Jakarta at the center of the (ASF) universe

Hi,

Earlier today I did some graphs on cross-pollination among Apache
projects and blogged a summary at [1]. Jakarta always ended up
dominating the graphs, so the version on my blog has Jakarta excluded.
If you're interested, there's a version with Jakarta in it at [2]. :-)

[1] http://jukkaz.wordpress.com/2007/11/16/the-apache-cloud/
[2] http://people.apache.org/~jukka/2007/asf5.png

BR,

Jukka Zitting

Re: Jakarta at the center of the (ASF) universe

On Nov 16, 2007, at 3:13 PM, Jukka Zitting wrote:

> Hi,
>
> Earlier today I did some graphs on cross-pollination among Apache
> projects and blogged a summary at [1]. Jakarta always ended up
> dominating the graphs, so the version on my blog has Jakarta excluded.

Why? W/o Jakarta, the diagrams don't make any sense. For example,
the Jakarta-free one has velocity's only relationship to DB (!), and
for Harmony, to DB and XML! Ant, arguably one of the most pervasive
projects, has no connection to anything else...

geir


>
> If you're interested, there's a version with Jakarta in it at [2]. :-)
>
> [1] http://jukkaz.wordpress.com/2007/11/16/the-apache-cloud/
> [2] http://people.apache.org/~jukka/2007/asf5.png
>

> BR,
>
> Jukka Zitting
>
> ------------------------------------------------------------ ---------
> To unsubscribe, e-mail: general-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: general-help@jakarta.apache.org
>

Re: Jakarta at the center of the (ASF) universe

Geir Magnusson Jr. wrote:
> Why? W/o Jakarta, the diagrams don't make any sense. For example, the
> Jakarta-free one has velocity's only relationship to DB (!), and for
> Harmony, to DB and XML! Ant, arguably one of the most pervasive
> projects, has no connection to anything else...
>
I agree with Geir. The graph that includes Jakarta looks much more
realistic than the other one.

Bye, Thomas.

Re: Jakarta at the center of the (ASF) universe

On Nov 18, 2007 12:07 PM, Geir Magnusson Jr. wrote:
>
> On Nov 16, 2007, at 3:13 PM, Jukka Zitting wrote:
>
> > Hi,
> >
> > Earlier today I did some graphs on cross-pollination among Apache
> > projects and blogged a summary at [1]. Jakarta always ended up
> > dominating the graphs, so the version on my blog has Jakarta excluded.
>
> Why? W/o Jakarta, the diagrams don't make any sense. For example,
> the Jakarta-free one has velocity's only relationship to DB (!), and
> for Harmony, to DB and XML! Ant, arguably one of the most pervasive
> projects, has no connection to anything else...

Ant as a piece of software is pervasive - but are the Ant committers
pervasive? Jukka's cloud shows community/commiter relationships rather
than software. The Jakarta one is interesting as it shows so much of
JavaLand at the ASF sprang from Jakarta. I agree with Jukka though -
it distorts the landscape.

Niall

> geir
>
> >
> > If you're interested, there's a version with Jakarta in it at [2]. :-)
> >
> > [1] http://jukkaz.wordpress.com/2007/11/16/the-apache-cloud/
> > [2] http://people.apache.org/~jukka/2007/asf5.png
> >
>
> > BR,
> >
> > Jukka Zitting

Re: Jakarta at the center of the (ASF) universe

On 11/16/07, Jukka Zitting wrote:
> Hi,
>
> Earlier today I did some graphs on cross-pollination among Apache
> projects and blogged a summary at [1]. Jakarta always ended up
> dominating the graphs, so the version on my blog has Jakarta excluded.
> If you're interested, there's a version with Jakarta in it at [2]. :-)
>


Nice pictures, thanks for sharing. FWIW, I like the one with Jakarta
(and Incubator).

