You are viewing a plain text version of this content. The canonical link for it is here.
Posted to jcp-open@apache.org by Jason van Zyl <ja...@maven.org> on 2005/11/07 20:36:14 UTC
Re: About joining JSR 279 and 280
Hi, I am responding to this on the JCP open list so others can chime in.
On Wed, 2005-10-26 at 19:20 -0400, Geir Magnusson Jr. wrote:
> On Oct 26, 2005, at 6:39 PM, Jason van Zyl wrote:
>
> > On Wed, 2005-10-26 at 13:58 -0400, Geir Magnusson Jr. wrote:
> >
> >> Yes, I applied to be the placeholder so Changshin could be on the EG.
> >>
> >> I was waiting for his response to engage on this.
> >>
> >
> > How did this practice of one person applying for everything happen? I
> > think this is a bad practice and know from first hand experience on
> > JSR277 that this method does not scale. How many JSRs are you involved
> > with now? I can barely keep up with one. I think this process needs to
> > go distributed.
>
> Jason,
>
> I've tried to explain this before. I'm not involved in these JSRs.
> I'm applying for membership in the EG on behalf of the ASF.
Ok, this is a practice I don't think really fits with the ASF. We should
decide on the lists who will participate and let that person do it. We
should vote like we do on all the other projects and then help that
person become part of the JSR.
> In this case, Changshin was identified as wanting to join the EG from
> the get go, and hasn't engaged, so they used me as a placeholder. I
> have no personal interest in these technologies and am just getting
> things going on behalf of the ASF.
I think this just creates unnecessary overhead for you. Again,
coordination on the list, a simple vote, let the person move forward. If
that person doesn't have enough get-up-and-go to organize themselves to
sync up with the spec lead then do we want them representing Apache?
> >
> > I think people who want to be involved should be interacting with the
> > spec leads directly and getting themselves setup. How did this ever
> > come
> > to be that one person at Apache is single point of entry for dealing
> > with JSRs and is it really of any benefit to Apache?
>
> I think so because it gives us oversight and accountability to who is
> representing the ASF on these EGs.
That is definitely not going to work with one person doing this. This
should be done on this list and discussed and we as a group should
provide oversight, not one person.
We did not vote on JSR277 because you were probably dealing with many
other issues at the time. This list should be the safety net.
> Nothing stops people from joining
> EGs as individuals, but if you want to be representing the ASF, then
> it's the ASF that is joining, and we need to be sure that we know who
> is doing what, and why.
So why would that be hindered by acting as a group and deciding in the
standard ASF fashion who will participate in what?
> There is great potential for misuse of this
> privilege that we (the ASF) have, and there is potential for damage
> to our reputation.
I think things like not voting on important JSRs is an example of
something damaging our reputation. It's not a question of you not trying
to do the right thing, you just can't do it all and it should be
distributed across the people wanting to participate.
> >
> > Just write a doc on how people can apply themselves, make an initial
> > introduction and that's it. Was there a problem with this before? Does
> > SUN only want to deal with one person?
>
> That is actually what happens. I apply on behalf of the ASF as the
> official point of contact, but thats where it ends. The person that
> wants to participate on the EG does all the work - they have the
> voice on the EG, they get the mail.
C'mon Geir, that is not how it worked with JSR277. Brett was the ideal
candidate for the lead and it literally took 6 weeks for us to get you
to agree to let Brett have the official lead position. You were seen as
the lead so it was incredibly difficult for Brett to participate. This
is something that I would like to prevent in the future.
I think that:
1) No JSR should be applied for until it has been decided on this list
who is going to represent Apache.
2) When the person is selected via the standard voting practices then
that person contacts the spec lead. Again if this person doesn't have
the motivation to contact the spec lead and get things moving then we
should not participate.
3) I think it would be fair say that one individual cannot possibly
handle more then one JSR. I can barely keep up with JSR 277.
> A doc is forthccoming.
Cool, have you managed to squeeze this out somewhere?
Also, where is the current list of who at Apache is representing us each
of the JSRs. Is there something in SVN?
> geir
>
--
jvz.
Jason van Zyl
jason at maven.org
http://maven.apache.org
Our achievements speak for themselves. What we have to keep track
of are our failures, discouragements and doubts. We tend to forget
the past difficulties, the many false starts, and the painful
groping. We see our past achievements as the end result of a
clean forward thrust, and our present difficulties as
signs of decline and decay.
-- Eric Hoffer, Reflections on the Human Condition
Re: About joining JSR 279 and 280
Posted by "Geir Magnusson Jr." <ge...@apache.org>.
On Nov 19, 2005, at 9:19 PM, Brett Porter wrote:
> Sorry for the late reply.
>
> +1 from me too, with some comments.
>
> Dain Sundstrom wrote:
>> +1 I would have thought this is how things already worked. It
>> seems like the Apache Way... of course I have a few tweaks :)
>> On Nov 7, 2005, at 11:36 AM, Jason van Zyl wrote:
>>> 1) No JSR should be applied for until it has been decided on this
>>> list
>>> who is going to represent Apache.
>
> ... unless the deadline is coming up and for some reason a decision
> hasn't been reached.
>
> I do think its important that the right person gets on there first,
> as we've seen in more than one case it takes some time to make a
> switch (it is possible, but it takes some time). It took about two
> months until I was able to participate effectively.
it's not always obvious who the right person is. But yeah, we
certainly want to get interested active people involved. 277 is a
special case, IMO, and I think we should be careful in how much we
hype it as a general problem.
>
>>>
>>> 2) When the person is selected via the standard voting practices
>>> then
>>> that person contacts the spec lead. Again if this person doesn't
>>> have
>>> the motivation to contact the spec lead and get things moving
>>> then we
>>> should not participate.
>> I would suggest that this list be copied on all "administrative"
>> communication with the spec leads or the EC. That way we have an
>> official record of what is going on. Further, if we don't
>> already have one, I would like to have a JCP website at Apache
>> that details our policies joining a EG and lists the people
>> representing Apache on the various specs. Sometimes, I want to
>> ask for a feature to be added to a spec, but don't know who is
>> representing Apache.
>
> The list of reps is in the foundation repository under JCP/jsr-
> reps.txt (members only). It looks out of date - I'm pretty sure
> some are missing.
>
> I'm not sure if its necessary to keep that there - it could be in
> the committers only repository or in the asylum when it is set up
> (in fact, it probably can be in public - there's nothing secret in
> there, though there isn't any benefit to this).
I'm going to put that info on the JCP site. There's nothing secret
about this. This is just historical that it is there.
>
> I agree with Dain on documenting the people on the JSRs and
> policies around them. I think it would be great to better promote
> the JCP within Apache - for instance, this list has about 12
> subscribers. There are a lot more on the members list. Sorting out
> who has access to what and making sure that people can get involved
> should be a priority.
I agree that better visibility will be good, but that said, this
little spurt of interest in the JCP with you three is good and
encouraging, and we should build on it. Historically, interest has
been light, but as there appears to be more interest in implementing
specs in open source (maybe because of commercial drivers? I dunno)
maybe we can get more interested and participation. I'm all for it.
>
>>> 3) I think it would be fair say that one individual cannot possibly
>>> handle more then one JSR. I can barely keep up with JSR 277.
>> I'm not sure on this one. Some specs don't have much volume and
>> others do. This is kind of like saying you can be heavily
>> involved in more than one Apache project. Of course, I don't
>> want an Apache rep that is so overwhelmed that they aren't
>> effectively representing us. Maybe we should ask the rep to
>> create a quarterly report, which we can attache to the board
>> report for this group. If they can't create a paragraph, it is a
>> good sign they aren't involved.
>
> Agreed. This shouldn't be a restriction, but by the same token if
> there are other equally qualified people then sharing the load is
> good.
>
Yep.
> So how do we take this forward?
Well, I'll get the JCP site up before apachecon, and as jason
suggested, we should certainly get together and talk about it.
geir
--
Geir Magnusson Jr +1-203-665-6437
geirm@apache.org
Re: About joining JSR 279 and 280
Posted by Jason van Zyl <ja...@maven.org>.
On Sun, 2005-11-20 at 13:19 +1100, Brett Porter wrote:
> So how do we take this forward?
I think with ApacheCon coming up fairly shortly I think we could have a
small get together and resolve the whole process we'd like to setup.
Maybe a BOF?
> Cheers,
> Brett
>
--
jvz.
Jason van Zyl
jason at maven.org
http://maven.apache.org
Re: About joining JSR 279 and 280
Posted by Brett Porter <br...@apache.org>.
Sorry for the late reply.
+1 from me too, with some comments.
Dain Sundstrom wrote:
> +1 I would have thought this is how things already worked. It seems
> like the Apache Way... of course I have a few tweaks :)
>
> On Nov 7, 2005, at 11:36 AM, Jason van Zyl wrote:
>
>> 1) No JSR should be applied for until it has been decided on this list
>> who is going to represent Apache.
... unless the deadline is coming up and for some reason a decision
hasn't been reached.
I do think its important that the right person gets on there first, as
we've seen in more than one case it takes some time to make a switch (it
is possible, but it takes some time). It took about two months until I
was able to participate effectively.
>>
>> 2) When the person is selected via the standard voting practices then
>> that person contacts the spec lead. Again if this person doesn't have
>> the motivation to contact the spec lead and get things moving then we
>> should not participate.
>
>
> I would suggest that this list be copied on all "administrative"
> communication with the spec leads or the EC. That way we have an
> official record of what is going on. Further, if we don't already have
> one, I would like to have a JCP website at Apache that details our
> policies joining a EG and lists the people representing Apache on the
> various specs. Sometimes, I want to ask for a feature to be added to a
> spec, but don't know who is representing Apache.
>
The list of reps is in the foundation repository under JCP/jsr-reps.txt
(members only). It looks out of date - I'm pretty sure some are missing.
I'm not sure if its necessary to keep that there - it could be in the
committers only repository or in the asylum when it is set up (in fact,
it probably can be in public - there's nothing secret in there, though
there isn't any benefit to this).
I agree with Dain on documenting the people on the JSRs and policies
around them. I think it would be great to better promote the JCP within
Apache - for instance, this list has about 12 subscribers. There are a
lot more on the members list. Sorting out who has access to what and
making sure that people can get involved should be a priority.
>> 3) I think it would be fair say that one individual cannot possibly
>> handle more then one JSR. I can barely keep up with JSR 277.
>
>
> I'm not sure on this one. Some specs don't have much volume and others
> do. This is kind of like saying you can be heavily involved in more
> than one Apache project. Of course, I don't want an Apache rep that is
> so overwhelmed that they aren't effectively representing us. Maybe we
> should ask the rep to create a quarterly report, which we can attache
> to the board report for this group. If they can't create a paragraph,
> it is a good sign they aren't involved.
Agreed. This shouldn't be a restriction, but by the same token if there
are other equally qualified people then sharing the load is good.
So how do we take this forward?
Cheers,
Brett
Re: Democracy? [was: About joining JSR 279 and 280]
Posted by "Geir Magnusson Jr." <ge...@apache.org>.
On Dec 9, 2005, at 2:10 PM, Roy T. Fielding wrote:
> On Dec 9, 2005, at 11:02 AM, Geir Magnusson Jr. wrote:
>
>> That was pre 2.5 though, IIRC.
>
> Yes, but just 2.5 is not enough to enable open source. The spec lead
> has to propose a license that is compatible as well. Sun, Day, and
> (I think) BEA and (maybe) IBM are the only ones that do that now.
