You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@thrift.apache.org by Esteve Fernandez <es...@sindominio.net> on 2008/12/23 20:10:16 UTC
more committers
Hi all, of the official committers listed in [1] only Mark, David, Kevin and
Jake are actually active. But I think we still need more, so I would like to
nominate Bryan as a new committer, he has demonstrated a lot of activity in
the Ruby and Java components, just look at how many tickets he has filed.
Although there are others taking care of those components, I think that it's
good to have some redundancy. Also, he would bring even more diversity to the
project (Facebook, Powerset, Third Rail, and now RapLeaf).
On a side note, quoting Bryan from an old e-mail: "It would also be nice to
have at least a clear point person for every library so there's someone to
email.". IMHO it's a bad signal that one of David's tasks is "Everything".
Not that I think David isn't knowledgeable enough or can't manage this
project, but it would be nice if tasks/components were more clearly defined:
General vision and implementation - Mark
Management - David
System admnistration - David
C++ - David, Mark
Java - David, Bryan?
Ruby - Kevin, Bryan?
Perl - Jake
Python - David
Erlang - ?
C# - ?
PHP - ?
Haskell - ?
Cocoa - ?
Smalltalk - ?
Haskell - ?
Cheers.
1 - http://incubator.apache.org/thrift/about/
Re: more committers
Posted by Bryan Duxbury <br...@rapleaf.com>.
>>
>> so I would like to nominate Bryan as a new committer
> +1. But just as a warning, Bryan, you're going to be hearing
> "please remove
> trailing whitespace from this patch" a lot. ;-)
I can take the heat.
-Bryan
Re: more committers
Posted by an...@gmail.com.
I dont mind reviewing, committing etc for C# if needed.
Im also interested in making thrift compile under Win32 (MS Visual
C++), so id like to eventually submit some patches for this as well.
On Wed, Dec 24, 2008 at 4:53 AM, Esteve Fernandez <es...@sindominio.net> wrote:
> On Tuesday 23 December 2008 20:31:19 David Reiss wrote:
>> > Erlang - ?
>>
>> Todd Lipcon (Toad on IRC) is really responsive to questions about the
>> Erlang mapping. Chris Piro or I can also look at these issues.
>
> Then, I think it would make sense to give Todd committer access. However I
> haven't seen much activity around Erlang in thrift-dev
>
>> > C# - ?
>>
>> I would also like to have a clear go-to person (or people) for C# issues.
>
> Yep, Ben (who knows C# and Python) appears to be MIA, THRIFT-209 is dead
> simple but can't be fixed because there aren't committers who can review the
> patch.
>
> I've contributed a couple of patches for C#, but I think Michael Greene and
> Anthony Abate surely know more C# than I.
>
>> > PHP - ?
>>
>> Mark and I can handle this. There has not been much interest in PHP
>> outside of Facebook, so it has been pretty low volume.
>
> Good. BTW, why does the Thrift compiler have special options for PHP
> (-php, -phpi, etc.) instead of using regular generator options (--gen xxx)?
>
>> > Haskell - ?
>> > Cocoa - ?
>> > Smalltalk - ?
>> > Ocaml - ?
>>
>> There has been very little interest in these languages outside of the
>> original authors, so I think they should still be considered the point
>> people.
>
> I'm sure I'm going to piss off some people, but since those languages don't
> have a mantainer and their paucity of tests, I would lower their priority (or
> label them as unsupported) when the next version of Thrift is released.
>
> Cheers.
>
--
-Anthony
Re: more committers
Posted by Jérémie BORDIER <ah...@gmail.com>.
I can't agree more... I'm quite new to thrift, submitting my first
patches but i was a bit disappointed by the lack of activity of the
committers / maintainers.. Patches are available and validated for
ages now without being applied, which makes all the people looking at
thrift doubtful about contributing and about the quality of the
current branch. Diving into the cpplib is quite straight forward but
some tricky points and important architecture decisions require
support from the cpp lib maintainers that arn't very available, making
things hard to be done..
--
Jérémie
On Mon, Jan 5, 2009 at 7:45 PM, Bryan Duxbury <br...@rapleaf.com> wrote:
> Perhaps more than anything, this thread should be viewed as a call to action
> for the existing PMC. Perhaps I don't know, since I'm also not in on the
> "private" discussions, but it certainly feels like nothing is being done at
> all. I wouldn't even be concerned about being a committer if it wasn't for
> the fact that I have to be very annoying to a David and Kevin to get my
> patches incorporated. Less tenacious individuals have clearly attached a
> patch or two and then disappeared for want of attention.
