| by Terry Schussler
Background
Welcome. This is the first in a series of one-on-one interviews
with the programmers who have brought you innovative products at
Macromedia. In this first interview, we talk to John Henry Michael
Thompson who is known by his friends as "JT." JT is a
34-year old programmer at Macromedia who has worked on many Macromedia
products, most particularly Macromedia Director. In fact, most of the
Lingo language was invented and developed by JT.
Today he tele-commutes from Philadelphia, PA, where he lives with
his wife and two sons. JT noted to me in the interview that his four
year old son Pharoah was born when Director 2.0 shipped, and his two
year old son Nile was born when Action! 1.0 shipped. JT is expecting a
third child in August, so we should safely assume that Director for
Windows will ship by then!
In college at MIT, the kids built their own networks, and everyone
had to use a short name to log in. JT was what I used and it sort of
stuck.
Academically, I’ve been involved with computers since high school
- the Bronx High School of Science in Bronx, NY. I concentrated in
mainly math, but I’ve always had a strong visual sense. I was either
going to go to the Bronx Science or Music and Art which is the main
art school there. I went on from there to MIT where I focused on
computer languages, mainly LISP and interpretive languages - I was
very interested in that - but I always had a strong need at that point
to reconnect with the visual world - the aesthetic world. So, after
about two years at MIT I took some time off and went back to New York
and study just straight drawing and painting for a while. And then I
went back to MIT and got refocused on integrating those two worlds,
the engineering and the art side, so I focused on visual arts.
Once I started doing it, I found that there was a wealth of
information and services at MIT to help in that direction so I got
connected with the Visible Languages Workshop (VLW). This was a small
group off to the side of MIT which was doing color prepress stuff,
this was back in 1981-82, this was before even the B&W Mac.
(laughs)
They’re now integrated into the Media Lab. Their focus was really
in line with what I was trying to do at that point which was to
develop tools that brings the computer power to the designer. They
were doing stuff that was pretty much on unique platforms, there was
only a few one-of-a-kind full color frame buffers back then. They were
running the same operating system as what ArcMac was running, which
was a predecessor to what the Media Lab uses today. It was a custom
built Unix-like workstation which you couldn’t find anywhere else in
the world. It was a really unique environment but their focus was
pretty much exactly where the Macintosh is today - making tools for
the designer, bypassing all the intermediate levels of programmers and
bringing all that horsepower as something useful, making the machine
more fluid.
I focused there initially on 3D stuff - there was a set of 3D tools
already at ArcMac which was really hard to use so I started doing a
WYSIWYG interface for them. It was one of my undergraduate projects
there.
My undergraduate thesis was in another information system type
thing, [(1983-84) - Newspeak or Nexis], which was basically the kind
of thing that you see coming up now where you have an agent going off
and collecting information from the online services and creating a
personalized newspaper for you. I wrote a visual map that showed the
interconnections of various places that you went to in your visual
database - your personal newspaper. It would actually go off and get
the articles and format it - really like a paper - so it wasn’t
screens full of text. It was color coded according to interest and
laid out spatially according to how much interest you had. Interest
articles got more space. The main feature was that it was dynamically
laided out and as easy to read as your regular paper.
My main work though, pretty much underground, was taking advantage
of the vlw and the film/video section there too that did personal
documentary. Their focus was not doing major motion picture but very
intimate documentaries - they tried to take the portable
cinematography technology, 16mm, 8mm, and make it very accessible -
and taught it from that perspective.
From that, my main personal work there was between those two
departments and after I had the 3D stuff running on the VLW computer I
got into realtime. I kind of hit the wall - you can spend all this
time - processing time - constructing animation - it wasn’t really
satisfying. So I thought of using the realtime image processing
capabilities of the frame buffer to do some music video type stuff. We
set up a link between the VLW computer and the high-end 1 inch
broadcast quality video equipment that the film/video section had. It
was basically modifying the frame buffer to output NSTC and creating a
controller for the 1 inch video equipment. Basically building stuff by
hand that now you can buy. It was fun, I did some animation and some
music videos. We integrated this with live performances and 3D. I
built some robots that would fly around to the music and break apart -
stuff like that.
After graduation, I hung around MIT for a while and then decided to
move to San Francisco because it was warmer. I got a job working at
Lucas Film working in DroidWorks project. This meshed my goals at the
time which was to learn more about the image making industry - film,
video - and from the industry point of view but without getting tied
up into it too much. I did work there mainly on the EditDroid system
which was a nonlinear video disc based editing system which they were
trying to productize.
It was a great experience for me because I got a beeline right into
all the high end video post productions places. They showed me how
they cut their films and how the while thing worked - and got a look
at another startup.
In parallel with all of this, I got to see the Mac from day one. We
saw the Mac, a small group of entrepreneurs I was working with, and we
said ‘yeah this is the machine, let’s clone it!’
