RAID: we are not supermen...(was: RE: Cocoon and Axis)

[marc fleury <ma...@jboss.org>]

|> Whee!  A fun philosophical debate!  ;-)
|
|I hear you--and am not shy in these things ;P

This is as applied as it gets, management of development is what Open Source
does.
I remember the head of Xerox PARC saying about Open Source that we puppies
didn't invent anything in the free-love department of sharing ideas and
bodily fluids but what we did do well was "cracking the complexity" which
leads me to the point below.


|active developer: Federico Barbieri.  What got the Avalon

Federico is a great guy but Avalon is superseeded by JMX imnsho.

But that is another thread.

|is actually doing well today.  You need some dynamic folks who are vocal
|and evangelize your project.

absolutely.

|I feel sometimes that my questions, input, and code fall on deaf ears in
|this community.  There are a handful of active committers, and each of them
|are hard at work on the respective pieces.  The frustration comes
|when there
|is a _useful_ patch that can help many people, and it is given lip service
|and never committed.
|
|Part of why Cocoon and Avalon are relatively healthy communities is that
|there is a responsiveness to ideas and patches.  Basically, first code
|wins--and not alot of time is spent "re-engineering" the patch.
|When people
|get ignored (in their perception), they figure they will go someplace else.
|Personally, I am stubborn and opinionated--but I do listen to other people.
|In a sense, I am not satisfied that I am wrong unless someone can give an
|intelligent reason why--and what would be BETTER.  I have
|publically conceded
|when I was in the wrong.


Ok, good analysis, wrong point.

The point here is that **I** don't even follow the patches on JBoss any
longer, I just don't have the time, I just don't have the knowledge, I just
don't stay on top of all the development in JBoss, it is **humanly
impossible**.  See the point above from PARC, it is not about selecting an
elitist few (although we do that very well as well, just not in RW/commit
rights) but creating an intelligent RAID that can take on the most complex
infrastructure development... something akin to the net, if some die, or are
arrested by the Men In Black, the "metacreation" as the RAID brain goes
on... right now it doesn't here in axis, you are vulnerable to people
dissapearing and you think this is the problem you need to solve? no! that
is a *given* people just don't care and will go, so you need to build a RAID
of people to make the project work.  I am not superman, Scott (CTO) is not
superman, even Rickard is not superman, none of us are, all of us admit to
public booboos but the RAID (Redundant Array of Independent Developers) is
superman, it is in fact "supra-human".  yes?

It seems you have enough interest (including us ;-) so time to distribute
those RW.  Watch the wave grow, manage the wave (fire people if they fuck up
too much, but remember the nodes fail and poopoo *it is normal*)

|To achieve this, the developers you have need to be responsive, and include
|new developers and users into the process of developing the
|project.  Instead

precisely.

marcf

"The problem with our design is
that the center did not hold!"
-- Ed Harris, The Stand --

Re: RAID: we are not supermen...(was: RE: Cocoon and Axis)

[Berin Loritsch <bl...@apache.org>]

marc fleury wrote:
> 
> |active developer: Federico Barbieri.  What got the Avalon
> 
> Federico is a great guy but Avalon is superseeded by JMX imnsho.

That is because you don't have the concept of what Avalon is, but
if you want to discuss this further, lets take it off the list.

> |I feel sometimes that my questions, input, and code fall on deaf ears in
> |this community.  There are a handful of active committers, and each of them
> |are hard at work on the respective pieces.  The frustration comes
> |when there
> |is a _useful_ patch that can help many people, and it is given lip service
> |and never committed.
> |
> |Part of why Cocoon and Avalon are relatively healthy communities is that
> |there is a responsiveness to ideas and patches.  Basically, first code
> |wins--and not alot of time is spent "re-engineering" the patch.
> |When people
> |get ignored (in their perception), they figure they will go someplace else.
> |Personally, I am stubborn and opinionated--but I do listen to other people.
> |In a sense, I am not satisfied that I am wrong unless someone can give an
> |intelligent reason why--and what would be BETTER.  I have
> |publically conceded
> |when I was in the wrong.
> 
> Ok, good analysis, wrong point.
> 
> The point here is that **I** don't even follow the patches on JBoss any
> longer, I just don't have the time, I just don't have the knowledge, I just
> don't stay on top of all the development in JBoss, it is **humanly
> impossible**.  See the point above from PARC, it is not about selecting an
> elitist few (although we do that very well as well, just not in RW/commit
> rights) but creating an intelligent RAID that can take on the most complex
> infrastructure development... something akin to the net, if some die, or are
> arrested by the Men In Black, the "metacreation" as the RAID brain goes
> on... right now it doesn't here in axis, you are vulnerable to people
> dissapearing and you think this is the problem you need to solve? no! that
> is a *given* people just don't care and will go, so you need to build a RAID
> of people to make the project work.  I am not superman, Scott (CTO) is not
> superman, even Rickard is not superman, none of us are, all of us admit to
> public booboos but the RAID (Redundant Array of Independent Developers) is
> superman, it is in fact "supra-human".  yes?

