Re: The Apache Way and good developers don’t like to communicate

[Adams Musa]

The submitter must acquire the skills to transmit reliable write-ups to
draw the attention of consumers.
On Nov 2, 2018 10:43 PM, "Dmitriy Pavlov"  wrote:

> Hi Craig,
>
> Ignite dev list is now quite busy-list. GitHub comments are not forwarded
> (for the main repository, it is only forwarded for an additional GitBox
> repository, and probably we need to disable it as well or create
> notifications list). Forwarding comments may create additional pressure to
> contributors because they need to filter out emails from contributions
> which are not interesting.
>
> Some Ignite developers use the list to find a reviewer, and reviewers reply
> to dev list/to JIRA about results. The reviewer mentions where to find
> comments.
>
> Community Developers, thank you for all your replies.
>
> I've forwarded the reference to this thread and to 'large code drops'
> thread to Ignite dev list. I'm sure it will help the participants to
> understand why it is an important thing to share their ideas/plans/results
> using the list.
>
> Sincerely,
> Dmitriy Pavlov
>
>
> пт, 2 нояб. 2018 г. в 18:31, Craig Russell :
>
> > +1 to what Myrle says.
> >
> > I especially like the idea of engaging the submitter to discuss a pull
> > request. Even if it's to say "Thanks for this. It looks perfect as is".
> But
> > often, there will be a bit of back and forth.
> >
> > One potential problem is the work flow using the github tools, which can
> > have the effect of dialog "off-list" between the submitter and the other
> > devs. I don't know how to solve this issue except for one of the other
> devs
> > to initiate contact via the dev list in addition to the issue tools.
> >
> > Are the github tools for the project configured to copy the dev list on
> > every interaction on a pull request?
> >
> > Craig
> >
> > > On Nov 2, 2018, at 8:23 AM, Myrle Krantz  wrote:
> > >
> > > If anyone on the PMC is uncomfortable with making someone a committer
> > > because of their uncommunicativeness, I'd be inclined to give it a
> little
> > > more time.  A "no" now, doesn't mean a "no" forever.  Maybe ask that
> > person
> > > a few questions about their PR's on the list and see if they can be
> > > encouraged to engage?  Maybe let the PMC member who's advocating for a
> > > particular contributor be the one to take point on that effort?
> > >
> > > Greets,
> > > Myrle
> > >
> >
> > Craig L Russell
> > Secretary, Apache Software Foundation
> > c...@apache.org  http://db.apache.org/jdo <
> > http://db.apache.org/jdo>
> >
>

Re: The Apache Way and good developers don’t like to communicate

[Dmitriy Pavlov]

Hi Craig,

Ignite dev list is now quite busy-list. GitHub comments are not forwarded
(for the main repository, it is only forwarded for an additional GitBox
repository, and probably we need to disable it as well or create
notifications list). Forwarding comments may create additional pressure to
contributors because they need to filter out emails from contributions
which are not interesting.

Some Ignite developers use the list to find a reviewer, and reviewers reply
to dev list/to JIRA about results. The reviewer mentions where to find
comments.

Community Developers, thank you for all your replies.

I've forwarded the reference to this thread and to 'large code drops'
thread to Ignite dev list. I'm sure it will help the participants to
understand why it is an important thing to share their ideas/plans/results
using the list.

Sincerely,
Dmitriy Pavlov


пт, 2 нояб. 2018 г. в 18:31, Craig Russell :

> +1 to what Myrle says.
>
> I especially like the idea of engaging the submitter to discuss a pull
> request. Even if it's to say "Thanks for this. It looks perfect as is". But
> often, there will be a bit of back and forth.
>
> One potential problem is the work flow using the github tools, which can
> have the effect of dialog "off-list" between the submitter and the other
> devs. I don't know how to solve this issue except for one of the other devs
> to initiate contact via the dev list in addition to the issue tools.
>
> Are the github tools for the project configured to copy the dev list on
> every interaction on a pull request?
>
> Craig
>
> > On Nov 2, 2018, at 8:23 AM, Myrle Krantz  wrote:
> >
> > If anyone on the PMC is uncomfortable with making someone a committer
> > because of their uncommunicativeness, I'd be inclined to give it a little
> > more time.  A "no" now, doesn't mean a "no" forever.  Maybe ask that
> person
> > a few questions about their PR's on the list and see if they can be
> > encouraged to engage?  Maybe let the PMC member who's advocating for a
> > particular contributor be the one to take point on that effort?
> >
> > Greets,
> > Myrle
> >
>
> Craig L Russell
> Secretary, Apache Software Foundation
> c...@apache.org  http://db.apache.org/jdo <
> http://db.apache.org/jdo>
>

Re: Discussions in pull requests vs. dev list (was: The Apache Way and good developers...)

[Kenneth Knowles]

This comes up a lot on Beam too. We have a practice where if a PR is a big
change, or if discussion veers into into design details rather than just
code, we expect our committers to say "let's move this to the dev list". We
also try to encourage everyone to announce on dev@ not just for the design
discussion, but also to circle back and close the loop once it is merged.
Otherwise the archive is a bunch of loose ends and you can't tell what was
done.

It takes constant work to maintain this culture. We are not nearly 100%
successful; people are sometimes surprised at what has happened that they
didn't know about. Partly, it is just that there is so much going on (10-20
PRs a day) but I would eagerly follow progress and contribute (ideally,
availability pending) to any ideas for (opt-in) culture or automation work
on this topic.

Kenn

On Fri, Nov 2, 2018 at 11:02 AM Christopher  wrote:

> In Accumulo and Fluo, we route to notifications@ also. If these went
> to dev@, it would be too spammy, and I suspect even fewer people would
> participate on important dev@ threads than they do now. Letting it go
> to notifications@, people can subscribe to all activity there, if they
> wish... or they can go to GitHub and subscribe/unsubscribe to GitHub
> notifications for the entire repo ("Watch"), or per-thread, based on
> their individual desires. They can also unsubscribe from
> notifications@ to avoid duplicate notices if they are already
> subscribed to GitHub's notifications (which, to be honest, are far
> more useful, easier to read, and better formatted, than emails from
> ASF with the same content; you can also automatically avoid notices
> from your own activity... something you can't do on list). Using
> notifications@ instead of dev@ and users choosing (or not) to get
> notifications from GitHub is the best of both worlds and provides
> maximal user choice.
>
> There may be room for suggested improvements and best practices, but I
> think it would be a mistake if ASF were to try to impose a more spammy
> workflow to dev@ onto every PMC, as a requirement. I understand the
> utility and universality of mailing lists... but there comes a point
> where our reliance on them for all activity, becomes a deterrent to
> ASF participation, especially for younger contributors who expect more
> convenience than threaded walls of text in their email inbox.
> On Fri, Nov 2, 2018 at 1:46 PM Joan Touzet  wrote:
> >
> > We also use them successfully on CouchDB and I don't see the problem
> here.
> > We do route these notifications to notifications@, not dev@.
> > My email client properly threads multiple comments on the same PR.
> >
> > Another option is to use the GitHub "watch" functionality on a
> repository, which can provide better formatted emails than the ASF Infra
> solution currently does. They also seem to thread better for some.
> >
> > It's important to note that our process forces PRs for all changes to
> master or any release branch (R-T-C, not C-T-R), and that PRs cannot be
> merged unless our full test suite passes. This is automatically enforced by
> GitHub for us.
> >
> > Automating an email to dev@ with what PRs have opened/merged/closed
> with numerical counts sounds useful, but not mandatory. People that care
> can just use GitHub's "watch" function, then put filters in their mail
> client to highlight the things they care about (for instance, UI changes,
> or changes to a specific file or directory).
> >
> > -Joan
> > - Original Message -
> > From: "Christofer Dutz" 
> > To: dev@community.apache.org
> > Sent: Friday, November 2, 2018 1:30:47 PM
> > Subject: Re: Discussions in pull requests vs. dev list (was: The Apache
> Way and good developers...)
> >
> > Hi,
> >
> > Well in the PLC4X project we are using GitHub pull requests. All
> comments are forwarded to the list and this is fine. You are able to follow
> what's going on without much problems.
> >
> > However when it comes to the PR reviews, things tend to get it of hand.
> Every now and then I open my mail client to find 50 new emails and all of
> these are related to a single PR and each mail being a single comment. This
> is extremely annoying
> >
> > Chris
> >
> > Outlook for Android herunterladen
> >
> > From: Dave Fisher 
> > Sent: Friday, November 2, 2018 4:59:43 PM
> > To: dev@community.apache.org
> > Subject: Re: Discussions in pull requests vs. dev list (was: The Apache
> Way and good developers...)
> >
> > HI Bertrand,
> >
> > YES! This is the big issue with GitHub based projects/podlings.
> >
> > > On Nov 2, 2018, at 9:14 AM, Bertrand Delacretaz <
> bdelacre...@apache.org> wrote:
> > >
> > > Hi,
> > >
> > > I'd like to address this specifically as I think it's a common issue
> > > in many of our projects today.
> > >
> > > On Fri, Nov 2, 2018 at 4:31 PM Craig Russell 
> wrote:
> > >> ...One potential problem is the work flow using the github tools,
> which can have the effect
> > >> of dialog "off-list" be

