Sometime ago we talked with Hajo, the old headmaster of Simutrans and this time we’re going to know the person who is now in charge of Simutrans project, Markus “Prissi” Pristovsek.
A little bit of history
First of all, we would like to know a little more about you. What’s your real name? Where are you from? What do you do?
My name is Dr. Markus Pristovsek. I am nearly 40, kind of assistant professor of applied (semiconductor) physics, currently at the technical university of Berlin, Germany. Berlin was also the place of birth and where I spent most of my life; I have been 2,5 years in japan, near Tsukuba, where I also met my japanese wife.
When did you join to the project?
Well, you do not join Simutrans, imho. I did lot of work for Simutrans, when I returned from japan in 2003, including the pak.japan (64er size). At that time the AI did only choose their engines based on power (960kW, 1920kW and 1280kW) and could not handle electrification. I asked Hajo, and he sent me the simplay.cc and I tried to make choice a little easier. And I included UTF-8 support, from my japanese dictionary for PalmOS.
When did you take over the leadership of the project?
In 2005 I was the only one left for programming. I did never claim leadership, I just do programming and, time permitting, work on improving pak64.
What was your relationship with Hajo before and after?
Not much either, mostly email. Hajo handed me everything quite quickly, before he took a long break, I had two months with him still contributing. In February 2004 I, Hsiegeln and Dario were the only ones left for programming. I was the most productive (even though a lot was try and error, even more than gdb was not running under window 98 and linux did not run properly on my PC – no networking).
Could you make a brief summary of the most important milestones in these 12 years of Simutrans?
No; but from 2004 I think those are:
- UTF8 support
- airplanes (which also introduced stop choosing)
- path based signalling
- climates
- scriptable interface
But most is just tuning on details and introducing flexibility. And enhancing performance; simutrans now consumes about 1/4 it consumed with 88.10.5 .
Management and leadership
Do you like Simutrans as it is now? What is the direction of the project? Would you change some of the decisions you took in the past?
Simutrans is quite mature. Its code base is less messy than it was five years ago, and cleaning up and debugging last issues still works. In hindsight it would have been better if I had more knowledge of C++ when I started to contribute. Some of the early code is really not good at all.
Having a mature game makes it difficult: On one side you cannot change much without offending many players. You should not introduce too many features or else no new player will surmount the frustration barrier. But it is a changing world, and Simutrans must change. Thus we are working on multiplayer, and maybe 3D is needed to survive in the future.
Simulationwise the engine is mostly complete. Currently I am pondering about taxes, which would make the account a useful indicator of success, as with a certain monthly profit your income would saturate at a certain value. A monthly income of 1 with 5% taxes would never exceed 20 total cash (+ vehicles). This could improve long term playing motivation and is also very useful for competitive network gaming.
Do you feel the pressure with every decision you take? As Peter Parker’s grand dad said, a great power involves a great responsibility. Do you regret somehow by having taken the leadership? What motivates you to keep on working on Simutrans?
I do programming as an art. As such there are few goals and work is rather based in my mood, e.g. if I am feeling like debugging, profiling, playing, pakset working or implementing new stuff. If I would feel responsible for everything I do I could never do something at all. The “sleepless nights” after releases are also mostly over, due to the nightly system.
What criteria do you follow to approve or discard new features, aside from common sense?
Do not underestimate common sense. An Occam’s razor : Leave away everything not needed to find the core. Simutrans simulation core depth is better (or rather deeper) than any other transport game. Any thing adding more depth most give an equal reward. Also features difficult to understand are not good, since they frustrate new players and lead to eternal support … Just remember the “why is there no route for my electric engine?” back when depots showed always electric stuff. New micromanagement is certainly a no-go, as games getting larger and more complex; on the other hand too much automation kills the gaming challenge. Eyecandy is also not too likely to be incorporated, if there are no good reasons (like matching way and speed in tunnels). Finally the reaction of the community gives also a good impression of what is desired.
Do you think over this time in the community has formed what we might call a Simutrans style, one common standard of what fits and what does not in the project?
There is an idea of “simutransish” out, which is constantly changing and evolving.
