Dancing Heart Rising, <dhr@iname.com>, writes:
> Indulging in simplification for a moment, the problem with mailing
> lists and newsgroups is their sequential nature. Various other
> models for online discussion forums, such as the many variations
> on hypertext, attempt to add "random access" to the medium.
>
> The www is roughly half of a hypertext system -- it's missing a
> commenting feature. Considering the uproar about the recent
> introduction of "postits" [1] (memory failure: I just saw this on wired
> news and don't remember what it's called), it seems unlikely that
> the www as a whole will embrace dynamic commenting any time
> soon. This doesn't mean that a site couldn't implement it, and
> maybe some have (research topic for later).
Yes, this is Third Voice, and the wired article is at http://www.wired.com/news/news/technology/story/19722.html (which has links to the main site). Presently it is only for Windows and only for Microsoft's Internet Explorer. It looks a bit limited judging from their demo.
Another system for annotating web pages is critlink, from crit.org, run by the Foresight (the nanotech people). This has the advantage that you don't have to use any special plug-ins with the browser. It works best if the web server runs the crit software, although you can piggyback off others like the one run at Foresight (which is pretty slow). There are a few sample hypertext-oriented discussions going at pages linked off of crit.org.
My experience with this system is that comments are added rather slowly, and it is hard to know when and where new points have been made. You can request to be notified by email when a particular page has a comment added, but if someone then adds a comment to that comment (as you might see in a debate situation) you don't get automatically notified of that.
There is also the related critmail project, which stores an email archive (like an archive for this list) in hyperlinked form. Several mailing lists use it, and it's nice because it provides back links when you include quoted text. A reader could click on a link in your text above that I've quoted, for example, and get back to the original. Something like this might have helped in the recent flame war when people were accusing each other of taking quotes out of context. (You can find sample critmail archives by following the critmail link from crit.org.)
Theoretically you can then annotate the critmail archive using critlink. However I don't think the systems are tightly coupled yet. What you might want is a sort of hybrid system of mailing list and crit suite. When somoene posts to the mailing list it goes into the critmail archive, automatically hyperlinked as usual. Then, when someone annotates part of the archive, it could send an automatic notification to the mailing list. This would include a link to the newly added annotation, perhaps even the first few lines. These notification messages would be recognized by the critmail software _not_ to be added to the hyperarchive, since they are redundant for that purpose.
Most people have mail readers which let them easily follow links in mail onto the web so for them this would be almost as convenient as reading the list by mail. (The new Microsoft mail software will actually include web pages *inline* in your displayed mail when links exist.) And for those who prefer to mostly follow the discussion by using critlink on the archive, they get automatic notification of when new links exist anywhere in the hypertext, which as I said doesn't seem to happen with the existing critlink.
A hybrid like this would provide some of the benefits of both conventional email lists and the annotations of the crit suite, perhaps at a cost of some convenience. End users don't have to run any special software, it's just the mail archive which needs to run the crit suite, and a number of archives are using critmail anyway because it's really nice.
Peter McCluskey is the developer of critmail and an occasional list member. I'll copy him on this email as I am not aware of the current status of the project.
Hal