jackrabbit-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Frédéric Esnault <f...@legisway.com>
Subject RE: Jackrabbit is dead (for us)
Date Mon, 09 Jul 2007 06:03:04 GMT

Wow, I have been misunderstood. I never thought even one second that Jackrabbit was not a
quality project. If it has been understood that I did (and it has) then I apologize. I have
a great respect for all the people creating this product.

My point was that Jackrabbit is like it is today due to choices that have been made, not because
it has been done incorrectly. These choices do not fit our needs, that's all. I'm really sorry
I let you feel I was disrespecting the huge amount of work that has been done on Jackrabbit.
Sorry again!

Of course, there was no competition and Jackrabbit has no interest in Legisway ;) But we were
testing it because we know that the work done in Jackrabbit is good and that choosing not
to use it would (will...) cost us time and efforts to develop, test and maintain parts already
present in Jackrabbit. We actually thought seriously about developing patches for Jackrabbit
and of course contributing them to the community if they seem valuable by Jackrabbit dev team.
But it would mean to review Jackrabbit core and modify it to be more flexible and still work
fast, add some admin functionalities, review the full search engine (this is our main issue),
and it has been decided that it would be better for us to go back to a RDBMS-based repository,
full-SQL support, fully customizable indexes, and so on (and btw, we actually need a repository,
not a CMS, sorry again, I think I was really tired, last Friday ;) )

This also answers a question from Thomas Mueller, who asked what other technologies we are
going to use. Well basically, a database, and something like SDO or (more probably) EMF for
in memory access, but choices are still not finalized.

The goal of my message was to give a feedback to Jackrabbit team, to tell them that Jackrabbit
was not suitable for us, and giving them the reasons. It may give valuable information to
them about what *some* companies need in such a product, and that was the only reason I wrote
this message. I again apologize to all that felt insulted or disrespected. Jackrabbit is a
good quality project, and this is why we spent time here evaluating it. It just doesn't fit
our specific needs. Keep up the good work !

Frédéric Esnault

-----Message d'origine-----
De : Brian Thompson [mailto:elephantium@gmail.com] 
Envoyé : vendredi 6 juillet 2007 16:43
À : users@jackrabbit.apache.org
Objet : Re: Jackrabbit is dead (for us)

Gee, did the Jackrabbit community have some vested interest in
legisway.comusing Jackrabbit as a repository for their current
project? Was there a
competition or something?  I wish someone had told me!

Also, since when is Jackrabbit classified as a CMS?  Everything I've read
suggests that Jackrabbit is a *repository*, not a full CMS.

That said, I tend to agree that some of Jackrabbit's current limitations are
vexing.  It would be REALLY nice if Jackrabbit supported editing node types
the way DBMSs support editing of table definitions.

However, it's also important to recognize the hard work that has gone into
making Jackrabbit a great technology.  As Bertrand noted, a bit of respect
would be appreciated.  Also, if Jackrabbit's current limitations cause
problems for you, why not develop patches for those limitations?  You'll
spend less development time and effort customizing Jackrabbit (or another
well-supported open source project) than in developing your own custom
repository and CMS.


On 7/6/07, Frédéric Esnault <fesn@legisway.com> wrote:
> Hi there,
> I'm sorry to inform you that we did not select Jackrabbit as the CMS of
> the platform we're currently developing (well designing right now).
> We'll develop our own CMS, of course not generic as Jackrabbit could be
> (and this is to be discussed...), but more suitable to our needs.
> Why we rejected Jackrabbit :
> -          lack of administration possibilities: It is currently
> impossible in JR to modify an existing node type, to add/modify/remove
> properties. Refactoring is important for us, and impossible in Jackrabbit.
> -          Strong constraints on the repository structure : we saw in the
> different mailing lists that Jackrabbit works (quite) well with a specific
> architecture, and that not following it induces very important and
> unacceptable performance loss (both for writing and searching);
> -          Search is also a (very) important feature for us, and currently
> Jackrabbit is much too limited in this area. SQL is not complete (well we
> don't need full SQL, but at least....joins...); and xpath is limited also,
> dereferencing is impossible or must be developed as an upper layer above
> Jackrabbit.
> Some details also maybe, but most important problems are the three listed
> above (admin, repository structure and search). I think Jackrabbit is good
> to create a blogging system, a forum or any article-based simple website,
> but it is definitely not suitable for professional, generic CMS. This is too
> bad, because some (who said most?) apache projects are really top quality
> projects that even professional softwares can rely on, and I thought
> Jackrabbit would follow this path. I agree that it's still a young project,
> maybe versions 2.5 or 3.0 will begin be mature enough, and powerful enough
> for demanding systems to rely on Jackrabbit.
> Thanks anyway for all your good and detailed answers ;-)
> Frederic Esnault

View raw message