On 9/18/13 12:12 PM, John F Sowa wrote:
> On 9/17/2013 11:50 AM, Jack Park wrote:
>> I would add to John's message a passing comment that blackboard
>> architectures, including Gelernter's tuplespace Linda, qualify as
>> coordination mechanisms in societies of agents architectures.
>> Blackboards provide an indirect message passing framework, in which
>> agents can subscribe, e.g. seti@home, to specific kinds of messages.
> Yes. The option of posting messages to a blackboard enables
> a message passing system to find links to agents whose identity
> is not known to the sender.
>
> But a direct link to a known agent is usually more efficient.
> Therefore, the FMF destination field supports both options:
> either the address of a known agent or an address of an agent
> whose primary duty is to manage a Linda-like blackboard.
>
> But any agent is allowed to provide "added value" by doing
> something beyond what the message requests (i.e., "cheat").
>
> The verbal distinction between "added value" and "cheating"
> is an example of the way people praise or deprecate things
> by their choice of words.
>
> If you like some kind of encapsulation, you call it a module.
> If you don't, you call it a silo.
>
> Re legacy: I also came across the following article:
>
>
>http://www.crn.com/news/data-center/240161390/ibm-pledges-1-billion-to-linux-effort-to-boost-power-system.htm
>
> IBM pledged a billion dollars to the Linux effort in 2000, and they're
> pledging another billion now.
>
> Linux is a legacy from one student's project to build a cheap Unix-like
> system. Unix is a legacy from an AT&T project to build a cheap version
> of the very big and complex Multics system, which MIT built on special-
> purpose hardware from GE, as a follow-on to a system that MIT built on
> special-purpose hardware from IBM in the early 1960s.
>
> Today, Linux runs on everything from smartphones (Android) to the
> biggest and fastest supercomputers. But you can trace the ideas,
> structures, and terminology (i.e., ontology) back to the 1960s.
>
> Moral of the story: Legacy software is too valuable to discard.
> If you kill the platform it runs on, it will be reborn on another
> platform. If you interoperate with it, you win. If not, you lose.
>
> John
>
> PS: The IBM 704 to 7094 had 36-bit words, but System/360 had 32/64-bit
> words. Many IBM customers defected to the GE 635, Univac 1108, and the
> Digital PDP-10, all of which had 36-bit words and the old data formats.
> Those customers who didn't defect ran 7094 emulators for many years.
> The emulators ran concurrently with the new instruction set, and
> they could share the same I/O devices. Interoperability!
>
>
John, (01)
There are different kinds of silos. When it comes to data, I am yet to
find any justification for a silo. (02)
My comment above doesn't imply open and unfettered access to data. It's
all about not having apps (including DBMS engines) sitting between me
and my data. (03)
BTW -- Ora Lassila explains the kind of data-de-silo-fication I seek in
a video interview [1]. (04)
Links: (05)
[1] http://bit.ly/T2aTNi -- loosely coupling applications and data
[2] http://slidesha.re/TBMT0Q -- size doesn't matter (if your data is in
a silo). (06)
-- (07)
Regards, (08)
Kingsley Idehen
Founder & CEO
OpenLink Software
Company Web: http://www.openlinksw.com
Personal Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca handle: @kidehen
Google+ Profile: https://plus.google.com/112399767740508618350/about
LinkedIn Profile: http://www.linkedin.com/in/kidehen (09)
smime.p7s
Description: S/MIME Cryptographic Signature
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J (01)
|