-Rahul


> [1] http://jukkaz.wordpress.com/2007/11/16/the-apache-cloud/
> [2] http://people.apache.org/~jukka/2007/asf5.png
>
> BR,
>
> Jukka Zitting
>

Re: Jakarta at the center of the (ASF) universe

On Nov 18, 2007, at 1:47 PM, Niall Pemberton wrote:

> On Nov 18, 2007 12:07 PM, Geir Magnusson Jr. wrote:
>>
>> On Nov 16, 2007, at 3:13 PM, Jukka Zitting wrote:
>>
>>> Hi,
>>>
>>> Earlier today I did some graphs on cross-pollination among Apache
>>> projects and blogged a summary at [1]. Jakarta always ended up
>>> dominating the graphs, so the version on my blog has Jakarta
>>> excluded.
>>
>> Why? W/o Jakarta, the diagrams don't make any sense. For example,
>> the Jakarta-free one has velocity's only relationship to DB (!), and
>> for Harmony, to DB and XML! Ant, arguably one of the most pervasive
>> projects, has no connection to anything else...
>
> Ant as a piece of software is pervasive - but are the Ant committers
> pervasive?

I'd guess certainly more than an island.

> Jukka's cloud shows community/commiter relationships rather
> than software. The Jakarta one is interesting as it shows so much of
> JavaLand at the ASF sprang from Jakarta. I agree with Jukka though -
> it distorts the landscape.

But that's the fact - that most of JavaLand sprang from jakarta...

geir


>
>
> Niall
>
>> geir
>>
>>>
>>> If you're interested, there's a version with Jakarta in it at
>>> [2]. :-)
>>>
>>> [1] http://jukkaz.wordpress.com/2007/11/16/the-apache-cloud/
>>> [2] http://people.apache.org/~jukka/2007/asf5.png
>>>
>>
>>> BR,
>>>
>>> Jukka Zitting
>
> ------------------------------------------------------------ ---------
> To unsubscribe, e-mail: general-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: general-help@jakarta.apache.org
>

Re: Jakarta at the center of the (ASF) universe

On Sun, Nov 18, 2007 at 01:58:29PM -0500, Geir Magnusson Jr. wrote:
>
> But that's the fact - that most of JavaLand sprang from jakarta...
>

Jukka's graph shows committer cross-polination, not *codebase*
cross-polination (as I understand it)... So yes, since most
committers for most ASF java projects were in Jakarta (since
those projects were *in* Jakarta, after all), I still think
that the non-Jakarta page provides a more accurate representation
of the "real" dynamics, by removing the artifical aspects of
Jakarta.

Of course, I could be wrong :)
--
============================================================ ===============
Jim Jagielski [|] jim@jaguNET.com [|] http://www.jaguNET.com/
"Great is the guilt of an unnecessary war" ~ John Adams

Re: Jakarta at the center of the (ASF) universe

On Nov 18, 2007 1:10 PM, Jim Jagielski wrote:
> On Sun, Nov 18, 2007 at 01:58:29PM -0500, Geir Magnusson Jr. wrote:
> >
> > But that's the fact - that most of JavaLand sprang from jakarta...
> >
>
> Jukka's graph shows committer cross-polination, not *codebase*
> cross-polination (as I understand it)...

then you'd expect Harmony and Geronimo would connect with Velocity via Geir...

i'm not sure what cross-pollination this graph refers to. Jukka,
could you clarify?

> So yes, since most
> committers for most ASF java projects were in Jakarta (since
> those projects were *in* Jakarta, after all), I still think
> that the non-Jakarta page provides a more accurate representation
> of the "real" dynamics, by removing the artifical aspects of
> Jakarta.
>
> Of course, I could be wrong :)
> --
> ============================================================ ===============
> Jim Jagielski [|] jim@jaguNET.com [|] http://www.jaguNET.com/
> "Great is the guilt of an unnecessary war" ~ John Adams
>
>
> ------------------------------------------------------------ ---------
> To unsubscribe, e-mail: general-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: general-help@jakarta.apache.org
>
>

