[topicmapmail] PSI repository

[Murray Altheim]

Alexander Johannesen wrote:
> Murray Altheim <m.altheim@open.ac.uk> wrote:
> 
>> I don't think "first-come, first-served" is a good policy. It just 
>> leads to grabbing.
> 
> As much as I understand your points, I think that a little
> good faith in your fellow topic mappers isn't too far
> fetched; we're not all evil and selfish, you know. :)
> 
> Would you be happy if the spec, program and site be licensed
> to GPL, the domain and site hosted at SF.net or similar, and the domain- 
> name owner making a statement (on an official site) the non-grabbing 
> nature of his ownership, and all people
> involved accept a good-will "I'm really all nice, and will get
> shunned if I do wrong" statement?

Faith in human nature should perhaps lead you away from the idea
that conflict resolution happens on its own. This has little to do
with how nice people are and a whole lot to do with territory, as
you should be able to gather by the DNS situation.

So, as in my previous example, we can expect various individuals,
companies, and organizations to perhaps all begin working on say,
some basis of formal-logic-for-topic-maps. The first entity that
grabs

     http://purl.org/psi/logic/

had better get it *completely* right and have the complete consensus
of the community, as they're gonna piss a lot of people off if they
don't. And, without going into the extremely long and complicated
discussions they've been having on the CG, SUMO, CL and other lists,
any such "generic" (i.e., non-domain-based) PSI set is going to be
authoritarian and coercive by nature. If you don't understand this
point of view I suggest reading the last month of either any of these
list archives. It's an extremely important point, and again, has
nothing whatsoever to do with human motivation or emotion. It's by
its nature coercive; only through process, evaluation, and consensus
can this be minimized.

> Or is your stance in this that *only* OASIS (or similar)
> committee can do this?

Yes. What you'll have otherwise is a committee of one or two
legislating what gets registered as "official" and the value of
it will be next to nil. Nobody would use it. That's why domain-based
ontologies are better: there's already a community making up that
domain that have organizations ready to argue through and come to
agreements. And agreements on how to communicate are what public
ontologies are all about. (E.g., nothing personal, but I'm not
going to let you decide for me how to talk).

At this point (absent any formal process) it would still be safe
to register subdomains under PSI/ that were themselves proprietary,
such as PSI/ONTOPIA or PSI/MONDECA or PSI/CERYLE (e.g., rather than
my CERYLE/PSI), but dangerous to register things like PSI/MEDICINE,
PSI/REALESTATE, PSI/BANKING, etc. as there are many competitive
players. It wouldn't be fair if say, Citibank registered their
ideas about banking prior to Bank of Scotland without some prior
agreement between banks. What would be safer (in this instance)
would be to register a banking organization as a subdomain and
let *them* administer it, say PSI/IBA (for some fictious banking
organization like the International Banking Association). Because
there are likely even different banking associations...

> I'm simply asking because setting up an OASIS committee
> is out of bounds for myself and sounds overly complex. yet,
> I want to put into motion the "regisitory". I remember
> someone (maybe even you?) complaining about the "bad feeling"
> they got from always using PSI's that were imaginary (meaning
> they had a local constant use, but no real address). I'm
> getting to that same point myself; I don't want to just
> cough up PSI's on the fly (even if they work fine), but want
> to use some shared PSI's so that exchange of TM's actually have some merit.

They won't have merit because of some DNS address, and you won't
(despite the desire) be able to rush this. If this is treated as
a hobbiest domain it will stay that way. The fact that Tony is
domain owner means only that he has the ability to use or abuse
the domain, just as someone with whitehouse.org can. It's an
important domain and if it is to actually be valuable it can't be
managed as a free-for-all. I'm not suggesting Tony would abuse it,
I'm merely thinking there should be something more than an informal
process. I don't necessarily think it has to be OASIS, but if you're
going to go around creating universal-anythings you're essentially
demanding consensus, not creating it.

As for repositories (which purl.org is decidedly not), that's a
different matter. The PURL machinery merely points a URL at some
other URL, so the PSI domain could point at any number of web
site PSI sets. It doesn't have to be one repository.

Murray

...........................................................................
Murray Altheim                         http://kmi.open.ac.uk/people/murray/
Knowledge Media Institute
The Open University, Milton Keynes, Bucks, MK7 6AA, UK                    .

     Why is it that they impeached Clinton when he lied about having sex
     with a consenting adult, but they aren't talking about impeaching
     Bush over having led the world into war by lying about the presence
     of WMDs in Iraq? Is the US more squeamish about sex than war?