[rendaw]

I was curious seeing this thread, and I just looked and don't get it either. AFAICT the CP/M references could have been entirely omitted and nothing in the narrative about TMP and TEMP would change.

[jodrellblank]

Except that DOS was made to have its first programs ported from CP/M, so it’s relevant to explain that there were no environment variables to inherit from CP/M and no developer habits or program standards to inherit from CP/M programs.

[drfloyd51]

Which is irrelevant to TMP or TEMP.

It could simply be: When envars were added to MSDOS…

[jodrellblank]

Multics had envars in the 1960s and Unix in the 1970s, why were they ‘added’ to DOS when it was so close to an older OS, why didn’t it inherit them from CP/M? Did it get TMP from CP/M and introduce TEMP because computers were bigger but then?

[elch]

Did Multics actually have something really similar to the much later envars introduced in Unix?

[imtringued]

Those questions appear awfully overfit to the current blog post.

[jodrellblank]

That comment feels awfully cherry-picked to the perfect untestable rebuttal based on what you want to be true.

How did you measure the fitness and decide it was 'over'?