Re: Discussions in pull requests vs. dev list (was: The Apache Way and good developers...)

[Christopher]

In Accumulo and Fluo, we route to notifications@ also. If these went
to dev@, it would be too spammy, and I suspect even fewer people would
participate on important dev@ threads than they do now. Letting it go
to notifications@, people can subscribe to all activity there, if they
wish... or they can go to GitHub and subscribe/unsubscribe to GitHub
notifications for the entire repo ("Watch"), or per-thread, based on
their individual desires. They can also unsubscribe from
notifications@ to avoid duplicate notices if they are already
subscribed to GitHub's notifications (which, to be honest, are far
more useful, easier to read, and better formatted, than emails from
ASF with the same content; you can also automatically avoid notices
from your own activity... something you can't do on list). Using
notifications@ instead of dev@ and users choosing (or not) to get
notifications from GitHub is the best of both worlds and provides
maximal user choice.

There may be room for suggested improvements and best practices, but I
think it would be a mistake if ASF were to try to impose a more spammy
workflow to dev@ onto every PMC, as a requirement. I understand the
utility and universality of mailing lists... but there comes a point
where our reliance on them for all activity, becomes a deterrent to
ASF participation, especially for younger contributors who expect more
convenience than threaded walls of text in their email inbox.
On Fri, Nov 2, 2018 at 1:46 PM Joan Touzet  wrote:
>
> We also use them successfully on CouchDB and I don't see the problem here.
> We do route these notifications to notifications@, not dev@.
> My email client properly threads multiple comments on the same PR.
>
> Another option is to use the GitHub "watch" functionality on a repository, 
> which can provide better formatted emails than the ASF Infra solution 
> currently does. They also seem to thread better for some.
>
> It's important to note that our process forces PRs for all changes to master 
> or any release branch (R-T-C, not C-T-R), and that PRs cannot be merged 
> unless our full test suite passes. This is automatically enforced by GitHub 
> for us.
>
> Automating an email to dev@ with what PRs have opened/merged/closed with 
> numerical counts sounds useful, but not mandatory. People that care can just 
> use GitHub's "watch" function, then put filters in their mail client to 
> highlight the things they care about (for instance, UI changes, or changes to 
> a specific file or directory).
>
> -Joan
> - Original Message -
> From: "Christofer Dutz" 
> To: dev@community.apache.org
> Sent: Friday, November 2, 2018 1:30:47 PM
> Subject: Re: Discussions in pull requests vs. dev list (was: The Apache Way 
> and good developers...)
>
> Hi,
>
> Well in the PLC4X project we are using GitHub pull requests. All comments are 
> forwarded to the list and this is fine. You are able to follow what's going 
> on without much problems.
>
> However when it comes to the PR reviews, things tend to get it of hand. Every 
> now and then I open my mail client to find 50 new emails and all of these are 
> related to a single PR and each mail being a single comment. This is 
> extremely annoying
>
> Chris
>
> Outlook for Android herunterladen
>
> From: Dave Fisher 
> Sent: Friday, November 2, 2018 4:59:43 PM
> To: dev@community.apache.org
> Subject: Re: Discussions in pull requests vs. dev list (was: The Apache Way 
> and good developers...)
>
> HI Bertrand,
>
> YES! This is the big issue with GitHub based projects/podlings.
>
> > On Nov 2, 2018, at 9:14 AM, Bertrand Delacretaz  
> > wrote:
> >
> > Hi,
> >
> > I'd like to address this specifically as I think it's a common issue
> > in many of our projects today.
> >
> > On Fri, Nov 2, 2018 at 4:31 PM Craig Russell  wrote:
> >> ...One potential problem is the work flow using the github tools, which 
> >> can have the effect
> >> of dialog "off-list" between the submitter and the other devs...
> >
> > I agree with that - in a project where most of the action happens on
> > GitHub (*) it can be hard or impossible to follow the action by just
> > subscribing to the project's dev list.
> >
> > I don't think there's anything wrong with discussions happening in
> > pull requests (PRs), that's the natural way of working in GitHub and
> > their tools are extremely convenient and efficient.
> >
> > However I think we need to stick to the principle that someone who's
> > just reading the dev list isn't missing out on anything. It is what
> > puts part-time contributors on an equal footing with people who work
> > full time on our projects which is IMO a key aspect of the Apache Way.
> >
> > So, how do we fix this? Blindly copying all GitHub PR comments to a
> > mailing list won't work IMO, as the result is not easy nor convenient
> > to follow. Copying is required for archival purposes (**) but not
> > practical for us humans IMO so I recommend copying to an "activity" or
> > "commits" l

Re: Discussions in pull requests vs. dev list (was: The Apache Way and good developers...)

[Joan Touzet]

We also use them successfully on CouchDB and I don't see the problem here.
We do route these notifications to notifications@, not dev@.
My email client properly threads multiple comments on the same PR.

Another option is to use the GitHub "watch" functionality on a repository, 
which can provide better formatted emails than the ASF Infra solution currently 
does. They also seem to thread better for some.

It's important to note that our process forces PRs for all changes to master or 
any release branch (R-T-C, not C-T-R), and that PRs cannot be merged unless our 
full test suite passes. This is automatically enforced by GitHub for us.

Automating an email to dev@ with what PRs have opened/merged/closed with 
numerical counts sounds useful, but not mandatory. People that care can just 
use GitHub's "watch" function, then put filters in their mail client to 
highlight the things they care about (for instance, UI changes, or changes to a 
specific file or directory).

-Joan
- Original Message -
From: "Christofer Dutz" 
To: dev@community.apache.org
Sent: Friday, November 2, 2018 1:30:47 PM
Subject: Re: Discussions in pull requests vs. dev list (was: The Apache Way and 
good developers...)

Hi,

Well in the PLC4X project we are using GitHub pull requests. All comments are 
forwarded to the list and this is fine. You are able to follow what's going on 
without much problems.

However when it comes to the PR reviews, things tend to get it of hand. Every 
now and then I open my mail client to find 50 new emails and all of these are 
related to a single PR and each mail being a single comment. This is extremely 
annoying

Chris

Outlook for Android herunterladen

From: Dave Fisher 
Sent: Friday, November 2, 2018 4:59:43 PM
To: dev@community.apache.org
Subject: Re: Discussions in pull requests vs. dev list (was: The Apache Way and 
good developers...)