>
> I think if we want this project to succeed in the long run, we need to be
> active and responsive. If the people in charge are unwilling or unable to do
> that, then I'd just appreciate it if they could hand it off to those who
> will dedicate the time.
>
> -Bryan
>
> On Jan 2, 2009, at 2:51 AM, Upayavira wrote:
>
>> On Mon, 2008-12-29 at 16:40 +0100, Esteve Fernandez wrote:
>>>
>>> On Sunday 28 December 2008 23:37:09 Ian Holsman wrote:
>>>>
>>>> In general in an apache project, the process for getting people commit
>>>> access is for an existing 'PMC' member to recognize the merits of an
>>>> individual and raise it on the private list, where it is discussed. It
>>>> is discussed in private as some people may feel that the person is not
>>>> ready yet, and saying that in public could discourage the individual.
>>>> "nominating" the person on a public list is generally seen as poor form,
>>>> as it might preempt the discussion, and lead to no response from the
>>>> project.
>>>
>>> Sorry if my actions caused any trouble. I'm not a PMC member and
>>> shouldn't
>>> have nominated Bryan or Todd before asking about the process.
>>
>> Not a problem. Given these discussions happen in private (one of the few
>> things that does/should in the ASF), there's no way for those outside to
>> know how they happen.
>>
>>> However, I'm still arguing for more reviewers/committers in general, I
>>> think
>>> Thrift's development is being hindered by the lack of people who can act
>>> on
>>> tickets, which leads to external people become frustrated and to not
>>> participate more actively. For example, THRIFT-129 (the most popular
>>> ticket)
>>> was reported on 2008-09-02 and last updated on 2008-10-23, but nobody
>>> committed the attached patch.
>>
>> You may be right that Thrift needs new committers. However, that is an
>> issue that I will leave to the Thrift committers to discuss/address.
>>
>> Upayavira
>>
>>
>
>
--
Jérémie 'ahFeel' BORDIER
Re: more committers
Posted by Doug Cutting <cu...@apache.org>.
Bryan Duxbury wrote:
> Perhaps more than anything, this thread should be viewed as a call to
> action for the existing PMC. Perhaps I don't know, since I'm also not in
> on the "private" discussions, but it certainly feels like nothing is
> being done at all.
As you guessed, you are not missing much. And an incubating project
with most of its committers from a single company should strive hard to
add new committers.
In other Apache projects that I'm involved with, we generally nominate
contributors to become committers when they have:
- contributed a series of high-quality, non-trivial patches; and
- demonstrated that they are easy to work with.
Looking at the contributions report (http://tinyurl.com/a2mrzy) I see a
few folks who've contributed lots of patches but who have not yet been
proposed as committers. From a glance at the interactions, these folks
do not appear to be hard to work with, but I cannot claim to have
followed Thrift closely enough to be sure, nor to nominate any myself.
But it does seem like some of the more active committers might strongly
consider nominating a few folks on thrift-private@.
Doug
Re: more committers
Posted by Bryan Duxbury <br...@rapleaf.com>.
Perhaps more than anything, this thread should be viewed as a call to
action for the existing PMC. Perhaps I don't know, since I'm also not
in on the "private" discussions, but it certainly feels like nothing
is being done at all. I wouldn't even be concerned about being a
committer if it wasn't for the fact that I have to be very annoying
to a David and Kevin to get my patches incorporated. Less tenacious
individuals have clearly attached a patch or two and then disappeared
for want of attention.
I think if we want this project to succeed in the long run, we need
to be active and responsive. If the people in charge are unwilling or
unable to do that, then I'd just appreciate it if they could hand it
off to those who will dedicate the time.
-Bryan
On Jan 2, 2009, at 2:51 AM, Upayavira wrote:
> On Mon, 2008-12-29 at 16:40 +0100, Esteve Fernandez wrote:
>> On Sunday 28 December 2008 23:37:09 Ian Holsman wrote:
>>> In general in an apache project, the process for getting people
>>> commit
>>> access is for an existing 'PMC' member to recognize the merits of an
>>> individual and raise it on the private list, where it is
>>> discussed. It
>>> is discussed in private as some people may feel that the person
>>> is not
>>> ready yet, and saying that in public could discourage the
>>> individual.
>>> "nominating" the person on a public list is generally seen as
>>> poor form,
>>> as it might preempt the discussion, and lead to no response from the
>>> project.
>>
>> Sorry if my actions caused any trouble. I'm not a PMC member and
>> shouldn't
>> have nominated Bryan or Todd before asking about the process.
