hawq-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Ting(Goden) Yao" <t...@pivotal.io>
Subject Re: HAWQ 795,796,797
Date Mon, 20 Jun 2016 21:14:20 GMT
If there's no intention to maintain our own copy, why would we have source
code incorporated in the code base?
Can we just refer to a specific version of Apache ORC?


On Fri, Jun 17, 2016 at 9:50 PM Lei Chang <lei_chang@apache.org> wrote:

> My understanding of the intention here is not to maintain our own copy and
> and the target is to contribute it back to ORC project. It is just for
> improving developer efficiency that might be introduced by delay from
> acceptance from another project.
>
> Hong and I had a offline discussion, I think we can have a better way for
> this. From initial development, we even do not need to change the library
> and if changes are needed, the proposal is to start JIRAs and submit pull
> requests on Apache ORC.
>
> Cheers
> Lei
>
>
> On Fri, Jun 17, 2016 at 8:32 AM, Roman Shaposhnik <roman@shaposhnik.org>
> wrote:
>
> > Guys,
> >
> > I'm looking at:
> >     https://github.com/apache/incubator-hawq/pull/701/
> >
> > And I've got to say -- this gives me a bit of a concern.
> >
> > Unless I'm missing something it looks like you're planning
> > on forking the entire Apache ORC codebase and maintaining
> > it inside of the Apache HAWQ codebase. If that's the case
> > I have to say with my mentor (and frankly just a developers')
> > hat on -- I'm a strong -1 on the approach.
> >
> > Can somebody please comment?
> >
> > Thanks,
> > Roman.
> >
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message