 We
were very proud to be one of the first to clone MacPaint for the
8086-based PC - it was a product called TelePaint. It was integrated
graphic telecommunications product.
Yes, but I’m more proud that we faithfully cloned MacPaint. We
had undo’s and flips and lassos - all that kind of stuff. We were
way ahead of the time - we had to pretty much build the windowing
system. I wrote all the graphic libraries for that and the image
encryption code.
Even though I was doing all of that workstation stuff, I still
wanted to do stuff on the Mac from that day one - I was just waiting
for the right opportunity. With the Mac I saw the merger of the
highend stuff that was happening at ArcMac and the VLW in terms of
ease of use and the graphic user interface with something that could
happen on the desktop without all that highend custom hardware. When
Lucas Film shut down the DroidWorks, I took my severance pay and
bought a Mac Plus.
I started coding on it, playing around and getting used to Think C
and thought ‘this is what I want to do.’ Through a friend and some
associates at DroidWorks I met Marc Canter. Then I started working on
the original Accelerator which was a predecessor to QuickTime - the
same concept - digital video stuff optimized to make playback of
Director movies faster.
No, I didn’t. [Marc Canter] showed it to me. Actually it was very
similar to the direction that we had been headed in the development of
the realtime processing stuff. Rather than look at stuff and build
tools which were based on the 3D metaphor where you go off, render,
then comeback and look at your effect, I was starting to do stuff
based on the limitations on what you can do in real time on the
machine. It was nice to see a tool which was like that - it was all
WYSIWYG.
I’ve always worked with small groups of entrepreneurs in that
kind of area - so [working on Accelerator] was right in line with what
I wanted to do. So we just hit it off, and kept on working together.
It was a small group then so we just worked on what needed to be done.
First Accelerator, then Color Accelerator, then integrating
Accelerator into Director...
Well, we never shipping it, but yes the first version just worked
in black and white. I started working on it when there was no Mac II
yet. I had to wait until I could physically get a Mac
II...(laughs)...so the first Mac II I got was used to build the color
Accelerator.
In the early part I was contracted to do the Accelerator. The Paint
program was the next project I got assigned. It evolved from code in
CheapPaint and GraphicsWorks - which were part of the Macromind
Utilities Disks that were sold as a product.
No, it was for Director 1.0. Actually what happened there was
first, there was VideoWorks and then in the transition from Chicago to
SF and getting their first injection of venture capital, instead of
releasing VideoWorks Professional they released Director 1.0 - which
had no interactivity. VideoWorks Interactive was still in this
never-ending beta where they just gave out to a few people for about
$500 and about two pages of documentation (laughs.)
During that transition from Chicago, I also started ripping out the
language and replacing it with a more sophisticated, more HyperTalk-like
language and replacing it for VideoWorks Interactive.
The original language was modelled after Basic . You had single
character variable names, a very limited vocabulary, no structure in
terms of handlers, no real string handling.
It was internal but it was sold one-on-one to other people. Apple
used it for all of their guided tours... - there were maybe a couple
of hundred people who were licensed the language - mainly people in
doing simulations or in education.
Right. I didn’t work on it until after that. The original
language was done by Jay Fenton. I started re-writing the language to
make it more HyperTalk-like and putting in XObjects. That was pretty
the main part of Director 2.0.
First they were going to ship a version of Director that was going
to be called Director Interactive Toolkit, or something wacky, which
was going to include the language and the XObject stuff but about a
month before it shipped they changed their minds and decided that was
a bad idea and decided to roll it into one product. That became
Director 2.0. That really didn’t get any planning until the very end
when the manuals needed to get put together. It was just me, Dan [Sadowski]
and Al [McNeil.] At that point the company was so small it could only
focus on one product at a time.
Yeah, the only other authoring system at that time was HyperCard.
Its benefits were its ease of use and its openness. The only way to to
really address it was to make the language as easy to use as HyperTalk
- hopefully borrow some of the ideas from it and also borrow some
ideas from other object oriented languages. At that point I was
researching object orientness in general - SmallTalk, Objective C -
and I had background in LISP in terms of the kind of things that you
can do interpretively, encapsulate data, recursion, that kind of
stuff. So those ideas got synthesized into Lingo and XObjects. The
notion of factories is borrowed from Objective C. The semantics of the
interpretive language is borrowed from SmallTalk and LISP.
It’s interesting, right after [Director 2.0] shipped, the company
began believing that it was at the end of its life cycle so they
started work on Action! I think that Director has survived because it
offers a more dynamic metaphor than the stacks concept of HyperCard.
Yes, the Lingo changes in Director 4.0 were somewhat driven by the
introduction of AppleScript. Another example of this is the birth and
ancestor command.
I think in the future we will put the Score under Lingo control.
This would allow users to create object-oriented templates which would
simplify production development.
I think ScriptX is a catch-up which deals with the substrate of
multimedia authoring. These tools are pointing in the same directions
- it comes down to performance and delivery - the issue of code
migration. It’s like C - there’s hundreds of C compilers but
it’s their performance and availability that distringuishes them.
