[fsflover]

What if they could do it, because people like you had quit?

[izacus]

Standards committees being completely divorced from reality of software engineering is why most of the standards are useless.

So the question is whether it was actually a loss.

[pjmlp]

Like POSIX, OpenGL, OpenCL, Vulkan, C, C++, JavaScript, TCP/IP,....

[rglullis]

Is there any example on your list where the standard came before the implementation?

[pjmlp]

Yes, Vulkan (Mantle was the idea), C (since C89), C++ (since C++89), OpenCL (after Apple gave it to Khronos).

[rglullis]

So, no. None of your examples are equivalent to OOXML. The implementations were first opened up and then standardized.

OOXML was the other way around: Microsoft had a standard and tried to enshrine into a standard and force others to waste time and resources to be compatible.

[pjmlp]

Only if you ignore what was standardised in PDF form and only later made available on existing implementations.

That is why I explicitly made references to specific versions as turning points, as I expected the usual FOSS advocacy replies.

[perching_aix]

What if then? Should they have just bodied the thing for the love of the game? So that people uncaring for their wellbeing then wouldn't have appreciated it as a sacrifice anyhow?

Quite often I find that if people stopped holding fundamentally broken dynamics together and just let the thing fail and fail hard, the overall long term outcome would be better off. Much to the opposite of your suggestion.

It's just that turns out, things being properly bodied or properly broken take coordinated action. People deciding one by one, one way or the other, is what actually enables and sustains pathological dynamics like this.

But then how does one single out any specific decision? Well, nohow, not with any rigor for sure.

[thomasfl]

I was new in the standards business. Believed this was common. Understand now that it wasn't.