I agree with your point--you missed mine.  You can't have a thriving community
when people who want to help CAN'T.  No one is superhuman.  That is what a
community is for.  Someone will know the answer to a question--they don't
even have to be a committer.  But for patches to be applied, the commetters
do need to stay on top of it.

For instance, Avalon has about a dozen committers--so far only a few actively
work on it.  Many times, something I start somebody else finishes.  Avalon is
5 subprojects--you can't keep track of all of them.  I am intimately familiar
with Framework and mostly familiar with Excalibur.  I have a rough idea on
LogKit, but am mostly in the dark regarding Cornerstone and Phoenix.  On the
other hand we have developers who focus on different aspects of the entire
project as a whole.  Somebody somewhere will know the answer.

The biggest thing that I brought to that project was once we had consensus,
I made us put it in writing.  If anyone had problems with it later, we had
a known base to work from.  Hopefully I can bring the same thing to this
project.

---------------------------------------------------------------------
To unsubscribe, e-mail: cocoon-dev-unsubscribe@xml.apache.org
For additional commands, email: cocoon-dev-help@xml.apache.org

Re: RAID: we are not supermen...(was: RE: Cocoon and Axis)

[Berin Loritsch <bl...@apache.org>]

marc fleury wrote:
> 
> |active developer: Federico Barbieri.  What got the Avalon
> 
> Federico is a great guy but Avalon is superseeded by JMX imnsho.

That is because you don't have the concept of what Avalon is, but
if you want to discuss this further, lets take it off the list.

> |I feel sometimes that my questions, input, and code fall on deaf ears in
> |this community.  There are a handful of active committers, and each of them
> |are hard at work on the respective pieces.  The frustration comes
> |when there
> |is a _useful_ patch that can help many people, and it is given lip service
> |and never committed.
> |
> |Part of why Cocoon and Avalon are relatively healthy communities is that
> |there is a responsiveness to ideas and patches.  Basically, first code
> |wins--and not alot of time is spent "re-engineering" the patch.
> |When people
> |get ignored (in their perception), they figure they will go someplace else.
> |Personally, I am stubborn and opinionated--but I do listen to other people.
> |In a sense, I am not satisfied that I am wrong unless someone can give an
> |intelligent reason why--and what would be BETTER.  I have
> |publically conceded
> |when I was in the wrong.
> 
> Ok, good analysis, wrong point.
> 
> The point here is that **I** don't even follow the patches on JBoss any
> longer, I just don't have the time, I just don't have the knowledge, I just
> don't stay on top of all the development in JBoss, it is **humanly
> impossible**.  See the point above from PARC, it is not about selecting an
> elitist few (although we do that very well as well, just not in RW/commit
> rights) but creating an intelligent RAID that can take on the most complex
> infrastructure development... something akin to the net, if some die, or are
> arrested by the Men In Black, the "metacreation" as the RAID brain goes
> on... right now it doesn't here in axis, you are vulnerable to people
> dissapearing and you think this is the problem you need to solve? no! that
> is a *given* people just don't care and will go, so you need to build a RAID
> of people to make the project work.  I am not superman, Scott (CTO) is not
> superman, even Rickard is not superman, none of us are, all of us admit to
> public booboos but the RAID (Redundant Array of Independent Developers) is
> superman, it is in fact "supra-human".  yes?

I agree with your point--you missed mine.  You can't have a thriving community
when people who want to help CAN'T.  No one is superhuman.  That is what a
community is for.  Someone will know the answer to a question--they don't
even have to be a committer.  But for patches to be applied, the commetters
do need to stay on top of it.

For instance, Avalon has about a dozen committers--so far only a few actively
work on it.  Many times, something I start somebody else finishes.  Avalon is
5 subprojects--you can't keep track of all of them.  I am intimately familiar
with Framework and mostly familiar with Excalibur.  I have a rough idea on
LogKit, but am mostly in the dark regarding Cornerstone and Phoenix.  On the
other hand we have developers who focus on different aspects of the entire
project as a whole.  Somebody somewhere will know the answer.

The biggest thing that I brought to that project was once we had consensus,
I made us put it in writing.  If anyone had problems with it later, we had
a known base to work from.  Hopefully I can bring the same thing to this
project.