You are viewing a plain text version of this content. The canonical link for it is here.
Posted to general@jakarta.apache.org by Ceki Gülcü <cg...@urbanet.ch> on 2001/01/31 13:21:48 UTC

Guidelines [Was: PMC Official Business]

Greetings,

I think the growth of Jakarta and other umbrella projects is a sign of 
success. Having too much is a nicer problem than having too little or being 
plain ignored. I am a newcomer to the ASF and do not know very much about 
the Apache organization. What I know is that Apache has an excellent 
reputation. However, reputation is an asset that melts faster than snow on 
a hot day. When log4j joined the Jakarta umbrella many hesitant users 
decided to adopt log4j as their logging library. Having log4j under Apache 
basically make the following statement: "there are many logging libraries 
but this one is Apache." Clearly our project benefited from this Apache 
endorsement.

I dare to hope that the inverse is also true. When good projects join 
Apache the reputation of the organization is strengthened. When 
jakarta-log4j was announced, one user wrote that "the best projects 
eventually became Apache projects." I can hardly imagine a better seal of 
approval. Inviting good projects to join our umbrella should be viewed as a 
win-win situation.

Coming back to the issue at hand, no one should underestimate the dangers 
of overstretching. I think that by letting the project commiters make most 
of the project related decisions the dangers of centralization and 
overstretching are alleviated. At the same time we need more cohesion. 
Guideline documents suggesting a project decision making process, a common 
project directory structure, a common project web look-and-feel, and 
solution to other commonly encountered problems could increase the cohesion 
between the different projects. I am beginning to realize that some of 
these guidelines are already there and other are being written. Hey, an 
hour ago I did not even know that this jakarta-general list existed. :-)

I feel that by being part of Jakarta I am part of a larger family of 
developers. As part of that family, I will use the tools produced by other 
members not because I have to but because I can. What I like most about OSS 
is the free exchange of ideas. Being part of the Apache community makes it 
easier to share ideas and that makes it worth the effort.

Changing the scope of the Jakarta project alone will not have any tangible 
effect. (Here, I am assuming that the intend is to extend the scope not 
restrict it.) Defining the Jakarta scope is just one step in the right 
direction. The scope is a bit like a constitution. We also need a body of 
supporting guideline documents (legislation).

Before sending flames, notice that I used the word guidelines and not 
rules. The last thing we need are more hard rules. Just my two centimes. Ceki

At 13:52 31.01.2001 +1100, you wrote:

Sam,
 > I'll take the liberty of excerpting another statement from Roy Fielding:
 >
 > So, in case anyone is wondering why these "umbrella projects" aren't
 > working the way we'd like, I'd say it is because they exceeded their
 > mandate. I think they did so with extremely good intentions, and the
 > board was kept
 > somewhat informed along the way, but it certainly wasn't by design.
I guess I don't have the full context here. When Roy says, "these projects
aren't working", I have to ask to what he is referring. How and for whom are
these projects not working? Is this coming from the project committers, from
the PMC itself, from users?
In the end I guess we are going to need to understand the role of the PMC.
When the guidelines say the PMC "is responsible for setting overall project
direction", what does that really mean in practice? How does it do carry out
that role?
 >
 > What I am trying to do now is determine if there is, in fact a "Jakarta
 > community". People working together towards common goals. It was my
 > assertion that there was, but I must say that at the moment I'm a bit
 > underwhelmed by the number of responses to this query.

If there is a Jakarta community, what are its common goals? If we say
something broad like Server side Java, then there are other candidate
technologies where there is little activity such as JMS, EJB, etc. When this
has come up, it has often been stated that there is no need for Jakarta to
pursue such technologies since there are other good OSS solutions for these.
That is certainly a reasonable statement but if this is true, then what is
to be Jakarta's charter?
Conor

----
Ceki Gülcü


Re: Guidelines [Was: PMC Official Business]

Posted by James Duncan Davidson <du...@x180.net>.
On 1/31/01 9:39 AM, "Ted Husted" <ne...@husted.com> wrote:

> Over in the Ant-Dev thread, most people seem to favor expanding the
> scope, so we may simply need to add something like "and related
> development tools".

I will reiterate that I think that it's a bad idea to increase the scope of
the charter to fit what we are doing. It would be better to factor out what
we can into bite size parts and drive coordination down into those pieces.
But that's just my opinion.

Ant has nothing specific to servers. It's not a "related development tool",
it's "a development tool".

-- 
James Duncan Davidson                                        duncan@x180.net
                                                                  !try; do()


Re: Guidelines [Was: PMC Official Business]

Posted by Ted Husted <ne...@husted.com>.
>  what is to be Jakarta's charter?

The stated mission (presumably based on the charter) is:

"Provide commercial-quality server solutions based on the Java Platform
that are developed in an open and cooperative fashion."

And it seems that perhaps the gist of ASF Chairman Roy Fielding's point


<
http://www.mail-archive.com/ant-dev%40jakarta.apache.org/msg01987.html
>

is that Ant is not a server solution.

Over in the Ant-Dev thread, most people seem to favor expanding the
scope, so we may simply need to add something like "and related
development tools".  Java is a component-based architecture, and to
provide commercial-quality (or better yet, "Apache-quality" ;0) server
solutions, we need related tools, like Ant, and other various gizmos
like XML digesters. So I would suggest that the charter reflect our
mission to 

"Provide for the Java platform, standards-based, production-grade
server solutions and related tools, developed in an open and
cooperative fashion."

I think the idea of being standards-based is implicit in the Apache
culture ("correct first"), but may need to be made plain to others.

Organizationally, I would suggest that Jakarta be product-centric. I
don't see any real benefit to catagorizing products or forming
additional PMCs. The PMC may wish to experiment with internal
sub-comittees to help watch over the various subprojects, and
catagorize them as an administrative convenience, but that's better
done by convention than proclamation.

But at all times, I think it's important that we organize for the
benefit of the end-user, which speaks to the "commercial-quality" or
"production-grade" part of our mission. This is not OSS for OSS's sake
-- we should be trying to build real products for mere mortals. With
the ultimate goal of a product becoming the de facto, if not actual,
reference implmentation for its breed.

Peter Donald perhaps said this better at: 
<
http://www.mail-archive.com/ant-dev%40jakarta.apache.org/msg02066.html
>

Meanwhile, I've also seen Jakarta referred to as an "incubator".
Moving forward, I would suggest that we actually start new subprojects
in an incubator area, until they have had their first stable release,
at which time the PMC would consider giving them "top-shelf" status.
This would give us the opportunity to try things, without putting the
prestige of Jakarta (and the ASF) on the line each time. 

-- Ted Husted, Husted dot Com, Fairport NY USA.
-- Custom Software ~ Technical Services.
-- Tel 716 425-0252; Fax 716 223-2506.
-- http://www.husted.com/about/struts/