>
> Not a problem. Given these discussions happen in private (one of
> the few
> things that does/should in the ASF), there's no way for those
> outside to
> know how they happen.
>
>> However, I'm still arguing for more reviewers/committers in
>> general, I think
>> Thrift's development is being hindered by the lack of people who
>> can act on
>> tickets, which leads to external people become frustrated and to not
>> participate more actively. For example, THRIFT-129 (the most
>> popular ticket)
>> was reported on 2008-09-02 and last updated on 2008-10-23, but nobody
>> committed the attached patch.
>
> You may be right that Thrift needs new committers. However, that is an
> issue that I will leave to the Thrift committers to discuss/address.
>
> Upayavira
>
>
Re: more committers
Posted by Upayavira <uv...@odoko.co.uk>.
On Mon, 2008-12-29 at 16:40 +0100, Esteve Fernandez wrote:
> On Sunday 28 December 2008 23:37:09 Ian Holsman wrote:
> > In general in an apache project, the process for getting people commit
> > access is for an existing 'PMC' member to recognize the merits of an
> > individual and raise it on the private list, where it is discussed. It
> > is discussed in private as some people may feel that the person is not
> > ready yet, and saying that in public could discourage the individual.
> > "nominating" the person on a public list is generally seen as poor form,
> > as it might preempt the discussion, and lead to no response from the
> > project.
>
> Sorry if my actions caused any trouble. I'm not a PMC member and shouldn't
> have nominated Bryan or Todd before asking about the process.
Not a problem. Given these discussions happen in private (one of the few
things that does/should in the ASF), there's no way for those outside to
know how they happen.
> However, I'm still arguing for more reviewers/committers in general, I think
> Thrift's development is being hindered by the lack of people who can act on
> tickets, which leads to external people become frustrated and to not
> participate more actively. For example, THRIFT-129 (the most popular ticket)
> was reported on 2008-09-02 and last updated on 2008-10-23, but nobody
> committed the attached patch.
You may be right that Thrift needs new committers. However, that is an
issue that I will leave to the Thrift committers to discuss/address.
Upayavira
Re: more committers
Posted by David Reiss <dr...@facebook.com>.
Sorry, I meant "depends on", not "conflicts with". I was referring to
textual conflicts.
David Reiss wrote:
> For what it's worth, THRIFT-129 conflicts with THRIFT-25, and I'd like
> to get the first committed first if possible.
>
> Esteve Fernandez wrote:
>> On Sunday 28 December 2008 23:37:09 Ian Holsman wrote:
>>> In general in an apache project, the process for getting people commit
>>> access is for an existing 'PMC' member to recognize the merits of an
>>> individual and raise it on the private list, where it is discussed. It
>>> is discussed in private as some people may feel that the person is not
>>> ready yet, and saying that in public could discourage the individual.
>>> "nominating" the person on a public list is generally seen as poor form,
>>> as it might preempt the discussion, and lead to no response from the
>>> project.
>> Sorry if my actions caused any trouble. I'm not a PMC member and shouldn't
>> have nominated Bryan or Todd before asking about the process.
>>
>> However, I'm still arguing for more reviewers/committers in general, I think
>> Thrift's development is being hindered by the lack of people who can act on
>> tickets, which leads to external people become frustrated and to not
>> participate more actively. For example, THRIFT-129 (the most popular ticket)
>> was reported on 2008-09-02 and last updated on 2008-10-23, but nobody
>> committed the attached patch.
>>
>> Cheers.
Re: more committers
Posted by David Reiss <dr...@facebook.com>.
For what it's worth, THRIFT-129 conflicts with THRIFT-25, and I'd like
to get the first committed first if possible.
Esteve Fernandez wrote:
> On Sunday 28 December 2008 23:37:09 Ian Holsman wrote:
>> In general in an apache project, the process for getting people commit
>> access is for an existing 'PMC' member to recognize the merits of an
>> individual and raise it on the private list, where it is discussed. It
>> is discussed in private as some people may feel that the person is not
>> ready yet, and saying that in public could discourage the individual.
>> "nominating" the person on a public list is generally seen as poor form,
>> as it might preempt the discussion, and lead to no response from the
>> project.
>
> Sorry if my actions caused any trouble. I'm not a PMC member and shouldn't
> have nominated Bryan or Todd before asking about the process.
>
> However, I'm still arguing for more reviewers/committers in general, I think
> Thrift's development is being hindered by the lack of people who can act on
> tickets, which leads to external people become frustrated and to not
> participate more actively. For example, THRIFT-129 (the most popular ticket)
> was reported on 2008-09-02 and last updated on 2008-10-23, but nobody
> committed the attached patch.
