I would suggest to cherry-pick the second commit "Remove pod2htmd.tmp in make clean"
to 1.1.1.

I have not observed the overlong line in windows make clean, for 1.1.1,
but instead of pod2htmd.tmp there are two of them: pod2htmd.tmp and pod2htmi.tmp
and the rmdir does not work if any cygwin is around.

Richard, you probably remember the issue with the "echo *" on windows, which
which was supposed to just print "*" in this case, but that causes a surprise
when cygwin's echo command is also in the path. then "*" is expanded as if it
was a linux echo. You wrote an echo.pl to fix that issue.
Now the rmdir is the same issue, I see no way how to make nmake use the builtin.
The other issue I had with windows is a command length in del $(GENERATED)
that happened only on master, 1.1.1 did not exceed the line length fortunately.
I saw no other solution than to use perl for that and use a temp file to pass the list.

I did test that on a windows 7 machine.
I use remote access to the our internal company network.
I can still do that, but I consider this not necessary.

so I don't feel I can give the final thumbs up.

Maybe you just trust me then?

so I don't feel I can give the final thumbs up.

Maybe you just trust me then?

I do, but if Richard is free to take a look, he's much more familiar with the build-system than I. I am not sure I be doing approvals after only cursory inspection. Though perhaps with the build-system I may not need to be quite so cautious. Ping me again in a couple of days if nobody else steps forward...

so I don't feel I can give the final thumbs up.

Maybe you just trust me then?

You know it quite well, reviewing is not a matter of trust or distrust. It‘s simply better if someone else checks the changes independently. As the proverb says: „Vertrauen ist gut, Kontrolle ist besser.“

Maybe I find some time to test it...

Just for reference, when I post a GCC patch,
I write in the e-mail

Bootstrapped and Reg-tested on x86_64-pc-linux-gnu
Is it OK for thunk?

bootstrap takes hours, and the test suite a day or more.
every reviewer trusts me, that I say that honestly,
the review is about how the code looks like nothing more nothing less.
If they start to not trust a person that is listed in "Write after Approval"
they have to duplicate all my effort, and moreover there is
no CI, it would not be usable since the test suite that runs for days.
Just automated test runs after a patch is applied.

Removing windows stuff for now.
I don't know if the nmake distclean was fixed by @levitte
but I dont care about windows any more.
In the moment I see the following files are leaking again,

$ git clean -ndx
Would remove crypto/include/
Would remove providers/fipsinstall.cnf
Would remove test-runs/

So since I started this PR new bugs were introduced,
nobody reviewed this, for a long time, if I spend time
to fix this PR again, it will not be reviewed again,
probably I better give up right now before I waste more time.