[Xesam] The Ontology Open Source Project and OSCAF - 2nd try
stefan.decker at deri.org
Sat May 30 09:23:12 PDT 2009
Resend due to mailing list block
----- Original Message -----
From: Decker, Stefan
To: Ivan Frade <ivan.frade at gmail.com>; Sebastian Trüg <strueg at mandriva.com>
Cc: xesam at lists.freedesktop.org <xesam at lists.freedesktop.org>; Leo Sauermann <leo.sauermann at dfki.de>; Antoni Mylka <antoni.mylka at dfki.uni-kl.de>; nepomuk-kde at semanticdesktop.org <nepomuk-kde at semanticdesktop.org>; Arnaud Laprevote <arnaud.laprevote at mandriva.com>; Stéphane Laurière <slauriere at mandriva.com>; Dr. Ansgar Bernardi <bernardi at dfki.uni-kl.de>; Browne, Claire
Sent: Sat May 30 16:02:12 2009
Subject: RE: [Xesam] The Ontology Open Source Project and OSCAF - 2nd try
Ivan Frade wrote:
> 2009/5/29 Sebastian Trüg <strueg at mandriva.com>
> Hi everybody,
> after the long discussion on the Xesam list which ended in mid air and a long
> private discussion with members of the former Nepomuk project and OSCAF we
> came up with a compromise. Let me mention the most important points:
> All this emails is based on the assumption that OSCAF must exists. There is no good reason for that (except burocracy overload, veto power for certain people, and pervert the meritocracy way to do the things).
I want to live in world in which I can collaborate with anyone, not just the users of one particular desktop system - in the same way that I can watch Web pages from everyone, not just pages which were created from users of a particular operating system.
We need to ensure that a future metadata and collaboration architecture does not get dependent on a particular desktop environment - otherwise we have interesting functions on one desktop, but continue to have the interoperability hell that we have now and a continuation of the current situation which we hoped to overcome: a world separated by system boundaries.
The Web is the role model here: it allows information exchange across platforms, but only for limited information and collaboration.
If we want to allow a similar interoperability on the desktop we need to ensure that the developed architecture is acceptable to everyone - of course the Open Source Community, but also companies like Apple, Microsoft, Oracle, and probably even more important smaller companies working in the space and producing products for the various operating systems.
We need to ensure interoperability across platforms.
But this does not fall from heaven and will not happen magically.
This requires hard work and agreement processes by the users and developers from different platforms. It requires someone or something that organises this agreement.
We today and now have the perhaps historic opportunity to enable people to work better together.
And that is the purpose of OSCAF because we did not see how to do with with existing instruments.
But it is also clear that we need all parties pulling in one direction.
> burocracy overload,
Every agreement process requires hard work and is painful because one own view is challenged by the views of others. I have particiapted in W3C standardsation groups, and yes, it is painful. However, if we want to avoid fragmentation I don't see another way other than engaging with each other. This engaging and challenging of each others assumptions is the real work. As with every organization some bureaucracy will be undeniably involved, and of course this need to be minimized.
>veto power for certain people, and pervert the meritocracy way to do the things).
An interesting property of agreement processes ala W3C is that _every participant in the discussion_ has veto power. However, if no agreement is reached everybody loses.
The group loses or wins together. The assumption is that everybody has an interest in
reaching an agreement. If that is not the case the process is doomed.
Wrt to meritocracy:
You are in so far right as an agreement process is often not driven my not merit but pragmatics.
The problem is how to determine merit: the most prevalent way is to let the the user community decide.
In this case the users for operating and collaboration software. This will take a long time (e.g. 20 years) and in the end of the day after hard competition fights one collaboration technology will win and own 80% of the market. With an agreement instead how much more progress could we achieve in 20 years!
The Web itself is not a meritocracy in all aspects. It is based on discussions and agreements organised by W3C, and so far it seems to work well - the Web is not yet fragmented like the operating system world, although many organisations (Microsoft, Netscape during the browser wars) have tried.
The reason the Web is not fragmented is that W3C organised the agreement process.
