Re: [ba-ohs-talk] Groove V2.0
On Monday, April 15, 2002, at 03:03 PM, John Maloney wrote:
> http://www.infoworld.com/articles/ap/xml/02/04/15/020415apgroove2.xml
>
> Noteworthy --
>
> ...Groove stays focused on the small teams that are collectively the
> engine
> of the knowledge economy. More often than not, these teams transcend
> companies and their IT infrastructures. ...Groove supports diverse group
> formation in a fluid and highly secure way.
>
> They figured out that people are the killer app. (01)
I also found the following of note: (02)
> Groove 2 happily addresses one of the major drawbacks of 1.x: its
> isolation from e-mail. Version 2 tackles this problem for Outlook users
> by enabling them to gather a complete e-mail thread plus attachments,
> create a shared space, and invite the thread's participants into that
> space. This fluid transition mirrors one of the more subtle but
> powerful aspects of Groove: the ability to move seamlessly from a
> fragmentary message-based dialogue into a shared space. (03)
This appears to be an instantiation of transforming a linear,
predominantly temporally-ordered message thread into a structured, more
topic-oriented shared space. (04)
Drawing on what Eugene described in his Directed Graphs paper
(http://www.eekim.com/ohs/papers/graphmodel/), an OHS analogue might
take the form of (1) yanking a message thread up into the metamodel, and
then transforming it into more of a node-oriented,
content/topic-directed graph. (05)
Once there, the messages could be (2) further edited down to a
paragraph-level granularity (which is what purple numbers seem to
support), so that the key content is both preserved and made more
available as building blocks for further work (for example, in IBIS
style). (06)
Participants could then (3) subscribe to such further work from that
space by way of RSS feeds. (07)
Ie, once you have addressability of documents, notes, and messages (a
core requirement for OHS, and also central to the REST approach (if
addressable means having a URI)), then if you have a metamodel you can
start developing clean tools for transforming nodes. Since everyone has
their favorite e-mail client, an enabling goal would be to make such
tools widely available for all such clients. A variation would be tools
for publishing (for example, by cc-ing or forwarding) lists to web pages
on a host, where a service performs transformations. And then you need
the shared space which can accept such node collections. (08)
XPath 2.0 and XSLT 2.0 both seem like key technologies here (cf recent
summaries What's New in XPath 2.0
(http://www.xml.com/pub/a/2002/03/20/xpath2.html) and What's New in XSLT
2.0 (http://www.xml.com/pub/a/2002/04/10/xslt2.html), particularly since
XPath can address into arbitrary document hierarchies and sequences. (09)
Cheers,
Mark (010)