#Varityper
11 messages in this thread
Stephen…you know, rumor has it that the Adobe hinting algorithm has or
shortly will be broken. There are three people working on throughout the
country – probably more, but three known to be good and likely to break it.
I think that not being able to use Adobe font is a limiting factor, yes,
but not a killer issue.
Bitstream has a larger library than Adobe (available to OEMs like Birmy)
and URW will have a giant library soon. Don't discount CG either. I think
that there are more people using Linos in limited environments than there
are service bureaus. Estimates suggest that there are between 1200 and 1900
PostScript Linos out there. You can bet that there aren't that many service
bureaus.
So it stands to reason that if someone has a Birmy (or Itek, etc.) they
could adapt rather easily to that environment. You have to also remember
that the folks making these super compatibles know very well the perceived
market limitations of not using Adobe fonts and have gone to great lengths
to (1) secure, at a minimum, 35 outlines that match the LW Plus set exactly
and (2) font downloaders that allow you to assign any names to fonts.
My shop will be running a machine just for Ventura users. All we have to do
is give out clients proper width tables. Screen fonts are nice (and will be
provided) but aren't necessary. The rest of the time the machine will be
used for our in-house book production and we know the Birmy will be great
for that… As soon as version 49.3 really makes it to the market I'll be
doing several benchmarks between all of the PS and PS-clone machines.
I don't disagree with your observations about Adobe versus clone fonts.
There are large non-Adobe font libraries available (and Adobe machines can
use them, while non-Adobe machines can't use Adobe fonts), and I respect
the craftsmanship that Bitstream, URW/The Company, and Compugraphic have
applied to creating these fonts. But it is important for prospective buyers
of expensive equipment to recognize the restrictions of non-Adobe
PostScript implementations, especially in the font area.
As far as the "cracking" of Adobe's hinting algorithm, I imagine that any
code can eventually be broken–and there's certainly a strong incentive for
anyone who can accomplish it! John Warnock of Adobe sounded quite confident
that it couldn't/wouldn't be done at last September's Seybold Seminar, and
offered to hire anyone who could do it on the spot.
I am generally in favor of rewarding innovators over "cloners". Unlike the
situation elsewhere in the computer world, the makers of laser printers,
typesetters, and other imaging devices are able to license Adobe's
PostScript for their own use. A company choosing to use a non-Adobe
implementation is either trying to save money or gain a performance edge. I
don't think that Adobe will continue to lag in performance against the
RIPS-based systems; Warnock stated his firm's intention to pursue whatever
hardware architectures and software techniques were necessary to improve
speed.
Also, I bought Adobe stock at the initial public offering two years ago.
It's doing very well….
Nicely put defense of Adobe. I think you're right on. With the pricing I've
seen on the new CG Postscript machines (which are using an Adobe-sanctioned
RIP) I think Birmy is in serious trouble. Why purchase a Birmysetter that
can't run the Adobe fonts (which whether David likes it or not are the
industry standard at the moment) when you can get a CG machine for about
the same money that will do both. Hopefully, this will force Linotype to
come down in price on their machines as well, resulting in even greater
savings for the end user.
Damn. I wish I had picked up that initial Adobe stock offering….
>Nicely put defense of Adobe. I think you're right on. With the pricing I've
>seen on the new CG Postscript machines (which are using an Adobe-sanctioned
>RIP) I think Birmy is in serious trouble. Why purchase a Birmysetter that
>can't run the Adobe fonts (which whether David likes it or not are the
>industry standard at the moment) when you can get a CG machine for about the
>same money that will do both.
Adobe PostScript fonts are, of course, the industry standard for
PostScript. After all, Adobe made PostScript. Adobe fonts are not,
however, industry standards for fonts. Far from it. Why Birmy (and other
soon to be announced clones)? They know there's a market. Why wait 120 days
for machine from Lino, CG, and Varityper? The Birmy is also 2 to 5 times
FASTER than an Atlas II RIP.
For me, that's important. You will also see a Birmy made specifically for
the MAC market that is so price competitive that…you'll see. Hey, I love
Adobe. Some of the cloners, though, like Birmy, Itek, etc. have chosen to
attack the *high-end*, high-speed market. Show me a real Adobe
implementation that can drive Ce Alphacomposer at 5300dpi at respectable
(usable)speeds?
A company in New Jersey is doing that. Just because there are clones
doesn't mean Adobe is in any way discredited. Quite the contrary. Also keep
in mind that just because someone licenses the real mccoy doesn't mean
their product will be great. One of the new licenses (name not mentioned)
produced output that when run under a densitometer doesn't come close to
spec! It's all relative…
Listen, I don't want to get into a big fight about this, but how can you
say Adobe isn't the industry standard for fonts in DTP? Name me a service
bureau that carries CG and Bitstream fonts that doesn't carry Adobe. Since
Adobe cut the deal where Linotype is now digitizing its faces, Adobe has
access to one of the largest font libraries around (I've got over 1,800
Merg/Lino faces on my 202s at last count). Likewise, Adobe's connection
with ITC is already reaping benefits (such as the Stone family being
recalled and reworked so it's now in line with ITC's specifications on cap
heights and weights). The CG font library? Yeesh. Ten years ago I scrapped
all my CG equipment and bought Alphatype because my clients wouldn't accept
Helios as Helvetica (they weren't crazy about Triumvirate, either).