> We should be voting "no" when the terms of use are not satisfactory.
Terms of use of the RI? The specs all come out under the 2.5+ spec
license, IIRC. I'll go back and check.
geir
--
Geir Magnusson Jr +1-203-665-6437
geirm@apache.org
Re: Democracy? [was: About joining JSR 279 and 280]
Posted by "Roy T. Fielding" <fi...@gbiv.com>.
On Dec 9, 2005, at 11:02 AM, Geir Magnusson Jr. wrote:
> That was pre 2.5 though, IIRC.
Yes, but just 2.5 is not enough to enable open source. The spec lead
has to propose a license that is compatible as well. Sun, Day, and
(I think) BEA and (maybe) IBM are the only ones that do that now.
We should be voting "no" when the terms of use are not satisfactory.
....Roy
Re: Democracy? [was: About joining JSR 279 and 280]
Posted by "Geir Magnusson Jr." <ge...@apache.org>.
That was pre 2.5 though, IIRC.
On Dec 9, 2005, at 2:00 PM, Roy T. Fielding wrote:
> On Dec 9, 2005, at 10:47 AM, Geir Magnusson Jr. wrote:
>>> * This list should discuss all votes on the EC and decide how
>>> Apache will cast it's vote
>>
>> Discussion is certainly good, but the ASF has never in my memory
>> cast a negative vote.
>
> Umm, many times, and we always vote NO on JSRs that cannot be
> implemented
> royalty-free in open source.
>
> ....Roy
>
--
Geir Magnusson Jr +1-203-665-6437
geirm@apache.org
Re: Democracy? [was: About joining JSR 279 and 280]
Posted by "Roy T. Fielding" <fi...@gbiv.com>.
On Dec 9, 2005, at 10:47 AM, Geir Magnusson Jr. wrote:
>> * This list should discuss all votes on the EC and decide how
>> Apache will cast it's vote
>
> Discussion is certainly good, but the ASF has never in my memory
> cast a negative vote.
Umm, many times, and we always vote NO on JSRs that cannot be
implemented
royalty-free in open source.
....Roy
Re: Democracy? [was: About joining JSR 279 and 280]
Posted by "Geir Magnusson Jr." <ge...@apache.org>.
On Nov 21, 2005, at 7:38 PM, Dain Sundstrom wrote:
> Geir, I believe you missed the meat of my proposal which was a
> proposal to make the JCP process at Apache democratic. It is my
> opinion that the process for JCP involvement at Apache is really a
> dictatorship and not democratic.
I'm not sure that either is a good word. First, the ASF is a
meritocracy. Everyone is entitled to speak their peace, but those
who make decisions are a subset of the entire community, just like in
all ares of the foundation's functioning.
In the case of ASF representation on expert groups, it's really been
very open - at first, anyone who was a member and was interested
could represent the ASF, and now, those who are committers are
eligible. I say "eligible" because I think we need to be careful.
Our participation in the JCP is really about process and community,
and not about technology - the ASF doesn't actually have specific
technology roadmaps as a foundation - we want "a thousand flowers to
bloom". The ability to participate in the technology activities -
namely the expert groups - is a secondary thing for us, and thus
we've formally kept the eligibility under formal approval (by me,
thus far). The reality has been that no one has ever been refused
participation, but we must certainly keep the option of saying no to
someone clearly available.
> Now, I get the feeling this is a benevolent dictatorship, and to
> be crystal clear I am not saying anything negative you or the
> current process.
C'mon. Fool me once...
> I would like to see the process changed to pure democracy, and to
> achieve this I think we need to change two areas, openness and voting.
>
> Openness
> * Publicly document the process to join an EG
> * Publicly document the requirements of an person representing Apache
> * Publicly document the people representing Apache on the various
> specs
> * All administrative communication between Apache and the JCP
> should be copied to this (or the private) list
> * An EG representative should give quarterly reports to this list
These are all good things, and all have been suggested before and are
or should be in process of happening.
>
> Voting
> * This list should discuss all votes on the EC and decide how
> Apache will cast it's vote
Discussion is certainly good, but the ASF has never in my memory cast
a negative vote. I'd also reserve for the VP, JCP the decision on
what the vote should be in the end. The responsibility has been
placed on that person's shoulders.
But sure - I'll start posting the weekly ballot info and if anyone is
interested, we can dig in.
> * This list should vote on who will represent apache on the EGs
This isn't so clear to me what you get out of this It's not obvious
why this list would know better than the specialists working in the
technology area - those volunteering to do the work - about who
should be the rep. Generally, we don't have multiple people wanting
to do it. When we do, I think we should always try for dual/treble/
etc representation to get the most people participating if the EG
will allow it, and if not, the individuals better be mature enough to
work this out themselves, or there's no way they should be on a JCP
expert group.
> * This list should discuss and vote to propose Apache sponsored
> specifications
Well, certainly people should be aware, but why would we ever act as
a gatekeeper preventing people from trying to take their technology
to the JCP? And sponsoring a spec is such a big deal, if anyone
signed up to do the work and do a spec, I'd be the first the bless
them and wish them luck, giving them any support they need. That
said, I'd be the first to try and shut them down if they were giving
the ASF a bad name, but I would consider that an unlikely,
extraordinary situation that the VP, JCP could just deal with.
> * This list should on a yearly basis propose to the board a
> individual that we would like to represent Apache in the EC (of
> course the board has the choice to choose whom ever they want, but
> we should offer the opinion of this group)
This makes little sense to me. It takes a bit of time (about a
year or so) to understand the political dynamics of the JCP enough to
be effective. The JCP is staffed by professional standards-bodies
representatives, and the politics is pretty tough. I think that
there is no upside for the ASF for this happening this way. So no, I
don't think that this should be turned over every year.
geir
> -dain
>
> On Nov 20, 2005, at 3:42 AM, Geir Magnusson Jr. wrote:
>
>>
>> On Nov 8, 2005, at 3:05 PM, Dain Sundstrom wrote:
>>
>>> +1 I would have thought this is how things already worked. It
>>> seems like the Apache Way... of course I have a few tweaks :)
>>>
>>> On Nov 7, 2005, at 11:36 AM, Jason van Zyl wrote:
>>>
>>>> 1) No JSR should be applied for until it has been decided on
>>>> this list
>>>> who is going to represent Apache.
>>>>
>>>> 2) When the person is selected via the standard voting practices
>>>> then
>>>> that person contacts the spec lead. Again if this person doesn't
>>>> have
>>>> the motivation to contact the spec lead and get things moving
>>>> then we
>>>> should not participate.
>>>
>>> I would suggest that this list be copied on all "administrative"
>>> communication with the spec leads or the EC. That way we have an
>>> official record of what is going on. Further, if we don't
>>> already have one, I would like to have a JCP website at Apache
>>> that details our policies joining a EG and lists the people
>>> representing Apache on the various specs. Sometimes, I want to
>>> ask for a feature to be added to a spec, but don't know who is
>>> representing Apache.
>>
>> No problem there. I've had the website planned, and this is a
>> good kick.
>>
>>>
>>>> 3) I think it would be fair say that one individual cannot possibly
>>>> handle more then one JSR. I can barely keep up with JSR 277.
>>>
>>> I'm not sure on this one. Some specs don't have much volume and
>>> others do. This is kind of like saying you can be heavily
>>> involved in more than one Apache project. Of course, I don't
>>> want an Apache rep that is so overwhelmed that they aren't
>>> effectively representing us.
>>
>> I think that it's up to individuals who volunteer if they want to
>> participate in more than one. I see no problem with it. If there
>> is a dormant rep, then certainly a person who wants to be active
>> should take over. But without any other person interested, a
>> dormant rep is fine. It happens all the time in the JCP.
>>
>>> Maybe we should ask the rep to create a quarterly report, which
>>> we can attache to the board report for this group. If they can't
>>> create a paragraph, it is a good sign they aren't involved.
>>
>> We have to be careful there - I've thought about monthly
>> "heartbeat" reports, but we have to be careful about information
>> from expert groups getting to public board reports. We have a
>> weird situation because our interests in openness need to square
>> with our responsibilities to keep EG confidential information
>> confidential (this varies EG by EG...)
>>
>> It's good to see some interest in this area.
>>
>> geir
>>
>> --
>> Geir Magnusson Jr +1-203-665-6437
>> geirm@apache.org
>>
>
--
Geir Magnusson Jr +1-203-665-6437
geirm@apache.org
Re: Democracy? [was: About joining JSR 279 and 280]
Posted by Brett Porter <br...@apache.org>.
Hi,
Some more thoughts on this.
Dain Sundstrom wrote:
> Geir, I believe you missed the meat of my proposal which was a proposal
> to make the JCP process at Apache democratic.
Something felt wrong about this word, and when prompted to think about
it I remembered the ASF is meritocratic, not democratic. I think its
probably what you mean anyway, you're just working on the assumption
that everyone here has merit, which seems a fair assumption at this point :)
Meritocracy vs democracy is an interesting distinction when it comes to
something such as this, but what I think it means is:
- all members with appropriate paperwork can participate (as they can
in any ASF project they are interested in, as I understand it)
- committers with proven skills in the relevant areas and with
appropriate paperwork can participate.
This would be the "committers list" of JCP@Apache. Importantly, there
should also be a way for any committer to jump in and learn about and
contribute to "prove" themselves too.
> It is my opinion that the
> process for JCP involvement at Apache is really a dictatorship and not
> democratic. Now, I get the feeling this is a benevolent dictatorship,
> and to be crystal clear I am not saying anything negative you or the
> current process. I would like to see the process changed to pure
> democracy, and to achieve this I think we need to change two areas,
> openness and voting.
In some ways I agree. It is a dictatorship in the respect that it's one
of the few offices not backed by a committee. From
http://www.apache.org/foundation/, I relate the first 4 with the board,
the projects with a PMC, then there is the PRC. The other two: legal
affairs (very new) and JCP are individual positions with no specific
committee, but discussion lists.
I'm sure Geir would argue here that others are welcome to participate as
we are doing here. I think this is just the chicken and egg problem you
would see anywhere. If the process is not documented or open, then you
don't tend to form a community around it.
Does it make sense to set up a PMC-like committee that can share some of
the responsibilities? I'm not exactly sure how the PRC is defined or
operates, but it might be similar. It seems to me that we should have
these lists:
- private. To be discussed only with those under NDA (not necessarily
members, if this is allowed - those that fit the 2 categories I listed
above)
- public. Only committers can subscribe, but a public archive
available. For discussing apache's process, participation and to provide
feedback on JSRs of interest. Kind of like legal-discuss. Would be good
to be as open as possible, without violating the trust of any EG's, so
members of the private group need to be prudent.
- jsrXXX-discuss. One per JSR where requested. Only NDA'd individuals,
but so more Apache people can discuss individual JSRs even if only one
can be on an EG.
>
> Openness
> * Publicly document the process to join an EG
> * Publicly document the requirements of an person representing Apache
> * Publicly document the people representing Apache on the various specs
> * All administrative communication between Apache and the JCP should be
> copied to this (or the private) list
+1 to these. I think we all agree on something like this and just need
to get it started.
> * An EG representative should give quarterly reports to this list
The JCP VP is meant to report to the board in Feb, May, Aug, Nov, though
I'm not aware of whether this happens. Geir, can you shed some light on
this? Anyway, it seems the perfect opportunity to provide feedback and
issues, to the extent possible given that board reports go public after
some time.
