Much of Jason's work seems to have filtered in to that general 'must have' group of programs that almost everybody has a copy of. Just now I have 3 programs running which were all or partly written by Jason; Argo Zip Driver, Long Filenames and the Voyager Internet Suite.
Acorn Arcade are eternally grateful to Jason, as he was the one who kindly paid to register us the domain name 'acornarcade.com'!
Acorn Arcade: A few years back now, there was talk of a new strategy game being programmed by you, WarZone. However, all has been quiet for some months now, what's happening?
Jason Tribbeck: The main issue has been time. When I started development on WarZone, I'd just finished the CD version of Dune II, so I was on a bit of a roll. I was working freelance, but mainly for the company which became Argo, so I had pretty much all the time to myself. However, when Argo was formed, they employed me full-time for my Internet skills, and gave me what amounted to a huge pay increase. There was one drawback in that I didn't know where I stood on developing my own software at home, so it went quiet. I also had bought a second-hand kit car that needed finishing off, so I did that for a while (it's been 18 months, and it still isn't driveable!) during the warm summer months. Now I know where I stand, which is the reason why a new version of LongFiles and also another program, VoiceCon, were released simultaneously, even though I had developed most of LongFiles 2 over a year ago.
There will be more releases soon. My next program - an IRC client - will hopefully be released in the next week or two.
Acorn Arcade: You haven't worked on any Acorn games for a while yet, is there any chance of any further Argo or even independant game releases?
Jason Tribbeck: Argo's main interest now is Internet and Internet software - it's unfortunate that an Acorn game will only sell between 2,000 and 3,000 units unless it's phenomenally good, and take around 6 months to develop properly. Argo's latest offering to ISPs/IAPs is a piece of software which retails for over 13,000 pounds a piece that took a similar amount of time to develop. You've only got to sell a minuscule amount to recoup the same amount in games sales. It's a policy that I understand, but I think it is a shame. However, many of the programmers were freelance programmers that have gone on to other games development.
Argo may do some game development, but it will be only for Network Computers, and networked multi-player, because it's in the field they're in.
Doing an independent release isn't beyond me - during my "quiet" period, I did do some playing around with different ideas, such as 3D texture shaded plotting routines that may find their way into a game sometime...
Acorn Arcade: What has been the most challenging or difficult program that you've ever set out to write? Many of us wouldn't know where to start on converting a game the size of DuneII or in writing an internet suite!!
Jason Tribbeck: Each project presents itself with challenges - with Voyager (Argo's Internet suite) it was learning how to create network connections, writing diallers, integrating various programs from other programmers, and so forth.
With Dune II, I was fairly pleased with the plotting routines - they do on-the-fly dithering of the images, effectively performing some of the functions as ChangeFSI in real-time. I was quite surprised as to how well they worked. You can see it in operation when a tank moves diagonally - they dithering remains in place, even though the tank is moving.
However, I think the most challenging program I've been asked to do was something that was never released. It was neither a game, nor a piece of software an end user would see. It was in the days of RISC OS2. A company called Simis (you may remember they wrote the flight simulator toolkit) developed a piece of hardware which was a separate graphic sub-system, and they wanted to have RISC OS windows running on it, so they contracted a friend of mine's company to do it, but they either didn't have the time, or the know-how, so they asked me. I had written a fast triangle plotting routine then, so they knew I would be able to plot triangles :) So, the task was to intercept all RISC OS calls and send them to this graphics card. It was slow, because we hadn't optimised it - we just wanted to show that it could be done. Unfortunately, two things happened: RISC OS 3 came out, and the routines did not work at all on the card (it was the sprite plotting), and also Simis dropped the project because it was taking too long. However, I think it was the first 24-bit RISC OS screen display, even if it did take 30 seconds to redraw a window!
Acorn Arcade: Most of your games have been conversions from other platforms. Do you prefer converting games or writing new games from scratch?
