On 3/19/12 9:26 AM, John F. Sowa wrote:
> Kingsley,
>
> Vannevar Bush in 1947 and Ted Nelson in 1963 developed the vision for
> Linked Open Data. (01)
I am in no way disputing that whatsoever. Neither is TimBL for that matter. (02)
> Ted coined the term 'hypertext'. (03)
Ditto, neither I nor TimBL dispute that. (04)
> Apple implemented
> a proprietary subset called HyperCard in 1987, which Tim B-L used. (05)
Ditto, neither I nor TimBL dispute that. (06)
> For the WWW project at Cern, Tim combined ideas from all of the above
> with a subset of SGML called HTML. (07)
Yes, and I never seen a TimBL presentation or talk, where when given the
opportunity he doesn't painstakingly make this point. Personally, I
think (as is always the case) you have to distinguish individuals from
organizations they build. TimBL != W3C . (08)
> His greatest contribution was to
> build a free and open source system on top of the Internet and make
> it available to everybody. (09)
Again, nobody is disputing that. (010)
>
> The huge volume of web pages provides the Open Data, and the http
> addresses in that data provide the links. The question is what
> methods should be used to add semantics to those links. (011)
And we all agree that we need a conceptual model with first-order logic
serving as the schema. (012)
I also think we all agree that you can use graphs as vehicles for
structured data using a variety of formats, rather than RDF's family of
formats, solely. (013)
>
>> none of the above [Apple, IBM, and Google] have produced systems in
>> line with the aspects of TimBL's vision which you like.
> They all hired R. V. Guha, the inventor of RDF, to work for them. (014)
So? (015)
Guha didn't deliver Linked Data, far from it. Yes, he delivered RDF but
that really isn't Linked Data, far from it. It's simply an option for
reaching said goal. (016)
> But none of them found that SW notations and tools for specifying
> and using semantics were suitable for their needs. (017)
They didn't find RDF/XML or anything around at the time suitable for
their goals. A lot of this has a lot do with skipping the need for a
Linked Data substrate en route to Web scale graphs with semantically
rich relations that leveraged the underlying architecture of the Word
Wide Web. (018)
> In fact, they
> didn't even think that RDF was sufficiently useful be included as
> a subset of their tools.
>
> IBM developed an XML-based representation and toolset, which they
> called UIMA (Unstructured Information Management Architecture),
> and they donated the source code to the Apache Foundation. (019)
Guess what? None if it works or is moderately practical re. scale of the
World Wide Web. They all lead to silos. (020)
Linked Data works, it scales, and is being widely adopted on massive
scales as I type. It's basically about structured data representation
via exploitation of: (021)
1. Entity-Attribute-Value model
2. de-referencable URIs -- where Names are distinct from Addresses
3. Content Negotiation -- explicit or implicit
4. Language Tags
5. Typed Literals. (022)
>
> Google adopted JSON as their replacement for RDF, and they developed
> a very large toolset, which they keep proprietary, although they
> make it available for developers who work with Google services. (023)
You are speaking about JSON syntax over RDFXML or Turtle. And by the
way, they are using HTML+Microdata, yet another format. They also
support (X)HTML+RDFa, yet another format. There are many syntaxes for
expressing graphs just as there are many serialization formats for
across the wire transmission between servers and user agents. (024)
>
>> Remember, since putting out that meme Linked Data has exploded on
>> the World Wide Web and the likes of Google have taken notice.
> Please reread the visions by Vannevar Bush and Ted Nelson. From the
> Wikipedia article about Vannevar B: (025)
See my comments above. You are misunderstanding the context of my
comments. I am not one of those folks to inaccurately positions Linked
Data or the Semantic Web as new ideas that dropped out of the sky. I
profoundly believe in innovation continuum that are devoid of absolute
new inventions. (026)
Linked Data is a tweak that leverages the ingenuity of URIs. (027)
URIs that function at InterWeb scales are a real innovation for which
TimBL does uniquely deserve immense credit. The information and data
space dimensions of the Web are the greatest showcases for URIs. (028)
>
>> Bush introduced the concept of what he called the memex (possibly derived
>> from "memory extension") during the 1930s, which he imagined as a
>> microfilm-based "device in which an individual stores all his books,
>> records, and communications, and which is mechanized so that it may be
>> consulted with exceeding speed and flexibility. It is an enlarged
>> intimate supplement to his memory." He wanted the memex to behave like
>> the "intricate web of trails carried by the cells of the brain"; essentially,
>> causing the proposed device to be similar to the functions of a human brain.
>> It was also important that it could be easily accessible '"a future device
>> for individual use... a sort of mechanized private file and library" in the
>> shape of a desk'. The important feature of the memex is that it ties two
>> pieces together. Any item can lead to another immediately.
> He also was involved with the early development of both analog and
> digital computers at MIT during the 1940s -- so he had some technical
> knowledge about how to construct such things, but the costs in 1947
> would have been prohibitive for personal use.
>
> So the "meme" isn't new. (029)
The meme is very new when you factor in: (030)
1. URIs
2. Ubiquity of the World Wide Web. (031)
> But there is more work to be done to develop
> better tools for implementing it. RDF is one useful idea, but there
> are others that some very smart people believe are better. (032)
I don't know how to respond to the comment above, because our
disagreements are somewhat artificial. By that I mean the production of
a position you take instinctively. For instance, somehow interpreting my
response as somehow undermining giants from the past. Remember, I have
berated the W3C's RDF narrative, repeatedly, for being thin on genealogy. (033)
Links: (034)
1. https://plus.google.com/112399767740508618350/posts/U4u4FBmLnvx -- an
old post about data objects (note reference to Ted Nelson) (035)
2.
http://www.openlinksw.com/weblog/public/search.vspx?blogid=127&q=ted%20nelson&type=text&output=html (036)
- some old blog posts of mine that include references to Ted Nelson. (037)
Kingsley (038)
>
> John
>
> _________________________________________________________________
> 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
>
> (039)
-- (040)
Regards, (041)
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 (042)
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)
|