> Voting
> * This list should discuss all votes on the EC and decide how Apache
> will cast it's vote
+1
> * This list should vote on who will represent apache on the EGs
An interesting one. Nowhere except the board do we vote one person
against another, nor do we allow self-nomination for votes, which is
what this seems to be about.
In theory I agree - it should be consensus of the community, I just
worry about the mechanics. I think we'd have to wait for the situation
to arise and see how it plays out. I'd find it unlikely that there will
be a lot of contention around a spot if it was discussed openly, and
anyone that easily offended when not selected because another is viewed
as having more merit in that instance, probably doesn't have the stomach
for it anyway :)
> * This list should discuss and vote to propose Apache sponsored
> specifications
+1
> * This list should on a yearly basis propose to the board a individual
> that we would like to represent Apache in the EC (of course the board
> has the choice to choose whom ever they want, but we should offer the
> opinion of this group)
I know this is something that is proposed and sometimes practiced
elsewhere. I'm not really comfortable with it in projects because it
creates unnecessary tension at that time and places more importance than
necessary on the chair. In a project, the chair is just reporting to the
board. They don't have any special powers, require any particular
technical expertise, and they aren't the designated leader of the
project. I feel that if they are doing a good job, the project is happy
with their reports to the board and handling of issues, and the board is
happy with the view of the project, then they should continue on as long
as they feel they have the time and energy.
The Apache EC member is actually quite different, and while they can
(and ideally should?) be the same, I wonder if it should be described as
a separate position to the VP position responsible to the board? I'm
certainly interested to hear more from Geir about what he thinks is
involved in that position and whether that is a fair distinction.
One real concern I have here is, currently, if the representative
decides that there are better things in life and decides to throw in the
towel, then it could be quite difficult to replace them. Does anybody
know enough about the processes, the current issues and relationships to
effectively do this?
Anyway, I hope we can get together at ApacheCon and discuss these things
(or anything else in the mean time). I'm looking forward to it.
Cheers,
Brett
Democracy? [was: About joining JSR 279 and 280]
Posted by Dain Sundstrom <da...@iq80.com>.
Geir, I believe you missed the meat of my proposal which was a
proposal to make the JCP process at Apache democratic. It is my
opinion that the process for JCP involvement at Apache is really a
dictatorship and not democratic. Now, I get the feeling this is a
benevolent dictatorship, and to be crystal clear I am not saying
anything negative you or the current process. I would like to see
the process changed to pure democracy, and to achieve this I think we
need to change two areas, openness and voting.
Openness
* Publicly document the process to join an EG
* Publicly document the requirements of an person representing Apache
* Publicly document the people representing Apache on the various
specs
* All administrative communication between Apache and the JCP
should be copied to this (or the private) list
* An EG representative should give quarterly reports to this list
Voting
* This list should discuss all votes on the EC and decide how
Apache will cast it's vote
* This list should vote on who will represent apache on the EGs
* This list should discuss and vote to propose Apache sponsored
specifications
* This list should on a yearly basis propose to the board a
individual that we would like to represent Apache in the EC (of
course the board has the choice to choose whom ever they want, but we
should offer the opinion of this group)
To me this is what Apache is all about, democracy.
-dain
On Nov 20, 2005, at 3:42 AM, Geir Magnusson Jr. wrote:
>
> On Nov 8, 2005, at 3:05 PM, Dain Sundstrom wrote:
>
>> +1 I would have thought this is how things already worked. It
>> seems like the Apache Way... of course I have a few tweaks :)
>>
>> On Nov 7, 2005, at 11:36 AM, Jason van Zyl wrote:
>>
>>> 1) No JSR should be applied for until it has been decided on this
>>> list
>>> who is going to represent Apache.
>>>
>>> 2) When the person is selected via the standard voting practices
>>> then
>>> that person contacts the spec lead. Again if this person doesn't
>>> have
>>> the motivation to contact the spec lead and get things moving
>>> then we
>>> should not participate.
>>
>> I would suggest that this list be copied on all "administrative"
>> communication with the spec leads or the EC. That way we have an
>> official record of what is going on. Further, if we don't already
>> have one, I would like to have a JCP website at Apache that
>> details our policies joining a EG and lists the people
>> representing Apache on the various specs. Sometimes, I want to
>> ask for a feature to be added to a spec, but don't know who is
>> representing Apache.
>
> No problem there. I've had the website planned, and this is a good
> kick.
>
>>
>>> 3) I think it would be fair say that one individual cannot possibly
>>> handle more then one JSR. I can barely keep up with JSR 277.
>>
>> I'm not sure on this one. Some specs don't have much volume and
>> others do. This is kind of like saying you can be heavily
>> involved in more than one Apache project. Of course, I don't want
>> an Apache rep that is so overwhelmed that they aren't effectively
>> representing us.
>
> I think that it's up to individuals who volunteer if they want to
> participate in more than one. I see no problem with it. If there
> is a dormant rep, then certainly a person who wants to be active
> should take over. But without any other person interested, a
> dormant rep is fine. It happens all the time in the JCP.
>
>> Maybe we should ask the rep to create a quarterly report, which we
>> can attache to the board report for this group. If they can't
>> create a paragraph, it is a good sign they aren't involved.
>
> We have to be careful there - I've thought about monthly
> "heartbeat" reports, but we have to be careful about information
> from expert groups getting to public board reports. We have a
> weird situation because our interests in openness need to square
> with our responsibilities to keep EG confidential information
> confidential (this varies EG by EG...)
>
> It's good to see some interest in this area.
>
> geir
>
> --
> Geir Magnusson Jr +1-203-665-6437
> geirm@apache.org
>
Re: About joining JSR 279 and 280
Posted by "Geir Magnusson Jr." <ge...@apache.org>.
On Nov 8, 2005, at 3:05 PM, Dain Sundstrom wrote:
> +1 I would have thought this is how things already worked. It
> seems like the Apache Way... of course I have a few tweaks :)
>
> On Nov 7, 2005, at 11:36 AM, Jason van Zyl wrote:
>
>> 1) No JSR should be applied for until it has been decided on this
>> list
>> who is going to represent Apache.
>>
>> 2) When the person is selected via the standard voting practices then
>> that person contacts the spec lead. Again if this person doesn't have
>> the motivation to contact the spec lead and get things moving then we
>> should not participate.
>
> I would suggest that this list be copied on all "administrative"
> communication with the spec leads or the EC. That way we have an
> official record of what is going on. Further, if we don't already
> have one, I would like to have a JCP website at Apache that details
> our policies joining a EG and lists the people representing Apache
> on the various specs. Sometimes, I want to ask for a feature to be
> added to a spec, but don't know who is representing Apache.
No problem there. I've had the website planned, and this is a good
kick.
>
>> 3) I think it would be fair say that one individual cannot possibly
>> handle more then one JSR. I can barely keep up with JSR 277.
>
> I'm not sure on this one. Some specs don't have much volume and
> others do. This is kind of like saying you can be heavily involved
> in more than one Apache project. Of course, I don't want an Apache
> rep that is so overwhelmed that they aren't effectively
> representing us.
I think that it's up to individuals who volunteer if they want to
participate in more than one. I see no problem with it. If there is
a dormant rep, then certainly a person who wants to be active should
take over. But without any other person interested, a dormant rep is
fine. It happens all the time in the JCP.
> Maybe we should ask the rep to create a quarterly report, which we
> can attache to the board report for this group. If they can't
> create a paragraph, it is a good sign they aren't involved.
We have to be careful there - I've thought about monthly "heartbeat"
reports, but we have to be careful about information from expert
groups getting to public board reports. We have a weird situation
because our interests in openness need to square with our
responsibilities to keep EG confidential information confidential
(this varies EG by EG...)
It's good to see some interest in this area.
geir
--
Geir Magnusson Jr +1-203-665-6437
geirm@apache.org
Re: About joining JSR 279 and 280
Posted by Dain Sundstrom <da...@iq80.com>.
+1 I would have thought this is how things already worked. It seems
like the Apache Way... of course I have a few tweaks :)
On Nov 7, 2005, at 11:36 AM, Jason van Zyl wrote:
> 1) No JSR should be applied for until it has been decided on this list
> who is going to represent Apache.
>
> 2) When the person is selected via the standard voting practices then
> that person contacts the spec lead. Again if this person doesn't have
> the motivation to contact the spec lead and get things moving then we
> should not participate.
I would suggest that this list be copied on all "administrative"
communication with the spec leads or the EC. That way we have an
official record of what is going on. Further, if we don't already
have one, I would like to have a JCP website at Apache that details
our policies joining a EG and lists the people representing Apache on
the various specs. Sometimes, I want to ask for a feature to be
added to a spec, but don't know who is representing Apache.
> 3) I think it would be fair say that one individual cannot possibly
> handle more then one JSR. I can barely keep up with JSR 277.
I'm not sure on this one. Some specs don't have much volume and
others do. This is kind of like saying you can be heavily involved
in more than one Apache project. Of course, I don't want an Apache
rep that is so overwhelmed that they aren't effectively representing
us. Maybe we should ask the rep to create a quarterly report, which
we can attache to the board report for this group. If they can't
create a paragraph, it is a good sign they aren't involved.
-dain
Re: About joining JSR 279 and 280
Posted by "Geir Magnusson Jr." <ge...@apache.org>.
On Nov 20, 2005, at 10:40 AM, Jason van Zyl wrote:
> On Sun, 2005-11-20 at 04:34 -0500, Geir Magnusson Jr. wrote:
>> Sorry about the delay - changed jobs...
>>
>> On Nov 7, 2005, at 2:36 PM, Jason van Zyl wrote:
>>
>>> Hi, I am responding to this on the JCP open list so others can
>>> chime in.
>>
>> Next time you take a mail from a private list to a public one, please
>> ask permission.
>
> Do you not remember that you are the one who told me to take the
> discussion regarding the JCP process within Apache to the jcp-open
> list
> in the first place some while back? Sorry if I misunderstood but I
> believe that is what you asked of me.
I actually don't recall, but I believe you and I apologize for being
snappish. It's been a bit rough lately.
geir
--
Geir Magnusson Jr +1-203-665-6437
geirm@apache.org
Re: About joining JSR 279 and 280
Posted by Jason van Zyl <ja...@maven.org>.
On Sun, 2005-11-20 at 04:34 -0500, Geir Magnusson Jr. wrote:
> Sorry about the delay - changed jobs...
>
> On Nov 7, 2005, at 2:36 PM, Jason van Zyl wrote:
>
> > Hi, I am responding to this on the JCP open list so others can
> > chime in.
>
> Next time you take a mail from a private list to a public one, please
> ask permission.
Do you not remember that you are the one who told me to take the
discussion regarding the JCP process within Apache to the jcp-open list
in the first place some while back? Sorry if I misunderstood but I
believe that is what you asked of me.
--
jvz.
Jason van Zyl
jason at maven.org
http://maven.apache.org
Three people can keep a secret provided two of them are dead.
-- Unknown
Re: About joining JSR 279 and 280
Posted by Brett Porter <br...@apache.org>.
Geir Magnusson Jr. wrote:
>
> Once he signs the NDA, you can share anything you wish with him. It's
> not up to the spec lead.
That's what I thought. Thanks for the confirmation.
Cheers,
Brett
Re: test-driven standards
Posted by Steve Loughran <st...@apache.org>.
