(Email sent to the XDG list at Freedesktop.org)

Recently I've found the discussion about a document standard proposal which really interested me. For many years I've been thinking how to solve the problem of maintaining e-libraries, specially about libre software.

I've tried to follow the archived discussion and I'd like to comment some things in order of trying to help with my experience in this fields.

First of all, I think we need the most general solution we can afford. Web libraries as TLDP.org or ES.TLDP.org share almost exactly the same problem as the personal document management on a desktop or in a multiuser server.

Other point, IMHO, is the unquestionable need of using XML/W3C technologies as our data base. I'd thought nobody would prefer other alternatives but found that .desktop proposal and, sincerely, seems to me it don't fit in the web view neither the system wide use. Having an XML data system we can be sure we are ready to the next future whatever it'll be.

And between the idea of choosing a pure XML or RDF alternative, I think the first one is more recommendable. We can express our meta-data on XML, which is very well supported in almost all present development platforms and well have the same use of both technologies at least at the first steps. Also, XML can be used in namespaces and, in the future, RDF could be used to relate to other metadata XML schemas or other new applications. BTW, I remember the times of the considering in the OMF working group about choosing XML or RDF, then, Paul Kendall wrote an analysis about why XML is preferred. Sadly I think the server which hosted the email list is dead now.

About the ScrollKeeper approach, I'm really happy with it, at least in the wide perspective. It seems to me to be very elegant and flexible and adaptable to the system-wide and distributed web use. For example you can use it to add _all_ the documentation in a present linux system into a comprehensive documentation manager. This is absolutely _great_ for compatibility with legacy systems (like the ancient man and info systems). The fail of a more wide use of it I think is in the distribution packagers, who could use the ScrollKeeper services no matter the application is, in the same way you can express in a RPM file what files are documentation using an special instruction.

I think ScrollKeeper can be used even for registering remote documents into the local doc management system, which is very suggesting for me.

About the problem that developers had had creating OMF.xml files, it is obvious for me a set of tools is needed to help in the creation and maintenance of them. That tools will work as an automated QA control. And I have the opinion that documentation and software publishing has lots of similarities, so, as you can suppose, a set of "auto-tools" for documentation are needed to, specially if you try to propose the use of conventions/restrictions in the creation and managing of documentation in order to have an almost 100% automatic management, with QA control and flexibility.

A very interesting thing for me these days is the need of a controlled vocabulary in the metadata info for the categorization of documents, since it'll let the automated publishing without any assistance. The use of Freedesktop previous standards seems to be very suggesting but I think it needs consider adding/merging with other points of view, for example re-using the years experience collected in *.TLDP.org. This brings me to one of my conclusions about the need of specifying a "publishing policy" standardized for libre software documentation, which could help documentation writers to focus on well defined types of documents. This brings to me too to the need of have an strong revision of the OMF spec. At this moment I don't have a ended proposal but I'll love to help with the present ideas I'm working on.

Finally I want to say that I'm very happy this effort is happening. Me myself was thinking some time ago about starting a similar one and considering to do it at Freedesktop.org. I think this is the momment and we are the people to do it. I hope I can know how to help in this beautiful work.

PD: Next week TLDP-ES is organizing a workshop and a meeting about our advances at the Open Source World Conference. If some of you plans to be there will be a pleasure talk in person about all of this. Happy hacking :-)

El mié, 08-10-2003 a las 10:53, Juantomás García escribió:

Los planes de MS para hacer el mundo de la gestión, 
su siguiente nicho de mercado y por supuesto privativo
están en marcha y podemos aprender mucho de ellos 
para adelantarnos a su jugada. 

Podéis ver los links de la noticia en:

http://news.com.com/2100-1012-5087925.html?part=dht&tag=ntop

¿Me equivoco al pensar que tenemos una ventana de casi dos años para reventar completamente su base de software y dejarles con el culo al aire? (Si es que tenemos alguna vez esa oportunidad).

Entiendo que lo que están preparando es un despliegue de tecnología del copón, una Gran Armada que lanzar a todos los frentes que planean de manera que nada ni nadie pueda resistirles. Supongo que han tomado buena nota de la experiencia de los últimos años y están cogiendo carrerilla para ofrecer grandes innovaciones tecnológicas, .Net por todas partes, aplicaciones seguras y en general un incremento tremendo de la calidad. Me consta que en EEUU son de los únicos que están contratando programadores a mantas, así que cuando el río suena...

Por otro lado, ésta nos la hicieron antes al menos con Exchange y NT (algunos recordarán aquello de «Not Today»), anunciártelo a bombo y platillo con muchos años de antelación para que el mercado se acojone y suspenda sus inversiones hasta la llegada de aquello del software providencial, que ya se sabe que Dios propone y el hombre dispone. Este pasado sucio es una gran arma comercial contra MS.

Así que, si no aprovechamos estos dos años para poner en la calle soft libre que pueda sustituir eficazmente a la actual base -en escritorios y redes locales, particulares, educación, administración y pyme- y con perspectivas de futuro, andamos jodidos. Pero viéndolo desde el otro punto de vista, tenemos dos años para conseguir la victoria más grande que hayan visto los siglos, al menos en lo que a informática se refiere ;-)

¿Quiénes están dispuestos a intentarlo? Los que vayan a trabajar, aprétense los machos. Los que no, por favor retiraos un poquito para dejar espacio libre a los demás.

Empieza el juego duro.