Re: [Monotone-devel] Solaris breakage

[William Uther]

On 24/07/2007, at 5:43 PM, Richard Levitte wrote:

> In message  
> <address@hidden> on  
> Mon, 23 Jul 2007 23:40:52 -0700, "Zack Weinberg" <address@hidden>  
> said:
>
> zackw> I will bet a fabulous no-prize that errno's gotten clobbered
> zackw> by the time we get round to printing it.
>
> I doubt that very much, but will happily stand corrected...

Well, I committed a change that makes sure the errnos were not  
clobbered, and it hasn't changed the error.  I also changed the dir  
we were trying to delete from test_dir1 to test_dirA in case the  
final letter was the problem (on linux you get EINVAL if "pathname  
has .  as last component").

I admit to being very confused still.  None of the straws I was  
clutching at have held any weight.

Can someone with a Solaris box do an ls -laR in the tester_dir/ 
clone_warning_with_multiple_heads/ directory after it has just  
failed?  Can they also do a "man 3 remove" and "man 2 rmdir" and see  
what could cause an "Invalid argument" error?

At this point I think we should not hold up the release for Solaris.   
With no way forward, we don't know how long this will take to fix,  
and the fact that it is not a major bug (it is just failing to clean  
up properly after a failure) I don't think it is worth waiting for.

How close is Windows to working?  I note that vista is still failing:

    charset:idna_encoding                         FAIL

and

362 revert_unchanged_file_preserves_mtime         FAIL (line 24)
445 ws_ops_with_wrong_node_type                   FAIL (line 12)

Be well all,

Will         :-}