From: Anders Sandberg (asa@nada.kth.se)
Date: Fri Aug 09 2002 - 15:55:24 MDT
On Fri, Aug 09, 2002 at 12:49:45PM -0700, Robert J. Bradbury wrote:
>
> A couple of suggestions.
Thanks! I'll see what can be incorporated. Right now the system has
a somewhat split personality, since it is intended to run both on a
desktop system and in VR; I think I can get the text and image sizes
reasonable once I have structured it a bit.
> It would be really nice if one could vary the color of the ball
> or perhaps the left-to-right shading depending on quantitatively
> whether the day-to-day post quantity was rising, falling or staying
> at an average level.
A nice idea! Overall, I think there are so many different things to
explore in this way that I need to make the program even more
interactive to find them all. Of course, one can put number of postings
as node size, the red component as the number of new postings, the green
as the number of responses to others and the blue to the delta in
posting (for example). But it would likely be rather hard to interpret,
so instead we need an efficient way of showing just a few factors at a
time and being able to switch between them.
> I have a suggestion. If you could convert the data into PDB format,
> you could use one of the standard molecule navigator programs to navigate
> through it. There are several browser plug-ins as well as pseudo-free
> software like ChemSketch that I think one could use. These have been
> tuned so you can rotate and zoom them fairly easily on a PC.
I took a look at PDB; it seemed both quite powerful but also terribly
chemistry specific and ancient (a format from 1971!). However, it
doesn't seem to support edges between the atoms as would be desirable.
Doing a VRML output function is fairly easy, but then we miss out the
chance to be come immortalized as atoms :-)
> It would be easy to automate this system so it ran every week or day
> so it could be continually available (like the wwwstats code for
> web server logs).
Exactly my thought. I already have a targa image print function, and it
would be doable to make the mail -> Maxm script -> xml -> Legba -> image
conversion pipeline and put it as a crontab job. The only problems are
1) the output in 3D often has occlusions, so finding the right view is
not trivial. Maybe we will have to do with a 2D representation instead,
make a few pictures from different angles or do a gif animation. 2) I
have horrendous amounts of work to do - not that it has stopped me this
week from working almost solely on Legba. But next week my professor is
back :-)
-- ----------------------------------------------------------------------- Anders Sandberg Towards Ascension! asa@nada.kth.se http://www.nada.kth.se/~asa/ GCS/M/S/O d++ -p+ c++++ !l u+ e++ m++ s+/+ n--- h+/* f+ g+ w++ t+ r+ !y
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 09:15:59 MST