I always know what kind of game I want to play, but it's overwhelming to set up, if even possible. It's tons of packages, forums, idk.. so I just end up playing vanilla, with maybe some rules among the players. But I wish there was a comprehensive interface to allow me to set all the stuff that I want.
From the top of my head (haven't played in over a year) it would be: cargo destination for people, cargo prices reflect proximity of other productiosn, mono/maglev only has passenger cabs, regular rail still gets upgrades, no planes I guess, "automatic" public transport within cities - no need to build stops, just buses driving around (transferring people to train stations). And probably more I can't think of now.
I’m the same. I once suggested on the Transport Tycoon forums that there be a mechanism for curated sets of NewGRFs that work well together to be selected from the main page and was resolutely told that it’s up to the user to put together their own sets as they like.
EDIT: it looks like something like this is now in v16 as “collections”! I wonder if these can be shared.
Half joking because if you are really into TTD The Simutrans interface feels terible(It looks like TTD but does not feel like TTD, this is often why photoshop people hate the gimp so much) but...
I started with simutrans, and have a hard time with ttd. Their systems disgust me in a way that they probably should not but it is that same bugbear, it looks like simutrans but does not feel like simutrans.
If anyone does try it out, beware pak128. The paks are more like mods that change the goods and price ratios rather than the mere graphics updates that their name would suggest. As someone who plays the hell out of pak64 I can't make a functioning network in pak128. I almost suspect it is broken. The price ratios just never work out. pak192comic however, is easy, fun and looks good.
If I had to describe the difference, it is that simutrans tries a little harder with it's systems, it take a little more care to make a functioning network in simutrans. Factories have specific destinations and storage sizes. for example iron ore to steel mill will jam full if you fail to deliver the steel to the car factory and no steel will be made if you fail to deliver the coal it needs. So freight network revenue is fragile until the entire chain is complete. Passenger service is tricky because the passengers want to go somewhere(exactwhere is not specified, you have to treat it statistically). If your network does not go there they won't take your trains. So the ideal is to have service everywhere, but you you have to start small and small pax networks make no money, thus why it is tricky. The money is made on the fast long distance routes, and you take a loss on the slow urban routes just to make sure you have their destination/origin covered.
It's actually not difficult to setup at all. The game just work out of the box without any mod. However there's a learning curve. Find a tutorial somewhere online and follow it thru then you're golden.
It took me maybe an hour to learn the first time I played it. It's worth it! :D
Tho after finishing the game a few times it gets old.
That’s not what they said. The game does work without mods, but the economy is weirdly nonsensical (AFAIK you can push unnecessary amounts of cargo over long distances, and the game rewards you anyway) and I personally find it kind of pointless, compared to something like Transport Fever 2.
Yeah cargodist's indeed broken. You could milk plenty amount of money by just setting up two airports far away.
Maybe try Simutrans. It's got actual economy simulation. I tried it and I didn't like it tho. I couldn't make a profit in that game and kept losing. :(
No. A coal plant will happily pay millions of dollars for coal hauled from the far side of the map, despite there being a coal mine right next door. Not like real life at all.
This would be awesome for OpenMW, too. Morrowind is kind of dated. But I like the original style. So a curated easy install of improvement mods as a service would be great.
I made version 12.2 my programming hobby project. First time I tackled it years ago was to have it run at real time instead of the default fast pace (even with the day length patch). Worked great including a feature to really schedule routes and have working yards for vehicles to wait. But when I wanted to change internal mechanics to add other features I hit the spaghetti code wall several times (and the code after 12.2 is just more complex with every version bump).
Took a break from it for a while and picked it up again, this time by not doing anything much feature-wise but pulling the code apart and de-C++-ing it. Bit by bit building a categorised code tree with single functions per file, deconstructing classes to plain functions and data structs, with the goal to convert from C++ to clear C. And then one day consolidate code to have for instance the same diagonal road options as current train tracks.
Might be pointless and take a long time to finish, but it's comforting to turn something overly complex into simple, small pieces and also create a unique code formatting style to visualise elements of a line of code for easier reading. Another fun part is reading about data design and coming up with a custom data container to replace the std::vector uses.
How fast do the trains and buses run on "real time" speed?
I suppose a 15 meter bus going 50km/h travels roughly its own length (13.40 meters to be exact) every second, so it takes 3 seconds to pass a grid square - checking out a YouTube video, it seems that's already the speed right now!