Dan Diephouse wrote:
>
> I will to try to help realize the dream because it is a dream of mine as
> well. There is no reason not to make the JSR source and tests available
> under the CDDL or some such license. I'm not sure how effective I will
> be through the ASF, but we'll see :-)
>
> - Dan
>
Well, I'm trying to overthrow the W3C's process, one TAG member at a
time. if the real standards bodies can set examples, there is no reason
why the rest can't.
I guess the problem w/ the JSR groups is that they often view tests as
strategic things to only be released to people who sign the NDAs, like
you have to do for TCKs. But in reality they are nothing but a means of
testing that an implementation meets the design, which is something
every end user should have the right to do.
-Steve
Re: test-driven standards
Posted by Dan Diephouse <da...@envoisolutions.com>.
Steve Loughran wrote:
> Dan Diephouse wrote:
>
>> Steve Loughran wrote:
>>
>>> On the subject of standards, its my opinion that standards groups
>>> need to adopt a lot more OSS thinking. That means
>>>
>>> -specs come with tests. Or the tests are the spec, to put it
>>> differently
>>> -team SCM repos of test+source are public
>>> -gump does a nightly build of everything
>>>
>>> see: see:
>>> http://cvs.sourceforge.net/viewcvs.py/*checkout*/deployment/deployment/doc/steve/test_driven_standards.doc
>>>
>>>
>>>
>>> I'm partway there with the spec I work on, but there is so much
>>> ignorance (we dont need tests) and unwillingness to put stuff into a
>>> shared SCM that we dont get the full benefits
>>>
> > I can't +1 this enough. Is there any kind of legal reason why current
> > JSRs are so closed about their TCKs and sources?
> >
>
>
> Dont know about legality, but I I kind of hint at other causes in this
> presentation on testing, where I worry about adoption being pushed
> back by management
>
> http://people.apache.org/~stevel/slides/testing.pdf
>
> Interestingly, junit testing is far more prevalent in java OSS than
> 'enterprise grade' apps that I often encounter. I havent yet come up
> with a good reason for this, but some hypotheses:
>
> -pressure from above to deliver code, no management pressure for tests.
> -more innovation in process is happening in OSS
> -in OSS-land, everybody knows not to trust an app without junit tests
> (or regular gump outages)
>
> I have a dream, and the dream is: every W3C standard, every JCP
> program has an SVN repo with a test suite, and gump builds every one,
> every night. Help me realise this dream... (*)
>
Already read your presentation from your blog, but thanks. I totally
agree with you. And FWIW your experience with unit testing in the
enterprise lines up completely with mine. I for the most part feel its
ingorance. A lot of organizations are isolated and live in their own
little world.
I will to try to help realize the dream because it is a dream of mine as
well. There is no reason not to make the JSR source and tests available
under the CDDL or some such license. I'm not sure how effective I will
be through the ASF, but we'll see :-)
- Dan
--
Dan Diephouse
Envoi Solutions LLC
http://netzooid.com
test-driven standards
Posted by Steve Loughran <st...@apache.org>.
Dan Diephouse wrote:
> Steve Loughran wrote:
>
>> On the subject of standards, its my opinion that standards groups need
>> to adopt a lot more OSS thinking. That means
>>
>> -specs come with tests. Or the tests are the spec, to put it differently
>> -team SCM repos of test+source are public
>> -gump does a nightly build of everything
>>
>> see: see:
>> http://cvs.sourceforge.net/viewcvs.py/*checkout*/deployment/deployment/doc/steve/test_driven_standards.doc
>>
>>
>>
>> I'm partway there with the spec I work on, but there is so much
>> ignorance (we dont need tests) and unwillingness to put stuff into a
>> shared SCM that we dont get the full benefits
>>
> I can't +1 this enough. Is there any kind of legal reason why current
> JSRs are so closed about their TCKs and sources?
>
Dont know about legality, but I I kind of hint at other causes in this
presentation on testing, where I worry about adoption being pushed back
by management
http://people.apache.org/~stevel/slides/testing.pdf
Interestingly, junit testing is far more prevalent in java OSS than
'enterprise grade' apps that I often encounter. I havent yet come up
with a good reason for this, but some hypotheses:
-pressure from above to deliver code, no management pressure for tests.
-more innovation in process is happening in OSS
-in OSS-land, everybody knows not to trust an app without junit tests
(or regular gump outages)
I have a dream, and the dream is: every W3C standard, every JCP program
has an SVN repo with a test suite, and gump builds every one, every
night. Help me realise this dream... (*)
-Steve
(*) I also have a dream that at the click of a button we can refile
bugreps of the type "Ant doesnt handle '--' in comments" as bugs with
the XML working group though some distributed bug database, but that one
is harder,
Re: About joining JSR 279 and 280
Posted by Dan Diephouse <da...@envoisolutions.com>.
Steve Loughran wrote:
> On the subject of standards, its my opinion that standards groups need
> to adopt a lot more OSS thinking. That means
>
> -specs come with tests. Or the tests are the spec, to put it differently
> -team SCM repos of test+source are public
> -gump does a nightly build of everything
>
> see: see:
> http://cvs.sourceforge.net/viewcvs.py/*checkout*/deployment/deployment/doc/steve/test_driven_standards.doc
>
>
>
> I'm partway there with the spec I work on, but there is so much
> ignorance (we dont need tests) and unwillingness to put stuff into a
> shared SCM that we dont get the full benefits
>
>
I can't +1 this enough. Is there any kind of legal reason why current
JSRs are so closed about their TCKs and sources?
--
Dan Diephouse
(616) 971-2053
Envoi Solutions LLC
http://netzooid.com
Re: About joining JSR 279 and 280
Posted by Steve Loughran <st...@apache.org>.
Geir Magnusson Jr. wrote:
>
> On Nov 20, 2005, at 5:42 PM, Brett Porter wrote:
>
>> Steve Loughran wrote:
>>
>>> Geir Magnusson Jr. wrote:
>>>
>>>> Sorry about the delay - changed jobs...
>>>
>>> so we hear. couldnt take working for the company that gave us XML1.1
>>> eh?
>>>
>>>>
>>>> (aside : I'm interested in how Brett is treating the OSGi issue,
>>>> as it's a very interesting political and IP minefield - from what
>>>> I hear, the dance is in full swing, as expected...)
>>>>
>>> I have to deal with some of the packaging bits from other standards,
>>> I'm interested in this too.
>>
>>
>> Steve - cool, good to hear! If you have any thoughts you'd like me to
>> forward on or keep in mind based on the initial description on the
>> site I'd be happy to do so. If I understand what you mean by
>> packaging, then it hasn't exactly been dealt with yet, so no rush.
>>
>> As a member, I believe there should be no problem with discussing it
>> with you, though you may need to sign an NDA and I'll ask that Geir
>> clarify that first. I'm sure it's fine, but I want to make sure we do
>> follow all appropriate confidentiality rules within the EGs.
>
>
> Yes - Steve must sign the NDA for any EG-private material. It can be
> found easily here :
>
> http://people.apache.org/~geirm/ApacheNDA.pdf or in the legal part of
> the SVN repository.
>
> Steve, once you have done that, fax to +1 203 665 6400 and I'll log and
> forward to Jim for safekeeping in the ASF Hall of Records.
As an HP employee, I probably already qualify for NDA access to the
data, I'd have to go that route, because I dont want to get contaminated
by any agreement that my lawyers didnt approve of.
I'll have to check.
On the subject of standards, its my opinion that standards groups need
to adopt a lot more OSS thinking. That means
-specs come with tests. Or the tests are the spec, to put it differently
-team SCM repos of test+source are public
-gump does a nightly build of everything
see: see:
http://cvs.sourceforge.net/viewcvs.py/*checkout*/deployment/deployment/doc/steve/test_driven_standards.doc
I'm partway there with the spec I work on, but there is so much
ignorance (we dont need tests) and unwillingness to put stuff into a
shared SCM that we dont get the full benefits
Re: About joining JSR 279 and 280
Posted by "Geir Magnusson Jr." <ge...@apache.org>.
On Nov 20, 2005, at 5:42 PM, Brett Porter wrote:
> Steve Loughran wrote:
>> Geir Magnusson Jr. wrote:
>>> Sorry about the delay - changed jobs...
>> so we hear. couldnt take working for the company that gave us
>> XML1.1 eh?
>>>
>>> (aside : I'm interested in how Brett is treating the OSGi issue,
>>> as it's a very interesting political and IP minefield - from
>>> what I hear, the dance is in full swing, as expected...)
>>>
>> I have to deal with some of the packaging bits from other
>> standards, I'm interested in this too.
>
> Steve - cool, good to hear! If you have any thoughts you'd like me
> to forward on or keep in mind based on the initial description on
> the site I'd be happy to do so. If I understand what you mean by
> packaging, then it hasn't exactly been dealt with yet, so no rush.
>
> As a member, I believe there should be no problem with discussing
> it with you, though you may need to sign an NDA and I'll ask that
> Geir clarify that first. I'm sure it's fine, but I want to make
> sure we do follow all appropriate confidentiality rules within the
> EGs.
Yes - Steve must sign the NDA for any EG-private material. It can be
found easily here :
http://people.apache.org/~geirm/ApacheNDA.pdf or in the legal part
of the SVN repository.
Steve, once you have done that, fax to +1 203 665 6400 and I'll log
and forward to Jim for safekeeping in the ASF Hall of Records.
>
> Certainly it would be good to set up the lists Geir has discussed
> in the past and be clear on the policies so we can do that in an
> easier manner. Geir mentioned having that done by and discussing it
> at ApacheCon, so it is not far off. If for any reason there are
> delays, poke me and I'll confirm with the spec lead what is
> admissible to share.
Once he signs the NDA, you can share anything you wish with him.
It's not up to the spec lead.
geir
--
Geir Magnusson Jr +1-203-665-6437
geirm@apache.org
Re: About joining JSR 279 and 280
Posted by Brett Porter <br...@apache.org>.
Steve Loughran wrote:
> Geir Magnusson Jr. wrote:
>
>> Sorry about the delay - changed jobs...
>
>
> so we hear. couldnt take working for the company that gave us XML1.1 eh?
>
>>
>
>> (aside : I'm interested in how Brett is treating the OSGi issue, as
>> it's a very interesting political and IP minefield - from what I
>> hear, the dance is in full swing, as expected...)
>>
>
> I have to deal with some of the packaging bits from other standards, I'm
> interested in this too.
>
Steve - cool, good to hear! If you have any thoughts you'd like me to
forward on or keep in mind based on the initial description on the site
I'd be happy to do so. If I understand what you mean by packaging, then
it hasn't exactly been dealt with yet, so no rush.
As a member, I believe there should be no problem with discussing it
with you, though you may need to sign an NDA and I'll ask that Geir
clarify that first. I'm sure it's fine, but I want to make sure we do
follow all appropriate confidentiality rules within the EGs.
Certainly it would be good to set up the lists Geir has discussed in the
past and be clear on the policies so we can do that in an easier manner.
Geir mentioned having that done by and discussing it at ApacheCon, so it
is not far off. If for any reason there are delays, poke me and I'll
confirm with the spec lead what is admissible to share.
Cheers,
Brett
Re: About joining JSR 279 and 280
Posted by Steve Loughran <st...@apache.org>.
Geir Magnusson Jr. wrote:
> Sorry about the delay - changed jobs...
so we hear. couldnt take working for the company that gave us XML1.1 eh?
>
> (aside : I'm interested in how Brett is treating the OSGi issue, as
> it's a very interesting political and IP minefield - from what I hear,
> the dance is in full swing, as expected...)
>
I have to deal with some of the packaging bits from other standards, I'm
interested in this too.
