[
https://issues.apache.org/jira/browse/HAMA-735?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13581749#comment-13581749
]
Suraj Menon commented on HAMA-735:
----------------------------------
This is a huge patch. Did we miss so many space formatting points?!
Thanks for removing the readState, writeState, create(Anything) clutter. I had
got some negative feedback from few of our users.
Can we make - V extends WritableComparable<?> into V extends
WritableComparable<? super V>. This would ensure that we are comparing V. We
can throw it for brainstorming. Would like to know pluses and minuses. I went
with this approach in the patch I uploaded yesterday. And I also carried the
same over to other classes like VerticeInfo and GraphJobMessage. I am still
split on Comparator for GraphJobMessage because only Vertice messages are
comparable.
I have not perused the whole patch yet.
> Tighten the graph API
> ----------------------
>
> Key: HAMA-735
> URL: https://issues.apache.org/jira/browse/HAMA-735
> Project: Hama
> Issue Type: Improvement
> Components: graph
> Affects Versions: 0.6.0
> Reporter: Thomas Jungblut
> Assignee: Thomas Jungblut
> Fix For: 0.6.1
>
> Attachments: HAMA-735_1.patch
>
>
> Tighten the API by the following improvements:
> - replace the need to provide a new object of vertex parts
> - readState/writeState shouldn't be forced to override (clutters code with
> empty methods)
> - add preconditions for not supplied readers
> - generalize the usage of vertex part classes and unify their usage
> - replace unuseful testcases / test methods
> - simplify graph logic by removing repair functionalities
> - formatting + organize of imports
> - add WritableComparable<?> to the vertex ID for efficient merging of
> vertices and joining of them with messages
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