HI Bertrand,

YES! This is the big issue with GitHub based projects/podlings.

> On Nov 2, 2018, at 9:14 AM, Bertrand Delacretaz  
> wrote:
>
> Hi,
>
> I'd like to address this specifically as I think it's a common issue
> in many of our projects today.
>
> On Fri, Nov 2, 2018 at 4:31 PM Craig Russell  wrote:
>> ...One potential problem is the work flow using the github tools, which can 
>> have the effect
>> of dialog "off-list" between the submitter and the other devs...
>
> I agree with that - in a project where most of the action happens on
> GitHub (*) it can be hard or impossible to follow the action by just
> subscribing to the project's dev list.
>
> I don't think there's anything wrong with discussions happening in
> pull requests (PRs), that's the natural way of working in GitHub and
> their tools are extremely convenient and efficient.
>
> However I think we need to stick to the principle that someone who's
> just reading the dev list isn't missing out on anything. It is what
> puts part-time contributors on an equal footing with people who work
> full time on our projects which is IMO a key aspect of the Apache Way.
>
> So, how do we fix this? Blindly copying all GitHub PR comments to a
> mailing list won't work IMO, as the result is not easy nor convenient
> to follow. Copying is required for archival purposes (**) but not
> practical for us humans IMO so I recommend copying to an "activity" or
> "commits" list, distinct from the dev list.
>
> One thing that I'd like to experiment in a podling (OpenWhisk) where I
> see this problem is to send weekly news to the dev list, listing which
> PRs are active, which modules are being actively worked on in
> multi-module project, etc. so that people know where to look in
> addition to the dev list.
>
> Ideally automated, but additional human comment certainly helps as well.

Regarding automation of this I think that the answer really lies in what can be 
automated out of GitHub / GitBox.

If you have a conversation with Infra about this, please let me know. I’d like 
to include some of other podlings in this experiment:
ECharts
Doris

Also, Dubbo is making a good switchover to using the dev list and is currently 
voting in a number of contributors as committers. It would be interesting to 
see what information they would find relevant.

Regards,
Dave

>
> I have mentioned that idea on the podling's dev list a while ago but
> didn't get any traction...I might try this myself and will report here
> if that happens. In the meantime, I'm interested in having other
> opinions on this topic.
>
> -Bertrand
>
> (*) which is fine given our https://gitbox.apache.org/ tooling
> (**) in case GitHub goes away - remember we are designing this
> Foundation for the next 50 years
>
> -
> To unsubscribe, e-mail: dev-unsubscr...@community.apache.org
> For additional commands, e-mail: dev-h...@community.apache.org
>


-
To unsubscribe, e-mail: dev-unsubscr...@communit

Re: Discussions in pull requests vs. dev list (was: The Apache Way and good developers...)

[Christofer Dutz]

Hi,

Well in the PLC4X project we are using GitHub pull requests. All comments are 
forwarded to the list and this is fine. You are able to follow what's going on 
without much problems.

However when it comes to the PR reviews, things tend to get it of hand. Every 
now and then I open my mail client to find 50 new emails and all of these are 
related to a single PR and each mail being a single comment. This is extremely 
annoying

Chris

Outlook for Android herunterladen

From: Dave Fisher 
Sent: Friday, November 2, 2018 4:59:43 PM
To: dev@community.apache.org
Subject: Re: Discussions in pull requests vs. dev list (was: The Apache Way and 
good developers...)

HI Bertrand,

YES! This is the big issue with GitHub based projects/podlings.

> On Nov 2, 2018, at 9:14 AM, Bertrand Delacretaz  
> wrote:
>
> Hi,
>
> I'd like to address this specifically as I think it's a common issue
> in many of our projects today.
>
> On Fri, Nov 2, 2018 at 4:31 PM Craig Russell  wrote:
>> ...One potential problem is the work flow using the github tools, which can 
>> have the effect
>> of dialog "off-list" between the submitter and the other devs...
>
> I agree with that - in a project where most of the action happens on
> GitHub (*) it can be hard or impossible to follow the action by just
> subscribing to the project's dev list.
>
> I don't think there's anything wrong with discussions happening in
> pull requests (PRs), that's the natural way of working in GitHub and
> their tools are extremely convenient and efficient.
>
> However I think we need to stick to the principle that someone who's
> just reading the dev list isn't missing out on anything. It is what
> puts part-time contributors on an equal footing with people who work
> full time on our projects which is IMO a key aspect of the Apache Way.
>
> So, how do we fix this? Blindly copying all GitHub PR comments to a
> mailing list won't work IMO, as the result is not easy nor convenient
> to follow. Copying is required for archival purposes (**) but not
> practical for us humans IMO so I recommend copying to an "activity" or
> "commits" list, distinct from the dev list.
>
> One thing that I'd like to experiment in a podling (OpenWhisk) where I
> see this problem is to send weekly news to the dev list, listing which
> PRs are active, which modules are being actively worked on in
> multi-module project, etc. so that people know where to look in
> addition to the dev list.
>
> Ideally automated, but additional human comment certainly helps as well.

Regarding automation of this I think that the answer really lies in what can be 
automated out of GitHub / GitBox.

If you have a conversation with Infra about this, please let me know. I’d like 
to include some of other podlings in this experiment:
ECharts
Doris

Also, Dubbo is making a good switchover to using the dev list and is currently 
voting in a number of contributors as committers. It would be interesting to 
see what information they would find relevant.

Regards,
Dave

>
> I have mentioned that idea on the podling's dev list a while ago but
> didn't get any traction...I might try this myself and will report here
> if that happens. In the meantime, I'm interested in having other
> opinions on this topic.
>
> -Bertrand
>
> (*) which is fine given our https://gitbox.apache.org/ tooling
> (**) in case GitHub goes away - remember we are designing this
> Foundation for the next 50 years
>
> -
> To unsubscribe, e-mail: dev-unsubscr...@community.apache.org
> For additional commands, e-mail: dev-h...@community.apache.org
>


-
To unsubscribe, e-mail: dev-unsubscr...@community.apache.org
For additional commands, e-mail: dev-h...@community.apache.org

Outlook for Android herunterladen

Re: Discussions in pull requests vs. dev list (was: The Apache Way and good developers...)

[Dave Fisher]

HI Bertrand,

YES! This is the big issue with GitHub based projects/podlings.

> On Nov 2, 2018, at 9:14 AM, Bertrand Delacretaz  
> wrote:
> 
> Hi,
> 
> I'd like to address this specifically as I think it's a common issue
> in many of our projects today.
> 
> On Fri, Nov 2, 2018 at 4:31 PM Craig Russell  wrote:
>> ...One potential problem is the work flow using the github tools, which can 
>> have the effect
>> of dialog "off-list" between the submitter and the other devs...
> 
> I agree with that - in a project where most of the action happens on
> GitHub (*) it can be hard or impossible to follow the action by just
> subscribing to the project's dev list.
> 
> I don't think there's anything wrong with discussions happening in
> pull requests (PRs), that's the natural way of working in GitHub and
> their tools are extremely convenient and efficient.
> 
> However I think we need to stick to the principle that someone who's
> just reading the dev list isn't missing out on anything. It is what
> puts part-time contributors on an equal footing with people who work
> full time on our projects which is IMO a key aspect of the Apache Way.
> 
> So, how do we fix this? Blindly copying all GitHub PR comments to a
> mailing list won't work IMO, as the result is not easy nor convenient
> to follow. Copying is required for archival purposes (**) but not
> practical for us humans IMO so I recommend copying to an "activity" or
> "commits" list, distinct from the dev list.
> 
> One thing that I'd like to experiment in a podling (OpenWhisk) where I
> see this problem is to send weekly news to the dev list, listing which
> PRs are active, which modules are being actively worked on in
> multi-module project, etc. so that people know where to look in
> addition to the dev list.
> 
> Ideally automated, but additional human comment certainly helps as well.

