[xsl] Re: Status of FOP

Subject: [xsl] Re: Status of FOP
From: Arved Sandstrom <Arved_37@xxxxxxxxxxxxxx>
Date: Wed, 25 Dec 2024 14:52:24 -0400
On Mon, Dec 25, Nikolai Grigoriev wrote:

Somewhat previous to that, Arved Sandstrom wrote:
>> What's gotten me a little bit riled up is the use of FOP as a basis of
>> comparison for commercial implementations. I think this is unfair to the user
>> community. FOP is not pretending to be ready, and I take it a little bit
>> amiss when it's used as the standard of comparison by commercial software.
>> I'm pretty thick-skinned, but to be honest I'd like it if the commercial
>> implementations compare to each other, not to FOP. I know it adds lustre
>> to the product marketing when you can say "our product X kicks Apache
>> XML FOP's ass", but this is not a proper comparison at the present. When
>> FOP announces a production release, then compare to FOP and let the fur fly.
>> :-)

> It's not always about marketing. FOP is by far the _best known_ engine -
> backed with Apache's strength, cited in all XSL FO tutorials I know,
> popularized by Sun people on conferences... No wonder it serves as a common
> qauge; for many, XSL FO and FOP are synonyms. FOP is not GPL, and is
> considered as an alternative to commercial implementations in many projects;
> why should I treat it in a special way in comparisons?

To the factors you cite above, one should also add that FOP is all-Java. We
run on Windows, MacOS and Linux, and make efforts to keep a useful subset of
functionality available for JDK 1.1.x. This makes it attractive to many. Also,
FOP has been around for quite a while in XSL-FO terms, when one counts its days
under James Tauber. But I agree, being an Apache XML project doesn't hurt.

It's not that FOP is APL and open-source and free, it's that the software is
not production-ready. We have never said that it is. It is not
feature-complete, for either basic or extended or complete conformance to the
spec, and we are holding off on optimizations (for the most part...you can now
compress) until the features are there. A production release is a good 3-4
months away. This is no fault of RenderX or anyone else, but folks ought not to
be using FOP for production. You, if I am not mistaken, have said so yourself.

People compare to Xerces and Xalan because they are held to be among the
front-runners in their niches. They also joined Apache XML as pretty mature
software, as compared to FOP, which was never shopped by James Tauber as
anything other than developmental stuff.

RenderX XEP, for example, can make a pretty good case independently on its own
merits. And I consider it entirely valid for you guys to say that XEP is the
most feature-complete FO processor around, compared to FOP, and others. IMO,
though, there is a right way to make that statement, and some not so right
ways. :-)

> To avoid marketing tricks and biased statements, the best thing would be an
> independent review of XSL FO processors. In my opinion, FOP can hardly be
> excluded from it.

I agree. Decent reviews usually discuss all the relevant factors.

> Merry Christmas,
> Nikolai Grigoriev
> RenderX

And the same to you, and all readers of the list.
Arved Sandstrom

 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list


Current Thread