From your point of view, what is and what’s not Simutrans? For example, it’s a transport simulator but it’s not a TTD clon, neither a city management game.
It is in the hand of the people what Simutrans is. Sometimes I use it as and electric aquarium, when I follow one of those wonderful complex worlds people have created. Some use it as a model railways. Some just play for the money. I think that Simutrans is much about freedom: Not only trains like TTD, not only nice graphics like locomotion, not flat lands only like transport giant. Easier to mod than any of those. All games have their advantages, but Simutrans does not try to beat those. It just tries not to fix your playing style, and neither your graphics you are using.
In percentage, how many ideas are left out by lack of time or technical support?
Maybe 80% of mine; I am full of ideas, but many of them either do not fulfill the above criteria or are cut you by my dwindling time.
If you could, would you like to dedicate more time to the project?
I would rather dedicate more time to my family and then to other stuff like my writing, guitar play, reading, cycling…
You have dedicated much time and effort, but this is no longer than hobby. You also see it from that point of view or it’s almost like a second job?
That moment I will leave it.
You dedicate most of you time to programming. Recently, several new programmers like Dwachs and Knightly have joined to the project, collaborating with patches and mods, how many active programmers are part of the development team at this moment?
Knightly and Dwachs. Honestly Knightly is very bright, imho even much better than me. Dwachs learned proper programming (I think), while I taught everything myself. (I started in 1984 with Pascal from a book; two years before I got my computer … )
Somebody at the forum said that he didn’t know what’s going on inside the development team. There are no ways of knowing or put in contact directly with them in order to exchange ideas. Is there a real background communication or each one works alone and then you put your ideas in common at the forum?
This is a hobby. Forcing somebody to work on something will kill it. But very recently there is the programmers corner, which saves the very little exchange behind the scenes for the afterworld. But 80-90% is in this forum visible for everybody.
Simutrans experimental
Let’s talk about SE. We will not ask for a technical analysis but I would like you to give your opinion about what you think this project has made or could make by Simutrans.
OpenSource projects branch, so SE is a branch. Let’s see how much it grows. It is certainly with a more steep learning curve than Simutrans standard.
Obviously, ST and SE are incompatible although they share some common elements in terms of code. But if possible, what things of SE would you like to include in ST?
Since SE incorporated mostly rejected patches, there is almost nothing.
From developer to player
With so much to do, besides your personal life, do you really have time to play Simutrans?
No, once a month or on a conference sometime.
Sometimes, not to play the game you’re developing makes you loose the player’s point of view (leisure and fun) to isolate the programmer’s point of view (work) Do you share this thought?
Of course. That’s what I do playing sessions once a month (usually one well developed game). That way gave you the traffic lights with customizable phases.
Do you really enjoy playing Simutrans as when you where just a player?
Yes; I also enjoy TTD, but this get boring after 30min, while Simutrans is much longer entertaining.
Have you tried any of the new paksets like pak96.comic, pak192.comic or pak.german? Do you like their graphic style? What about the economy system and balance?
pak.german was maintained and pa(c)ked by myself before giving it over due to lack of time. I cannot draw even if my live depend on it.
About pak96 comic, I like the houses of Alexander (see interview in the german blog). I am not so much satisfied about balancing and vehicle choices; but then it was a few month since I tried it out. Any size larger than 96 is too large for my liking. I do not like pak128, too complex and unbalanced economy, too many vehicles, too small capacity for trucks …
As a player, what do you love most and what you dislike most about Simutrans?
I love the depth of simulation and I dislike graphics (not for pak96), crappy sound support, too small font sizes on large monitors
You also manage the simutrans official pakset, pak64, which is slowly growing, slowly but steadily, adding new elements. What would you remark from the pakset against others more or less buoyant at this moment? What do you think is missing in the pakset?
Consistency, but this is hard to obtain over 10 years …
#1 by misterkanister on 2010-03-12 - 17:51
Great interview!
#2 by SimFAN99 on 2010-03-17 - 08:38
niceee, never knew!
#3 by daffi on 2010-04-01 - 04:57
Absolutely great…
It would be greatfull if you had Prissi's pictures on it