> You say "compromise". I guess that happens when you have two solutions and find a middle term. In >this case we have: a good solution (open source developement) and a second solution that is the same, >but adding a layer of useless burocracy on top (that does no actual work and gets money). Either you >clarify what is the value that OSCAF add or it shouldn't be added.
See above. My core assumption is we are all interested in interoperability, getting more and more parties signing up and making their systems interoperable.
That is what OSCAF needs to do - organize the agreement process.
How this is organized is a different story - it may make perfect sense to develop first and agree later, which would mean to change software if the agreement (which always involved changes) is derived later). But we need the willingness and insight of all parties that agreement and discussion is necessary, otherwise there is no point in starting and we lose an historic opportunity to bring the different worlds together.
WRT to money - I assume you did not check the facts.
I paid for the domain OSCAF.org 400 Euro out of my personal funds which I will never get back. I am also willing to donate the SemanticDesktop.org domain, which I personally paid for already for several years. OSCAF is a not-for profit and I don't see who would earn any money with it - nor is the intention to earn money with it. The intention is to use the resources for professional organisational work, which I would not want to and can not do on a daily basis.
As a university professor my salary is paid by the Irish taxpayer , the Nepomuk developments were paid by European taxpayer and by the companies involved (e.g. IBM, HP etc).
However, I am willing to personally pay much more from my personal funds if there is the slightest chance that we can bring the different worlds together and enable people to collaborate without artificial system boundaries. We can do for the desktop what the Web did for text documents, but we can only do this together.
Ivan, if you or anybody know a better way than OSCAF to organise and start this agreement I am curious.
OSCAF is based on my own work in W3C standardization groups, and I discussed OSCAF with Tim Berners-Lee. He suggested to also create a link to W3C, which I think is a good idea.
But you need a entity which is not biased by any specific vested interests and which is credible enough to get different parties together.
So, how would you do this?
All the best,
> 1. OSCAF
> There was a lot of concern about OSCAF. However, OSCAF has always been
> intended to be an open and non-profit organization to give an "official" face
> to the desktop ontology maintenance. It is not driven by a specific company,
> nor will it hold any copyright over the ontologies. You can look at it as the
> KDE e.V. for the desktop ontologies. The "scary" texts on the homepage will be
> changed, the semanticdesktop.org domain will be transferred to OSCAF.
> The latter is important since we need the domain to stay with an impartial
> XESAM maintaining nepomuk otologies:
> 1) There is ontology maintenance
> 2) It is driven by meritocracy = the people who does the work
> 3) The copyright works as in any other open source project
> 4) No fees/burocracy -> just work
> 5) No "scary" texts, or legal subterfuges.
> Besides, KDE e.V. (or GNOME Foundation) doesn't have veto power over the contents of the respective projects.
> 2. The actual development
> The actual development will happen on freedesktop.org. We can reuse existing
> development facilities such as an svn, mailing lists, task trackers, and so
> on. Whenever a release is to be made the new version will be uploaded to the
> OSCAF server (might not be that important to "us" desktop developers at the
> moment but is for semantic web compatibility).
> As i said few mails ago, The open source community does the work and put the resources, and OSCAF "tag" a release (and gets the money). Sounds unfair.
> 3. Copyright
> The ontologies will be released under a free licence. Contributors will keep
> their copyright. We propose a dual MIT/CCBY licensing since ontologies can be
> seen as creative work rather than real source-code.
> Not sure about this. Not sure even if it is relevant at all. The current Nepomuk license is open enough to allow a open source developement.
> 4. Maintenance
> Within the Nepomuk project tools have been developed to ensure the quality and
> the validity of the ontologies. We propose to install these on the development
> server (freedesktop) to ensure that
> - commits do not break backwards-compatibility
> - commits do not introduce contradictions
> - etc.
> Probably those tools are open source already. i dont see the big deal here.
> Is this a compromise everybody can live with?
> Please comment.
> I dont like it. We are here to build ontologies (or improve the existent ones). For that we just need people working
> and Infrastructure; everything else is superfluous. OSCAF is not providing any of those ingredients (and adding problems on top). The conclusion is easy.
More information about the Xesam