Hello. In order to understand and be able, for me, to help other future contributers, please help me understand the issue at hand better. Mainly i want to know how to go about things that happened to me recently:
I submitted a bug report about extremely poor performance of the Spin operator Link to the bug report. The issue was closed on the grounds of it being a performance improvement issue and not a bug. My view of this is such that the lack of performance in the Spin operator can go so low, to the point of being unusable when even spinning just 1 vert, is becoming a bug. After that i double checked with the page Making Good Bug Reports, even though one could argue that Functionality which works as intended but could be improved can be applied here, on the other hand was the performance degradation an intended design decision/side_effect or a bug in the implementation.
After all, as an example, nobody would say that a poor viewport performance with just a couple of simple objects is not a bug since everything works as intended, not a bug, it’s just slow.
Please help me understand who’s being more reasonable here and how can i help to advanced the fix/patch.
I pretty much want to contribute, but my c++ skills are subpar, in the process of learning, and i was wondering if there’s a possibility for me to contribute any other way? I also wanted a clarification on what kind of performance degradation is considered a bug what isn’t
Typically if there’s a performance regression compared to a previous version, and the performance deficit isn’t expected, then that’s a bug.
However if performance is low, and it’s expected to be low based on the design of the code/feature and the hardware you have, then it’s not a bug and is either closed or marked as a “known issue”.
Contributing to this particular problem will be difficult if not impossible without coding experience.
But in general for Blender as a whole, there are many ways you can contribute. Translation, Documentation, creating artwork for release notes, helping with triaging bugs etc. See the Developer Handbook - Blender Developer Documentation for all kinds of information. Please let me know if you have further questions, I am happy to help you out.
No further questions yet . Thank you for trying to help me out.
This is exactly the source of my confusion. Maybe you can help me out understand this issue from a developer’s standpoint. I just can’t understand why would there be any expected performance drop in the spin operator if the only transformed geometry is a couple of vertices, literally… but the perf drops with respect to the amount of verts the object has in total, geometry that in not being manipulated upon.
It just feels like that there’s a buggy for loop that goes through each vertex in a given object, doing some useless checks, instead of calculating just selected ones.
My knowledge of the code in this area of Blender is limited, so it’s probably best if someone else with the relevant knowledge answers this question.
