Content
Interviews
Click a photo to read one of the exclusive interviews
Attention to Detail - Fred Gill and Martin Green
As the guys responsible for writing the SDK, so that other programmers
may be able to create games more freely and not have to re-invent the wheel
each time, Attention to Detail were pivotal in turning the Multi-system into
something that would have been viable; And they certainly did.
As this was (strictly speaking) only their second paying job as a software
house, the Multi-system is something that they can now look back on with the
benefit of hindsight - it's interesting to hear their views, they are a
mixture of the benefit of wisdom, fond memories of their fledgling careers
and quite revealing.
Interview:
Slipstream
Could you tell me what projects you were involved in for the Konix
Multi-system?
Martin Green
We did some demos, then created a development toolkit including an art
package, graphics package and I think some other parts.
Fred Gill
We also did the audio tools and audio driver (Martin's department); the main
audio chip was a DSP, with complete flexibility on how it was programmed and
used; Martin came up with a completely flexible way of exposing the
parameters of the driver to tools so the audio team could do what they wanted
with it. We also helped develop the copy protection system. The graphic
package had its own script language, which was pretty cool (my bit of the art
package), allowing developers to save and load custom formats, as well as
extend the art package functionality.
Slipstream
How were you or your company approached to develop on this machine?
Martin Green
Jon Dean was acting as our agent / manager and he got the contract. I'm not
sure how he did this...
Fred Gill
Wyn Holloway contacted Jon - not sure how.
Slipstream
Did you ever see or get to play with a finished Konix machine?
Martin Green
I'm not really sure. I think we saw some quite close to final versions.
Fred Gill
Not final, final - but as Martin says, very close.
Slipstream
What do you think were the best features of the machine from a programming
point of view?
Martin Green
I had a lot of fun programming the DSP... but I don't really remember
otherwise.
Fred Gill
The DSP was great fun from the audio side; I had great fun on the polygon
side; the blitter supported a list mode where you could draw N lines, and
vary the distance between each line, and the length of each line drawn - the
basic building block of a polygon. There was also a cool "remember the last
pixel colour" mode, which meant if you were very clever you could draw
polygons almost for free, by clearing the display to the "remember last"
colour, then just drawing the lhs of each - some horrendous sorting problems
to solve though to make this work...
Slipstream
Very few of the games know about utilised the bike/car/plane controls, what
are your thoughts about this.
Martin Green
I think developers didn't have the kit for long enough to figure out how to
use it.
Fred Gill
I thought it was a great idea.
Slipstream
Was it comparatively difficult to program for this machine?
Martin Green
I don't think so....
Fred Gill
It was very easy to program for.
Slipstream
Was being involved in the KMS any different to coding for other consoles or
computers?
Martin Green
Consoles are always different to PCs, but a lot of the features on the KMS
are similar to later consoles - e.g. separate processors for DSP, Blitters
etc.
Slipstream
What do you think could have made the machine better? (More memory, different
processor, higher resolution, different storage, faster Blitting etc.)
Martin Green
All of the above! But I'm pretty sure we identified some weaknesses in the
design that we thought could have been fixed.
Fred Gill
All of the above! We did manage to convince Flair to put a palette in for the
256 colours; originally 254 colours were fixed, with only 2 that were
programmable with a palette! And we also helped push a change in the core CPU
from Z80 to a variation of the 8086...
Slipstream
On the change to an 8086 variant: this is an unusual choice for a console
processor as the Motorola 68000 seemed such a popular 16 bit processor, why
choose the Intel part, was it cost or did it offer something the Motorola
didn't?
Fred Gill
We pushed for something better and faster than Z80, but the final choice came
down to Flare and Wyn; my guess is that it offered the bigger address space,
but at lower cost / integration cost than the 68000.
Slipstream
Were you interested, fascinated or excited in any way by the machine?
Martin Green
I'm pretty sure we were - at the time, I think it was pretty powerful.
Fred Gill
Excited - it was pretty powerful at the time.
Slipstream
Did you have any direct involvement with Wyn Holloway or the guys from Flare
or were you communicating your observations or concerns through Jon Dean?
Martin Green
We worked very closely with the hardware guys at both MSU and Flare - we paid
for Robert Kent's honeymoon in a room over a pub :)
Slipstream
How do you think the machine would have done if it did get released - If you
are willing to speculate, do you think it would have been successful?
Martin Green
I doubt it would have worked, but that's mostly based on lack of games, and
lack of marketing money.
Fred Gill
Unfortunately, it would have failed - there wasn't enough software (early
adopters fell by the wayside as time went on), and the reliability of the
controller was not as good as it should have been.
Slipstream
Did your experience with the Konix act as a good foundation for programming
the Jaguar?
Fred Gill
It helped a little, although the chipset on Jaguar was much more powerful- in
addition to the 68000 there were 2 co-processors, one of which could be
dedicated to audio and the other which was used to drive the Blitter chip
(freeing up the 68000) - allowing for some concurrency...