>
> Cheers.
Re: more committers
Posted by Esteve Fernandez <es...@sindominio.net>.
On Sunday 28 December 2008 23:37:09 Ian Holsman wrote:
> In general in an apache project, the process for getting people commit
> access is for an existing 'PMC' member to recognize the merits of an
> individual and raise it on the private list, where it is discussed. It
> is discussed in private as some people may feel that the person is not
> ready yet, and saying that in public could discourage the individual.
> "nominating" the person on a public list is generally seen as poor form,
> as it might preempt the discussion, and lead to no response from the
> project.
Sorry if my actions caused any trouble. I'm not a PMC member and shouldn't
have nominated Bryan or Todd before asking about the process.
However, I'm still arguing for more reviewers/committers in general, I think
Thrift's development is being hindered by the lack of people who can act on
tickets, which leads to external people become frustrated and to not
participate more actively. For example, THRIFT-129 (the most popular ticket)
was reported on 2008-09-02 and last updated on 2008-10-23, but nobody
committed the attached patch.
Cheers.
Re: more committers
Posted by Ian Holsman <li...@holsman.net>.
Esteve Fernandez wrote:
> On Wednesday 24 December 2008 23:31:30 David Reiss wrote:
>
>>>>> Erlang - ?
>>>>>
>>>> Todd Lipcon (Toad on IRC) is really responsive to questions about the
>>>> Erlang mapping. Chris Piro or I can also look at these issues.
>>>>
>>> Then, I think it would make sense to give Todd committer access.
>>>
>> I don't think that is necessary. As long as a committer can check in
>> patches that are approved, I don't know of any reason that the primary
>> reviewer needs to be a committer.
>>
>
> The problem is, who approves and applies them? There's still a bunch of
> tickets with patches that have been voted up and that nobody committed.
>
> Someone (I think it was Doug) said that giving reviewers committer status
> motivates people to keep reviewing and to get more involved. I can't agree
> more with that statement and I think it's important to build a healthy
> community (one of the goals of the Apache brand).
>
>
I won't discuss the merits the of the individual, just the process. I'm
not subscribed to the private thrift list, so don't know what has been
discussed (if anything).
In general in an apache project, the process for getting people commit
access is for an existing 'PMC' member to recognize the merits of an
individual and raise it on the private list, where it is discussed. It
is discussed in private as some people may feel that the person is not
ready yet, and saying that in public could discourage the individual.
"nominating" the person on a public list is generally seen as poor form,
as it might preempt the discussion, and lead to no response from the
project.
and correct, it's not just developers who can get committer status in
apache projects. Anyone who actively contributes to the project is
eligible. The same goes for ASF membership. We have several ASF members
who have never written a line of code in their life, but their
contributions in other ways are seen as useful.
Regards
Ian
Re: more committers
Posted by Esteve Fernandez <es...@sindominio.net>.
On Wednesday 24 December 2008 23:31:30 David Reiss wrote:
> >>> Erlang - ?
> >>
> >> Todd Lipcon (Toad on IRC) is really responsive to questions about the
> >> Erlang mapping. Chris Piro or I can also look at these issues.
> >
> > Then, I think it would make sense to give Todd committer access.
>
> I don't think that is necessary. As long as a committer can check in
> patches that are approved, I don't know of any reason that the primary
> reviewer needs to be a committer.
The problem is, who approves and applies them? There's still a bunch of
tickets with patches that have been voted up and that nobody committed.
Someone (I think it was Doug) said that giving reviewers committer status
motivates people to keep reviewing and to get more involved. I can't agree
more with that statement and I think it's important to build a healthy
community (one of the goals of the Apache brand).
> I think this is also the de facto case. I'd be fine with making it
> slightly more official by breaking into first-tier and second-tier
> supported languages.
BTW, are there any plans for a new release?
Cheers.
Re: more committers
Posted by Todd Lipcon <tl...@gmail.com>.
On Wed, Dec 24, 2008 at 5:31 PM, David Reiss <dr...@facebook.com> wrote:
> >>> Erlang - ?
> >> Todd Lipcon (Toad on IRC) is really responsive to questions about the
> >> Erlang mapping. Chris Piro or I can also look at these issues.
> >
> > Then, I think it would make sense to give Todd committer access.
> I don't think that is necessary. As long as a committer can check in
> patches that are approved, I don't know of any reason that the primary
> reviewer needs to be a committer.