Re: About joining JSR 279 and 280
Posted by Brett Porter <br...@apache.org>.
Geir Magnusson Jr. wrote:
> There no doubt that you are both seen as doing the right thing, as well
> as actually doing the right thing. I don't actually know what's going
> to happen w/ 277, but there be dragons. I'm on your side here,
> watching your back. My only vested interest here is being able to use
> Harmony as a lever for what we want to do.
I'm sure we'll get more of a chance to talk about this in person soon,
but I'm interested to hear more on this topic. I think I have an idea,
at least in part. But it's also suitably cryptic :)
> Agreed that I would be a corncob were I the EG rep given my time
> constraints, but we can't reject the model out of hand, as we may have
> to do the same thing with you as the conduit, and actually will for now
> (for example, Steve's interest...)
Absolutely - I'm more than happy to be guinea pig in this regard too.
> I see. That's our mission in the EC, and something we've continually
> worked on. However, it's delicate. The latest progress we've made
> JCP-wide was the so called "JCP 2.6" process document which actually
> stated encouragement for transparency and open-source techniques for
> doing expert groups. However, I don't think it will ever be mandated.
Thanks for the rest of these details - I've followed this from what was
available publically for a while and its good to get some more insight.
This'd be a good start on some content for other interested ASF reps.
> For fun, read this
>
> http://www.jcp.org/aboutJava/communityprocess/JSPA2.pdf
Heh, been there. I was being asked to sign it a while back :-/
Ok, looks like we've got some good things to move forward with here
between the content here and the original mails from Jason and Dain.
Cheers,
Brett
Re: About joining JSR 279 and 280
Posted by "Geir Magnusson Jr." <ge...@apache.org>.
On Nov 20, 2005, at 7:00 PM, Brett Porter wrote:
> Geir Magnusson Jr. wrote:
>> If there is someone that wants to do things, then there is no
>> need for placeholder. However, I'll certainly continue to do
>> that if we don't have volunteer. I do agree that I can do a
>> better job in publicizing, and will.
>
> Sounds good.
>
>> Brett, I have no doubt that you represent the ASF, and not Maven,
>> and currently there aren't any issues surrounding this. Just
>> know that the JCP is an *awfully* political place, so be prepared
>> for crap, especially since both you and Jason are such
>> significant members of the Maven project *and* significant
>> employees of Mergere. People always seem to jump to the worst of
>> a set of conclusions when there's doubt, so be aware.
>
> I understand. I had hoped the ASF would be a safehaven from that
> sort of crap, and I guess I misinterpreted what you were saying as
> "what some people might think", to what you actually thought.
Ah. No. I have absolutely no doubts, and I can assure that so far,
I have nothing at all negative from anyone. I'll let you know if I
hear anything. But there is some real and significant contention
surrounding 277. Be careful when elephants walk.
>
> It's important to me personally that not only am I seen to do the
> right thing by the ASF, but that I actually do that. In that way,
> hopefully, there can be no doubt when it comes to drawing conclusions.
There no doubt that you are both seen as doing the right thing, as
well as actually doing the right thing. I don't actually know what's
going to happen w/ 277, but there be dragons. I'm on your side here,
watching your back. My only vested interest here is being able to
use Harmony as a lever for what we want to do.
>
>> I certainly had an interest in participation, and still have that
>> interest. However, as VP of the JCP, I would never put myself
>> first though, and I'm actually insulted that you would think that.
> > ...
>> What was my intention?
>
> This read back too strongly. I felt that you had intended to put
> yourself as the EG representative, and let others go through you to
> communicate with the group. I never meant to imply you would stop
> others from participating deliberately. I'm just naturally
> concerned that if that was how things were going to operate, that
> it is not workable.
Agreed that I would be a corncob were I the EG rep given my time
constraints, but we can't reject the model out of hand, as we may
have to do the same thing with you as the conduit, and actually will
for now (for example, Steve's interest...)
>
> Thanks for clarifying it - I certainly take your word that it
> wasn't your intention, and I apologise for causing any insult.
Thank you, and I apologize for any frustration and stress that you
felt getting the situation resolved. It wasn't my intention at all.
>
>> Thanks - it should be no problem. I am excited about this
>> interest, and lookforward to moving forward and refining things.
>
> +1
>
>>> I appreciate the vote of confidence. I thought you were still on the
>>> list, at least in ability to read the messages?
>> I have to check. My understanding is that I was cut off, and that
>> it was all Brett, all the time :)
>
> If you aren't inundated with messages, then you were cut off:)
>
> They didn't seem to mind that as a first step, so maybe we can
> enable that again.
>
>>> I do think any future rep would need to be getting involved in
>>> advance, and able to at least read the list and discuss
>>> internally. I'm happy to follow up with the lead as the current
>>> rep if you think that is best. I think it might be better to see
>>> if we can do something JCP wide, if at all possible. Again, just
>>> let me know how I can help.
>> Oh, I think that it would be a horror show to switch reps :) I'm
>> really annoyed with Sun about not letting multiple people on, but
>> this is the prerogative of the spec lead.
>> Also, I don't understand. What do you mean "JCP wide"?
>
> As I understand it, there is a governing body for the JCP as a
> whole. I know Apache has done a lot of good work in opening that
> up. Maybe it is a topic we can continue to take up at that level
> rather than attacking each spec that comes along?
I see. That's our mission in the EC, and something we've continually
worked on. However, it's delicate. The latest progress we've made
JCP-wide was the so called "JCP 2.6" process document which actually
stated encouragement for transparency and open-source techniques for
doing expert groups. However, I don't think it will ever be
mandated. There are those in the process that will always want to
have control over EG size and scope, and wish to leave it up to the
individual spec leads. I can sorta sympathize, actually, because
unlike our style of OSS development, EG participants *are* companies,
and they want the ability to work in private for bunches of reasons,
including IP issues - that IP can be offered and withdrawn w/in the
working EG, and is only required to be licensed to implementors via
the final spec.
For fun, read this
http://www.jcp.org/aboutJava/communityprocess/JSPA2.pdf
These IP rules are actually very much an ongoing concern because
while the benefit to the community is obvious - that IP flows to
compatible implementations - it can and has a chilling effect on
contributions because some organizations are so scared of a
representative of theirs on an EG saying the wrong thing and thus
effectively giving away IP accidentally, they don't participate in
EGs in a meaningful way. So far, no one has thought of a good
solution to this problem.
There is also tremendous resistance to changing the JSPA, because the
last revision was so painful and slow. It's accepted as current
conventional wisdom that the JSPA won't be changing, and we have to
somehow find ways to operate in and around it.
>
>> Yep, although that one of my missions as EC rep is to help keep
>> pounding wedges into the JCP to open it up. I'd want to try to
>> get those multiple people on the EG as our first try if we had
>> that situation...
>
> Sure, let us know if you need more hammers.
The most effective set of hammers that I can think of are open source
led JSRs because they should not only show the JCP community that the
sky won't fall if they work this way, but also it takes on the subtle
incompatibilities between the JSPA and the commonly understood "open
source working model" in an actionable, material way. The general
problem with this is that it's an awful time sink and people don't
see the benefit.
The first problem with OSS-led JSRs is that it's hard to identify
projects that make sense as a JSR, because a lot of the things we do
are great innovative singleton implementations (e.g. Maven) but offer
no benefit to standardization for the purpose of multiple competing
implementations. (I mean, do you really want to see products which
are copies of Maven from BEA, IBM, Borland.... No - they are better
off just using maven and contributing to the community...) Sure,
there are aspects from these projects for which we could create JSRs
(like standardizing on repository structure from Maven, etc), but
your average volunteer doesn't wish to go spend the time leading a
JSR and locking down the activity of the OSS project when they could
be doing and creating more. Clearly you aren't the average volunteer
because you are doing this in 277, but we also didn't start this JSR,
and I think it's an interesting question of if we would have. What's
the upside for Maven? So far, Maven is ubiquitous....
The second problem is that lots of things we do innovate in the same
areas of the standard, generally because we've identified at a grass
roots level that the standard, well, sucks. Good examples of this
are things like Velocity and Spring, which effectively were responses
to developer usability issues in JSPs and J2EE respectively. The
problem you run into is that the entrenched interests dig in against
you when you suggest that a JSR is the way to go - we discovered that
first hand with Velocity when we tried to do a JSR, and lots of the
things learned in Spring will make it into J2EE (for example, EJB3
arguably has elements), but I don't think the JCP vendor interests
will ever want to confuse customers by offering what is effectively a
competitor to the J2EE programming model as a supported specification.
The only two JSRs that I can recall are JDOM and Groovy. JDOM just
petered out because Jason eventually so no upside in doing the JSR
work, I suppose, and Groovy is still going. We'll see.
The second set of hammers, and most economical, are open source
participation in JSRs - this exposes the JCP community to us - which
is a big help because it combats the perception that we're a bunch of
yahoo cowboys... - but unless we can be successful in influencing the
spec lead to change the working rules to that of OSS, our "community
activism" effect is limited, and we are "just another" technical
contributor. That too is still very important, but less effective
towards our higher-level goal of changing the JCP. It's slow :)
geir
>
> Cheers,
> Brett
--
Geir Magnusson Jr +1-203-665-6437
geirm@apache.org
Re: About joining JSR 279 and 280
Posted by Brett Porter <br...@apache.org>.
Geir Magnusson Jr. wrote:
> If there is someone that wants to do things, then there is no need for
> placeholder. However, I'll certainly continue to do that if we don't
> have volunteer. I do agree that I can do a better job in publicizing,
> and will.
Sounds good.
> Brett, I have no doubt that you represent the ASF, and not Maven, and
> currently there aren't any issues surrounding this. Just know that the
> JCP is an *awfully* political place, so be prepared for crap,
> especially since both you and Jason are such significant members of the
> Maven project *and* significant employees of Mergere. People always
> seem to jump to the worst of a set of conclusions when there's doubt,
> so be aware.
I understand. I had hoped the ASF would be a safehaven from that sort of
crap, and I guess I misinterpreted what you were saying as "what some
people might think", to what you actually thought.
It's important to me personally that not only am I seen to do the right
thing by the ASF, but that I actually do that. In that way, hopefully,
there can be no doubt when it comes to drawing conclusions.
> I certainly had an interest in participation, and still have that
> interest. However, as VP of the JCP, I would never put myself first
> though, and I'm actually insulted that you would think that.
>
> ...
>
> What was my intention?
This read back too strongly. I felt that you had intended to put
yourself as the EG representative, and let others go through you to
communicate with the group. I never meant to imply you would stop others
from participating deliberately. I'm just naturally concerned that if
that was how things were going to operate, that it is not workable.
Thanks for clarifying it - I certainly take your word that it wasn't
your intention, and I apologise for causing any insult.
> Thanks - it should be no problem. I am excited about this interest,
> and lookforward to moving forward and refining things.
+1
>> I appreciate the vote of confidence. I thought you were still on the
>> list, at least in ability to read the messages?
>
>
> I have to check. My understanding is that I was cut off, and that it
> was all Brett, all the time :)
If you aren't inundated with messages, then you were cut off:)
They didn't seem to mind that as a first step, so maybe we can enable
that again.
>> I do think any future rep would need to be getting involved in
>> advance, and able to at least read the list and discuss internally.
>> I'm happy to follow up with the lead as the current rep if you think
>> that is best. I think it might be better to see if we can do
>> something JCP wide, if at all possible. Again, just let me know how I
>> can help.
