BLITs vs. PCs vs normal terminals

[Doug Gwyn]

In article <1139 at psivax.UUCP> friesen at psivax.UUCP (Stanley Friesen) writes:
>	I *think* what he is really complaining about is the need to
>download the software *every* time you turn th system on, this takes
>*time*. It would be nice to just turn the BLIT on and have it work,
>just like an ordinary terminal. At least the host computer has a disk
>drive to keep the programs on-line when you power it down. What I
>would like to see is a BLIT with a bubble memory or battery-powered
>static RAM or something like that which retained the downloaded
>software when the power is turned off.

The lack of agreement about this point is probably due to different
ideas about the right way to use a Blit/DMD.  The usual use requires
cooperation from the host that would be difficult or impossible to
re-establish once the communication link is broken.  Neville's
application seems to be a BitGraph emulator, and other suggestions
have sounded much like Sun workstations.  I think we would all agree
that for such applications, terminal-resident code is an advantage.
(In fact, the 8;7;5 firmware upgrade provided the DMD with layersys
in ROM; it used to be downloaded when starting layers mode.)  Some
of us simply do not think the DMD is best used as a BitGraph or Sun.

P.S.  AT&T and/or Teletype have made it very hard to develop stand-
alone applications of that kind, since they refuse to provide source
code for either "layersys" or the resident terminal ROM.  This is
very short-sighted on their part; we had an application that would
have resulted in acquisition of many more DMDs than we already have,
except we couldn't obtain adequate information and so we abandoned
the idea of using DMDs for the project.  I suspect sales to the
BitGraph emulator folks are down, too.