, which causes an assertion failure in GLG3D. Click for the full debug/error output.
- A non-fatal G3D assertion failure can happen while adding tubes to the model even when all the NVIDIA stuff is okay. Click for the full debug/error output.
- The cutting plane uses the wrong raster scan width on certain datasets, so in some directions the image on the plane is just translated sideways, while in others it looks totally scrambled. My bet is that the program is using hard-coded image dimensions rather than reading the image dimensions from the parameters file on disk. Use
$G/src/brainapp/obj/brainapp-gcc3-d -f $G/data/diffusion/edinburgh/tumor_fullstudy.2009.11.23/tumor1/tubes.2009.11.25/brainapp/settings.cfg -f $G/src/brainapp/settings-desktop.cfgto reproduce.
Feature Requests
- Overhaul boolean operations on ROIs with color-coding and multiple selection.