Regarding automation of this I think that the answer really lies in what can be 
automated out of GitHub / GitBox.

If you have a conversation with Infra about this, please let me know. I’d like 
to include some of other podlings in this experiment:
ECharts
Doris

Also, Dubbo is making a good switchover to using the dev list and is currently 
voting in a number of contributors as committers. It would be interesting to 
see what information they would find relevant.

Regards,
Dave

> 
> I have mentioned that idea on the podling's dev list a while ago but
> didn't get any traction...I might try this myself and will report here
> if that happens. In the meantime, I'm interested in having other
> opinions on this topic.
> 
> -Bertrand
> 
> (*) which is fine given our https://gitbox.apache.org/ tooling
> (**) in case GitHub goes away - remember we are designing this
> Foundation for the next 50 years
> 
> -
> To unsubscribe, e-mail: dev-unsubscr...@community.apache.org
> For additional commands, e-mail: dev-h...@community.apache.org
> 


-
To unsubscribe, e-mail: dev-unsubscr...@community.apache.org
For additional commands, e-mail: dev-h...@community.apache.org

Discussions in pull requests vs. dev list (was: The Apache Way and good developers...)

[Bertrand Delacretaz]

Hi,

I'd like to address this specifically as I think it's a common issue
in many of our projects today.

On Fri, Nov 2, 2018 at 4:31 PM Craig Russell  wrote:
> ...One potential problem is the work flow using the github tools, which can 
> have the effect
> of dialog "off-list" between the submitter and the other devs...

I agree with that - in a project where most of the action happens on
GitHub (*) it can be hard or impossible to follow the action by just
subscribing to the project's dev list.

I don't think there's anything wrong with discussions happening in
pull requests (PRs), that's the natural way of working in GitHub and
their tools are extremely convenient and efficient.

However I think we need to stick to the principle that someone who's
just reading the dev list isn't missing out on anything. It is what
puts part-time contributors on an equal footing with people who work
full time on our projects which is IMO a key aspect of the Apache Way.

So, how do we fix this? Blindly copying all GitHub PR comments to a
mailing list won't work IMO, as the result is not easy nor convenient
to follow. Copying is required for archival purposes (**) but not
practical for us humans IMO so I recommend copying to an "activity" or
"commits" list, distinct from the dev list.

One thing that I'd like to experiment in a podling (OpenWhisk) where I
see this problem is to send weekly news to the dev list, listing which
PRs are active, which modules are being actively worked on in
multi-module project, etc. so that people know where to look in
addition to the dev list.

Ideally automated, but additional human comment certainly helps as well.

I have mentioned that idea on the podling's dev list a while ago but
didn't get any traction...I might try this myself and will report here
if that happens. In the meantime, I'm interested in having other
opinions on this topic.

-Bertrand

(*) which is fine given our https://gitbox.apache.org/ tooling
(**) in case GitHub goes away - remember we are designing this
Foundation for the next 50 years

-
To unsubscribe, e-mail: dev-unsubscr...@community.apache.org
For additional commands, e-mail: dev-h...@community.apache.org

Re: The Apache Way and good developers don’t like to communicate

[Craig Russell]

+1 to what Myrle says.

I especially like the idea of engaging the submitter to discuss a pull request. 
Even if it's to say "Thanks for this. It looks perfect as is". But often, there 
will be a bit of back and forth.

One potential problem is the work flow using the github tools, which can have 
the effect of dialog "off-list" between the submitter and the other devs. I 
don't know how to solve this issue except for one of the other devs to initiate 
contact via the dev list in addition to the issue tools. 

Are the github tools for the project configured to copy the dev list on every 
interaction on a pull request?

Craig

> On Nov 2, 2018, at 8:23 AM, Myrle Krantz  wrote:
> 
> If anyone on the PMC is uncomfortable with making someone a committer
> because of their uncommunicativeness, I'd be inclined to give it a little
> more time.  A "no" now, doesn't mean a "no" forever.  Maybe ask that person
> a few questions about their PR's on the list and see if they can be
> encouraged to engage?  Maybe let the PMC member who's advocating for a
> particular contributor be the one to take point on that effort?
> 
> Greets,
> Myrle
> 

Craig L Russell
Secretary, Apache Software Foundation
c...@apache.org  http://db.apache.org/jdo 

Re: The Apache Way and good developers don’t like to communicate

[Myrle Krantz]

+1 on everything Chris says here.

If anyone on the PMC is uncomfortable with making someone a committer
because of their uncommunicativeness, I'd be inclined to give it a little
more time.  A "no" now, doesn't mean a "no" forever.  Maybe ask that person
a few questions about their PR's on the list and see if they can be
encouraged to engage?  Maybe let the PMC member who's advocating for a
particular contributor be the one to take point on that effort?

Greets,
Myrle


On Fri, Nov 2, 2018 at 1:43 PM Christofer Dutz 
wrote:

> Hi Dimitriy,
>
> well I guess I simply forgot about the option the PMC != Committers ;-)
>
> So I agree, in some cases I could probably support someone not able to
> communicate because of such reasons a committer, but definitely not to
> become a PMC.
> But if it's just laziness or unwillingness (for whatever reasons) I would
> not support them even becoming a committer, as I think these contributions
> would just be
> uncommented code-drops, which I don't think have a positive impact on the
> community (See other thread here)
>
> Chris
>
>
> Am 02.11.18, 13:19 schrieb "Dmitriy Pavlov" :
>
> Dear ASF Fellows,
>
> I strongly appreciate all your replies. I believe there is no just one
> correct answer. Which is why I need opinions of folks from other
> projects.
>
> Myrle, Apache Ignite has 26 PMC members and 38 committers, so PMCs is a
> subset of committers set.
>
> About collaboration: I guess these contributors were communicating with
> someone of community within a company they work for, in person/or,
> probably, Skype. So maybe the code was good. And they became effective
> code
> contributors without valuable communication on lists and without
> contributing to the community.
>
> Chris, About the subject: it is translated version of the argument I
> hear
> about contributors, who are not often present on dev/user list.
>
> So if a person does not like to communicate, let's say, afraid of
> society/publicity, can he or she be a committer in Apache?
>
> Sincerely,
> Dmitriy Pavlov
>
> пт, 2 нояб. 2018 г. в 14:31, Myrle Krantz :
>
> > Hi Dmitriy,
> >
> > Is Ignite a PMC = committer community or a PMC ⊂ committer community?
> >
> > You may have different requirements for communication level
> depending on
> > which of these your community is.  But I don't believe it is
> possible to
> > write very good code without being willing to talk with others about
> it.
> >
> > Still, different communities have different "bars".  And I've come
> to be
> > convinced by Greg Stein, that a lower committer bar is better for
> > attracting contributions.  People might feel more comfortable
> communicating
> > once they've been given the committer bit?
> >
> > Regards,
> > Myrle
> >
> > On Fri, Nov 2, 2018 at 9:53 AM Dmitriy Pavlov  >
> > wrote:
> >
> > > Dear ASF Fellows,
> > >
> > > I am PMC member of Apache Ignite, but I joined PMC relatively
> recently. I
> > > need help from you again in regarding the Apache Way.
> > >
> > > Question is related to comittership for community members,
> > >
> > > - who are not visible on dev/user list, have a couple of threads
> they
> > > participated
> > >
> > > - but contributed a significant feature or many fixes.
> > >
> > > Usually, such contributors work for a commercial company with
> sufficient
> > > product expertise, so they probably collaborate with experts, but
> outside
> > > space of Apache.
> > >
> > >
> > > Several guides and policies
> > >
> > > https://www.apache.org/foundation/how-it-works.html#meritocracy
> > >
> > > http://community.apache.org/newcommitter.html
> > >
> > > and others say that PMC member needs to evaluate communication and
> > > cooperative work with peers, ability to be a mentor, behavior in
> > > disagreement.
> > >
> > >
> > > Communication is required by Apache Ignite guide
> > >
> > >
> >
> https://cwiki.apache.org/confluence/display/IGNITE/Committership+Bar+Guidance
> > >
> > > Simultaneously
> > >
> https://community.apache.org/contributors/#contributing-a-project-copdoc
> > >
> > > contains a mention someone who contributed sufficiently to ‘ANY’
> area may
> > > become a committer. So why can't we count code only contribution
> without
> > > contribution to community/project?
> > >
> > > There are several cases when I may disagree with other PMC members.
> > >
> > > I insist candidate should communicate in ASF space because A)
> > > community-first and motto: B) “If it didn’t happen on the mailing
> list it
> > > didn’t happen.” For such cases then contributors collaborate
> outside
> > Apache
> > > space we can still accept a contribution, still appreciate
> contributor’s
> > > effort

