Re: Not reinventing the wheel ...

Markus Hoenicka a écrit :
> Stéphane Téletchéa <[email protected]> was heard to say:
> 
>> Not sure about the future, but actually it already works with Bibus
>> (that's the reason i've looked towards it).
>>
>> You can get a rather complete description for OpenOffice.org connection:
>> http://bibus-biblio.sourceforge.net/bibus_doc/html/en/usingOOo.html
>>
> 
> 
> As far as I understand the document there are two ways to interact with
> OpenOffice:
> 
> 1) ODBC
> This simply pulls the data out of a SQL table via ODBC. This requires that there
> is a 1:1 mapping of the fields that OO expects and the fields in the database
> table. RefDB reference databases are relational, that is this simple mapping is
> not going to work.

Right.

> 2) External control of OO through the UNO interface
> Sure you can program OO (and Word, fwiw) using their APIs, but this requires
> each bibliography tool to write one interface per supported word processor (n
> tools times m word processors makes n*m interfaces).

Right, but that also means you can take advantage of the existing work 
done in bibus, this is my major argument, in the end.

> The OO bib project (Bruce, correct me if I'm wrong) tries to implement a
> mechanism that allows OO to retrieve cited data through the SRU/SRW
> (http://www.loc.gov/standards/sru/) interface from any database that provides
> such an interface. This would result in n+m interfaces and is the better
> approach as soon as you're looking at more than two word processors and more
> than two data sources.

Sorry to not follow you argument but i've been on their mailing list 
long time ago (around 2002 IIRC) and this feature is *far* from being 
the priority. We're now close to 2007 and this is still a work in 
progress. What we need is somthing working for the moment, and may be 
another implementation later if they provide a better way.

Second that does not solve the Word problem, bibus does.

> regards,
> Markus

I'm not against any idea here, this is just i'm looking for solutions 
since 2000/2001 and so far, only little pieces are available, nothing 
integrated.

My suggestions are :

- Take refdb as the backend system (for data manipulation, import, 
export and mangament),
- use exising code (from refbase) for the php interface since it seems 
more advanced that what we have in refdb for now, at least this will 
help both sides i presume
- take the bibus connectors for Word/OpenOffice.org

The problem lies in the compaction of it. Do we need to integrate all in 
refdb (that would be a must) or let them work flawlessly together (i was 
thinking of getting a correspondance between tables and/or renaming some) ?

I hope i'm clearer now.

Cheers,

Stéphane


-- 
Stéphane Téletchéa, PhD.                  http://www.steletch.org
Unité Mathématique Informatique et Génome http://migale.jouy.inra.fr/mig
INRA, Domaine de Vilvert                  Tél : (33) 134 652 891
78352 Jouy-en-Josas cedex, France         Fax : (33) 134 652 901

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642