You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@mrql.apache.org by Leonidas Fegaras <fe...@cse.uta.edu> on 2013/03/24 20:26:15 UTC
[DISCUSS] MRQL dev process
Dear MRQL-workers (miracle-workers),
Thank you for volunteering to support the incubation of the MRQL
project. I think it's time to start discussing some of the details.
First, I would like to discuss the idea of adopting the
Review-Then-Commit policy for MRQL, which requires consensus approval
before a commit (ie, at least three binding +1 votes and no vetos).
The initial committers would be those listed in the MRQL Incubation
proposal. Let's discuss this during the next few days and then I will
open it for vote.
Thank you
Leonidas Fegaras
Re: [DISCUSS] MRQL dev process
Posted by Alex Karasulu <ak...@apache.org>.
On Sun, Mar 24, 2013 at 9:26 PM, Leonidas Fegaras <fe...@cse.uta.edu>wrote:
> Dear MRQL-workers (miracle-workers),
>
Hehehe, I like the pun :). Second I love the name, and that 300 movie!
RTC sounds cool especially with all the crap code I have been seeing on
many projects. Sounds like you guys are a lot more serious about getting it
right. That's pleasant to see.
I just saw Ed's post too and it makes sense since not that many people have
enough time to review and the community is just starting off right now and
it's small. He has a good point for now at least to keep it at one other
committer approving. Maybe you can start off with one other +1 and see how
that works, and later you can move up to 3 +1 if that's at all needed.
--
Best Regards,
-- Alex
Re: [DISCUSS] MRQL dev process
Posted by "Edward J. Yoon" <ed...@apache.org>.
On Wed, Mar 27, 2013 at 12:15 AM, Leonidas Fegaras <fe...@cse.uta.edu> wrote:
> From the discussion so far, it seems that we can adopt the
> review-then-commit policy that requires at least two +1 from committers and
> no -1 from committer. If the contributor is already a committer, then it
> requires just one extra +1 and no -1 from committers. This may change latter
> when we recruit more committers. Should we vote for this?
Yes, this would be our first vote. But I think It's a little early for
us to start voting. Maybe next week?
> Also, it seems that git could be a better choice than svn. Does the majority
> of ASF project use svn? It's not hard to switch to svn. I either case,
Git seems already used for many Apache projects (S4, Crunch, Tajo, ..,
etc.). We might want to check with infrastructure@apache.org once
again.
> somebody must create a git or svn repo and a wiki page.
> Here is the progress so far:
> 1) Ed has graciously volunteered to convert the MRQL html doc to wiki.
> Thanks Ed!
> 2) I have fixed the Copyright info in the source: I have replaced source
> file headers, LICENSE and NOTICE files based on ASF policy.
> 3) I have made the Makefile more generic so it can compile any file dropped
> in the src directory. Do you think it's a better idea to switch to Maven? I
> have never used it.
+1 for maven. It is simplest way to manage dependencies, deploy the
site, and build artifacts.
> I am planning to split the source files to smaller logically independent
> files and write a developer's roadmap -- this may take few weeks.
> Leonidas
>
>
> On Mar 24, 2013, at 7:54 PM, Karthik Kambatla wrote:
>
>> +1 for review-then-commit.
>>
>> +1 on one "+1" from another committer and no "-1"s if the contributor is a
>> committer. For the case where contributor is not a committer, is it still
>> one "+1" from a committer?
>>
>> On Sun, Mar 24, 2013 at 3:16 PM, Edward J. Yoon <ed...@apache.org>
>> wrote:
>> +1 for R-T-C.
>>
>> > before a commit (ie, at least three binding +1 votes and no vetos).
>>
>> IMO, one "+1" from another committer is enough.
>>
>> On Mon, Mar 25, 2013 at 4:26 AM, Leonidas Fegaras <fe...@cse.uta.edu>
>> wrote:
>> > Dear MRQL-workers (miracle-workers),
>> > Thank you for volunteering to support the incubation of the MRQL
>> > project. I think it's time to start discussing some of the details.
>> > First, I would like to discuss the idea of adopting the
>> > Review-Then-Commit policy for MRQL, which requires consensus approval
>> > before a commit (ie, at least three binding +1 votes and no vetos).
>> > The initial committers would be those listed in the MRQL Incubation
>> > proposal. Let's discuss this during the next few days and then I will
>> > open it for vote.
>> > Thank you
>> > Leonidas Fegaras
>> >
>>
>>
>>
>> --
>> Best Regards, Edward J. Yoon
>> @eddieyoon
>>
>
--
Best Regards, Edward J. Yoon
@eddieyoon
Re: [DISCUSS] MRQL dev process
Posted by Leonidas Fegaras <fe...@cse.uta.edu>.