Slipstream
Do you think the Slipstream controller design would have helped sell the
machine or hindered it in any way? People may have considered it a
novelty.
Martin Green
I suspect it was a novelty... controller design is very hard to get good
gameplay from.
Fred Gill
I echo Martin's views here - at the time we thought it was amazing, but
looking back...
Slipstream
What did you think of the Power Chair and the other peripherals such as the
light gun with re-coil and the suggested ski controller?
Martin Green
Wyn had lots of ideas - the problem was making them happen. These things
exist now, I guess - but each one is a major product development in it's own
right. You need lots and lots of cash ...
Fred Gill
Ground breaking at the time; the problem was manufacturing them to a high
enough quality, given the price point the market could support.
Slipstream
What did you think of the floppy disk as a storage medium? Would you have
preferred some other form of storage such as ROM Cartridge or CDROM? Did it
limit your design?
Martin Green
One point to note is that floppy was good from a production point of view -
cartridge consoles had huge problems with manufacturing delays and minimum
orders - publishers hated them. I remember this being the big reason why
floppy was considered good.
Fred Gill
I agree with Martin.
Slipstream
On paper (and in some of the tech demo's) the machine seemed quite capable at
handling 3D, lots of large colourful sprites and had great music
capabilities. In reality was it actually capable of putting all these
together to form a game? Could it have run games that its contemporaries
were?
Martin I'm not sure - Fred was more into it than I was. I think the DSP / Blitter combination was pretty good.
Fred Gill
80% yes - there were some tricks you could do on the Amiga that would have
pushed it very hard.
Slipstream
The Super Nintendo and Megadrive both had additional hardware released which
allowed them to run low poly 3D. The KMS seemed capable of this from day one.
Do you think given this that it could have been a viable competitor or maybe
even had an advantage in this area?
Martin Green
Clearly it should have been an advantage.
Slipstream
How do you think the machine would have faired elsewhere in the world, for
example the US and Japan? Would there have been a stigma attached because the
machine was British?
Fred Gill
Unfortunately, yes - although it was nearly purchased by an American
publisher (but that's another story).
Slipstream
What was the atmosphere like regarding the KMS in the development community,
when you talked about your project with friends were they interested or
intrigued by the KMS or could they see it failing?
Martin Green
I remember quite a lot of enthusiasm :)
Slipstream
How did you hear that the KMS wasn't going to be released?
Martin Green
I think we got to the show and found the stand empty!
Slipstream
Do you have any anecdotes or funny stories regarding the Konix?
Martin Green
I think Jon Dean had to remove me from the room at an early meeting because I
was being too honest about some aspect of the design... I was very young and
thought some part of the design wasn't very good.... and I told Wyn as much
rather too directly :)
Slipstream
Do you have any documentation or code at all that you would be willing to
share for this site. Simple concept art or level designs for a Konix game or
maybe source code would make a great contribution copyright permitting?
Martin Green
I don't think I have anything...
Fred Gill
I will try to dig out the original demos (don't have a floppy in any of my
PCs though!)
Slipstream
As authors of the SDK, did you have to talk with other KMS developers? What
(if you did) were their concerns, issues or opinions of the machine?
Martin Green
We presented the SDK to the other developers at some meetings. I'm afraid I
can't remember their views..
Slipstream
Looking at Jon Dean's videos, I believe I saw a sneak peek of Tunnel of Doom
running on a PC. Was it anything other than a tech demo? Was it intended to
be a full game (I only mention this, because it looked really good!)
Martin Green
I think it was just a demo - and I remember Fred put it together very, very
quickly...
Fred Gill
Started as a technical demo for a game design; this was successfully pitched
to a publisher but we weren't involved - Jon Dean took it forward with
another developer, but I don't think it was ever published
Slipstream
Was there just one dev kit for your entire team, did you have to take turns to
use it?
Martin Green
I think we had a few..
Fred Gill
We had 2-3; some for tools creation, and others for demo creation.
Slipstream
Who was responsible for the different parts of the SDK.
Martin Green
I worked on the low-level parts of the art package and wrote an FM
synthesizer module for the music program which I don't think was ever
finished.
Fred Gill
Chris & Martin worked on the art package; I did the scripting language
for it. Jim Torjussen helped on the audio side. I did demos, and so did Jon
Steele.
Slipstream
Did the whole team contribute to the demos like the big heads demo?
Martin Green
I think the demos were split amongst the team, but I'm not sure...
Fred Gill
Demos were split between Jon Steele and myself.
Slipstream
Was your involvement with the KMS your first job as a dev studio?
Martin Green
Yes.
Fred Gill
Not strictly true - we did some Atari VCS (2600) prototypes before that (they
were never published).