>
>
> Oh, I think that it would be a horror show to switch reps :) I'm
> really annoyed with Sun about not letting multiple people on, but this
> is the prerogative of the spec lead.
>
> Also, I don't understand. What do you mean "JCP wide"?
As I understand it, there is a governing body for the JCP as a whole. I
know Apache has done a lot of good work in opening that up. Maybe it is
a topic we can continue to take up at that level rather than attacking
each spec that comes along?
> Yep, although that one of my missions as EC rep is to help keep
> pounding wedges into the JCP to open it up. I'd want to try to get
> those multiple people on the EG as our first try if we had that
> situation...
Sure, let us know if you need more hammers.
Cheers,
Brett
Re: About joining JSR 279 and 280
Posted by "Geir Magnusson Jr." <ge...@apache.org>.
On Nov 21, 2005, at 7:49 PM, Dain Sundstrom wrote:
> On Nov 20, 2005, at 5:15 PM, Geir Magnusson Jr. wrote:
>
>> On Nov 20, 2005, at 5:40 PM, Brett Porter wrote:
>>
>>>> Sometimes we're invited to be a part, and I say yes and be the
>>>> placeholder to let things move forward, and when someone shows
>>>> up, that person becomes the rep. When someone indicates they
>>>> want to be a part of a JSR, I notify the JCP that the ASF will
>>>> be on the EG, and the person that indicated they wished to
>>>> participate is the rep. There's paperwork that people have to
>>>> go through, and me being the placeholder doesn't put any work
>>>> on me, but just lets the JCP wheels turn faster for us.
>>>
>>> Ok, this clearly isn't what happened with 277. If it is just a
>>> one off
>>> as you've suggested, then that's fine - I just want to make sure
>>> we're
>>> learning some lessons and are set up correctly so it doesn't
>>> happen again.
>>
>> 277 was, IMO, a one off, one we won't repeat.
>
> Another one off to mention is JEE 5. I have been asking to be on
> that since July and only since this discussion got going have I
> seen any progress. At the same time, I asked about EJB 3 since one
> of the co-spec leads asked me to join to replace our dormant
> representative Jeremy Boynes, and again I have only seen progress
> since these resent discussions got going. EJB 3 is a different
> case, since we had a rep that went dormant, but it is a similar
> problem because Geir is the only one that can make the change. My
> participation in EJB has become some sort of internal Sun issue,
> but I really don't know where it stands since all of that
> administrative communication is private.
There have been updates. I really don't want to hash this out here,
but I'll note that your attitude towards the ASF has changed in a
very positive way very recently (something on the order of 180 degrees).
geir
--
Geir Magnusson Jr +1-203-665-6437
geirm@apache.org
Re: About joining JSR 279 and 280
Posted by Dain Sundstrom <da...@iq80.com>.
On Nov 20, 2005, at 5:15 PM, Geir Magnusson Jr. wrote:
> On Nov 20, 2005, at 5:40 PM, Brett Porter wrote:
>
>>> Sometimes we're invited to be a part, and I say yes and be the
>>> placeholder to let things move forward, and when someone shows
>>> up, that person becomes the rep. When someone indicates they
>>> want to be a part of a JSR, I notify the JCP that the ASF will
>>> be on the EG, and the person that indicated they wished to
>>> participate is the rep. There's paperwork that people have to
>>> go through, and me being the placeholder doesn't put any work on
>>> me, but just lets the JCP wheels turn faster for us.
>>
>> Ok, this clearly isn't what happened with 277. If it is just a one
>> off
>> as you've suggested, then that's fine - I just want to make sure
>> we're
>> learning some lessons and are set up correctly so it doesn't
>> happen again.
>
> 277 was, IMO, a one off, one we won't repeat.
Another one off to mention is JEE 5. I have been asking to be on
that since July and only since this discussion got going have I seen
any progress. At the same time, I asked about EJB 3 since one of the
co-spec leads asked me to join to replace our dormant representative
Jeremy Boynes, and again I have only seen progress since these resent
discussions got going. EJB 3 is a different case, since we had a rep
that went dormant, but it is a similar problem because Geir is the
only one that can make the change. My participation in EJB has
become some sort of internal Sun issue, but I really don't know where
it stands since all of that administrative communication is private.
Regardless of any bad experiences any of us have had, I think it is
important for us to strive for democracy. Even if things were
working perfectly, I don't think this is the Apache way.
-dain
Re: About joining JSR 279 and 280
Posted by "Geir Magnusson Jr." <ge...@apache.org>.
On Nov 20, 2005, at 5:40 PM, Brett Porter wrote:
> Hi Geir,
>
> Geir Magnusson Jr. wrote:
>> Next time you take a mail from a private list to a public one,
>> please ask permission.
>
> This is not public, but less private (committers vs members).
I actually did set this up as a public list.
>
> Still, I agree its generally poor form to do so, though in this
> case no
> harm done as you've still kept the original quotes so there is
> obviously
> nothing sensitive in it.
It's over.
>
>> That's not the issue. It's not about overhead, but being sure
>> that people aren't independently claiming to represent the ASF.
>
> I don't think what Jason proposed is any different. Interested
> individual arrives on this list, everyone agrees/another person also
> volunteers to represent and we sort it out here in a semi-public way.
> End result: someone that the ASF chooses to have represent applies to
> the EG (as they have to do this anyway), and with an introductory mail
> from the current VP of JCP (ie, you) to the spec lead so that they
> know
> that it is who the ASF wants.
>
> Doing the "placeholder representative" thing has been problematic on
> more than one occasion, and is not done globally now anyway. Unless we
> get blanket permission to have multiple participants, it will continue
> to be problematic, I think.
If there is someone that wants to do things, then there is no need
for placeholder. However, I'll certainly continue to do that if we
don't have volunteer. I do agree that I can do a better job in
publicizing, and will.
>
>> Sometimes we're invited to be a part, and I say yes and be the
>> placeholder to let things move forward, and when someone shows
>> up, that person becomes the rep. When someone indicates they
>> want to be a part of a JSR, I notify the JCP that the ASF will
>> be on the EG, and the person that indicated they wished to
>> participate is the rep. There's paperwork that people have to go
>> through, and me being the placeholder doesn't put any work on me,
>> but just lets the JCP wheels turn faster for us.
>
> Ok, this clearly isn't what happened with 277. If it is just a one off
> as you've suggested, then that's fine - I just want to make sure we're
> learning some lessons and are set up correctly so it doesn't happen
> again.
277 was, IMO, a one off, one we won't repeat.
>
>>> C'mon Geir, that is not how it worked with JSR277. Brett was the
>>> ideal
>>> candidate for the lead and it literally took 6 weeks for us to
>>> get you
>>> to agree to let Brett have the official lead position. You were
>>> seen as
>>> the lead so it was incredibly difficult for Brett to participate.
>>> This
>>> is something that I would like to prevent in the future.
>> I think you are oversimplifying.
>> We were trying to get multiple ASF reps, and that took a while.
>> Then we flipped it over to Brett when that failed. At no time
>> did I ever try to stop or inhibit Brett's participation. And 277
>> is an interesting JSR because as far as I can tell, it is unique
>> in that it covers at least 3 projects here at the ASF- Maven,
>> Felix and Harmony. Some might even argue that Maven is over-
>> represented because you are on the EG independently, and there
>> are other projects here that could be interested (now that
>> Harmony has classlibrary code activity, it is a candidate...)
>
> I think this is completely unfair. I am not on the group
> representing Maven, I'm representing the ASF. I follow the Felix and
> Harmony lists to ensure I know what is going on there.
Brett, I have no doubt that you represent the ASF, and not Maven, and
currently there aren't any issues surrounding this. Just know that
the JCP is an *awfully* political place, so be prepared for crap,
especially since both you and Jason are such significant members of
the Maven project *and* significant employees of Mergere. People
always seem to jump to the worst of a set of conclusions when there's
doubt, so be aware.
Also, it's _theoretically_ possible that we get into a situation
where the technical interests of project A and project B are
orthogonal (I'm not picking on any project here...). It will be
interesting to deal with that if it ever comes up.
>
> Doug Lea is also a listed participant in Harmony, and a member of
> the EG. Richard Hall is a participant in Felix, and a member of the
> EG.
>
> On the other hand, Jason and I have a lot of different views about
> Maven
> itself and worked on different areas within it that are both relevant.
>
> I don't believe Maven is over-represented, and besides - that would be
> the spec lead's call and he certainly knew of both Jason and I's
> affiliation with the project.
if the spec lead came back and said that you couldn't be because of
Maven, I'd fight that. We get to choose who the spec lead is. It's
you :)
>
> I also don't think Jason was over simplifying. Quite frankly, I'm
> a bit annoyed at the way you are trying to represent it and would
> like a
> chance to explain how I saw it play out.
Fair enough.
>
> Looking back now, I'd put to you that being interested in the JSR
> yourself, you decided that you would be the representative of the
> ASF. That may be your perogative as the VP of JCP - I'm not sure -
> but I don't think it is how things should work. You yourself said
> it is about "being sure that people aren't independently claiming
> to represent the ASF". If I'm wrong, please correct me.
I certainly had an interest in participation, and still have that
interest. However, as VP of the JCP, I would never put myself first
though, and I'm actually insulted that you would think that.
>
> To recap, I contacted you about this very early - before the thread
> on Harmony (and this is before Felix/Oscar existed), and just after
> I had posted to the Maven list about it. I suggested Apache should
> be involved, because at that point, Apache was not listed as a
> supporting member, and in particular as Maven had been mentioned in
> the proposal and we had some interested folk. You told me that you
> had applied to the EG. You also later replied to harmony-dev on the
> topic:
> http://mail-archives.apache.org/mod_mbox/incubator-harmony-dev/
> 200506.mbox/%3c2952AFF9-E64C-45DE-B175-E49056094E20@apache.org%3e
> saying "I'm on the EG".
>
> You did separately clarify in both instances that it was on behalf of
> the ASF, and on harmony-dev said that I had also expressed
> interest, but I now think now that this gave away your intention.
What was my intention?
>
> When it originally came up, Jason took your advice and applied as
> an individual. As far as I know, after that I'm the only person who
> put their hand up to represent Apache on the EG other than
> yourself. Yet a month later at the end of August (over two months
> since the process started), I got no replies to my requests for
> information about where we were at, the group had started with you
> on it, and there was no way for anyone to participate.
>
> It took a week to find you on IRC and get any info on the progress
> (which was about 50 messages you were able to drop to me), and another
> week for you to attempt to get multiple representatives on the list
> (15 Sep). It took two weeks for me to catch up and I started
> sending you
> messages to forward. It was only then that you pursued making me
> rep, on the 1st of Oct. I was all set up by 4th of Oct.
>
> I was very patient at the time, and I understand you had more
> important things to deal with in your personal life especially over
> those last couple of weeks.
>
> However, I'm offended that you'd now say that you didn't inhibit my
> participation, and blame Sun for things that took only the last 2
> weeks of nearly 4 months of set up.
Let me put it this way - I had no intention of inhibiting your
participation. You can believe me or not. I will certainly go back
and review the data. I don't recall this timing, because I know
that I logged your NDA for participation in early september... But I
will go review.
>
> If I've missed anything, please let me know. I was obviously very
> frustrated by the way this panned out and felt like I was being
> kept in the dark. If I wasn't so inclined to participate, I would
> have given up long before I got a chance to. I would really like to
> be set straight if that has clouded my view, but it concerns me
> because it really doesn't fit with how Apache is supposed to operate.
