[
https://issues.apache.org/jira/browse/IVY-1486?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14237886#comment-14237886
]
Pieter Goddijn edited comment on IVY-1486 at 12/8/14 3:27 PM:
--------------------------------------------------------------
Edit: this patch (both versions) doesnt work completely for our case.
The difference is probably that we use a lastest strategy on the main
dependency (latest.revision)
When usings a specific revision it works, but when using latest it doesnt
(the ivy-report task does show the correct version, but the actual delivered
ivy.xml doesnt
Edit2: i apear to have repsonded to a wrong issue (getting the wrong revisions
in deliver, but not related to exclude.
was (Author: [email protected]):
Edit: this patch (both versions) doesnt work completely for our case.
The difference is probably that we use a lastest strategy on the main
dependency (latest.revision)
When usings a specific revision it works, but when using latest it doesnt
(the ivy-report task does show the correct version, but the actual delivered
ivy.xml doesnt
> Order-dependent resolution: exclude not respected
> -------------------------------------------------
>
> Key: IVY-1486
> URL: https://issues.apache.org/jira/browse/IVY-1486
> Project: Ivy
> Issue Type: Bug
> Components: Core
> Affects Versions: trunk
> Reporter: David Turner
> Attachments: 0001-Fix-IVY-1486.patch,
> 0001-Make-doesCallersExclude-check-only-callers-of-this-v.patch, ivy-4sq.xml,
> ivy-exclude-finagle.xml, ivy-min2.xml
>
>
> To repro, create an ivy.xml which looks like this (see the attached one for
> an real-world example that you can test out):
> dependency A 1.1
> dependency X 1.2
> exclude Y
> -------
> dependency A 1.1's depends on:
> dependency X 1.1
> Dependency X 1.1's depends on:
> dependency Y 1.1
> Dependency X 1.2's depends on:
> dependency Y 1.2
> ------------------------------
> In this case, we will get Y 1.2. But if we swap the order of the
> dependencies in our ivy.xml, we'll get Y 1.1. I think 1.1 is correct,
> because we've excluded 1.2.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
