Please change the "language" question in Community Polls

CodeSmile

The community polls have been asking this question:

Which is your most used programming language?

Assuming users can really only choose one, then the question masks the reality of multi-language use.

It makes no difference whether someone uses a language 51% of their time or 100% - both count as 100%. This causes users' supporting languages to not be weighted at all. A distribution collapsed into a binary choice creates a winner-takes-all bias. It may be reinforcing to see high GDScript values but the question is meaningless. Users aren't electing a president.

Where it shows: C# didn't even jump by 3% between the 2023 and 2024 polls during the Unity developer exodus. I find this particularly unlikely to reflect the reality.

Please consider changing this question, even if it breaks historical comparability. Readers expect a fairly accurate, unbiased representation of language usage distribution.

DaveTheCoder

CodeSmile Please consider changing this question

This forum has nothing to do with the Community Polls.

I suggest re-posting your feedback as an Issue or Discussion in the godot-proposals repository:
https://github.com/godotengine/godot-proposals#godot-improvement-proposals

That might not be the best place, but at least the people who create the polls will see it.

If you do that, please post the link to your issue or discussion here.

kuligs2

CodeSmile I cant remember seeing this paul, but its not a presidential electory question. Its more for noobs who just starting out, to make decision, wether learn proprietary language laike C# and be C# andy, or use gdscript and become gigachad with nice jawline.

The result of the poll would be the obvious decision to make for noobs who know no better on which langvich to choose frum and stoit makin gaymes.

CodeSmile

kuligs2 Its more for noobs who just starting out, to make decision

Nothing in that poll suggests "hey, this stat is only for ~~noobs~~ beginners to consider".

The poll results are forming public opinion on any skill level, across all media, for many things - including resource allocation.

The "presidential election" analogy is not meant to rebuff user's choices. It's an analogy for the winner-takes-all effect that many can relate to: It suffices to win one state by a small margin, but winner gets all electoral votes => a distribution collapsed into a binary result.

In case of the language question it means a single user who uses GDScript 55% of the time will send 100% of his vote to GDScript, while the remaining 45% other languages are being nullified. And they all get an equal amount of 0% - which is to say: we lose track how much each of these individual languages are actually being used - could be 45% C#, might be C++, perhaps even Rust - we don't know, that data is lost.

Fixing this biased question means GDScript would still lead by a huge margin but we'd be able to better assess what the supporting language uses really are.

DaveTheCoder If you do that, please post the link to your issue or discussion here.

Thanks! I'll post it there soon.