Re: The Apache Way and good developers don’t like to communicate

[Ishamisham Ishamisham]

Pada 2 Nov 2018 20:49, "Rafael Weingärtner" 
menulis:

Hey Christofer, you just said what I was going to comment.

How is it possible for someone to contribute code (opening a PR, and so on)
without engaging with the community? I mean, don’t you have a reviewing
process to accept/merge code?

We had this situation in the past in CloudStack, and you still can find the
open PRs of these people around. That happens because questions/requests
were made during the review process, and they (these questions/requests)
were never answered.

On Fri, Nov 2, 2018 at 9:43 AM Christofer Dutz 
wrote:

> Hi Dimitriy,
>
> well I guess I simply forgot about the option the PMC != Committers ;-)
>
> So I agree, in some cases I could probably support someone not able to
> communicate because of such reasons a committer, but definitely not to
> become a PMC.
> But if it's just laziness or unwillingness (for whatever reasons) I would
> not support them even becoming a committer, as I think these contributions
> would just be
> uncommented code-drops, which I don't think have a positive impact on the
> community (See other thread here)
>
> Chris
>
>
> Am 02.11.18, 13:19 schrieb "Dmitriy Pavlov" :
>
> Dear ASF Fellows,
>
> I strongly appreciate all your replies. I believe there is no just one
> correct answer. Which is why I need opinions of folks from other
> projects.
>
> Myrle, Apache Ignite has 26 PMC members and 38 committers, so PMCs is
a
> subset of committers set.
>
> About collaboration: I guess these contributors were communicating
with
> someone of community within a company they work for, in person/or,
> probably, Skype. So maybe the code was good. And they became effective
> code
> contributors without valuable communication on lists and without
> contributing to the community.
>
> Chris, About the subject: it is translated version of the argument I
> hear
> about contributors, who are not often present on dev/user list.
>
> So if a person does not like to communicate, let's say, afraid of
> society/publicity, can he or she be a committer in Apache?
>
> Sincerely,
> Dmitriy Pavlov
>
> пт, 2 нояб. 2018 г. в 14:31, Myrle Krantz :
>
> > Hi Dmitriy,
> >
> > Is Ignite a PMC = committer community or a PMC ⊂ committer
community?
> >
> > You may have different requirements for communication level
> depending on
> > which of these your community is.  But I don't believe it is
> possible to
> > write very good code without being willing to talk with others about
> it.
> >
> > Still, different communities have different "bars".  And I've come
> to be
> > convinced by Greg Stein, that a lower committer bar is better for
> > attracting contributions.  People might feel more comfortable
> communicating
> > once they've been given the committer bit?
> >
> > Regards,
> > Myrle
> >
> > On Fri, Nov 2, 2018 at 9:53 AM Dmitriy Pavlov  >
> > wrote:
> >
> > > Dear ASF Fellows,
> > >
> > > I am PMC member of Apache Ignite, but I joined PMC relatively
> recently. I
> > > need help from you again in regarding the Apache Way.
> > >
> > > Question is related to comittership for community members,
> > >
> > > - who are not visible on dev/user list, have a couple of threads
> they
> > > participated
> > >
> > > - but contributed a significant feature or many fixes.
> > >
> > > Usually, such contributors work for a commercial company with
> sufficient
> > > product expertise, so they probably collaborate with experts, but
> outside
> > > space of Apache.
> > >
> > >
> > > Several guides and policies
> > >
> > > https://www.apache.org/foundation/how-it-works.html#meritocracy
> > >
> > > http://community.apache.org/newcommitter.html
> > >
> > > and others say that PMC member needs to evaluate communication and
> > > cooperative work with peers, ability to be a mentor, behavior in
> > > disagreement.
> > >
> > >
> > > Communication is required by Apache Ignite guide
> > >
> > >
> >
>
https://cwiki.apache.org/confluence/display/IGNITE/Committership+Bar+Guidance
> > >
> > > Simultaneously
> > >
> https://community.apache.org/contributors/#contributing-a-project-copdoc
> > >
> > > contains a mention someone who contributed sufficiently to ‘ANY’
> area may
> > > become a committer. So why can't we count code only contribution
> without
> > > contribution to community/project?
> > >
> > > There are several cases when I may disagree with other PMC
members.
> > >
> > > I insist candidate should communicate in ASF space because A)
> > > community-first and motto: B) “If it didn’t happen on the mailing
> list it
> > > didn’t happen.” For such cases then contributors collaborate
> outside
> > Apache
> > > space we can still accept a 

Re: The Apache Way and good developers don’t like to communicate

[Jim Jagielski]

I agree that it's hard to be a critical and crucial part of a community without 
being *IN* the community.

More than anything, the PMC is the entity within the ASF that is the community 
representative for a project. They "run" the project for the community with the 
understanding that they understand the community best. If someone is not IN the 
community, nor active in the community, how can they serve that role?

> On Nov 2, 2018, at 5:56 AM, Bertrand Delacretaz  
> wrote:
> 
> On Fri, Nov 2, 2018 at 10:22 AM Christofer Dutz
>  wrote:
>> ... So if someone is an awesome coder, but doesn't communicate, I would not 
>> invite him or vote for inviting...
> 
> I agree with that when it comes to inviting people on a PMC.
> 
> I think it's fine to accept code contributions from people who are not
> visible on the project's lists, but I would not invite them on a PMC.
> 
> Having *committers* who are not visible on the project's lists might
> be ok, but the PMC is a *management* committee, and that requires open
> communications, not just code.
> 
> -Bertrand
> 
> -
> To unsubscribe, e-mail: dev-unsubscr...@community.apache.org
> For additional commands, e-mail: dev-h...@community.apache.org
> 


-
To unsubscribe, e-mail: dev-unsubscr...@community.apache.org
For additional commands, e-mail: dev-h...@community.apache.org

Re: The Apache Way and good developers don’t like to communicate

[Ishamisham Ishamisham]

Pada 2 Nov 2018 20:49, "Bertrand Delacretaz" 
menulis:

On Fri, Nov 2, 2018 at 1:43 PM Christofer Dutz
 wrote:
> ...I think these contributions would just be uncommented code-drops...

Code speaks louder than words sometimes ;-)

But in general I agree with you.


-Bertrand

-
To unsubscribe, e-mail: dev-unsubscr...@community.apache.org
For additional commands, e-mail: dev-h...@community.apache.org

Re: The Apache Way and good developers don’t like to communicate

[Bertrand Delacretaz]

On Fri, Nov 2, 2018 at 1:43 PM Christofer Dutz
 wrote:
> ...I think these contributions would just be uncommented code-drops...

