Process for Contributing to Thrift?

[Jonathan Seidman <jo...@opendatagroup.com>]

We've developed a Thrift service that provides a common client interface to
distributed file systems, similar to the HDFS-APIs project which is part of
the Hadoop contrib package, but designed to support a broader array of
DFS's. Currently we support Hadoop and Sector, and given time and resources
would like to implement support for KFS.

We plan to open source this project, and would like to explore contributing
it to Thrift. We've searched the wiki and mailing lists for info on the
process to contribute a project to Thrift, but haven't been able to find
anything. If this is something that folks feel would have value as a
contributed project, can somebody provide us with details on the process we
need to follow?

Thanks.

Jonathan

-- 
Jonathan Seidman
Open Data Group

Re: Process for Contributing to Thrift?

[Jonathan Seidman <jo...@opendatagroup.com>]

Thanks much for the suggestions - we'll go ahead and release this through
one of the project hosting services and provide a link when it's up.

Thanks again.

Jonathan

On Mon, May 18, 2009 at 10:35 AM, Todd Lipcon <to...@cloudera.com> wrote:

> Hi Jonathan,
>
> Unlike Hadoop, thrift doesn't have much of a concept of a contrib/ dir. If
> you make changes to Thrift itself (ie the code generator or the language
> binding libraries) then you would open a JIRA and post a patch.
>
> It sounds like your project is standalone and just happens to depend on
> Thrift. In that case I think you'd be best off trying to open source it
> separately, probably using a service like Google Code or GitHub (my
> personal
> preference)
>
> Hope that helps,
> -Todd
>
> On Mon, May 18, 2009 at 8:28 AM, Jonathan Seidman <
> jonathan.seidman@opendatagroup.com> wrote:
>
> > We've developed a Thrift service that provides a common client interface
> to
> > distributed file systems, similar to the HDFS-APIs project which is part
> of
> > the Hadoop contrib package, but designed to support a broader array of
> > DFS's. Currently we support Hadoop and Sector, and given time and
> resources
> > would like to implement support for KFS.
> >
> > We plan to open source this project, and would like to explore
> contributing
> > it to Thrift. We've searched the wiki and mailing lists for info on the
> > process to contribute a project to Thrift, but haven't been able to find
> > anything. If this is something that folks feel would have value as a
> > contributed project, can somebody provide us with details on the process
> we
> > need to follow?
> >
> > Thanks.
> >
> > Jonathan
> >
> > --
> > Jonathan Seidman
> > Open Data Group
> >
>



-- 
Jonathan Seidman
Open Data Group

Re: Process for Contributing to Thrift?

[Bryan Duxbury <br...@rapleaf.com>]

I agree with Todd's suggestion. Put it in some other open-source  
system so that you can have maximum ease of control over the project.  
I think that we would happily add a link to your stuff to our  
website, though.

-Bryan

On May 18, 2009, at 8:35 AM, Todd Lipcon wrote:

> Hi Jonathan,
>
> Unlike Hadoop, thrift doesn't have much of a concept of a contrib/  
> dir. If
> you make changes to Thrift itself (ie the code generator or the  
> language
> binding libraries) then you would open a JIRA and post a patch.
>
> It sounds like your project is standalone and just happens to  
> depend on
> Thrift. In that case I think you'd be best off trying to open  
> source it
> separately, probably using a service like Google Code or GitHub (my  
> personal
> preference)
>
> Hope that helps,
> -Todd
>
> On Mon, May 18, 2009 at 8:28 AM, Jonathan Seidman <
> jonathan.seidman@opendatagroup.com> wrote:
>
>> We've developed a Thrift service that provides a common client  
>> interface to
>> distributed file systems, similar to the HDFS-APIs project which  
>> is part of
>> the Hadoop contrib package, but designed to support a broader  
>> array of
>> DFS's. Currently we support Hadoop and Sector, and given time and  
>> resources
>> would like to implement support for KFS.
>>
>> We plan to open source this project, and would like to explore  
>> contributing
>> it to Thrift. We've searched the wiki and mailing lists for info  
>> on the
>> process to contribute a project to Thrift, but haven't been able  
>> to find
>> anything. If this is something that folks feel would have value as a
>> contributed project, can somebody provide us with details on the  
>> process we
>> need to follow?
>>
>> Thanks.
>>
>> Jonathan
>>
>> --
>> Jonathan Seidman
>> Open Data Group
>>

Re: Process for Contributing to Thrift?

[Todd Lipcon <to...@cloudera.com>]

Hi Jonathan,

Unlike Hadoop, thrift doesn't have much of a concept of a contrib/ dir. If
you make changes to Thrift itself (ie the code generator or the language
binding libraries) then you would open a JIRA and post a patch.

It sounds like your project is standalone and just happens to depend on
Thrift. In that case I think you'd be best off trying to open source it
separately, probably using a service like Google Code or GitHub (my personal
preference)

Hope that helps,
-Todd

On Mon, May 18, 2009 at 8:28 AM, Jonathan Seidman <
jonathan.seidman@opendatagroup.com> wrote:

> We've developed a Thrift service that provides a common client interface to
> distributed file systems, similar to the HDFS-APIs project which is part of
> the Hadoop contrib package, but designed to support a broader array of
> DFS's. Currently we support Hadoop and Sector, and given time and resources
> would like to implement support for KFS.
>
> We plan to open source this project, and would like to explore contributing
> it to Thrift. We've searched the wiki and mailing lists for info on the
> process to contribute a project to Thrift, but haven't been able to find
> anything. If this is something that folks feel would have value as a
> contributed project, can somebody provide us with details on the process we
> need to follow?
>
> Thanks.
>
> Jonathan
>
> --
> Jonathan Seidman
> Open Data Group
>