I really am upset at the idea that you think that I was in some way
personally motivated to do this to you or anyone - at least that is
what I am reading into this. The intention is to open things up and
include people. I'm really sorry too that you decided to do this on
an open and public list, but you certainly have your reasons, I suppose.
I will go review my emails.
>
> With all that said, I'm more than happy to put it behind us and get
> working together on both this JSR and the JCP in general.
Yes, I think that would be far more productive.
>
> I'd really like to see the jsr lists set up for discussion, and the
> policies and representatives posted on a website as discussed. I'll
> again offer my help under your direction on this - I can put other
> things aside if necessary.
Thanks - it should be no problem. I am excited about this interest,
and lookforward to moving forward and refining things.
>
> Let's also clarify the meaning of the two jcp lists and rename them
> if necessary.
>
>> (aside : I'm interested in how Brett is treating the OSGi issue,
>> as it's a very interesting political and IP minefield - from what
>> I hear, the dance is in full swing, as expected...)
>
> Yep, I'm treating it as the EG has been asked to treat it. From
> Apache's end, Felix is very new and has its work cut out
> implementing the R4 spec
> itself (I'm not on the PPMC, so I don't know how they are dealing with
> that minefield, if at all, but certainly the incubator will sort that
> out). I doubt it is particularly relevant to Apache's representation
> other than to ensure that Felix is not disadvantaged in any way,
> which I
> don't see happening. I'm sure that Richard would pick up on anything I
> might happen to miss in that regard.
>
>> I disagree, because there is no downside to the current process.
>> With one exception, 277 has been the only JSR in which there's
>> been more than one person interested in being the rep, and what I
>> was trying for, if you remember, was multi-representation - to
>> have more people involved. It didn't work, and it didn't work
>> fast (as is the way with Sun), and for now, Brett is the rep, and
>> I'm sure he's doing a fine job.
>
> I appreciate the vote of confidence. I thought you were still on the
> list, at least in ability to read the messages?
I have to check. My understanding is that I was cut off, and that it
was all Brett, all the time :)
>
>> There are more groups at the ASF interested in 277, so it's
>> probable that the rep will rotate or be shared in the future, and
>> I don't think that we need or even can sometimes work all of that
>> ahead of time either.
>
> I agree on this. Though, so far, I haven't seen anyone with more
> than a
> cursory interest.
>
> I'll also be the first to put my hand up if time constraints don't
> permit me from working effectively on it.
>
> I do think any future rep would need to be getting involved in
> advance, and able to at least read the list and discuss internally.
> I'm happy to follow up with the lead as the current rep if you
> think that is best. I think it might be better to see if we can do
> something JCP wide, if at all possible. Again, just let me know how
> I can help.
Oh, I think that it would be a horror show to switch reps :) I'm
really annoyed with Sun about not letting multiple people on, but
this is the prerogative of the spec lead.
Also, I don't understand. What do you mean "JCP wide"?
>
>> We get into the JSR as soon as we can w/ or w/o a person that's
>> interested. If someone comes along later, they get to do it
>> unless there's contention, and then I'd suggest those that are
>> interested in doing the work sort it out.
>
> I agree if its the case of someone coming along later that's
> interested.
> When one or more people are interested from the start, then we should
> sort it out then.
Yep, although that one of my missions as EC rep is to help keep
pounding wedges into the JCP to open it up. I'd want to try to get
those multiple people on the EG as our first try if we had that
situation...
>
>> Remember, the ASF participating in the JCP is a interesting
>> political balance. Our interests at the highest level is about
>> continuing to ensure transparency and openness are part of the
>> JCP processes. The ASF as an organization doesn't really have a
>> technical agenda, but so far we've been able to have technical
>> participation by domain experts from projects without any
>> problems. That said, every now and then we get an oddball.
>
> I think the situation of 277 will become less odd as we get more
> projects. A recent example - it looks imminent that Geronimo will
> accept
> a number of incubated subprojects in the EE space. This will surely
> mean
> more projects interested in related JSRs. I'm sure the same would
> be the
> case with anything to do with Harmony as it spans the whole of
> JavaSE, and will happen in and with other projects. I'd hate for
> the same experience to befall anyone else.
The more the merrier.
>
>> The ASF can't be seen as a way to get people onto JSRs by
>> companies (to overload representation), and sometimes we may have
>> conflicting or multiple technical agendas as is possible with 277.
>
> I agree. Hopefully we are being crystal clear about how members of the
> ASF participate as individuals affiliated with the ASF. We love our
> hats :)
It's actually very hard, and I've discovered that people tend to jump
to the worst conclusions about people when there's even a possibility
of multiple hats... ;)
>
>> In foundations/JCP and probably out of date. That's something
>> I'm going to put on the upcoming JCP website, as there is no
>> problem with that being public information. It's just there for
>> historical reasons.
>
> Ok, looking forward to working on this with you all.
>
> Thanks,
> Brett
--
Geir Magnusson Jr +1-203-665-6437
geirm@apache.org
Re: About joining JSR 279 and 280
Posted by Brett Porter <br...@apache.org>.
Hi Geir,
Geir Magnusson Jr. wrote:
> Next time you take a mail from a private list to a public one, please
> ask permission.
This is not public, but less private (committers vs members).
Still, I agree its generally poor form to do so, though in this case no
harm done as you've still kept the original quotes so there is obviously
nothing sensitive in it.
> That's not the issue. It's not about overhead, but being sure that
> people aren't independently claiming to represent the ASF.
I don't think what Jason proposed is any different. Interested
individual arrives on this list, everyone agrees/another person also
volunteers to represent and we sort it out here in a semi-public way.
End result: someone that the ASF chooses to have represent applies to
the EG (as they have to do this anyway), and with an introductory mail
from the current VP of JCP (ie, you) to the spec lead so that they know
that it is who the ASF wants.
Doing the "placeholder representative" thing has been problematic on
more than one occasion, and is not done globally now anyway. Unless we
get blanket permission to have multiple participants, it will continue
to be problematic, I think.
> Sometimes we're invited to be a part, and I say yes and be the
> placeholder to let things move forward, and when someone shows up, that
> person becomes the rep. When someone indicates they want to be a part
> of a JSR, I notify the JCP that the ASF will be on the EG, and the
> person that indicated they wished to participate is the rep. There's
> paperwork that people have to go through, and me being the placeholder
> doesn't put any work on me, but just lets the JCP wheels turn faster
> for us.
Ok, this clearly isn't what happened with 277. If it is just a one off
as you've suggested, then that's fine - I just want to make sure we're
learning some lessons and are set up correctly so it doesn't happen again.
>> C'mon Geir, that is not how it worked with JSR277. Brett was the ideal
>> candidate for the lead and it literally took 6 weeks for us to get you
>> to agree to let Brett have the official lead position. You were seen as
>> the lead so it was incredibly difficult for Brett to participate. This
>> is something that I would like to prevent in the future.
>
>
> I think you are oversimplifying.
>
> We were trying to get multiple ASF reps, and that took a while. Then
> we flipped it over to Brett when that failed. At no time did I ever
> try to stop or inhibit Brett's participation. And 277 is an
> interesting JSR because as far as I can tell, it is unique in that it
> covers at least 3 projects here at the ASF- Maven, Felix and Harmony.
> Some might even argue that Maven is over-represented because you are on
> the EG independently, and there are other projects here that could be
> interested (now that Harmony has classlibrary code activity, it is a
> candidate...)
I think this is completely unfair. I am not on the group
representing Maven, I'm representing the ASF. I follow the Felix and
Harmony lists to ensure I know what is going on there.
Doug Lea is also a listed participant in Harmony, and a member of the
EG. Richard Hall is a participant in Felix, and a member of the EG.
On the other hand, Jason and I have a lot of different views about Maven
itself and worked on different areas within it that are both relevant.
I don't believe Maven is over-represented, and besides - that would be
the spec lead's call and he certainly knew of both Jason and I's
affiliation with the project.
I also don't think Jason was over simplifying. Quite frankly, I'm
a bit annoyed at the way you are trying to represent it and would like a
chance to explain how I saw it play out.
Looking back now, I'd put to you that being interested in the JSR
yourself, you decided that you would be the representative of the ASF.
That may be your perogative as the VP of JCP - I'm not sure - but I
don't think it is how things should work. You yourself said it is about
"being sure that people aren't independently claiming to represent the
ASF". If I'm wrong, please correct me.
To recap, I contacted you about this very early - before the thread on
Harmony (and this is before Felix/Oscar existed), and just after I had
posted to the Maven list about it. I suggested Apache should be
involved, because at that point, Apache was not listed as a supporting
member, and in particular as Maven had been mentioned in the proposal
and we had some interested folk. You told me that you had applied to the
EG. You also later replied to harmony-dev on the topic:
http://mail-archives.apache.org/mod_mbox/incubator-harmony-dev/200506.mbox/%3c2952AFF9-E64C-45DE-B175-E49056094E20@apache.org%3e
saying "I'm on the EG".
You did separately clarify in both instances that it was on behalf of
the ASF, and on harmony-dev said that I had also expressed interest, but
I now think now that this gave away your intention.
When it originally came up, Jason took your advice and applied as an
individual. As far as I know, after that I'm the only person who put
their hand up to represent Apache on the EG other than yourself. Yet a
month later at the end of August (over two months since the process
started), I got no replies to my requests for information about where we
were at, the group had started with you on it, and there was no way for
anyone to participate.
It took a week to find you on IRC and get any info on the progress
(which was about 50 messages you were able to drop to me), and another
week for you to attempt to get multiple representatives on the list
(15 Sep). It took two weeks for me to catch up and I started sending you
messages to forward. It was only then that you pursued making me rep, on
the 1st of Oct. I was all set up by 4th of Oct.
I was very patient at the time, and I understand you had more important
things to deal with in your personal life especially over those last
couple of weeks.
However, I'm offended that you'd now say that you didn't inhibit my
participation, and blame Sun for things that took only the last 2 weeks
of nearly 4 months of set up.
If I've missed anything, please let me know. I was obviously very
frustrated by the way this panned out and felt like I was being kept in
the dark. If I wasn't so inclined to participate, I would have given up
long before I got a chance to. I would really like to be set straight if
that has clouded my view, but it concerns me because it really doesn't
fit with how Apache is supposed to operate.
With all that said, I'm more than happy to put it behind us and get
working together on both this JSR and the JCP in general.
I'd really like to see the jsr lists set up for discussion, and the
policies and representatives posted on a website as discussed. I'll
again offer my help under your direction on this - I can put other
things aside if necessary.
Let's also clarify the meaning of the two jcp lists and rename them if
necessary.
> (aside : I'm interested in how Brett is treating the OSGi issue, as
> it's a very interesting political and IP minefield - from what I hear,
> the dance is in full swing, as expected...)
Yep, I'm treating it as the EG has been asked to treat it. From Apache's
end, Felix is very new and has its work cut out implementing the R4 spec
itself (I'm not on the PPMC, so I don't know how they are dealing with
that minefield, if at all, but certainly the incubator will sort that
out). I doubt it is particularly relevant to Apache's representation
other than to ensure that Felix is not disadvantaged in any way, which I
don't see happening. I'm sure that Richard would pick up on anything I
might happen to miss in that regard.
> I disagree, because there is no downside to the current process. With
> one exception, 277 has been the only JSR in which there's been more
> than one person interested in being the rep, and what I was trying for,
> if you remember, was multi-representation - to have more people
> involved. It didn't work, and it didn't work fast (as is the way with
> Sun), and for now, Brett is the rep, and I'm sure he's doing a fine job.