Code speaks louder than words sometimes ;-)

But in general I agree with you.

-Bertrand

-
To unsubscribe, e-mail: dev-unsubscr...@community.apache.org
For additional commands, e-mail: dev-h...@community.apache.org

Re: The Apache Way and good developers don’t like to communicate

[Rafael Weingärtner]

Hey Christofer, you just said what I was going to comment.

How is it possible for someone to contribute code (opening a PR, and so on)
without engaging with the community? I mean, don’t you have a reviewing
process to accept/merge code?

We had this situation in the past in CloudStack, and you still can find the
open PRs of these people around. That happens because questions/requests
were made during the review process, and they (these questions/requests)
were never answered.

On Fri, Nov 2, 2018 at 9:43 AM Christofer Dutz 
wrote:

> Hi Dimitriy,
>
> well I guess I simply forgot about the option the PMC != Committers ;-)
>
> So I agree, in some cases I could probably support someone not able to
> communicate because of such reasons a committer, but definitely not to
> become a PMC.
> But if it's just laziness or unwillingness (for whatever reasons) I would
> not support them even becoming a committer, as I think these contributions
> would just be
> uncommented code-drops, which I don't think have a positive impact on the
> community (See other thread here)
>
> Chris
>
>
> Am 02.11.18, 13:19 schrieb "Dmitriy Pavlov" :
>
> Dear ASF Fellows,
>
> I strongly appreciate all your replies. I believe there is no just one
> correct answer. Which is why I need opinions of folks from other
> projects.
>
> Myrle, Apache Ignite has 26 PMC members and 38 committers, so PMCs is a
> subset of committers set.
>
> About collaboration: I guess these contributors were communicating with
> someone of community within a company they work for, in person/or,
> probably, Skype. So maybe the code was good. And they became effective
> code
> contributors without valuable communication on lists and without
> contributing to the community.
>
> Chris, About the subject: it is translated version of the argument I
> hear
> about contributors, who are not often present on dev/user list.
>
> So if a person does not like to communicate, let's say, afraid of
> society/publicity, can he or she be a committer in Apache?
>
> Sincerely,
> Dmitriy Pavlov
>
> пт, 2 нояб. 2018 г. в 14:31, Myrle Krantz :
>
> > Hi Dmitriy,
> >
> > Is Ignite a PMC = committer community or a PMC ⊂ committer community?
> >
> > You may have different requirements for communication level
> depending on
> > which of these your community is.  But I don't believe it is
> possible to
> > write very good code without being willing to talk with others about
> it.
> >
> > Still, different communities have different "bars".  And I've come
> to be
> > convinced by Greg Stein, that a lower committer bar is better for
> > attracting contributions.  People might feel more comfortable
> communicating
> > once they've been given the committer bit?
> >
> > Regards,
> > Myrle
> >
> > On Fri, Nov 2, 2018 at 9:53 AM Dmitriy Pavlov  >
> > wrote:
> >
> > > Dear ASF Fellows,
> > >
> > > I am PMC member of Apache Ignite, but I joined PMC relatively
> recently. I
> > > need help from you again in regarding the Apache Way.
> > >
> > > Question is related to comittership for community members,
> > >
> > > - who are not visible on dev/user list, have a couple of threads
> they
> > > participated
> > >
> > > - but contributed a significant feature or many fixes.
> > >
> > > Usually, such contributors work for a commercial company with
> sufficient
> > > product expertise, so they probably collaborate with experts, but
> outside
> > > space of Apache.
> > >
> > >
> > > Several guides and policies
> > >
> > > https://www.apache.org/foundation/how-it-works.html#meritocracy
> > >
> > > http://community.apache.org/newcommitter.html
> > >
> > > and others say that PMC member needs to evaluate communication and
> > > cooperative work with peers, ability to be a mentor, behavior in
> > > disagreement.
> > >
> > >
> > > Communication is required by Apache Ignite guide
> > >
> > >
> >
> https://cwiki.apache.org/confluence/display/IGNITE/Committership+Bar+Guidance
> > >
> > > Simultaneously
> > >
> https://community.apache.org/contributors/#contributing-a-project-copdoc
> > >
> > > contains a mention someone who contributed sufficiently to ‘ANY’
> area may
> > > become a committer. So why can't we count code only contribution
> without
> > > contribution to community/project?
> > >
> > > There are several cases when I may disagree with other PMC members.
> > >
> > > I insist candidate should communicate in ASF space because A)
> > > community-first and motto: B) “If it didn’t happen on the mailing
> list it
> > > didn’t happen.” For such cases then contributors collaborate
> outside
> > Apache
> > > space we can still accept a contribution, still appreciate
> contributor’s
> > 

Re: The Apache Way and good developers don’t like to communicate

[Christofer Dutz]

Hi Dimitriy,

well I guess I simply forgot about the option the PMC != Committers ;-)

So I agree, in some cases I could probably support someone not able to 
communicate because of such reasons a committer, but definitely not to become a 
PMC.
But if it's just laziness or unwillingness (for whatever reasons) I would not 
support them even becoming a committer, as I think these contributions would 
just be 
uncommented code-drops, which I don't think have a positive impact on the 
community (See other thread here)

Chris


Am 02.11.18, 13:19 schrieb "Dmitriy Pavlov" :

Dear ASF Fellows,

I strongly appreciate all your replies. I believe there is no just one
correct answer. Which is why I need opinions of folks from other projects.

Myrle, Apache Ignite has 26 PMC members and 38 committers, so PMCs is a
subset of committers set.

About collaboration: I guess these contributors were communicating with
someone of community within a company they work for, in person/or,
probably, Skype. So maybe the code was good. And they became effective code
contributors without valuable communication on lists and without
contributing to the community.

Chris, About the subject: it is translated version of the argument I hear
about contributors, who are not often present on dev/user list.

So if a person does not like to communicate, let's say, afraid of
society/publicity, can he or she be a committer in Apache?

Sincerely,
Dmitriy Pavlov

пт, 2 нояб. 2018 г. в 14:31, Myrle Krantz :

> Hi Dmitriy,
>
> Is Ignite a PMC = committer community or a PMC ⊂ committer community?
>
> You may have different requirements for communication level depending on
> which of these your community is.  But I don't believe it is possible to
> write very good code without being willing to talk with others about it.
>
> Still, different communities have different "bars".  And I've come to be
> convinced by Greg Stein, that a lower committer bar is better for
> attracting contributions.  People might feel more comfortable 
communicating
> once they've been given the committer bit?
>
> Regards,
> Myrle
>
> On Fri, Nov 2, 2018 at 9:53 AM Dmitriy Pavlov 
> wrote:
>
> > Dear ASF Fellows,
> >
> > I am PMC member of Apache Ignite, but I joined PMC relatively recently. 
I
> > need help from you again in regarding the Apache Way.
> >
> > Question is related to comittership for community members,
> >
> > - who are not visible on dev/user list, have a couple of threads they
> > participated
> >
> > - but contributed a significant feature or many fixes.
> >
> > Usually, such contributors work for a commercial company with sufficient
> > product expertise, so they probably collaborate with experts, but 
outside
> > space of Apache.
> >
> >
> > Several guides and policies
> >
> > https://www.apache.org/foundation/how-it-works.html#meritocracy
> >
> > http://community.apache.org/newcommitter.html
> >
> > and others say that PMC member needs to evaluate communication and
> > cooperative work with peers, ability to be a mentor, behavior in
> > disagreement.
> >
> >
> > Communication is required by Apache Ignite guide
> >
> >
> 
https://cwiki.apache.org/confluence/display/IGNITE/Committership+Bar+Guidance
> >
> > Simultaneously
> > https://community.apache.org/contributors/#contributing-a-project-copdoc
> >
> > contains a mention someone who contributed sufficiently to ‘ANY’ area 
may
> > become a committer. So why can't we count code only contribution without
> > contribution to community/project?
> >
> > There are several cases when I may disagree with other PMC members.
> >
> > I insist candidate should communicate in ASF space because A)
> > community-first and motto: B) “If it didn’t happen on the mailing list 
it
> > didn’t happen.” For such cases then contributors collaborate outside
> Apache
> > space we can still accept a contribution, still appreciate contributor’s
> > effort and say thank you; but not promote as a committer. But I may
> > over-estimate the role of collaboration in the ASF. I may be too strict
> in
> > understanding ASF principles.
> >
> > But PMCs who suggest such comittership candidates may counter-argument
> >
> > - those cool developers don't like to communicate (they may be a little
> bit
> > uncomfortable with public communications/tries to avoid spam/any other
> > reasons they have).
> >
> > - If he or she will communicate often, then he or she will never have
> time
> > to write a code.
> >
> > So what do you think? Is it required to communicate with the rest of the
> > community publicly more than a co

