Re: [ba-unrev-talk] Continuation of Doug's Colloquium
On Wednesday, January 22, 2003, at 11:58 AM, Jack Park wrote:
> You said "I have my own opinions and ideas..."
>
> We've been sharing our ideas here for quite a while now. Would you
> care to share yours with us? (01)
I would prefer to present working code that implements a basic subset
of my ideas, then use this as the platform for discussion. Without up
front code, it would be difficult to describe my thoughts and we could
end up talking about completely different things while using the same
words. (02)
In a nutshell, I would like to program a text editor which has live
code running on highlighted regions. This code would interface to CVS
for versioning and show live changes like IRC. So while you're editing
the document, you can see other people's cursors moving around and they
can enter comments, questions and proposed changes which you can then
CVS merge into your own document or ignore. (03)
I would then like to put a community rating system on top so changes
can work through the regular channels and trickle upwards to the main
release based on who and how many people think it's a good idea. The
live regions can run different kinds of code, so a section of a
document could be marked for chat, another section could be a Wiki,
another section could be a RSS feed, etc. I understand Emacs implements
some of this already, though on a buffer by buffer basis. (04)
Obviously this is a fairly major undertaking, so I don't actually have
details fleshed out to the point where I could debate one way or
another. I'm expecting to implement some basic aspects then emergent
properties will define further development. This is why I moved to the
University environment, since I have the ability to make 2000 students
use my programs. This helps me avoid the Open Source problem of how to
attract attention in sufficient numbers to actually be able to test
some of these theories. (05)
In response to Henry; I think Purple Numbers are interesting, but I
think explicitly coded locations is going to be difficult to implement.
Another option is to refer to data by location and time, which is how
I'm planning to tackle keeping the highlighted regions co-ordinated
with the data. (06)
--
spwhite@chariot.net.au (07)