info-vax (12/08/82)
>From reid%Shasta@SU-Score Tue Dec 7 22:24:26 1982
Mail-From: ARPANET host SANDIA rcvd at 5-Dec-82 2201-PST
Mail-From: ARPANET site SU-SCORE rcvd at 5-Dec-82 2258-MST
Mail-From: SU-NET host Shasta rcvd at 5-Dec-82 2153-PST
To: info-vax@Sandia, G.Ryland@SCORE
Remailed-Date: 6 Dec 1982 2241-PST
Remailed-From: the tty of Geoffrey S. Goodfellow <Geoff5 at SRI-CSL>
Remailed-To: Info-VAX@SRI-CSL: ;
I stand by my statement that the 2 Imagen printers in the Stanford
Computer Systems lab are precariously close to being useless junk. It
is beginning to appear to me that perhaps other Imagen customers have
gotten better treatment; this means that I should switch from public
flaming about Imagen to private flaming at Imagen. Perhaps one of the
sites whose Imagen printers work properly could forward me a copy of
the controller prom contents and the Unix support software for it.
Given that most of the software was developed on Stanford's VAX, it is
very ironic that we didn't end up with working copies of it.
info-vax (12/08/82)
>From CPR@MIT-XX Wed Dec 8 05:18:48 1982
Mail-From: ARPANET host SANDIA rcvd at 7-Dec-82 1033-PST
Mail-From: ARPANET site MIT-XX rcvd at 7-Dec-82 1133-MST
To: info-vax@SANDIA
Remailed-Date: 7 Dec 1982 2252-PST
Remailed-From: the tty of Geoffrey S. Goodfellow <Geoff5 at SRI-CSL>
Remailed-To: Info-VAX@SRI-CSL: ;
Brian, since you are still flaming in public, I will respond in public:
you have contacted IMAGEN about one particular problem exactly once (that's
1 time), and never about the issues you mentioned in your earlier complaint.
Perhaps if you contact the IMAGEN customer support group you could get your
problems solved. Other sites seem to find that to be the case.
I believe the irony of your situation is lost when you consider the
situation in this light: how can you get support if you don't ask for it?
-------
info-vax (12/10/82)
>From MCLINDEN@RUTGERS Thu Dec 9 18:05:35 1982
Mail-From: ARPANET host SANDIA rcvd at 8-Dec-82 2139-PST
Mail-From: ARPANET site RUTGERS rcvd at 8-Dec-82 2242-MST
To: Info-Vax@SANDIA
Remailed-Date: 9 Dec 1982 0039-PST
Remailed-From: the tty of Geoffrey S. Goodfellow <Geoff5 at SRI-CSL>
Remailed-To: Info-VAX@SRI-CSL: ;
We have had an Imagen Imprint-10 for about 3 months and have the
following comments:
1. Print Quality: Having access to both a Xerox 9700 and the
Imagen, I can honestly say that considering the difference in
price, the difference in print quality is negligible. On the
minus side for the Imagen, the Canon is advertised as a "plain
paper copier"; in fact, this isn't quite true since high grade
bond paper does not do well in the Canon environment. (This is
something that Imagen readily told us when we purchased the
printer.) On the plus side (and a minus for Xerox), we almost
never have stray marks on pages printed with the Imagen, some-
thing which is NOT true for the 9700. If you don't mind using
shiny paper the Imagen is acceptable.
2. Price: The original request was for inexpensive printers;
at last glance the Xerox 9700 was $500K. Xerox does offer a
more modestly price 2700, which they compare with the Imagen.
The 2700 is disappointingly rigid; more promising is the
printer which they sell with the Star set-up, but as someone
else pointed out, you have to get the whole package which
is over $40K. I have no knowledge of the Symbolics, although
we considered them for a time. My "personal opinion" is they
are aiming at the Lisp machine market for the time being, and
we were going in a different direction.
