xml-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Martin Stricker <shu...@gmx.de>
Subject Re: xml.apache.org charter proposal
Date Thu, 05 Apr 2001 23:14:23 GMT
Snipping together some mails (after one week of holidy, whew! ;-)) )

I hope I don't violate any spamming guideline, I just think this is
better than a dozen individual responses fired out at once. I'll try to
restrain myself in not to answer to every interesting thought... ;-))

=======================================================================

Arved Sandstrom wrote at Thu, 29 Mar 2001 21:44:40 -0400:
> I would personally be somewhat against something that is essentially
> finished, and is just going to be maintained - even if it is a
> kick-ass application.

I'm completely against this! I'm using apache.org as a great (maybe the
best) repository of good software. I'd like to some coding, too, but in
the mext few months I'll have no time for it. I think dropping a good
and ready-to-use software just because there won't be much more fun with
development is just *stupid*. Just be happy getting an easy-to-maintain
software which won't give you much of a headache! ;-)) By the way: Your
statement would lead to kicking out any xml.apache.org software which is
finished!

========================================================================

Ted Leung wrote at Thu, 29 Mar 2001 21:42:57 -0800:
> 
> ----- Original Message -----
> From: <Scott_Boag@lotus.com>
> To: <general@xml.apache.org>
> Sent: Thursday, March 29, 2001 7:31 PM
> Subject: RE: xml.apache.org charter proposal
> 
> >
> > Arved Sandstrom <Arved_37@chebucto.ns.ca> wrote:
> > > Scott, what did you mean by 'a standard that is based on XML'?
> > > I don't think the Ant format is a 'Standards-based' thing - but
> > > it's great to have available components that consume this and
> > > do useful work.
> >
> > Ant is a really good example to discuss.  It fits most of the
> > criteria I named.  Whether or not it's a standard could be debated,
> > but I really think it's on its way to becoming a defacto standard,
> > so it fits the bill in my book.
> 
> Right.  But when Ant started as a Jakarta project, there was no OASIS,
> W3C, IETF or any other standard (except for make).  It only became a
> defacto standard after the fact.  So if we adopted a "standards" based
> charter, we'd never do any development of things like Ant.  That's why
> I'm not excited about a charter that involves using standards as the
> exclusion mechanism.

Exactly my thoughts. Restricting ourselves to "standards" (which needs
to be defined: SAX is not a standard issued by any authority but just
widely used, whereas DOM is a W3C standard. I like the DOM approach
more, but SAX is easier to use so I use SAX.) is restricting our
possibilities! I have to fight this thinking at work (where these
managers think of windoze as a standard where a unix server is much more
powerful), it is restricting my possibilities to run things best
significantly!

=======================================================================

Scott Boag wrote at Thu, 29 Mar 2001 22:31:32 -0500:
> I wouldn't think Ant would be a good candidate for xml.apache.com.
>   Why? Because it is not a "general" ML... it is for a specific
> purpose... it could just as easily be ZXY language, if it worked as
> well as ant does. But maybe you could say the same thing about FOs?

It works great for many purposes. So to hell with all this "standards"
talk! And remember: It's used widely, so one could call it a
de-facto-standard... ;-))

> As I said, maybe this is the wrong start.  Maybe we should work on
> what our vision is... what is it that xml.apache.org is trying to
> build beyond riding on the XML buzz word?

As far as I am concerned xml.apache.org is trying to build good,
reusable and interoperatable software for all people who want to use the
benefits of XML. These people might or might not care about standards
whatsoever, so we should have
a) implementations of each relevant standard and
b) any additional applications that make our lives with XML easier.
Preferably all these applications are done in a modularized way so each
developer can easily choose her/his favorite modules and build her/his
seamless-fitting solution quick and easy.

At least that's the way I do my software.

> My vision: We exist to enable clean information to be easily
> exchanged, transformed, presented, and used in a great variety of
> configurations, and we believe that XML is the best protocol for that
> information.  We want to define a set of pluggable components that
> exchange or deal with XML information sets that plug into each other
> using standard APIs, that are very high performance, reliable, and
> easy to use.  These components should be part of a underlying
> architectural orchistration that will allow them to work together
> without major negotiations or breakages.  We believe that these
> components should be vendor neutral and useable as core components for
> all.  We believe that the best way to define this XML information
> exchange architecture is by having both individuals and corporations
> collaborate on the best possible infrastructure, APIs, code, testing,
> and release cycles.  We believe that these XML information components
> will help to change the world in a positive way by making knowledge
> and data as they exist on computer networks much more powerful that
> they are currently.

I do fully agree. BTW thje only time you use "standard" here is in
"standard APIs" which are (that's how I see it) our self-made APIs used
for communication between the different xml.apache.org projects and with
the individual application framework. If someone comes up with a new,
non-standard approach we should help her/him, not reject her/him! And
from a developer's point of view to code a new approach is always much
more fun than re-code that old "standard" approach for the nth time...

> Can other's articulate why they are here, and what they would like to
> build on xml.apache.org???

I'm here because I like the general idea of XML. I did some minor
projects for several companies and it was fun. I like to use apache
software (not restricted to xml) because it's usually well-written and,
most important, works nicely. The one thing I'd like to code at
xml.apache.org is a true XML-related database. For that project I would
even skip some of my other activities to free the necessary time. I'd
like to propose this XML-database as a new project if enough others
volunteer to help me. I never did a thing as big as this...