From the discussion so far, it seems that we can adopt the review-
then-commit policy that requires at least two +1 from committers and
no -1 from committer. If the contributor is already a committer, then
it requires just one extra +1 and no -1 from committers. This may
change latter when we recruit more committers. Should we vote for this?
Also, it seems that git could be a better choice than svn. Does the
majority of ASF project use svn? It's not hard to switch to svn. I
either case, somebody must create a git or svn repo and a wiki page.
Here is the progress so far:
1) Ed has graciously volunteered to convert the MRQL html doc to wiki.
Thanks Ed!
2) I have fixed the Copyright info in the source: I have replaced
source file headers, LICENSE and NOTICE files based on ASF policy.
3) I have made the Makefile more generic so it can compile any file
dropped in the src directory. Do you think it's a better idea to
switch to Maven? I have never used it.
I am planning to split the source files to smaller logically
independent files and write a developer's roadmap -- this may take few
weeks.
Leonidas
On Mar 24, 2013, at 7:54 PM, Karthik Kambatla wrote:
> +1 for review-then-commit.
>
> +1 on one "+1" from another committer and no "-1"s if the
> contributor is a committer. For the case where contributor is not a
> committer, is it still one "+1" from a committer?
>
> On Sun, Mar 24, 2013 at 3:16 PM, Edward J. Yoon
> <ed...@apache.org> wrote:
> +1 for R-T-C.
>
> > before a commit (ie, at least three binding +1 votes and no vetos).
>
> IMO, one "+1" from another committer is enough.
>
> On Mon, Mar 25, 2013 at 4:26 AM, Leonidas Fegaras
> <fe...@cse.uta.edu> wrote:
> > Dear MRQL-workers (miracle-workers),
> > Thank you for volunteering to support the incubation of the MRQL
> > project. I think it's time to start discussing some of the details.
> > First, I would like to discuss the idea of adopting the
> > Review-Then-Commit policy for MRQL, which requires consensus
> approval
> > before a commit (ie, at least three binding +1 votes and no vetos).
> > The initial committers would be those listed in the MRQL Incubation
> > proposal. Let's discuss this during the next few days and then I
> will
> > open it for vote.
> > Thank you
> > Leonidas Fegaras
> >
>
>
>
> --
> Best Regards, Edward J. Yoon
> @eddieyoon
>
Re: [DISCUSS] MRQL dev process
Posted by Karthik Kambatla <ka...@cloudera.com>.
+1 for review-then-commit.
+1 on one "+1" from another committer and no "-1"s if the contributor is a
committer. For the case where contributor is not a committer, is it still
one "+1" from a committer?
On Sun, Mar 24, 2013 at 3:16 PM, Edward J. Yoon <ed...@apache.org>wrote:
> +1 for R-T-C.
>
> > before a commit (ie, at least three binding +1 votes and no vetos).
>
> IMO, one "+1" from another committer is enough.
>
> On Mon, Mar 25, 2013 at 4:26 AM, Leonidas Fegaras <fe...@cse.uta.edu>
> wrote:
> > Dear MRQL-workers (miracle-workers),
> > Thank you for volunteering to support the incubation of the MRQL
> > project. I think it's time to start discussing some of the details.
> > First, I would like to discuss the idea of adopting the
> > Review-Then-Commit policy for MRQL, which requires consensus approval
> > before a commit (ie, at least three binding +1 votes and no vetos).
> > The initial committers would be those listed in the MRQL Incubation
> > proposal. Let's discuss this during the next few days and then I will
> > open it for vote.
> > Thank you
> > Leonidas Fegaras
> >
>
>
>
> --
> Best Regards, Edward J. Yoon
> @eddieyoon
>
Re: [DISCUSS] MRQL dev process
Posted by "Edward J. Yoon" <ed...@apache.org>.
+1 for R-T-C.
> before a commit (ie, at least three binding +1 votes and no vetos).
IMO, one "+1" from another committer is enough.
On Mon, Mar 25, 2013 at 4:26 AM, Leonidas Fegaras <fe...@cse.uta.edu> wrote:
> Dear MRQL-workers (miracle-workers),
> Thank you for volunteering to support the incubation of the MRQL
> project. I think it's time to start discussing some of the details.
> First, I would like to discuss the idea of adopting the
> Review-Then-Commit policy for MRQL, which requires consensus approval
> before a commit (ie, at least three binding +1 votes and no vetos).
> The initial committers would be those listed in the MRQL Incubation
> proposal. Let's discuss this during the next few days and then I will
> open it for vote.
> Thank you
> Leonidas Fegaras
>
--
Best Regards, Edward J. Yoon
@eddieyoon