I appreciate the vote of confidence. I thought you were still on the
list, at least in ability to read the messages?
> There are more groups at the ASF interested in 277, so it's probable
> that the rep will rotate or be shared in the future, and I don't think
> that we need or even can sometimes work all of that ahead of time either.
I agree on this. Though, so far, I haven't seen anyone with more than a
cursory interest.
I'll also be the first to put my hand up if time constraints don't
permit me from working effectively on it.
I do think any future rep would need to be getting involved in advance,
and able to at least read the list and discuss internally. I'm happy to
follow up with the lead as the current rep if you think that is best. I
think it might be better to see if we can do something JCP wide, if at
all possible. Again, just let me know how I can help.
> We get into the JSR as soon as we can w/ or w/o a person that's
> interested. If someone comes along later, they get to do it unless
> there's contention, and then I'd suggest those that are interested in
> doing the work sort it out.
I agree if its the case of someone coming along later that's interested.
When one or more people are interested from the start, then we should
sort it out then.
> Remember, the ASF participating in the JCP is a interesting political
> balance. Our interests at the highest level is about continuing to
> ensure transparency and openness are part of the JCP processes. The
> ASF as an organization doesn't really have a technical agenda, but so
> far we've been able to have technical participation by domain experts
> from projects without any problems. That said, every now and then we
> get an oddball.
I think the situation of 277 will become less odd as we get more
projects. A recent example - it looks imminent that Geronimo will accept
a number of incubated subprojects in the EE space. This will surely mean
more projects interested in related JSRs. I'm sure the same would be the
case with anything to do with Harmony as it spans the whole of JavaSE,
and will happen in and with other projects. I'd hate for the same
experience to befall anyone else.
> The ASF can't be seen as a way to get people onto JSRs
> by companies (to overload representation), and sometimes we may have
> conflicting or multiple technical agendas as is possible with 277.
I agree. Hopefully we are being crystal clear about how members of the
ASF participate as individuals affiliated with the ASF. We love our hats :)
> In foundations/JCP and probably out of date. That's something I'm
> going to put on the upcoming JCP website, as there is no problem with
> that being public information. It's just there for historical reasons.
Ok, looking forward to working on this with you all.
Thanks,
Brett
Re: About joining JSR 279 and 280
Posted by "Geir Magnusson Jr." <ge...@apache.org>.
Sorry about the delay - changed jobs...
On Nov 7, 2005, at 2:36 PM, Jason van Zyl wrote:
> Hi, I am responding to this on the JCP open list so others can
> chime in.
Next time you take a mail from a private list to a public one, please
ask permission.
>
> On Wed, 2005-10-26 at 19:20 -0400, Geir Magnusson Jr. wrote:
>
>> On Oct 26, 2005, at 6:39 PM, Jason van Zyl wrote:
>>
>>
>>> On Wed, 2005-10-26 at 13:58 -0400, Geir Magnusson Jr. wrote:
>>>
>>>
>>>> Yes, I applied to be the placeholder so Changshin could be on
>>>> the EG.
>>>>
>>>> I was waiting for his response to engage on this.
>>>>
>>>>
>>>
>>> How did this practice of one person applying for everything
>>> happen? I
>>> think this is a bad practice and know from first hand experience on
>>> JSR277 that this method does not scale. How many JSRs are you
>>> involved
>>> with now? I can barely keep up with one. I think this process
>>> needs to
>>> go distributed.
>>>
>>
>> Jason,
>>
>> I've tried to explain this before. I'm not involved in these JSRs.
>> I'm applying for membership in the EG on behalf of the ASF.
>>
>
> Ok, this is a practice I don't think really fits with the ASF.
Sure it does. I'm the rep from the ASF to the JCP organization, and
I do this on behalf of the ASF, not myself.
> We should
> decide on the lists who will participate and let that person do it. We
> should vote like we do on all the other projects and then help that
> person become part of the JSR.
These aren't mutually exclusive, as we can easily switch.
I think you are making a mountain out of a molehill here. This
hasn't stopped anyone who wants to participate. I work hard to get
anyone who wants to get involved.
>
>
>> In this case, Changshin was identified as wanting to join the EG from
>> the get go, and hasn't engaged, so they used me as a placeholder. I
>> have no personal interest in these technologies and am just getting
>> things going on behalf of the ASF.
>>
>
> I think this just creates unnecessary overhead for you. Again,
> coordination on the list, a simple vote, let the person move
> forward. If
> that person doesn't have enough get-up-and-go to organize
> themselves to
> sync up with the spec lead then do we want them representing Apache?
>
That's not the issue. It's not about overhead, but being sure that
people aren't independently claiming to represent the ASF.
Sometimes we're invited to be a part, and I say yes and be the
placeholder to let things move forward, and when someone shows up,
that person becomes the rep. When someone indicates they want to be
a part of a JSR, I notify the JCP that the ASF will be on the EG,
and the person that indicated they wished to participate is the rep.
There's paperwork that people have to go through, and me being the
placeholder doesn't put any work on me, but just lets the JCP wheels
turn faster for us.
>
>>>
>>> I think people who want to be involved should be interacting with
>>> the
>>> spec leads directly and getting themselves setup. How did this ever
>>> come
>>> to be that one person at Apache is single point of entry for dealing
>>> with JSRs and is it really of any benefit to Apache?
>>>
>>
>> I think so because it gives us oversight and accountability to who is
>> representing the ASF on these EGs.
>>
>
> That is definitely not going to work with one person doing this. This
> should be done on this list and discussed and we as a group should
> provide oversight, not one person.
I have no problem with that, and I'm more than happy to have more
people involved. Until now, there has been little interest. My
goals as VP JCP have always been to continue to open things and get
more participation w/in the ASF community. Now that there's some
interest, great.
>
> We did not vote on JSR277 because you were probably dealing with many
> other issues at the time. This list should be the safety net.
No. It was a simple mistake.
>
>
>> Nothing stops people from joining
>> EGs as individuals, but if you want to be representing the ASF, then
>> it's the ASF that is joining, and we need to be sure that we know who
>> is doing what, and why.
>>
>
> So why would that be hindered by acting as a group and deciding in the
> standard ASF fashion who will participate in what?
>
I suspect that people won't care as a group, and that we're happy to
have people volunteer. This is all around a singular instance with
JSR-277 which is exceptional, IMO. With 279 and 280, the subject of
the thread, do you want to challenge that Ias is doing it? You are
welcome to, and I'll take any concerns of yours seriously. I didn't
"decide" to place Ias there - he asked, I jumped, got us hooked in,
and he didn't follow up immediately. He then did, he's on, and
that's that.
>
>> There is great potential for misuse of this
>> privilege that we (the ASF) have, and there is potential for damage
>> to our reputation.
>>
>
> I think things like not voting on important JSRs is an example of
> something damaging our reputation. It's not a question of you not
> trying
> to do the right thing, you just can't do it all and it should be
> distributed across the people wanting to participate.
Jason, you are talking one outlier example and trying to turn it into
something it isn't. The ASF has an excellent voting and attendance
record in the JCP. We were the "JCP Member of the Year" this year,
remember? Yes, I missed vote for the one JSR that you decided you
are interested in, but that doesn't indicate a pattern of any sort,
and it was followed up with a note indicating that "yes" was our vote
to make our intention clear.
>
>
>>>
>>> Just write a doc on how people can apply themselves, make an initial
>>> introduction and that's it. Was there a problem with this before?
>>> Does
>>> SUN only want to deal with one person?
>>>
>>
>> That is actually what happens. I apply on behalf of the ASF as the
>> official point of contact, but thats where it ends. The person that
>> wants to participate on the EG does all the work - they have the
>> voice on the EG, they get the mail.
>>
>
> C'mon Geir, that is not how it worked with JSR277. Brett was the ideal
> candidate for the lead and it literally took 6 weeks for us to get you
> to agree to let Brett have the official lead position. You were
> seen as
> the lead so it was incredibly difficult for Brett to participate. This
> is something that I would like to prevent in the future.
I think you are oversimplifying.
We were trying to get multiple ASF reps, and that took a while. Then
we flipped it over to Brett when that failed. At no time did I ever
try to stop or inhibit Brett's participation. And 277 is an
interesting JSR because as far as I can tell, it is unique in that it
covers at least 3 projects here at the ASF- Maven, Felix and
Harmony. Some might even argue that Maven is over-represented
because you are on the EG independently, and there are other projects
here that could be interested (now that Harmony has classlibrary code
activity, it is a candidate...)
(aside : I'm interested in how Brett is treating the OSGi issue, as
it's a very interesting political and IP minefield - from what I
hear, the dance is in full swing, as expected...)
>
> I think that:
>
> 1) No JSR should be applied for until it has been decided on this list
> who is going to represent Apache.
>
I disagree, because there is no downside to the current process.
With one exception, 277 has been the only JSR in which there's been
more than one person interested in being the rep, and what I was
trying for, if you remember, was multi-representation - to have more
people involved. It didn't work, and it didn't work fast (as is the
way with Sun), and for now, Brett is the rep, and I'm sure he's doing
a fine job.
There are more groups at the ASF interested in 277, so it's probable
that the rep will rotate or be shared in the future, and I don't
think that we need or even can sometimes work all of that ahead of
time either.
We get into the JSR as soon as we can w/ or w/o a person that's
interested. If someone comes along later, they get to do it unless
there's contention, and then I'd suggest those that are interested in
doing the work sort it out.
> 2) When the person is selected via the standard voting practices then
> that person contacts the spec lead. Again if this person doesn't have
> the motivation to contact the spec lead and get things moving then we
> should not participate.
>
> 3) I think it would be fair say that one individual cannot possibly
> handle more then one JSR. I can barely keep up with JSR 277.
>
I don't agree here. If an individual chooses to do so, they can. If
we can get multiple people participating, that should be a goal for
us. If we have a situation where the spec lead refuses to allow that
(like 277 did), then we have to reconsider and take participation
factors into account to ensure that people have a chance.
Remember, the ASF participating in the JCP is a interesting political
balance. Our interests at the highest level is about continuing to
ensure transparency and openness are part of the JCP processes. The
ASF as an organization doesn't really have a technical agenda, but so
far we've been able to have technical participation by domain experts
from projects without any problems. That said, every now and then we
get an oddball. The ASF can't be seen as a way to get people onto
JSRs by companies (to overload representation), and sometimes we may
have conflicting or multiple technical agendas as is possible with 277.
>
>> A doc is forthccoming.
>>
>
> Cool, have you managed to squeeze this out somewhere?
>
> Also, where is the current list of who at Apache is representing us
> each
> of the JSRs. Is there something in SVN?
In foundations/JCP and probably out of date. That's something I'm
going to put on the upcoming JCP website, as there is no problem with
that being public information. It's just there for historical reasons.
geir
>
>
>> geir
>>
>>
> --
> jvz.
>
> Jason van Zyl
> jason at maven.org
> http://maven.apache.org
>
> Our achievements speak for themselves. What we have to keep track
> of are our failures, discouragements and doubts. We tend to forget
> the past difficulties, the many false starts, and the painful
> groping. We see our past achievements as the end result of a
> clean forward thrust, and our present difficulties as
> signs of decline and decay.
>
> -- Eric Hoffer, Reflections on the Human Condition
>
>
--
Geir Magnusson Jr +1-203-665-6437
geirm@apache.org