[
https://issues.apache.org/jira/browse/HADOOP-7557?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13204735#comment-13204735
]
Doug Cutting commented on HADOOP-7557:
--------------------------------------
Suresh> Doing this code where mandatory keys are enforced and optional keys are
ignore is not easy.
Why is that hard? The implementation either requires the keys (mandatory) or
first checks whether they're specified (optional).
I provided a patch that's equivalent in functionality to Sanjay's protobuf
patch yet is about half the size, disregarding the protobuf generated code in
his patch. I don't think adding some code to make some items optional would
not make it much bigger or complicated.
Do you have real examples of RPC headers that would make this unmanageable
using the approach I presented in that patch?
Sanjay> Would adding this field be a reasonable compromise?
Seems like added complexity without much added value.
Why not just get rid of Writable and RpcEngine and hardwire protobuf from the
top down? That would be easier to support and would be easier for folks to
implement in other languages. Why have an implementation that pretends to be
serialization independent? What advantages does that provide?
> Make IPC header be extensible
> -------------------------------
>
> Key: HADOOP-7557
> URL: https://issues.apache.org/jira/browse/HADOOP-7557
> Project: Hadoop Common
> Issue Type: Sub-task
> Reporter: Sanjay Radia
> Assignee: Sanjay Radia
> Attachments: HADOOP-7557.patch, IpcHeader.proto, ipcHeader1.patch,
> ipcHeader2.patch
>
>
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