Jason Tribbeck: A bit of both. With conversions, you get the source code, the graphics and the sound, so all you have to do is convert the source code, plot the graphics and play the sound. However, there is no scope for creativeness. James Pond II was my first conversion, and it was quite a bit flaky - I'd been doing 68000 at university, and so I knew a little of the assembly code there was, but instead of trying to work out what routines did what, I worked out what instruction did what, and converted it instruction by instruction. It took a long time, but I got there in the end. If I was to do it now, I would do it the proper way, ie. write whole routines - the game would have been much faster, and also smaller.
When you're developing a game from scratch, you don't get any of the backup support, such as working version to play with (which I'd have to admit slowed the Dune II conversion by no end!). However, you can get some nice touches in the game which you wouldn't normally be allowed to put in because of licensing restrictions.
If anyone's interested, there are three ways you can perform a conversion:
- You get a copy of the game, and then you write the routines that does the same as the game.
- You get a copy of the game, the source code (written hopefully in C) which you then compile, and modify if necessary.
- You get a copy of the game, the source code, and you convert it line-by-line.
1) tends to be the slowest development time, because it takes longer to develop. However, the end product does end up being smaller and faster than the other two.
2) tends to be the quickest development time, because most of the code is already done. However, if there's any bugs in the original, they're also in the conversion.
3) tends to be the slowest execution speed, as the routines are not optimal.
Xenon II was converted using method 1, Dune II using method 2 and JamesPond II with method 3.
Acorn Arcade: Did you make the switch from Games programming to 'serious' application work because of the money or did you just get bored of converting games with source-codes of up to 176Mb (DuneII)?
Jason Tribbeck: It wasn't really the money, it was the friendship I have with Andrew Foyle (who founded the original company) and Richard Jelbert, and their development of VTi into ArgoNet, and then into Argo Interactive PLC. They kind of led me into the path of serious programming since I'd finished Dune II, mainly because I am perceived as being a fast coder - I wrote Argo's FTP server in about 14 hours, which included the time to learn how to use Internet sockets. I'd also become the server administrator, mainly because I'd actually used a Unix machine before, whereas they hadn't and it would have been expensive to get one properly.
Acorn Arcade: A lot of your more recent work has been in the writing of frontends/music players for other people's games, would you prefer to be programming your own games?
Jason Tribbeck: A lot of my work has always been writing front-ends and the like. Quite often, it is a games programmer who doesn't know anything about WIMP programming, or music, and they need a hand in finishing the game off. I've been WIMP programming for several years now, and music programming since the days of the BBC Micro - plus the fact that music is one of my hobbies.
I have even started to produce a CD with some of my tunes on it, if anyone wants it... Please?
Acorn Arcade: A copy of your long filenames patch, LongFiles, must be present on many if not almost all Acorn computers. However, do you yourself use the program or do you still not trust it?
Jason Tribbeck: That should not have been published in the Acorn User magazine interview - I didn't get a copy of the text before it was published - but I did say that, so I can't use the "they misquoted me" reply...
LongFiles does do unusual things to the filing system structure. In RISC OS, the structure goes like this: "FileSwitch" -> "Filing System" (FileSwitch for anyone that doesn't know performs many filing system operations that each filing system would have to perform, such as wild-card scanning and so forth into filing system calls). With LongFiles, it goes "FileSwitch" -> "LongFiles" -> "Filing System". As a result, it is a bit obtrusive. However, with the latest version, it is easier to find a file if LongFiles gets confused and forgets the long filename.
I will be using LongFiles 2, especially when it is finished and combined with the FileSuite which will perform many functions, such as more than 77 entries per directory, hidden file compression and so forth. I already have a compression algorithm that learns file types in order to get the best compression available.
Acorn Arcade: Thank you for doing this interview, I wish you all the best for the future!
Jason Tribbeck: And I wish you luck for you and everyone who visits Acorn Arcade.