The Birmysetter is a fine machine (I saw a demo by the Answers, Etc. crew
at the Folio show in New York a couple of months ago). Yet I can't see why
anyone would lock themselves into a machine that won't run Adobe fonts,
especially now that Adobe-sanctioned machines are coming out at competitive
prices (hopefully Linotype will respond with a much needed price
reduction). Personally, I think there will be a shaking out of typeface
libraries, with Adobe/Linotype providing complete text faces and Bitstream
handling primarily headline faces (kind of the VGC of Postscript). Maybe
I'm wrong, but I just don't see the service bureaus willing to continue to
duplicate font libraries.
One more time…and I don't think our *public* debate is a fight. If
anything, I think it lends itself to being recorded for posterity. The
facts we are both bringing to the forum are, in themselves, valuable. About
fonts – I know how significant the Adobe fonts are. But I also believe that
the desktop publishing marketing has yet to capture a significant part of
the typesetting business. Here in Washington, D.C. there are still MANY
typesetting firms, all making big bucks, doing things the very traditional,
pre-DTP way. Poor folks, we both know. It is this market that is larger
than all of DTP and Adobe fonts. Surely the curve is shifting.
For me, and many of my customers, buying a machine that is significantly
less expensive to purchase and maintain and can interpret postscript
language operators is something well worth considering. Lino can dip only
so far on price. They have a big organization, their equipment is carefully
made in Germany (changes in currency can impact their profits), and their
recorder (L300) uses some expensive components. The L200 is a step in the
right direction for them and allows them to compete extremely aggressively.
Varityper and CG don't want to get into a price war — there's business
for everyone.
Let's give Birmy some time before being too harsh a skeptic. FYI, the
technology used in the L200 was invented/perfected by Walter Hanson at
Ultre*Setter (LI, New York). The Ultre*Setter is the recorder used in the
Birmysetter. Let's continue the dialog. Perhaps we should poll the members
of this forum for their thoughts on fonts.
I've been connected with the typesetting industry for years. Typesetting
shops using Compugraphic, Varityper, or other non-Linotype equipment have
had to struggle in the market against the understandable desire of their
customers to have "true" fonts–which in almost all cases means Linotype
faces with Linotype tradenames. My clients with CG shops are looking
forward with great anticipation to being able to sell Adobe/Linotype type!
No one goes looking for Triumvirate (which is a pretty good copy of
Helvetica, but no improvement); at best they settle for it.
So who are the three working on this? 😉 You can tell me in Easyplex if
you like. From what I've heard, the hints are built by hand. That makes
the process inherently labor-intensive, to say the least, and therefore any
Adobe competitor needs to put in at least as much work as Adobe has, to get
the same results, so more power to them…
I thought you might be interested in this quote from an excellent new book,
Real World PostScript, just out, regarding Adobe's "hints". (Edited by
Stephen F. Roth, Addison Wesley, 1988). It's from an essay, "Building Smart
Fonts" by Michael Mace (founder of Century Software).
"So how does Adobe do it? Does the company have 'that special Adobe font
magic' (as one employee put it)? Actually there's no magic involved, but
there is a fair amount of obfuscation. Adobe's fonts are not outlines;
they're algorithms.
"The details of Adobe's system are secret, but apparently their fonts use
an outline that mathematically modifies itself, depending on the font size
and printer resolution, to fit the pixel grid perfectly….
[Mace then produces an algorithm to do the same thing, but says the
resulting outline is bulky to store and slow to execute.]
"How slow? An algorithm much like this one doubles the print time of a font
I tested it on. Adobe's fonts are speedy in part because they are written
in assembly language, not PostScript. They can thus access the processor
very efficiently, allowing them to do more calculation than any third-party
font could."
He concludes: "What the world is waiting for is a computer whiz to cook up
an algorithm that executes quickly and can be applied automatically by a
program like Fontographer. Then we'd be talking about PostScript's font
magic, not just Adobe's."
At last September's Seybold Seminar, John Warnock of Adobe pointed out that
a good font scaling algorithm could be written in PostScript and used by
non-Adobe fonts. In fact, he said that Adobe's algorithm was originally
created and tested in PostScript code; it was implemented at a lower level
in order to improve rasterizing speed in the "special case" of fonts.
Creators of clone interpreters will eventually include their own
(proprietary) scaling "hints", at least for moderate-resolution machines.
The "big secrets" are how to create an Adobe-style font, complete with the
hooks into Adobe's hinting algorithm, that can run on Adobe interpreters,
and how to create a clone interpreter that can read Adobe fonts and take
advantage of the hints. Warnock claims that Adobe's font encryption will
make these both very difficult.
I read that passage in the bookstore already. 🙂 The way I've heard it,
the Adobe hints are decisions to ignore certain strokes at small sizes.
Obviously, there has to be some feedback between the character and the
ratio of the proposed charactsize to stroke, and the density of the output
bitmap. I heard a story once, probably apocryphal, that said a lot of the
hints are done by hand, with a blue pencil, on large versions of the
letters, and that this information controls a set of higher-than-beizer
order equations whose parameters change as that bitmap/size ratio changes.