xmlgraphics-fop-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jeremias Maerki <...@jeremias-maerki.ch>
Subject Re: Question about status of JEuclid and possible inclusion in FOP
Date Wed, 03 May 2006 12:13:11 GMT
I forgot to CC the jeuclid list when I replied to Max. Here's the link:
Just in case...

The whole thing about bringing JEuclid into the XML Graphics project is
not that simple. As I said before, this would be a new subproject which
means that it would have to go through the Incubator and not only the IP
clearance process. However, without a live community (at least >2 active
developers) around the thing it has practically no chance to succeed. I
have to clarify my position here: I'm not actively using MathML and
that's why I don't have much time to help here. I'm willing to help
faciliating a migration if the preconditions are met. The problem is
that I don't see they are. JEuclid is not unlike Barcode4J (my baby)
which is also not going to come here mostly because there's no live
developer community there. Last fall, I was hoping that JEuclid may gain
some new momentum, but as far as I can see the whole thing was a single
commit to CVS in December and that was it. Please be aware that such a
migration is A LOT OF WORK and takes considerable energy. The legal work
alone is a project in itself.

So, IMO you should work towards a JEuclid release from within the
SourceForge project. JEuclid is better served that way. People
interested in MathML should gather there and strengthen JEuclid from
within first. FOP can still use JEuclid for MathML handling. What I can
help with right now is to do a review of JEuclid and update the MathML
extension for FOP Trunk. This extension can either live here in FOP or
in JEuclid. I don't care so much. When the JEuclid release is available
and the FOP extension updated we can bundle JEuclid with FOP as we've
already decided last year. Same for Barcode4J, BTW.

If there are strong voices from within the XML Graphics project to adopt
JEuclid we can reevaluate but until then I don't see JEuclid coming to
the ASF. As a prebuilt JAR under a compatible license, yes, but not as
source code.

On 03.05.2006 13:17:04 Gennadiy Tsarenkov wrote:
> Dear Max,
> results  of  our  work  (not the latest one, but still good enough for
> publishing) is available from cvs.sourceforge.net/cvsroot/jeuclid. Now
> I'm  waiting  for feedback from anyone of Apache FOP members on review
> the  JEuclid code. At the moment, I have no idea whether some work has
> been done in this regard.
> Present status of our work (we still performing some regression tests,
> but its 99% already done) is the following:
> - we have merged JEuclid with FOP 0.20.5 (version available under sf.net
> is based on one of the pre 0.20.4).
> - we remove all references to AWT library, so now JEuclid is able to
> work on "screenless" servers without XWindows configuration.
> - support "badly" fonts (e.g. Arial Unicode MS, which we are enforced
> to use for MathML rendering).
> Due to last two issues we extended metric files with extended
> information about symbol location within the glyph.
> Plans for merging into 0.90 and later are not exists so far because we
> do have to improve the following functionality of the FOP itself to
> satisfy our needs (these features are done by us for FOP 0.20.5).
> 1. Support of embedded PDF's using itext library.
> 2. Support of the specific logging into files without any use of
> console (from my point of view, this is absolutely dummy requirement
> and we can use log4j, but I'm not in charge to modify it).
> 3. Support of Arial Unicode MS as TYPE1 font, and not as TYPE0 font.
> 4. Support of "phantom" attribute on page-sequence. To hide particular
> pages in PDF result file but take them into account in page numbering.
> 5. Support both JDK 1.3 and 1.4. This requires us to integrate support
> of SAXON parser for JDK 1.4 due to nasty bug in xalan shipped with JDK
> 1.4. However for JDK 1.3 we still have to keep support of xalan.
> 6. Improving text layout within the merged table cells for lists
> (I'guess that for FOP 0.90 it is not the issue anymore).
> 7. Implement specific DTD lookup algorithm (e.g. search for DTD not
> only in the XML location but also in current directory).
> >From one hand we would like to have the JEuclid integrated into FOP
> ASAP, however, we cannot follow release schedule of the FOP due to
> very major changes of FOP due to features above.
> As  far  as  I understood Jeremias, he recommended us wait until FOray
> will be integrated into main branch, but this work still not completed
> (or even canceled???).
> Concerning  licensing  issue.  We  have contacted the JEuclid original
> publishers  and  got their permission to re-publish JEuclid to the AFL
> license  as required for FOP integration. Internally, from our side we
> got all the permissions to publish our work as well. We can commit all
> necessary  statement as required by ASF once our work will be accepted
> for publication.
> Concerning your issues:
> - I would say this is good idea to include jeuclid into xmlgraphics.
> However, we need some FOP developers comments on the extended metric
> file. Otherwise, output will be useless in producing paper-ready
> PDF's.
> - We already doing that, however only for the TIFF format. This can be
> easily done by means of fo:external-graphics in the stylesheet (if I
> understood your work right).
> - This feature we are mostly interested in. At the moment this
> implemented by means of xsl:copy-of due to misinterpretation of
> character entities in xsl:copy implementation.
> You are absolutely free to re-use our results in any way under terms
> of AFL.
> Best regards,
> Gennadiy
> Saturday, April 29, 2006, 10:33:17 PM, you wrote:
> MB> Dear Gennady,
> MB> Dear developers,
> MB> I've just recently played around with mathml and tried to include  
> MB> that in my fop documents. I've found several tools, and among others  
> MB> "jeuclid". jeuclid is very complete, it is just missing a few adapter
> MB> classes. I've written a small one to convert mml to svg and it works  
> MB> just fine.
> MB> I've then found out that there was work done merging jeuclid into  
> MB> fop / xmlgraphics. What is the current status of this? What are the   
> MB> license / technical issues? Is this desired at all?
> MB> Here  is what I would like to see:
> MB> - include jeuclid in xmlgraphics
> MB> - add code to fop to support the inclusion of "mml" documents as  
> MB> external images.
> MB> - add code to fop to support mml embedded within fo documents
> MB> i would be willing to provide the first two items, if it is legal to  
> MB> do so...
> MB> Max
> -- 
> Best regards,
>  Gennadiy                            mailto:g.tsarenkov@sam-solutions.net

Jeremias Maerki

View raw message