Re: Jakarta at the center of the (ASF) universe

On Nov 18, 2007 1:14 PM, Nathan Bubna wrote:
> On Nov 18, 2007 1:10 PM, Jim Jagielski wrote:
> > On Sun, Nov 18, 2007 at 01:58:29PM -0500, Geir Magnusson Jr. wrote:
> > >
> > > But that's the fact - that most of JavaLand sprang from jakarta...
> > >
> >
> > Jukka's graph shows committer cross-polination, not *codebase*
> > cross-polination (as I understand it)...
>
> then you'd expect Harmony and Geronimo would connect with Velocity via Geir...
>
> i'm not sure what cross-pollination this graph refers to. Jukka,
> could you clarify?

ah. i RTFA. a connection requires 5 committers in common. i'd be
curious to see the graph with the threshold set to 3 (as that is more
of a magic number in Apache community stuff). :)


>
> > So yes, since most
> > committers for most ASF java projects were in Jakarta (since
> > those projects were *in* Jakarta, after all), I still think
> > that the non-Jakarta page provides a more accurate representation
> > of the "real" dynamics, by removing the artifical aspects of
> > Jakarta.
> >
> > Of course, I could be wrong :)
> > --
> > ============================================================ ===============
> > Jim Jagielski [|] jim@jaguNET.com [|] http://www.jaguNET.com/
> > "Great is the guilt of an unnecessary war" ~ John Adams
> >
> >
> > ------------------------------------------------------------ ---------
> > To unsubscribe, e-mail: general-unsubscribe@jakarta.apache.org
> > For additional commands, e-mail: general-help@jakarta.apache.org
> >
> >
>

Re: Jakarta at the center of the (ASF) universe

On Nov 18, 2007, at 4:10 PM, Jim Jagielski wrote:

> On Sun, Nov 18, 2007 at 01:58:29PM -0500, Geir Magnusson Jr. wrote:
>>
>> But that's the fact - that most of JavaLand sprang from jakarta...
>>
>
> Jukka's graph shows committer cross-polination, not *codebase*

I don't think that anyone confused codebase and committer. I thought
that many of the ant committers had much influence in what followed,
since ant was one of the early arrivals in Jakarta as it was the build
system for tomcat... therefore the linkages are meaningful, IMO.

I think that the jakarta node represents meaningful information.

For example, Velocity came from core Turbine people, and you can't get
any sense of that from the Jakarta-free graph. Maybe that's the
problem - that history isn't represented in current committer lists,
and thus when you drop Jakarta, information is lost.

>
> cross-polination (as I understand it)... So yes, since most
> committers for most ASF java projects were in Jakarta (since
> those projects were *in* Jakarta, after all), I still think
> that the non-Jakarta page provides a more accurate representation
> of the "real" dynamics, by removing the artifical aspects of
> Jakarta.

I guess it comes down to what Jukka's trying to show....


geir

Re: Jakarta at the center of the (ASF) universe

Jukka is not subsribed, but the reason there are 5 is to kind of limit the size of the image (1
results in a huge image)

Mvgr,
Martin

Nathan Bubna wrote:
> On Nov 18, 2007 1:14 PM, Nathan Bubna wrote:
>> On Nov 18, 2007 1:10 PM, Jim Jagielski wrote:
>>> On Sun, Nov 18, 2007 at 01:58:29PM -0500, Geir Magnusson Jr. wrote:
>>>> But that's the fact - that most of JavaLand sprang from jakarta...
>>>>
>>> Jukka's graph shows committer cross-polination, not *codebase*
>>> cross-polination (as I understand it)...
>> then you'd expect Harmony and Geronimo would connect with Velocity via Geir...
>>
>> i'm not sure what cross-pollination this graph refers to. Jukka,
>> could you clarify?
>
> ah. i RTFA. a connection requires 5 committers in common. i'd be
> curious to see the graph with the threshold set to 3 (as that is more
> of a magic number in Apache community stuff). :)
>
>
>>> So yes, since most
>>> committers for most ASF java projects were in Jakarta (since
>>> those projects were *in* Jakarta, after all), I still think
>>> that the non-Jakarta page provides a more accurate representation
>>> of the "real" dynamics, by removing the artifical aspects of
>>> Jakarta.
>>>
>>> Of course, I could be wrong :)
>>> --
>>> ============================================================ ===============
>>> Jim Jagielski [|] jim@jaguNET.com [|] http://www.jaguNET.com/
>>> "Great is the guilt of an unnecessary war" ~ John Adams
>>>
>>>
>>> ------------------------------------------------------------ ---------
>>> To unsubscribe, e-mail: general-unsubscribe@jakarta.apache.org
>>> For additional commands, e-mail: general-help@jakarta.apache.org
>>>
>>>
>
> ------------------------------------------------------------ ---------
> To unsubscribe, e-mail: general-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: general-help@jakarta.apache.org
>
>
>

