Elias Sinderson wrote:
> Aleksander Slominski wrote:
>
>
>> Michael Champion wrote:
>>
>>
>>
>>> On 3/15/06, Mark Baker <[EMAIL PROTECTED]> wrote:
>>>
>>>
>>>
>>>
>>>> I'd be interested to know what characteristics of Linda caused you to
>>>> mention it in the context of enterprise integration though, because
>>>> based on my experiences with both TBSs and the Web, I suspect the Web
>>>> can provide *at least* those same characteristics.
>>>>
>>>>
>>>>
>>> The question was not directed to me, but I've been thinking about this
>>> for awhile (and FWIW talked about it at XML 2005
>>> http://www.idealliance.org/proceedings/xml05/slides/champion.ppt ). I
>>> have a slide on the similarities / differences between the Web and
>>> tuple spaces that boils down to a) Web resources are located by their
>>> identity, tuples located by value; b) links make the Web work, queries
>>> make tuple spaces work; and c) the Web has no intrinsic notion of
>>> queries -- search engines are necessary, but not part of the
>>> infrastructure whereas querying is fundamental to tuple spaces.
>>>
>>>
>>>
>> ATOM Store (ATOM 1.0 + Atom Publishing Protocol aka APP) may be a more
>> successful reincarnation of this idea [...] ATOM Store as XML Tuple Spaces
>> 2.0 :-)
>>
>>
>>
> Yikes, that's surprisingly close to my dissertation topic, although
> missing a few crucial details... :-) A brief abstract is posted here,
> if interested:
> <http://www.commerce.net/events/?post=/2006/04/131600.a684eceee76fc522773286a895bc8436.html>
>
> Regarding the differences that Michael Champion noted above between
> tuplespaces and the web:
> a) Yes, of course, however it should be noted that one is simply a
> degenerate form of the other, in which URL is the only field matched
> against.
>
hi Elias,
i think URLs can be very powerful in matching. for example one could use
xpath as part of expression (similarly to XQL used in Ruple). another
possibility i implemented and fond to work very well is to extract
triples (RDF) from XML content and its metadata (as contained in
atom:content and atom:entry) and then allow queries to select content
based on triples describing it. both approaches are very close to tuple
matching
> b) There is no reason why tuples cannot contain hyperlinks to other
> tuples, by GUID or otherwise, within their structure. Indeed, this is a
> commonly implemented pattern within tuplespace systems.
> c) Agreed, however there has been progress on that front, most recently
> the DASL draft prepared by the WebDAV WG, which introduces an extensible
> method, SEARCH, to the HTTP protocol stack. Although not defined
> explicitly in this draft, the query grammar used with SEARCH could be,
> for example, XQuery or similar -- allowing server-side searches of XML
> resources within the provenance of the search arbiter.
>
and of course nothing really prevents one to implement 'traditional'
tuple matching: send XML that represent template tuple (could be just
any XML), analyze its structure, and then return URLs for XML tuples
that matches the template ...
best,
alek
--
The best way to predict the future is to invent it - Alan Kay
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/service-orientated-architecture/
<*> To unsubscribe from this group, send an email to:
[EMAIL PROTECTED]
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
