incubator-ivy-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Gilles Scokart (JIRA)" <>
Subject [jira] Commented: (IVY-493) buildlist broken
Date Mon, 07 May 2007 21:06:15 GMT


Gilles Scokart commented on IVY-493:

While working on the test, I also found an unexpected behavior when we have a loop.

When we have an (indirect) dependencies from a module A to a module B, and the path from A
to B is not included in a cycle, I expected to have B before A.

This is currently not the case if A is included in a cycle (not including B).

For example (inspired from testCircularDependency2):
M2 depends on M3
M2 depends on M1
M3 depends on M2
M4 depends on M3

can produce the sequence : M3, M1, M2, M4 with some particular initial order
But there is a path M3->M2->M1 not included in any circular loop.  So, I think this
result is not valid.  All ordered sequence should have M1 before M3.

If you want, I can include a fix for that in my patch as well.


> buildlist broken
> ----------------
>                 Key: IVY-493
>                 URL:
>             Project: Ivy
>          Issue Type: Bug
>          Components: Core
>    Affects Versions: 2.0.0-alpha-1
>            Reporter: Gilles Scokart
>            Priority: Critical
>             Fix For: 2.0.0-alpha-2
>         Attachments: IVY-493.patch
> buildlist doesn't sort the modules anymore in 2.0-alpha1 when local ivy files doesn't
contains a revision.  
> I traced back the problem to the ModuleDescriptorSorter that was asuming that the revision
is null in such a case, while it is now 'working@hostname'.
> I will post the patch soon.

This message is automatically generated by JIRA.
You can reply to this email to add a comment to the issue online.

View raw message