Slipstream
How did you go from working on the KMS to the Jaguar - obviously the Jaguar
was in some ways an evolution of the Flare Slipstream design.
Martin Green
I think we were introduced to Jaguar by the Flare team - which was very kind
of them, of course.
Slipstream
If the hardware was a pure design that existed in it's final form and hadn't
have evolved quite so much (i.e. changes like different processors and more
memory being added) the machine may have faired better. Was it fit for
purpose as a machine that was to take on the 16bit home computers and games
consoles of the time?
Martin Green
It's very hard to say. I think the basic architecture definitely had the
power to take on other machines of the time, but the problem was always
getting to market with enough games ready to go. If you look how much cash
Microsoft burnt through to get Xbox launched, you see that the main problem
is cash - OK, back with the KMS it would be less important, but still it was
the main problem.
Slipstream
It's been written that the machine's hardware had one major flaw, in that
processes had to contend for bus time. This seemed to mean that the machine
was capable of producing great music, colourful 2D graphics with lots of
pixels or 3D, but not all at the same time. Any thoughts?
Martin Green
All designs have compromises, and most machines are capable of doing much
better demos than full games. Flare produced some very strong designs, and
I'm sure they had horrible deadlines. I remember them biting their nails
waiting to hear whether a particular revision of the chip had passed testing.
But I felt that they were a little aloof from the low-level world of games
programming - they didn't completely understand what we needed, and I thought
this meant that there were unnecessary holes in the design.
Just for example, go back to the ZX Spectrum. All they needed to do was add a
single register to change the starting screen address, and scrolling games
would have been many times faster - but I don't think they saw this.
Slipstream
The resolution seemed a bit strange; most arcade games had a 320 x 200
resolution generally with 256 colours. It seems that it would have been
difficult to port arcade games of the time to this console, not only would
programmers had to have jumped through hoops to learn to make the most of the
machine, but they would then need to get quite inventive with the games
assets to mimic the arcade game. How true is this?
Martin Green
I'm not sure, but certainly major changes in resolution aren't easy to
handle.
Slipstream
Do you think this machine would have been capable of a game such as
Starwing/fox? The SNES used a Ben Cheese designed DSP and had a fairly weedy
(relatively speaking) CPU in comparison to other machines. On paper the
Multi-system looked like it could have handled it, in reality, was it up to
the job of drawing flat shaded polygons at a reasonable (i.e. 15fps or
greater) frame rate?
Martin Green
Fred? - I don't remember the machine specifics enough....
Fred Gill
I reckon you'd have gotten about 5fps with Starwing / Fox.
Slipstream
It must have been very disappointing for your hard work in producing the SDK,
and a number of demos to not be used in a saleable product, however it must
have gone some way towards helping your careers. Do you think the failure of
the console to get to market tarnished or adversely affected your careers or
helped in any way?
Martin Green
If the console had been a huge success, then I'm sure it would have boosted
our careers - but of course, you can play what-if as much as you like. ATD
continued for quite a number of years, and we made a reasonable living - so
we didn't do too badly. On the other hand, people we knew from those days -
like Jez San - went on to do extremely well for themselves... but I still
believe that the games business was mostly about luck.
Fred Gill
I agree with Martin, but it also helped us appreciate how important tools are
to game development in later years.
Slipstream
After the crash of the Konix machine, Martin Brennan and Wyn Holloway (As
MSU) took on a contract to further develop the Slipstream ASIC and produced a
prototype for another games machine for a Chinese company. The machine was
now to use a 386 processor and have a CDROM interface. Were you approached to
help produce an SDK, games or demo's for this machine?
Martin Green
I think we heard rumours, but I don't think we were approached
Slipstream
Flare sold their ASIC to Bellfruit, who used it as the basis for a few
interactive video pub quiz machines such as Telly addicts, Radio Times, Dave
Lee Travis's Treble Top and Top of the pops. Were you in any way involved as
A.T.D. in producing code for these quiz games or modifying the SDK in any way
to better suit what Bellfruit wanted?
Martin Green
Yes - we had a long involvement with Bellfruit, and even designed a
subsequent games machine for them... but that's another story.
Fred Gill
Yes - InQuizitor, which was their first Flare-based game, went straight to #1
- the front-end / attract loop to that game was created by ATD - raising the
standard for all subsequent quiz machines from all manufacturers. We did
several products with Bellfruit.
Slipstream
In my recent interview with Martin Brennan of Flare, he told me that the
Flare One was intended to take on the Amiga but be cheaper to produce because
so much functionality was packed into the one custom IC. What would you say
to this? Can you see where they were coming from in their design and do you
think that it managed to achieve that goal?
Martin Green
I think the goal was mostly achieved, but I can't remember the capabilities
of each machine well enough to be certain.
Fred Gill
I think the key phrase here is "mostly achieved" - the Amiga had quite a few
tricks up it's sleeves...