Re: The Apache Way and good developers don’t like to communicate

[Dmitriy Pavlov]

Dear ASF Fellows,

I strongly appreciate all your replies. I believe there is no just one
correct answer. Which is why I need opinions of folks from other projects.

Myrle, Apache Ignite has 26 PMC members and 38 committers, so PMCs is a
subset of committers set.

About collaboration: I guess these contributors were communicating with
someone of community within a company they work for, in person/or,
probably, Skype. So maybe the code was good. And they became effective code
contributors without valuable communication on lists and without
contributing to the community.

Chris, About the subject: it is translated version of the argument I hear
about contributors, who are not often present on dev/user list.

So if a person does not like to communicate, let's say, afraid of
society/publicity, can he or she be a committer in Apache?

Sincerely,
Dmitriy Pavlov

пт, 2 нояб. 2018 г. в 14:31, Myrle Krantz :

> Hi Dmitriy,
>
> Is Ignite a PMC = committer community or a PMC ⊂ committer community?
>
> You may have different requirements for communication level depending on
> which of these your community is.  But I don't believe it is possible to
> write very good code without being willing to talk with others about it.
>
> Still, different communities have different "bars".  And I've come to be
> convinced by Greg Stein, that a lower committer bar is better for
> attracting contributions.  People might feel more comfortable communicating
> once they've been given the committer bit?
>
> Regards,
> Myrle
>
> On Fri, Nov 2, 2018 at 9:53 AM Dmitriy Pavlov 
> wrote:
>
> > Dear ASF Fellows,
> >
> > I am PMC member of Apache Ignite, but I joined PMC relatively recently. I
> > need help from you again in regarding the Apache Way.
> >
> > Question is related to comittership for community members,
> >
> > - who are not visible on dev/user list, have a couple of threads they
> > participated
> >
> > - but contributed a significant feature or many fixes.
> >
> > Usually, such contributors work for a commercial company with sufficient
> > product expertise, so they probably collaborate with experts, but outside
> > space of Apache.
> >
> >
> > Several guides and policies
> >
> > https://www.apache.org/foundation/how-it-works.html#meritocracy
> >
> > http://community.apache.org/newcommitter.html
> >
> > and others say that PMC member needs to evaluate communication and
> > cooperative work with peers, ability to be a mentor, behavior in
> > disagreement.
> >
> >
> > Communication is required by Apache Ignite guide
> >
> >
> https://cwiki.apache.org/confluence/display/IGNITE/Committership+Bar+Guidance
> >
> > Simultaneously
> > https://community.apache.org/contributors/#contributing-a-project-copdoc
> >
> > contains a mention someone who contributed sufficiently to ‘ANY’ area may
> > become a committer. So why can't we count code only contribution without
> > contribution to community/project?
> >
> > There are several cases when I may disagree with other PMC members.
> >
> > I insist candidate should communicate in ASF space because A)
> > community-first and motto: B) “If it didn’t happen on the mailing list it
> > didn’t happen.” For such cases then contributors collaborate outside
> Apache
> > space we can still accept a contribution, still appreciate contributor’s
> > effort and say thank you; but not promote as a committer. But I may
> > over-estimate the role of collaboration in the ASF. I may be too strict
> in
> > understanding ASF principles.
> >
> > But PMCs who suggest such comittership candidates may counter-argument
> >
> > - those cool developers don't like to communicate (they may be a little
> bit
> > uncomfortable with public communications/tries to avoid spam/any other
> > reasons they have).
> >
> > - If he or she will communicate often, then he or she will never have
> time
> > to write a code.
> >
> > So what do you think? Is it required to communicate with the rest of the
> > community publicly more than a couple of times to become a committer?
> >
> > Sincerely,
> >
> > Dmitriy Pavlov
> >
>

Re: The Apache Way and good developers don’t like to communicate

[Myrle Krantz]

Hi Dmitriy,

Is Ignite a PMC = committer community or a PMC ⊂ committer community?

You may have different requirements for communication level depending on
which of these your community is.  But I don't believe it is possible to
write very good code without being willing to talk with others about it.

Still, different communities have different "bars".  And I've come to be
convinced by Greg Stein, that a lower committer bar is better for
attracting contributions.  People might feel more comfortable communicating
once they've been given the committer bit?

Regards,
Myrle

On Fri, Nov 2, 2018 at 9:53 AM Dmitriy Pavlov  wrote:

> Dear ASF Fellows,
>
> I am PMC member of Apache Ignite, but I joined PMC relatively recently. I
> need help from you again in regarding the Apache Way.
>
> Question is related to comittership for community members,
>
> - who are not visible on dev/user list, have a couple of threads they
> participated
>
> - but contributed a significant feature or many fixes.
>
> Usually, such contributors work for a commercial company with sufficient
> product expertise, so they probably collaborate with experts, but outside
> space of Apache.
>
>
> Several guides and policies
>
> https://www.apache.org/foundation/how-it-works.html#meritocracy
>
> http://community.apache.org/newcommitter.html
>
> and others say that PMC member needs to evaluate communication and
> cooperative work with peers, ability to be a mentor, behavior in
> disagreement.
>
>
> Communication is required by Apache Ignite guide
>
> https://cwiki.apache.org/confluence/display/IGNITE/Committership+Bar+Guidance
>
> Simultaneously
> https://community.apache.org/contributors/#contributing-a-project-copdoc
>
> contains a mention someone who contributed sufficiently to ‘ANY’ area may
> become a committer. So why can't we count code only contribution without
> contribution to community/project?
>
> There are several cases when I may disagree with other PMC members.
>
> I insist candidate should communicate in ASF space because A)
> community-first and motto: B) “If it didn’t happen on the mailing list it
> didn’t happen.” For such cases then contributors collaborate outside Apache
> space we can still accept a contribution, still appreciate contributor’s
> effort and say thank you; but not promote as a committer. But I may
> over-estimate the role of collaboration in the ASF. I may be too strict in
> understanding ASF principles.
>
> But PMCs who suggest such comittership candidates may counter-argument
>
> - those cool developers don't like to communicate (they may be a little bit
> uncomfortable with public communications/tries to avoid spam/any other
> reasons they have).
>
> - If he or she will communicate often, then he or she will never have time
> to write a code.
>
> So what do you think? Is it required to communicate with the rest of the
> community publicly more than a couple of times to become a committer?
>
> Sincerely,
>
> Dmitriy Pavlov
>

Re: The Apache Way and good developers don’t like to communicate

[Bertrand Delacretaz]

On Fri, Nov 2, 2018 at 10:22 AM Christofer Dutz
 wrote:
>... So if someone is an awesome coder, but doesn't communicate, I would not 
>invite him or vote for inviting...

