[announce] Jackrabbit based Open Source DMS: jLibrary

[Martin Perez <mp...@gmail.com>]

First of all, my apologizes if someone considers that this is not the best
site to inform about this. But, I thought, that being an open source
document management system application, and being based on jackrabbit maybe
it could of interest for someone.

I have released jLibrary 1.0beta4 (http://jlibrary.sourceforge.net ,
http://www.sourceforge.net/projects/jlibrary ). Well, I'm not going to talk
here about the features, possibilities, use cases, bla bla bla. If you want
to see it, simply browse to the web. I'm going to post simply comments about
the project that maybe some could find useful, and so this post won't be
only marketing spam :D

* Apache jackrabbit has fulfilled all my expectactives and even more. When
first I choose it to replace my old backend, I never had thought of such
stability, performance, and robustness. The change has been really a good
surprise and a great experience.

* I have migrated jLibrary from a
hibernate+ehcache+lucene+filters(poi,pdfbox,textmining) to a Jackrabbit
based backend. It has been a really success. Hibernate was a really
nightmare in comparison to jackrabbit. All that nasty session management,
all the persistence work done by hand, .. people, believe me, if you want a
hibernate based backend that suits good to a repository managed
architecture, then, do the change! The performance, is, for my tests, much
better. I can assure from this fact that jackrabbit is faster than
hibernate, it would be an idiot if I said that. But I can say that for my
case, it runs faster.

* I had the honour to contribute some text filters to jackrabbit. Those text
filters were originaly on jLibrary and now I have externalized them to
Jackrabbit. I hope to do more contributions as I have liberated many stress
with this release :)

* The WebDAV server works fine for me. jLibrary uses it, and I have tried
several webdav clients. Well, it isn't the better WebDAV client on the world
as it has been discussed on this mailing list, but it do the simple work.

* The most important problem I'm facing currently is that some text
filtering process are really slow. Specifically, PDFBox library is really
slow, and so when someone tried to index a file, the index process takes
some time. Marcel did a great hack on this to avoid jackrabbit freezing on
node indexing. I think that the bug that makes jackrabbit to index four!
times a document has not been solved yet, but I think that with the recent
system index externalization it will be solved soon.

Well, I think that I have no more to say. Thanks to the jackrabbit team for
the help. I will continue to spamize this mailing list as I want to release
a final version concurrently with Jackrabbit final version. I hope you like
jLibrary as much as I like Jackrabbit.

Regards,

Martin