[Date Prev] [Date Next] [Thread Prev] [Thread Next] Indexes: Main | Date | Thread | Author

Re: [ba-ohs-talk] Node Sequencing [Was: **** Instant Outlining !!! ***]

Eric Armstrong wrote:    (01)

> Murray Altheim wrote:
>>I'm trying to figure out how to participate in
>>the OPML design process without joining Yahoo Groups, since I
>>won't encourage getting spammed ad infinitum by giving them
>>my email address.
> I wish I had been that smart. My spam levels have reached
> egregious proportions.    (02)

Well, yesterday I received my first Yahoo-based spam, due to Dave
Winer posting the private message I'd sent him to the Yahoo Groups
list, a message which ironically included my complaints about their
abuse of people's privacy. I've decided not to have anything to do
with OPML, as I'll not contribute to Mr. Winer's success with it.
Not only did he post my private message without asking, he included
my *password* in joining the OPML/Userland group. I had to go and
change my password in a bunch of places because of this. Duh.    (03)

>>I was trying to figure out how to add multiple dimensions (ala
>>Hytime) when I gave up to concentrate on other things (I think
>>about that era I got transferred into JavaSoft).
>>has an <outline> element whose content is attribute-based, which I
>>think is really abusive of what is an "attribute", and problematic.
>>I'd recommend remaking that as element content so that it can
>>contain markup
> That strikes me as eloquent, and necessary    (04)

I went back and looked at the project I'd started last year, which    (05)

turns out I'd named the XML Sequence Map (XSQM) markup language.
I can now see there's a few things I'd want to add, but the discussion
here (and some private emails) led me to think there was something
essential going on.    (06)

More below...     (07)

>>(I'd likely write up an XHTML+OPML DTD for this) or
>>do something similar to my earlier SeqML language by having the
>><outline> element be a link to a content-bearing node. This latter
>>idea *really* opens up OPML as a possible organizing outline for
>>something like NODAL's nodes.
> Sounds promising, from here.
> ----Original Message----
>>Eric Armstrong wrote:
>>>play machine wrote:
>>>>... a lack of expressive flexibility of all hierarchically
>>>>structured presentations, be they supported by tools
>>>>or not. It may sound trivial, but this exactly is the
>>>>reason why I don't like outliners. Not in NNTP, not in
>>>>Notes, and also not in Groove. They are too rigid:
>>>>just try to reassign a text snippet to another train
>>>>of thought.
>>>Absolutely correct. Full hierarchies fail in a lot of ways.
>>>For this, categories and more maleable node-graphs are
>>>required. A steamshovel is better than the manual version,
>>>too. But a hand shovel still beats the hell out of shoveling
>>>dirt with your hands.
>>>>There are also no simultaneous tracks
>>>>(crosspostings are strictly forbidden here), plain
>>>>trees only, no meshes, no rhizomes.    (08)

Aha! you got me. What are meshes and rhizomes? I can imagine the
former, but the latter has me really curious.    (09)

>>>Valid criticism about things that definitely need to be addressed.
>>>However, there will always be two ways to look at the
>>>content in such a system. One is TouchGraph / Tinderbox
>>>mechanism, with graphs. The other will be a hierchical
>>>projection of the underlying data. I suspect that the interface
>>>in use here -- with the potential for collaborative sharing --
>>>will remain in use, even when deeper semantics are implemented
>>>underneath.    (010)

Yes, this was what I'd come to last year, following some discussions
with Eliot Kimber and others, ideas from Hytime, thoughts about
Cortazar's "Hopscotch", Durrell's "The Alexandria Quartet", the
movie "Magnolia", etc. and some gin-soaked arguments with a close
friend over hypertext and non-linear fiction. Something is gelling.    (011)

But I don't think characterizing that second view as "hierarchical"
is quite right. One of the things quite obvious to us in "everyday"
reality is that there's a *lot* of things going on simultaneously.
I think a truly representational hypertext system (and the systems
that store such content) should be able to represent the richness
of that everyday reality. Like Hytime, it should be able to contain
more than one dimension, and register events along any of those
dimensions. Events should be able to overlap in any dimension, say,
time and space, just like everyday reality. If one were designing
an outliner (or an outlining representation language), this level
of potential overlap would be necessary, or the system fails as you
say to be sufficiently expressive.    (012)

I don't know if I will be successful, but the idea with XSQM was to
couple the topic (ie., subject) based view of content with an event
based view. There would even be some interesting inferenced links
between those two views. If one thinks of each node as an event
(which I admit is a bit of a distortion) this even gets more
interesting.    (013)

Murray    (014)

BTW, you can buy Czech-made absinthe here in the UK. Even more strange,
you can get it at Tesco, which is the grocery equivalent to WalMart.
Murray Altheim                  <http://kmi.open.ac.uk/people/murray/>
Knowledge Media Institute
The Open University, Milton Keynes, Bucks, MK7 6AA, UK    (015)

      In the evening
      The rice leaves in the garden
      Rustle in the autumn wind
      That blows through my reed hut.  -- Minamoto no Tsunenobu    (016)