[ 
https://issues.apache.org/jira/browse/IVYDE-64?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12543573
 ] 

Xavier Hanin commented on IVYDE-64:
-----------------------------------

Thanks for this patch, it actually simplifies the resolve process in 
IvyClasspathContainer. There's only one thing I think is still necessary that 
your patch removes: the check to see if we should use previous resolve if it 
exist. Indeed, Ivy itself doesn't do such a check when you call resolve. What 
Ivy does is use its cache to avoid downloading unrequired artifacts and 
metadata, so if you use only static versions and have everything in cache, the 
resolve process will be only local and very quick. But what we do when 
usePreviousResolveIfExists is true in IvyDE is use the last xml resolve report, 
which leads to no actual resolve process. This can be useful when you call 
resolve from Ant, and only want to reuse the same resolve result within IvyDE. 
Hence to preserve this we still need the report parsing operation in this case, 
so I don't know if using the ResolveReport instead of the report when we 
actually do a resolve is really simpler or not.

WDYT?

> Simplify the resolve process
> ----------------------------
>
>                 Key: IVYDE-64
>                 URL: https://issues.apache.org/jira/browse/IVYDE-64
>             Project: IvyDE
>          Issue Type: Improvement
>          Components: classpath container
>    Affects Versions: 1.3.0
>         Environment: trunk r595956
>            Reporter: Nicolas Lalevée
>         Attachments: IVYDE-64-r595956.patch
>
>
> The current trunk, the resolve job is first checking manually if the artifact 
> has already been resolved. If not in cache, it let ivy resolve it. And 
> finally it parses the xml reports in the cache.
> So first, as far as I have understood, the manual checking against the cache 
> is not needed, ivy already does it.
> And then ivy produce in Java some resolve report that already contain every 
> needed information, there is no need to parse the reports in cache.

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

Reply via email to