>
While I'm flattered to be nominated, I'll agree with David here --
especially given the git-friendly workflow here, I've never had any issues
getting patches in on the Erlang bindings or elsewhere.
>
> > However I
> > haven't seen much activity around Erlang in thrift-dev
> I think that is fine. As long as it is being used and people are happy
> with it, I don't think it is necessary for improvements to be actively
> made to it.
>
Exactly - we use thrift for 5 or 6 erlang applications at Amie Street and
just haven't had any occasion to modify the bindings at all since writing
them almost a year ago (after a couple of bug fixes and improvements thanks
to the team at FB). Kudos here are owed to Mark and the other original
designers of Thrift - it's simple enough that bindings just work and low
activity is a sign of stability rather than staleness or abandonware.
Thanks and hope everyone has enjoyed their holidays!
-Todd
Re: more committers
Posted by David Reiss <dr...@facebook.com>.
>>> Erlang - ?
>> Todd Lipcon (Toad on IRC) is really responsive to questions about the
>> Erlang mapping. Chris Piro or I can also look at these issues.
>
> Then, I think it would make sense to give Todd committer access.
I don't think that is necessary. As long as a committer can check in
patches that are approved, I don't know of any reason that the primary
reviewer needs to be a committer.
> However I
> haven't seen much activity around Erlang in thrift-dev
I think that is fine. As long as it is being used and people are happy
with it, I don't think it is necessary for improvements to be actively
made to it.
>>> PHP - ?
>> Mark and I can handle this. There has not been much interest in PHP
>> outside of Facebook, so it has been pretty low volume.
>
> Good. BTW, why does the Thrift compiler have special options for PHP
> (-php, -phpi, etc.) instead of using regular generator options (--gen xxx)?
I just haven't gotten around to converting it because there were a bunch
of options and I want to make sure that we don't break any build scripts.
>>> Haskell - ?
>>> Cocoa - ?
>>> Smalltalk - ?
>>> Ocaml - ?
>> There has been very little interest in these languages outside of the
>> original authors, so I think they should still be considered the point
>> people.
>
> I'm sure I'm going to piss off some people, but since those languages don't
> have a mantainer and their paucity of tests, I would lower their priority (or
> label them as unsupported) when the next version of Thrift is released.
I think this is also the de facto case. I'd be fine with making it slightly
more official by breaking into first-tier and second-tier supported languages.
--David
Re: more committers
Posted by Esteve Fernandez <es...@sindominio.net>.
On Tuesday 23 December 2008 20:31:19 David Reiss wrote:
> > Erlang - ?
>
> Todd Lipcon (Toad on IRC) is really responsive to questions about the
> Erlang mapping. Chris Piro or I can also look at these issues.
Then, I think it would make sense to give Todd committer access. However I
haven't seen much activity around Erlang in thrift-dev
> > C# - ?
>
> I would also like to have a clear go-to person (or people) for C# issues.
Yep, Ben (who knows C# and Python) appears to be MIA, THRIFT-209 is dead
simple but can't be fixed because there aren't committers who can review the
patch.
I've contributed a couple of patches for C#, but I think Michael Greene and
Anthony Abate surely know more C# than I.
> > PHP - ?
>
> Mark and I can handle this. There has not been much interest in PHP
> outside of Facebook, so it has been pretty low volume.
Good. BTW, why does the Thrift compiler have special options for PHP
(-php, -phpi, etc.) instead of using regular generator options (--gen xxx)?
> > Haskell - ?
> > Cocoa - ?
> > Smalltalk - ?
> > Ocaml - ?
>
> There has been very little interest in these languages outside of the
> original authors, so I think they should still be considered the point
> people.
I'm sure I'm going to piss off some people, but since those languages don't
have a mantainer and their paucity of tests, I would lower their priority (or
label them as unsupported) when the next version of Thrift is released.
Cheers.
Re: more committers
Posted by David Reiss <dr...@facebook.com>.
> so I would like to nominate Bryan as a new committer
+1. But just as a warning, Bryan, you're going to be hearing "please remove
trailing whitespace from this patch" a lot. ;-)
> Erlang - ?
Todd Lipcon (Toad on IRC) is really responsive to questions about the Erlang
mapping. Chris Piro or I can also look at these issues.
> C# - ?
I would also like to have a clear go-to person (or people) for C# issues.
> PHP - ?
Mark and I can handle this. There has not been much interest in PHP
outside of Facebook, so it has been pretty low volume.
> Haskell - ?
> Cocoa - ?
> Smalltalk - ?
> Ocaml - ?
There has been very little interest in these languages outside of the original
authors, so I think they should still be considered the point people.
--David