Solaris package

Matt Mackall mpm at selenic.com
Mon Aug 3 14:44:43 CDT 2009


On Mon, 2009-08-03 at 11:32 +0200, Martin Geisler wrote:
> "rupert.thurner" <rupert.thurner at gmail.com> writes:
> 
> I'm sending this reply to the mercurial-devel list since we've moved
> from user questions into development questions. Please make sure to send
> your reply there too.
> 
> > the remaining errors are:
> > Failed test-convert-cvs-branch: output changed and returned error code
> > 1
> > Failed test-convert-cvs-synthetic: output changed and returned error
> > code 1
> > Failed test-convert-cvsnt-mergepoints: output changed
> > Failed test-convert-git: output changed
> > Failed test-hook: output changed
> > Failed test-import: output changed
> > Failed test-paths: output changed
> >
> > we do not have cvsnt installed, so probably this test should be
> > skipped. as we use the current standard 1.11.23 of the stock cvs we
> > wonder why the tests fail.
> >
> > just to take out one error, how does one read that output currectly?
> 
> The output of each test is compared with previously recorded output:
> test-foo stores its output in test-foo.out. You see the differences in
> the output:
> 
> > ERROR: test-hook output changed
> > --- Expected output
> > +++ Test output
> >  -1,29 +1,29 @@
> > -commit hook: HG_NODE=29b62aeb769fdf78d8d9c5f28b017f76d7ef824b
> > HG_PARENT1=0000000000000000000000000000000000000000
> > +commit hook: HG_NODE=29b62aeb769fdf78d8d9c5f28b017f76d7ef824b
> > HG_PARENT1=0000000000000000000000000000000000000000 HG_PENDING=true
> >
> > does this mean the HG_PENDING=true is there despite it is not
> > expected? why it is like this?
> 
> Yes, the above diff means that the HG_PENDING=true part should not be
> there. I don't know why it is there.

That definitely shouldn't happen. We only pass pending to the pretxn
hooks. Is this repeatable? Does it happen with 1.2.1?

-- 
http://selenic.com : development and support for Mercurial and Linux




More information about the Mercurial-devel mailing list