I agree with that when it comes to inviting people on a PMC.

I think it's fine to accept code contributions from people who are not
visible on the project's lists, but I would not invite them on a PMC.

Having *committers* who are not visible on the project's lists might
be ok, but the PMC is a *management* committee, and that requires open
communications, not just code.

-Bertrand

-
To unsubscribe, e-mail: dev-unsubscr...@community.apache.org
For additional commands, e-mail: dev-h...@community.apache.org

Re: The Apache Way and good developers don’t like to communicate

[Christofer Dutz]

Have to admit it was the title which triggered me ... but I guess it was 
intentionally ;-)

Well for me it is clear:

Communication == Community

Code == Code

And: Community over Code.

So if someone is an awesome coder, but doesn't communicate, I would not invite 
him or vote for inviting.
In the projects I'm currently involved in, I usually even contact the potential 
committer and tell him/her that
their contributions are highly welcome, but in order to become a part of the 
community, they have to become part of it
and the only option to do this, is to communicate with the rest.

Regarding communication ... I would say communication takes very little time, 
if you stay focused and get to the point. 
Writing drama-emails (1-5 Screen length long emails) does, but that's not 
needed and I would even say: Not wanted 
(At least as soon as email discussions explode with screenheight over 
screenhight long emails, that's when I start filtering.

Chris



Am 02.11.18, 09:53 schrieb "Dmitriy Pavlov" :

Dear ASF Fellows,

I am PMC member of Apache Ignite, but I joined PMC relatively recently. I
need help from you again in regarding the Apache Way.

Question is related to comittership for community members,

- who are not visible on dev/user list, have a couple of threads they
participated

- but contributed a significant feature or many fixes.

Usually, such contributors work for a commercial company with sufficient
product expertise, so they probably collaborate with experts, but outside
space of Apache.


Several guides and policies

https://www.apache.org/foundation/how-it-works.html#meritocracy

http://community.apache.org/newcommitter.html

and others say that PMC member needs to evaluate communication and
cooperative work with peers, ability to be a mentor, behavior in
disagreement.


Communication is required by Apache Ignite guide

https://cwiki.apache.org/confluence/display/IGNITE/Committership+Bar+Guidance

Simultaneously
https://community.apache.org/contributors/#contributing-a-project-copdoc

contains a mention someone who contributed sufficiently to ‘ANY’ area may
become a committer. So why can't we count code only contribution without
contribution to community/project?

There are several cases when I may disagree with other PMC members.

I insist candidate should communicate in ASF space because A)
community-first and motto: B) “If it didn’t happen on the mailing list it
didn’t happen.” For such cases then contributors collaborate outside Apache
space we can still accept a contribution, still appreciate contributor’s
effort and say thank you; but not promote as a committer. But I may
over-estimate the role of collaboration in the ASF. I may be too strict in
understanding ASF principles.

But PMCs who suggest such comittership candidates may counter-argument

- those cool developers don't like to communicate (they may be a little bit
uncomfortable with public communications/tries to avoid spam/any other
reasons they have).

- If he or she will communicate often, then he or she will never have time
to write a code.

So what do you think? Is it required to communicate with the rest of the
community publicly more than a couple of times to become a committer?

Sincerely,

Dmitriy Pavlov

Re: The Apache Way and good developers don’t like to communicate

[Malcolm Upayavira Holmes]

I cannot see how someone could be a an effective community member without 
joining the community's communication channel. That just seems a practical 
necessity. Especially if someone is to be given the keys, to become a 
committer. They need to be there to respond, e.g. if they were to break things.

Upayavira

On Fri, 2 Nov 2018, at 8:53 AM, Dmitriy Pavlov wrote:
> Dear ASF Fellows,
> 
> I am PMC member of Apache Ignite, but I joined PMC relatively recently. I
> need help from you again in regarding the Apache Way.
> 
> Question is related to comittership for community members,
> 
> - who are not visible on dev/user list, have a couple of threads they
> participated
> 
> - but contributed a significant feature or many fixes.
> 
> Usually, such contributors work for a commercial company with sufficient
> product expertise, so they probably collaborate with experts, but outside
> space of Apache.
> 
> 
> Several guides and policies
> 
> https://www.apache.org/foundation/how-it-works.html#meritocracy
> 
> http://community.apache.org/newcommitter.html
> 
> and others say that PMC member needs to evaluate communication and
> cooperative work with peers, ability to be a mentor, behavior in
> disagreement.
> 
> 
> Communication is required by Apache Ignite guide
> https://cwiki.apache.org/confluence/display/IGNITE/Committership+Bar+Guidance
> 
> Simultaneously
> https://community.apache.org/contributors/#contributing-a-project-copdoc
> 
> contains a mention someone who contributed sufficiently to ‘ANY’ area may
> become a committer. So why can't we count code only contribution without
> contribution to community/project?
> 
> There are several cases when I may disagree with other PMC members.
> 
> I insist candidate should communicate in ASF space because A)
> community-first and motto: B) “If it didn’t happen on the mailing list it
> didn’t happen.” For such cases then contributors collaborate outside Apache
> space we can still accept a contribution, still appreciate contributor’s
> effort and say thank you; but not promote as a committer. But I may
> over-estimate the role of collaboration in the ASF. I may be too strict in
> understanding ASF principles.
> 
> But PMCs who suggest such comittership candidates may counter-argument
> 
> - those cool developers don't like to communicate (they may be a little bit
> uncomfortable with public communications/tries to avoid spam/any other
> reasons they have).
> 
> - If he or she will communicate often, then he or she will never have time
> to write a code.
> 
> So what do you think? Is it required to communicate with the rest of the
> community publicly more than a couple of times to become a committer?
> 
> Sincerely,
> 
> Dmitriy Pavlov

-
To unsubscribe, e-mail: dev-unsubscr...@community.apache.org
For additional commands, e-mail: dev-h...@community.apache.org

The Apache Way and good developers don’t like to communicate

[Dmitriy Pavlov]

Dear ASF Fellows,

I am PMC member of Apache Ignite, but I joined PMC relatively recently. I
need help from you again in regarding the Apache Way.

Question is related to comittership for community members,

- who are not visible on dev/user list, have a couple of threads they
participated

- but contributed a significant feature or many fixes.

Usually, such contributors work for a commercial company with sufficient
product expertise, so they probably collaborate with experts, but outside
space of Apache.


Several guides and policies

https://www.apache.org/foundation/how-it-works.html#meritocracy

http://community.apache.org/newcommitter.html

and others say that PMC member needs to evaluate communication and
cooperative work with peers, ability to be a mentor, behavior in
disagreement.


Communication is required by Apache Ignite guide
https://cwiki.apache.org/confluence/display/IGNITE/Committership+Bar+Guidance

Simultaneously
https://community.apache.org/contributors/#contributing-a-project-copdoc

contains a mention someone who contributed sufficiently to ‘ANY’ area may
become a committer. So why can't we count code only contribution without
contribution to community/project?

There are several cases when I may disagree with other PMC members.

I insist candidate should communicate in ASF space because A)
community-first and motto: B) “If it didn’t happen on the mailing list it
didn’t happen.” For such cases then contributors collaborate outside Apache
space we can still accept a contribution, still appreciate contributor’s
effort and say thank you; but not promote as a committer. But I may
over-estimate the role of collaboration in the ASF. I may be too strict in
understanding ASF principles.

But PMCs who suggest such comittership candidates may counter-argument

- those cool developers don't like to communicate (they may be a little bit
uncomfortable with public communications/tries to avoid spam/any other
reasons they have).

- If he or she will communicate often, then he or she will never have time
to write a code.

So what do you think? Is it required to communicate with the rest of the
community publicly more than a couple of times to become a committer?

Sincerely,

Dmitriy Pavlov