Hi, all
I get the feeling that there is some uncertainty about who is doing what
with FOP, and where it might be headed. So I'd like to address some of those
concerns, real or imagined. I can only speak for myself, not for any of the
other committers or developers.
Number one, it's important to understand the status of FOP, and to
understand the role of developers and committers. Like any other Apache
project (XML or Jakarta or httpd or whatever), people become involved by
submitting patches or making other contributions. This makes them
developers. People who have been contributing long enough, often enough, and
have had their contributions accepted (more on this in a bit) are qualified
to become committers. Existing committers must consider each candidate
entirely on their merits - there are no quotas, and no nominating procedure,
and if you think you ought to be a committer then we have to consider your
case in good faith. We try to be as transparent and fair about this as we can.
Committers have an obligation to guide the project, FOP in this case. That
means that we do exercise some judgment as it pertains to the suitability of
code: does it work, and does it fit within the guiding software vision that
we (the committers) are in fact tasked to maintain. But we cannot reject
submissions on a whim, and we are obliged to make an effort to ensure that
non-committers are given equal access to the CVS repository, in terms of
getting good-quality code into the codebase.
FOP was originally written by one man - James Tauber, who gave the codebase
to XML Apache when that project first started up. The current software is
still strongly shaped by his original architectural and design decisions -
many of them have had some excellent longevity, others are causing us some
grief. This is not to criticize James - IMO he wrote a fine piece of
software and it's up to us others to continue on with it. FOP is somewhat of
an odd duck among Apache projects insofar as almost all (if not all) of the
committers and developers are "real" volunteers, that is, we all have real
jobs, and FOP ain't it. There are no companies contributing bodies to this
project _under the auspices of this project_; there have been some code
contributions but that is not the same thing.
It may look like FOP is dead in the water at the moment, although a few
hardy committers and developers are chugging away. :-) [Let me hasten to add
that I am not currently one of the committers who is getting a whole lot
done on FOP.] I think it would be fair to say that in the big scheme of
things - where is FOP in terms of being a processor that reliably and
robustly processes FO that conforms to the full Extended XSL feature set -
FOP has a long way to go. And in terms of what needs to be done, if I can
say this without detracting from some excellent work that has been done this
year, yes, in the big scheme of things we _have_ been sort of dead in the
water this year.
Right now I think all the FOP committers are just very busy. In fact, I know
they are. I spend most of my time wrt FOP trying to answer some email and
trying to commit people's code. If I can put in a special plea about the
latter, can you all please strive to submit patches that you have tested
(i.e. you know that they work cleanly when fed to "patch") and that are
diffs against very recent CVS? Because I am not joking when I say that I
currently squeeze out maybe 5 hours a week for doing FOP - one patch
consisting of 3 or 5 files that I end up having to hand-patch can eat up
half of that time. That means 2 or 3 hours that I am not spending on other
matters pertaining to FOP, which I'll touch on now. I don't want to sound
like I'm whining - I'm not - that is simply the way things are.
What would _I_ like to be doing more of? It's been happening in the
background for a few months now, but when Karen Lease and I get the chance
we have been working on an architectural/high-level redesign of FOP. Why are
we doing this? If you have looked at the code anytime in the past 6 months
or more you probably found it increasingly hard to follow, too complex, too
interdependent, and too difficult to modify. All true. The original core
design of FOP is basically on its last legs, and has been for quite a while.
We are trying to fix that. I hope that once that is done we can move forward
expeditiously.
In the interim, we are (for the most part) trying to not waste time doing
work on code that is going in the garbage after the redesign anyway. This
weekend I plan to take uncommitted patches from Seshadri GK (aka
FOP-Support) and fold them in; one patch is a combined effort between the 2
of us that makes images inline again. Shortly after that, depending on what
all the committers want, we will very likely see a FOP-0.19 release. The
purpose of this is entirely to have a FOP that keeps people marginally happy
for a while. After that I suspect we will see no more releases for a few
months as we basically tear FOP down and build it back up.
I continue to be very enthusiastic about this project. We have some very
good developers contributing expertise to FOP. FOP is going through a slow
period and some growing pains, but it is not going away.
I hope this addresses some concerns and answers some questions. Again, I
stress that this is just one committer's opinion, and I hope we hear from
others. I'll also extend an invitation to anyone out there who is teetering
on the brink that if you wish to participate feel free. There are no
obstacles here, except those that occur because we are stretched for time.
Thanks for your patience so far.
Regards,
Arved Sandstrom
Fairly Senior Software Type
e-plicity (http://www.e-plicity.com)
Wireless * B2B * J2EE * XML --- Halifax, Nova Scotia
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, email: [EMAIL PROTECTED]
