What is your development/ debugging process?

Everything about development and the OpenMW source code.
Post Reply
Vakath
Posts: 5
Joined: 30 Jul 2017, 18:06

What is your development/ debugging process?

Post by Vakath »

Hi All,

This is a general question to developers on how you generally work/ workflow. I've never worked on an application of this scale before and I'm having trouble figuring out the best way to go about working!

How do you get the game to a state where a bug is testable etc?

Currently I am attempting to resolve a few bugs and so I am writing code in visual studio, building in debug mode, playing the game and trying it out.

I have seen there is a RelWithDebInfo build option too in Visual Studio which I will be using to (I assume) improve performance through compiler optimisations while keeping the debug symbols?

My whole process at the moment is obviously verrrrry slow and I'm keen to figure out the best way to move forward!

This is clearly a very open question, but if someone has the patience to explain their workflow it would be very helpful. :geek:

Thanks,
JP
User avatar
raevol
Posts: 3093
Joined: 07 Aug 2011, 01:12
Location: Caldera

Re: What is your development/ debugging process?

Post by raevol »

I guess it depends on the bug?

The most important thing when reporting a bug is to give steps for how to reproduce it through normal gameplay. So, you write how you can reliably get the but to happen 100% of the time, and describe what goes wrong and what should happen instead. Then when a dev codes a solution, they can follow your instructions to ensure that the expected behavior happens with their build before committing.
User avatar
Frik
Posts: 124
Joined: 20 Dec 2016, 05:22

Re: What is your development/ debugging process?

Post by Frik »

Vakath wrote: 02 Aug 2017, 22:14 My whole process at the moment is obviously verrrrry slow and I'm keen to figure out the best way to move forward!
Which part of the process is slow, compiling?

I don't use visual studio, just make & a text editor (pluma), but once you have a clean compile, you should only be compiling/linking modules that have changed. You can make use of multiple CPUs using:

make -j`getconf _NPROCESSORS_ONLN`

Perhaps I'm not understanding which part of the process is slowing you down. My flow is update the source, run make, test changes. Repeat.
Post Reply