Re: Universal network graphics language

Jonathan Abbey (
Fri, 28 Jan 1994 09:21:31 -0600

I'm really surprised that everyone is jumping off from Dr. Sleator's
initial comment to the land of virtual reality and autonomous agents..

I think that simply having a graphical system (ideally with unicode support)
that you can use to telnet with would go a long way towards giving the
Internet a much fluid and flexible face.

Here are some candidates for the job:

NAPLPS. I've not yet seen the actual specs for the protocol, but I've read
some of Dave Hughes' writings on the things he has accomplished with it, and
it sounds at least flexible enough to handle foreign character sets (such as
Turkish, and other non left-to-right writing systems) and basic graphics. It
also is supposed to provide support mathematical writing and music. No word
here on whether it allows you to define fonts and/or objects to be manipulated.

RIPscrip. This is a display graphics protocol currently growing in popularity
on the PC BBS platform. It has the ability to do high-resolution graphics
with drawing instructions rather than bitmaps, and it supports some kind
of mouse interaction.

Other existing candidates: the X protocol, PostScript, Display PostScript.
I'd rule out the X protocol because there is no defined way to be able to save
an "X document", to preserve the graphics and formatting information across an
entire interaction, and because it requires a high-bandwidth, low-latency
connection to function effectively. PostScript I'd rule out for not having
the kind of interactive features that we'd need. This leaves Display
PostScript along with NAPLPS and RIPscrip and suchlike things.

Does anyone out there have any pointers to NAPLPS standards information,
and/or pointers to any NAPLPS-compatible freely distributable UNIX software?

Jonathan Abbey
Applied Research Laboratories The University of Texas at Austin