Troubleshooting Methodologies
Do any of the major programs (autodesk, algorithmic, unreal) standards for trobleshooting their application? The reason I ask is because I noticed a lot of people (amittingly myself as well when I'm in a hurry) will either do one of two things when stuck on a problem. They will either hit the message boards immediately or they will brute force attack the problem with the most common solutions on the to of their "to try" list. I believe much of this can be avoided.
Are there any common troubleshooting steps you guys attempt over and over again? I have a series of things I end up doing over and over again regardless of the issue. I call it the game of "eliminating as many variables as humanly possible." I don't know that much about 3D but often times a super deep technical understanding is non essential IMO. The way I see it there are 5 components to a problem.
-Application
-Settings
-File/Scene
-Tool
-Mesh (with components)
I always try to isolate which component the problem exist in before actually finding the problem.
Scene
-Save scene and restart application
-Make sure you are on the latest version or service pack of whatever you bought.
Settings
-If at all possible, back up the preferences folder and reset everything to default.
-If the issue still persist, then copy it back over
File/Scene
-See if the issue occurs with other files. Texture not showing up? Load a different test texture.
-Try different formats
-If test textures and test formats work, then try an exported version of your scene with ONLY the essentials.
Tool
-Ensure the tool is working the way you expect it to by experimenting on more basic meshes or content.
Mesh/Content
-Save a copy of the file as a test
-Slowly delete parts if your mesh to eliminate the number of variables that effect each-other. For example, if you have an xNormal issue, cut the mesh in half and test both halves. If only one half breaks, cut that on half until the component is identified.
Replies
If I'm doing something and it's not working, I'll usually pull up the documentation page to ensure what I should expect.
If it's not working, I might then try it on a simple model, if it works, I need to extrapolate the differences between my simple model and the more complex one.
If that still doesn't work I might fiddle with the settings for said tools.
In various orders Etc. Etc.
In my opinion troubleshooting should be a bigger part of your skillset than your googling skills one day. It just takes some time to get there.
On a side note, a common solution people are quick to throw is the reset xforms one. I think Mark or EQ called it the "nuke" option, which I got a laugh at. Telling someone this is a good temporary solution to some problems but doesn't help beginners understand what's gone wrong, and it's definitely not a "fix me" button. It's very often that they also might want to retain their transformation values. It's not a good troubleshooting step.