[gimme_treefiddy]

It does seem vague though. It’s almost akin to saying “I could’ve done this in half the code”. Well add your PR then.

[HideousKojima]

They did, they implemented a crazy fast terminal renderer (that still supports advanced features like right-to-left languages in Unicode, etc.) from scratch in about 3,000 lines of code just to prove their point:

https://github.com/cmuratori/refterm

[zamalek]

Why not just create a PR, instead of a from-scratch grandstanding repo?

This is a situation where tu quoque really doesn't smell like a logical fallacy. There is a lot of fuss about "being right" over "being helpful."

[HideousKojima]

I'm assuming he thought his PR would be rejected, given their reaction to his initial questions about the issue:

https://github.com/microsoft/terminal/issues/10362

[zamalek]

A closed PR would have been even more damning for Microsoft. Especially given that Casey could have provided apples-apples benchmark results - not, what amounts to, a micro-benchmark.

From my perspective, there are two reasons why someone would build an entire project instead of opening a PR:

* Merely to be able to point and it and say, "see? see? I was right!" How obnoxiously childish.

* Implementing his ideas into a mature/non-greenfields codebase wasn't in-fact "extremely simple."

[db48x]

He simply didn't want to license his modifications under the terms required by Microsoft.

[dogleash]

Probably doesn't want to work for microsoft for free. Microsoft is a 2 trillion dollar company; doing their job for them would be a total chump move.