=======================================================================

Scott Boag wrote at Fri, 30 Mar 2001 09:49:46 -0500:
> I was trying to say much more than that.  I was trying to articulate
> anarchitectual vision based on pipe components, through which flows
> the XMLinformation set.  And was trying to state some concrete goals
> as to what weshould be trying to achieve.

I agree with you. I just don't want to be restricted by so-called
standards in achieving this goal.

> xml.apache.org -->  should be a set of loosely coupled components, or
> building blocks, that talk via standards based on the W3C XML
> information set.

Right. And we should create the "standard" (API) with which they are
talking to each other.

> I'm not sure we should be about "cool".  I think that's more Jakarta.
> I like creating strong building blocks, and having other people
> assemble those blocks in various ways. I would rather our central
> thought be about information exchange, presentation, and access
> through XML, and clean loosly coupled componentry, rather than "cool"
> or even broad-level "innovation".  In a lot of ways, implementing
> standards is the opposite of innovation.

I agree: XML is about information interchange and information
presentation. But I don't think this doesn't allow us to be innovative.
XML itself is innovative yet based on the old standard SGML (which took
nearly 20 years to bocome a real standard ;-)) ). If we want to provide
useful software we have to be innovative, or we really would be nothing
else but W3C's implementation facility. There are many possibilities for
XML not yet discovered, and we could be the first to address them! And
our innovative, non-standard approach could easily evolve to a de-facto
standard as time goes by. If we don't innovate we'll become rolled over
by others who are.

=======================================================================
Ted leung wrote at Fri, 30 Mar 2001 09:45:27 -0800:
> I can't believe that the only good thoughts about informaction
> exchange, presentation, and access through XML are being thought of by
> W3C, OASIS, etc. We have some very good thinkers here. I want this to
> be an environment where some of those thoughts can grow.

Exactly my point. Use standards as ecessary guidelines, but don't let us
be restricted by them.

=======================================================================
Scott Boag wrote at Fri, 30 Mar 2001 13:13:06 -0500:
> If we could build a clear picture of an architectural vision,
> innovation could occur quite nicely without making everything look
> like a jumble-bag of dubious "cool" new stuff.

How about:
"All xml.apache.org projects must be centered around XML and must be
useful for a broad audience of developers. All projects have to
implement a standard xml.apache.org API (yet to be developed?) for easy
communication and interaction of the different project's software. The
overall goal of xml.apache.or is to provide a flexible framework of
independend yet interoperating applications for quick and easy
development of individual XML-based solutions."

At least that's what I would like to have.

=======================================================================

Arved Sandstrom wrote at :
> I'm not against the latter approach - I don't think I was saying
> anything in earlier posts other than really trying not to be limited
> to implementation of standards.

Exactly my point: Standards are necessary, but we shouldn't be
constrained by them.

=======================================================================

Scott Boag wrote at Fri, 30 Mar 2001 21:23:38 -0500:
> For the record, in case it hasn't been clear, I have been convinced of
>this, and agree with it, though I think you'll agree with me that
> standards should still be a central and important part of our work.

Absolutely agreed. ;-))

=======================================================================

Fotis Jannidis wrote at Tue, 3 Apr 2001 12:41:49 +0200:
> But there is no reason why we couldn't define two independent criteria
> to decide the inclusion of projects. One based on the 'standard'
> argument, the other one based on a softer set of arguments like

> - should be using xml in a central position
> - should be solving a immediate need (like ant)
> etc.

Good Idea! I'd like to add one:
- fits in a hole in the xml.apache.org applications framework (like a
set of interfaces or APIs for communication or data transfer between the
other applications)

=======================================================================

========================================================================


Some clarification:

Standards are necessary and important. Without them interaction between
applications are nearly impossible. If I decide to implement a standard
this implementation has to be complete and correct. Any additions above
the standard are dangerous (best example: the browser wars, it's really
hard work to develop a complex website working nicely with both big
browsers).

But there is another face of the medal: Standards need to evolve with
the needs and possibilities future innovations bring in. This can either
be done by a (mostly slow progressing) standards authority like the W3C
or by developers just implementing what they need. Of couse the latter
approach has a drawbach: It might be necessary to change this
implementation when the standard catches up and uses a different
approach. But you can be in luck and the standard adopts your work. Of
course standard and non-standard parts of the implementation have to be
strictly divided from each other for easy replacement of the
non-standard part.

And then there are new necessities where no standards whatsoever apply.
Now you could wait (maybe forever) that a standards authority develops a
standard for your problem - or you just go on and create your own
standard. Of course this has to be compliant to any related existing
standard, and it has to be well-thought-trough (sorry, couldn't think
about a fitting English word, I'm a natve German speaker). If you do
your standard well it might become a real de-facto standard, maybe even
a standards authority will pick it up. SAX became a standard that way.

What I want to say with all this: I don't want to throw away existing
standards, but I also don't want to be restricted by them when I try to
do something new, something innovative.

So I would include in the charter something like
"xml.apache.org wants to implement a corresponding framework of
applications which enable developers of all kinds to easily work with
all the benefits XML provides. To gain this goal xml.apache.org is
committed to the existing standards but will ste it's own standards
where the existing ones don't reach. For this purpose and to provide
innovations to developers experimental projects can be part of
xml.apache.org. If standards change or new standards get created our
preojects will adopt them."

How does this sound?

Best regards,
Martin Stricker

---------------------------------------------------------------------
In case of troubles, e-mail:     webmaster@xml.apache.org
To unsubscribe, e-mail:          general-unsubscribe@xml.apache.org
For additional commands, e-mail: general-help@xml.apache.org


Mime
View raw message