Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Plain View
Flume >> mail # dev >> API Docs and PDFs on the website


+
Brock Noland 2012-12-04, 16:16
+
Ralph Goers 2012-12-04, 17:09
+
Brock Noland 2012-12-04, 17:12
Copy link to this message
-
Re: API Docs and PDFs on the website
That's there for Flume 0.9 backwards compatibility, thus being in legacy
sources. It's already marked as @deprecated.

Someone who is an expert with Thrift may have more insight into whether we
can maintain backwards compat while changing the package/class names, but
my understanding is that it was left that way for that reason.
On Tue, Dec 4, 2012 at 9:12 AM, Brock Noland <[EMAIL PROTECTED]> wrote:

> Hi,
>
> It looks like it's from the legacy thrift source:
>
>
> https://github.com/apache/flume/tree/trunk/flume-ng-legacy-sources/flume-thrift-source/src/main/java
>
> Does anything know if there is a valid reason it's not org.apache? I
> assume it was just missed.
>
> Brock
>
> On Tue, Dec 4, 2012 at 11:09 AM, Ralph Goers <[EMAIL PROTECTED]>
> wrote:
> >
> > On Dec 4, 2012, at 8:16 AM, Brock Noland wrote:
> >
> >> Hi,
> >>
> >> I assume the PDFs and API Docs on the release page
> >> http://flume.apache.org/releases/1.3.0.html are just manually
> >> generated and then committed directly to the prod website?
> >
> > Yes - it looks like you did it correctly.   Unfortunately, the
> maven-pdf-plugin doesn't support RST so you have to do it manually.
> >
> > I noticed the API Javadoc contains a reference to
> com.cloudera.flume.handlers.thrift.  Why is that?
> >
> > Ralph
> >
> >
> >>
> >> I just want to confirm as I am trying to put together a guide on how
> >> to edit the website and what items need to be updated as part of a
> >> release.
> >>
> >> Brock
> >
>
>
>
> --
> Apache MRUnit - Unit testing MapReduce -
> http://incubator.apache.org/mrunit/
>