Re: Jakarta at the center of the (ASF) universe

On Nov 18, 2007 10:20 AM, Thomas Vandahl wrote:
> Geir Magnusson Jr. wrote:
> > Why? W/o Jakarta, the diagrams don't make any sense. For example, the
> > Jakarta-free one has velocity's only relationship to DB (!), and for
> > Harmony, to DB and XML! Ant, arguably one of the most pervasive
> > projects, has no connection to anything else...
> >
> I agree with Geir. The graph that includes Jakarta looks much more
> realistic than the other one.
>

It also pretty clearly illustrates what happens when splitting up
Jakarta was a deliberate choice, not a random activity. In other
words, the resulting connectivity afterwards is more of the "well,
duh" variety. It is effect, not cause.

Craig


> Bye, Thomas.
>
>
> ------------------------------------------------------------ ---------
> To unsubscribe, e-mail: general-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: general-help@jakarta.apache.org
>
>

Re: Jakarta at the center of the (ASF) universe

On Nov 19, 2007 12:44 AM, Craig McClanahan wrote:
> On Nov 18, 2007 10:20 AM, Thomas Vandahl wrote:
> > Geir Magnusson Jr. wrote:
> > > Why? W/o Jakarta, the diagrams don't make any sense. For example, the
> > > Jakarta-free one has velocity's only relationship to DB (!), and for
> > > Harmony, to DB and XML! Ant, arguably one of the most pervasive
> > > projects, has no connection to anything else...
> > >
> > I agree with Geir. The graph that includes Jakarta looks much more
> > realistic than the other one.
> >
>
> It also pretty clearly illustrates what happens when splitting up
> Jakarta was a deliberate choice, not a random activity. In other
> words, the resulting connectivity afterwards is more of the "well,
> duh" variety. It is effect, not cause.

How would deliberate/random look differently?

Looks to me that it clearly illustrates what happens when people are
left in the svn karma when their subproject goes tlp.

Hen

Re: Jakarta at the center of the (ASF) universe

I would tend to agree with Jim. The commit rights never attached to
"Jakarta" but only to a specific subproject under the Jakarta
umbrella. There has never been any such thing as a Jakarta committer,
only committers to current and former Jakarta subprojects. Likewise,
there is no such thing as an ASF committer. The codebase commit rights
attach to specific TLPs.

As mentioned elsewhere, the first graph might be even more interesting
if the threshold was lowered, even to just one committer in common.

-Ted.

On Nov 18, 2007 4:10 PM, Jim Jagielski wrote:
> On Sun, Nov 18, 2007 at 01:58:29PM -0500, Geir Magnusson Jr. wrote:
> >
> > But that's the fact - that most of JavaLand sprang from jakarta...
> >
>
> Jukka's graph shows committer cross-polination, not *codebase*
> cross-polination (as I understand it)... So yes, since most
> committers for most ASF java projects were in Jakarta (since
> those projects were *in* Jakarta, after all), I still think
> that the non-Jakarta page provides a more accurate representation
> of the "real" dynamics, by removing the artifical aspects of
> Jakarta.
>
> Of course, I could be wrong :)
> --
> ============================================================ ===============
> Jim Jagielski [|] jim@jaguNET.com [|] http://www.jaguNET.com/
> "Great is the guilt of an unnecessary war" ~ John Adams