cygwin: Documentation not built, which causes test failures

[williamstein]

sage -t  "devel/sage/sage/misc/sagedoc.py"                  
**********************************************************************
File "/home/wstein/sage-4.4.3/devel/sage/sage/misc/sagedoc.py", line 891:
    sage: len(search_doc('tree', interact=False).splitlines()) > 2000
Expected:
    True
Got:
    False
**********************************************************************
File "/home/wstein/sage-4.4.3/devel/sage/sage/misc/sagedoc.py", line 496:
    sage: 'abvar/homology' in _search_src_or_doc('doc', 'homology', 'variety', interact=False)
Expected:
    True
Got:
    False

CC: @jpflori

Component: porting: Cygwin

Reviewer: Karl-Dieter Crisman, Jean-Pierre Flori

Issue created by migration from https://trac.sagemath.org/ticket/9171

[kcrisman]

comment:1

I usually get these errors if the documentation isn't built. Is that's what's going on here?

[kcrisman]

comment:2

I get the second failure, but not the first one, on a recent build on XP.

[kcrisman]

comment:3

I'm getting these failures by hand, though. And I checked - the documentation is not automatically built. So let's change this title.

[kcrisman]

comment:4

JP says that the doc does now build (since Maxima now works with #9167).

[jpflori]

comment:6

Yup the doc built fine for me and the test passes.
If you can reproduce that, let's close this ticket.

[kcrisman]

comment:7

This is really frustrating for me - I simply cannot build the doc, or (once again) even start Sage, even though it just was working a few days ago. All the usual can't map foo.dll to the same address stuff. I've rebased several times, no luck, though sometimes different files that can't map...

I'd like to try this again, though - yet another build from scratch, hopefully. There is no rush to close this, after all, as there is no patch or spkg required.

[jpflori]

comment:8

I once again built the doc succesfully, so let's close this one.
I doubt XP/7 32bits/64bits will make any difference.

[kcrisman]

comment:9

No, I think you are right. Assuming one can get Sage to start reliably, this should now be okay.

[jpflori]

Reviewer: Karl-Dieter Crisman, Jean-Pierre Flori

[jpflori]

comment:10

Ok, so I'm putting this as positive_review/wontfix

[kcrisman]

comment:11

I guess it would be worth confirming this doctest passes :) though I'm sure it will now.

[jpflori]

comment:12

It passed on my 5.5.rc0 and 5.6.rc0 'make ptest' and just did so on my 5.7.beta1.