Re: nvm.lua-5.2 failing on mingw

Stephen Leake <[email protected]>
Newsgroups gmane.comp.version-control.monotone.devel
Message-ID <[email protected]>
Richard Levitte <[email protected]> writes:

> In message <[email protected]> on Mon, 07 May 2012 12:36:10 +0100, Stephen Leake <[email protected]> said:
>
> stephen_leake> Stephen Leake <[email protected]> writes:
> stephen_leake> 
> stephen_leake> > Completely reinstalling mingw, using an older installer that has g++
> stephen_leake> > 4.5.2, fixed this particular problem (running 'make check' now).
> stephen_leake> >
> stephen_leake> > So apparently g++ 4.6.2 has problems with exceptions on Mingw. Cygwin is
> stephen_leake> > still at 4.5.3; I don't think Debian is at 4.6 yet. So perhaps this is
> stephen_leake> > not surprising.
> stephen_leake> 
> stephen_leake> 'make check' now completed.
> stephen_leake> 
> stephen_leake> tester and unit tests all pass.
> stephen_leake> 
> stephen_leake> func tests had one failure; netsync_key_hooks didn't get the expected
> stephen_leake> return value from the server when sent SIGTERM. I think this is a known
> stephen_leake> bug in MinGW, so we should skip that test.
> stephen_leake> 
> stephen_leake> All of the extra tests failed, as on the previous release. The failure
> stephen_leake> is 'attempt to call global 'mtn_setup' (a nil value)', so there's
> stephen_leake> something fundamental wrong in extra test driver.
> stephen_leake> 
> stephen_leake> In sum; as far as Cygwin and MinGW are concerned, nvm.lua-5.2 can be
> stephen_leake> merged to main.
>
> Wait, you're saying all of "extra" is failing?  I definitely don't get
> that on my (Linux) box...

Right, this failure is specific to MinGW; it has been there since extra
was introduced.

> I'm also a bit surprised that we consider failures with the "extra"
> part ok.

The failure appears to be in the test suite itself; it is not invoking
mtn properly. No one has had the time to investigate.

I don't use any of the tools tested in extra (on any platform), and
apparently no one has complained that they don't work on MinGW (either
no one else uses them, or they do work).

> Other than that, I agree that nvm.lua-5.2 should be merged into nvm.

The test failure occurs on nvm as well, so no need to hold the merge on
that account.

-- 
-- Stephe
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.