3 Fonts: Well, for all intents and purposes there is only
one: Computer Modern (which is hardly my favorite). We weren't
to thrilled about that so we wrote a program which converted
the vfonts from Berkeley Unix to Impress fonts, and life has
been much better since. We also have a little font editor which
is crude but bearable. In general, I must say that I was dis-
appointed in the software people at Imagen (see below). Their
strategy seems to be to go for breadth rather than depth, so
that I don't believe that font support is high on their priority
list. I understand that other people are working on converting
TFM (TEX) fonts to Impress style raster fonts, but I haven't
seen any of that.
4. Software (in general). As long as people are going to hack
away at C/A/T code from TROFF, trying to make it fit their
printer, the quality and flexibility of the printer driver is
going to be severely limited. Currently, troff output is filtered
through 2 additional filters before the output is in Impress
formast. The code for these operations has been developed at
at least three different sites across the country and for three
different purposes. The result is that no matter how much you
hack it, troff never quite works the way you think it should
and the problem is with the driver. We have been told, by the
people at Imagen, that a totally new and original driver is in
the works: GOOD! We are seriously considering device-independent
troff if things get much worse. On the positive side, you can
produce reasonably good looking papers with the current software,
but much improvement is needed before we'll have the flexibility
we expected when we bought the machine. Oh, and while we're on
the subject, a word of warning. The current filter for troff
output loads the fonts at compile time. The total number of
font slots is 7 (or 8 depending on what you need). Since bold,
italics, and symbols (like math), are separate "fonts" in the
Impress terminology, almost all of the slots are filled by
the characters for a single font style. The result is that if
you want to use other fonts or font families, you must compile
a separate filter for EACH font style you wish to use. This
is only true if you want to use troff. The result is, of
course, that it doesn't do you any good to have 40 font families
unless you want to compile 40 filters. I am told that this
situation might be improved in the future, by loading fonts at
execute time. Since the fonts are compiled anyway, this would
not be hard to do, and by the time you see this, we may have
done it ourselves, already.
5. Hardware and General reliability. My experience has not been
that of Brian Reid's. We have had no hardware problems with our
machine and we are running at 19K from the DZ-11 with only minor
problems (the big one being that the printer sends interrupts to
the Vax when it wants something. This, of course, serves to remind
everyone when the printer is up and running.) There are two possible
solutions to this. One is to get the additional memory, which is
ridiculously priced right now. The other may lie in an Ethernet
link which we are contemplating (assiuming 4.2 is ever released).
In general, I have found the service and software people very
accommodating, my general impression is that this will be a good
product in the near future. For now, I would have to say that I
think that they marketed the product a bit too early, and that
many of us are now finding the bugs. On the positive side we are
ever more hopeful that this will turn out to be a good investment,
something of which we were not sure in the first few weeks.
Sean McLinden
Decision Systems Lab
University of Pittsburgh
School of Medicine
-------
info-vax (02/28/83)
>From GEOFF@SRI-CSL Sun Feb 27 18:07:49 1983
Received: by UCBVAX.ARPA (3.314/3.5)
id AA04068; 27 Feb 83 18:10:50 PST (Sun)
Mail-From: ARPANET host SANDIA rcvd at 31-Dec-82 1534-PST
Mail-From: ARPANET site SUMEX-AIM rcvd at 31-Dec-82 1637-MST
Mail-From: SU-NET host Shasta rcvd at 29-Dec-82 0936-PST
To: mark.umcp-cs@Udel-relay
Cc: info-vax@Sandia
Remailed-Date: 27 Feb 1983 0857-PST
Remailed-From: the tty of Geoffrey S. Goodfellow <Geoff5 at SRI-CSL>
Remailed-To: Info-VAX@SRI-CSL.ARPA: ;
With respect to the message that I sent around to this mailing list a
month ago about inexpensive laser printers, I think it is only fair to
say that two people from Imagen came to Stanford last week and spent
the afternoon overhauling our machine, and brought us actual
documentation. It hasn't failed or crashed since, and it's been more
than a week. It seems that my generalizations about Imagen printers
made from our experience were somewhat wrong. I think that what was
really going on is that we were Imagen's first customer, and that they
hadn't really gotten their act together when they delivered our
machines. (This kind of uneven service at the beginning is pretty
typical of startup companies, actually).
Brian Reid
Stanford