It’s the implementation of the language not the definition that
makes the difference.
I would like to see it easy to switch back and forth between
multiple visual metaphors and the underlying procedural code
environment. There’s a need to provide a sense of flow between them.
If there were a continuum of levels, the one you would use would
depend on your level of commitment and level of expertise. No one view
is appropriate for all tasks.
I think that there will always be a need to get "under the
hood" - every view will have compromise, there will be a need to
get closer to the machine. More people will get familiar with using
the tools to get their ideas implemented. They will then need to get
underneath the metaphor presented by the tool. The demand for
procedural programming skills will be very high. In fact the ability
to work iteratively is a great strength of Director because the
metaphor is very accessible.
Every line of code was touched because it went from Pascal to C -
but that was just translation. In terms of new code, most of Lingo is
new, which is about a third of Director’s code. The core of Lingo
got rebuilt. All the new windows of the user interface got a new
object-oriented framework. All of the text window, text editing, and
the Cast window code is new.
(thinks for a while)
It’s tough to say because I haven’t been really blown away by
anything - some of the earlier animations were very eye-opening
though...
(thinks some more)
Well, the most interesting has been Iron Helix, but I don’t play
games on the computer that much so I’ve only just gone through a few
scenes of it. I appreciated the response and stuff like that. It’s a
not a single thing I see, a single piece, it’s more the variety that
I see. Different people doing different things.
The kid’s products - WordTales, Yearn2Learn PEANUTS - because I
have kids I spend more time playing with those.
I don’t think that the potential has really been tapped. I spend
a lot of time looking at these titles and theres still a long ways to
go. It’s hard to be awed by any one title given their limitations
and the potential that I have in my mind. Most of the time I look at
these titles and think ‘oh, they should have done this, and this,
and this...’
That’s still pretty vague. It could evolve into doing that. its
really not set up to do that right now. Where it has the most benefit
is that people can prototype new interfaces very quickly to get a
preview of those kinds of things until they have the actual underlying
structure for two way communications.
My little two cents on interactive tv is that it still like a lot
of the titles - they’re still wedded to the old metaphors. Just like
HyperCard is thinking in the old static metaphor they still think that
‘oh you have interactive tv so that means there’s one producer
thats spewing out all this stuff into the home’
As opposed to ‘there’s a media, just like the telephone
conversation we’re having now, you and I, and the important thing is
that the carrier can facilitate it or communicate it.’ There’s no
way in hell that you can create the hours of interactive tv that they
project. Plus that type of media is not interesting. People are more
interested in interacting with other peoples than with the computer.
People can carried away with the technology so much that they forget
that really evolves people is when they can use the technology to be
in touch with more people, instead of more computers. (laughs)
Interactive tv should focus on remote access to other real world
environments rather than synthetic worlds. I’d rather sit at home
getting even just a camera view of a real beach rather than just a
synthesized view. And with interactive tv, I’d rather have an
interactive two-way voice and video communication with my family than
interactive shopping. The nice thing there is that you’re not
spending more time thinking about new "content" issues but
rather how to make this use of the technology more transparent.
I like to spend family time away from the computer. I’m in front
of the computer so much I prefer to take them to the park, to the
playground, to dance class. Real world stuff. I’d rather have [my
kids] building video games than playing them.
I teach "Interactive Video" at the Interactive
Telecommunications Program at Tisch School of the Arts at New York
Univerity. My long term objective is to create tools that make the
creative process more accessible to the computer. I want to teach
people how to do this as well. I’m particularly interested in
increasing people’s media literacy.
An added advantage of teaching is it helps me keep in touch with
Director users and how the product is actually used.
 I
wanted to design cars. I was doing isometric design drawings of cars -
even though I hated being in them because I got carsick- because my
father worked at Ford Motor Company painting cars. He would bring home
those brochures with the futuristic cars. Then I wanted to be an
architect and in junior high school I wanted to build computers - I
wanted to get a PhD in computer science. I also wanted to be an artist
too - a traditional artist - but this is my niche here.
I’ll be teaching...I’m not sure...it’s not even that far away
is it? I’d like to spend more time in Jamaica. I grew in two
geographically different cultures (England and Jamaica) and I’d like
my kids to spend time growing up in Jamaica and then ultimately
Africa.
One of my visions is to create the Encyclopaedia Africanus. Just
like you have the Encyclopaedia Brittannica, I’d like to create one
on and about Africa. I’d like to travel around and have everyone say
hello in their native language - there’s 1,500 different dialects.
It’s a very rich opportunity in that it’s only now that Africans
are talking for themselves, so there’s a lot of potential for fresh
material.
The long term mission is to be a part of that somehow. To go back
to the third world and bring some of this technology there and also
come back with something that I can present to other people like
travelogues or whatever.
dimage@apple.applelink.com
top |
