Solaris package

rupert.thurner rupert.thurner at gmail.com
Wed Aug 5 03:43:55 CDT 2009


On Aug 4, 9:06 pm, Mads Kiilerich <m... at kiilerich.com> wrote:
> On 04/08/2009, at 16.05, "rupert.thurner" <rupert.thur... at gmail.com>  
> wrote:
>
>
>
>
>
> > ERROR: test-hook output changed
> > --- Expected output
> > +++ Test output
> > @@ -1,29 +1,29 @@
> > precommit hook: HG_PARENT1=0000000000000000000000000000000000000000
> > pretxncommit hook: HG_NODE=29b62aeb769fdf78d8d9c5f28b017f76d7ef824b
> > HG_PARENT1=0000000000000000000000000000000000000000 HG_PENDING=true
> > 0:29b62aeb769f
> > -commit hook: HG_NODE=29b62aeb769fdf78d8d9c5f28b017f76d7ef824b
> > HG_PARENT1=0000000000000000000000000000000000000000
> > -commit.b hook: HG_NODE=29b62aeb769fdf78d8d9c5f28b017f76d7ef824b
> > HG_PARENT1=0000000000000000000000000000000000000000
> > +commit hook: HG_NODE=29b62aeb769fdf78d8d9c5f28b017f76d7ef824b
> > HG_PARENT1=0000000000000000000000000000000000000000 HG_PENDING=true
> > +commit.b hook: HG_NODE=29b62aeb769fdf78d8d9c5f28b017f76d7ef824b
> > HG_PARENT1=0000000000000000000000000000000000000000 HG_PENDING=true
>
> Do other solaris users see the same failures?
>
> Perhaps it could be interesting to see the actual value of HG_PENDING  
> - perhaps you could hack the code to show that? IIRC solaris setenv  
> differs from most other unices (also other BSDs?) by not strdup'ing  
> its value. Python should hide that, but ...

not sure what you mean... here: http://hg.intevation.org/mercurial/crew/file/1adabc0c05a2/tests/test-hook
is no HG_PENDING. who sets it and where in the code should be the
problem when setting it? will try to output it.

could it be that the other test case failures follow a similar
structure? if we take the test-import test case, the result is:
ERROR: test-import output changed
--- Expected output
+++ Test output
 -148,7 +148,6 @@
 updating working directory
 2 files updated, 0 files merged, 0 files removed, 0 files unresolved
 applying patch from stdin
-abort: empty commit message
 % hg export in email, should use patch header
 requesting all changes
 adding changesets

this means the expected result is an error because of an empty commit
message, and it anyway commited?

and the test-paths testcase:

ERROR: test-paths output changed
--- Expected output
+++ Test output
 -1,2 +1,4 @@
 updating working directory
 0 files updated, 0 files merged, 0 files removed, 0 files unresolved
+comparing with /tmp/hgtests.7DZZjc/test-paths/b
+comparing with /tmp/hgtests.7DZZjc/test-paths/b
!....................................................................................

rupert



More information about the Mercurial-devel mailing list