I calculated it long ago but I can't find the original notes. It can be extrapolated from the default 74 ticks a day (and the distance represented by 1 tile) to ((33 ticks per cpu second) * 86400) in my patch, and I think it came about a tile size of around 8 metres or so. Too long ago to be sure, but I found a time speedup of 4x compared to realtime was a good compromise. It made transport by slow vehicles viable to keep factories running, or a lot less were needed to keep up deliveries.
The newer setting for minutes per tick made realtime possible eventually. I had it set to 11, I think.
It was also quite easy to limit production (or passengers) in the code from 06:00 to 20:00 a day for a little extra realism with either emptying a station during the night or have no traffic servicing buildings. Even weekends are possible. Warehouses would be a nice addition.
Deploying and building OpenTTD if my favourite exercise when i am trying to learn new CI tool or orchestrator or packaging format. Simple enough, yet fun.
Sorry if it's a meta post, or a faq, or if I'm just missing some HN lore here, but I finally have to ask : why is it that, everytime there's a new release for OpenTTD (and 0ad, for that matters), it makes it to the front page ?
I can kinda get why there are often news about Factorio, for example (it's not subtly related to programming), and there is nothing wrong with open source projects getting visibily even when they're not "yet another web framework" or "the last open weight LLM wrapper".
But why those games in particular ? Are the authors regulars of HN ? Is there an interesting forking history / backstory ?
I speak only for myself, but: I can't think of any other computer game that I've spent time on, and that I look forward to playing again, over such a long span of time.
While I don't play it often, I have been playing it for over 20 years.
It's impressive (again, at least to me) for its longevity if nothing else. Just seeing that it continues to progress makes me happy in dumb ways.
It has the competitiveness of games like Civ, The Settlers and so on. But with the world building aspects of games like Sim City.
There aren’t many other games that strike this balance as well.
I first started playing the original DOS game in the 90s and I honestly think OpenTTD looks and plays just as well now a the original did 30+ years ago. The level of detail in the pixel art is ideal for the type of game it is.
OpenTTD and 0ad are open-source games that have been under development for many years, which tickles the hacker sensibility. I think these two games fit into the HN mindset much more than Factorio, which is a newer, proprietary game.
Because Transport Tycoon was back then a HUGE success, and it was developed by Chris Sawer, a some-how game-dev-celebrity because of his former successes
I think it's just that it's a cool game that's popular with HNers. Same with 0ad and other software like GnuCash. Always interesting to see what new features they have introduced and even just read how others play the game.
For some people it's the game. Like I played the original (not even TTD, just TT) and this is still a very fun game.
In a nutshell it's an optimization sandbox since the beginning, with more and more ways to invent and customize solutions for the troubles you make yourself.
The flip, I think, is realistic enough. It could be just slower and the speed of turnaround reflect whether the loco needs to turn around. But the slow reverse that can be solved with two locos is a nice upgrade too.
Wonder if the Yogscast Jingle Jam will use v16 in this year's stream. Might stop Duncan crashing so many trains if they're all driving slower backwards.
hah I played for an hour this new beta. It's fun. I haven't played for years.
I couldn't workout the train signals though, there are 4 of them and they have a digit as well with them, not sure what's going on there. I'm probably not the target audience :)
For most signalling you want the path signals. They come in 2 flavours: both kinds only do the path in one direction, one kind is non-blocking to trains running the opposite way, the other kind is blocking, one way only.
This combo lets you set up an upline and downline with a terminal station: you have one way only signals on the line, and non-blocking signals on the station entrances pointing into the station, and a crossing between. Trains will get to the end of the upline, and wait to get a path to the end of the station. Then they'll turn around and wait to get a path to the downline.
>This combo lets you set up an upline and downline with a terminal station:
That actually is nothing to do with path based signals, and could be done quite simply with purely vanilla TTD signals.
What path-based signaling allows is for multiple trains to make non-conflicting moves through a single block at the same time, which basic signals wouldn't.
That's not quite true. The original signals would let a train into the cross over block when no station lines were available because they didn't have that pass through. Depending on the layout it would either get stuck or end up heading in the wrong direction. Pre-signals could solve it but they weren't in the original.
You just put a signal at both ends of the passing loop. Just the way they’re signaled in real life - one signal to authorize a train into the loop, one to authorize movements out of the loop.
Path based signals (nor pre-signals for that matter) didn’t exist in plain original TTD and constructing such track layouts was easy and straightforward.
> Separating locomotives from trains next?
Someone wrote a patch for version 8 or 9 years ago which enabled coupling and decoupling, although one needed to set it up in a specific order to avoid failing couplings or train crashes. It's a shame the devs never picked it up.
> And then, finally, hump yards?
I doubt a hump yard will ever be a feature because as far as I see it, it would need a complete overhaul of the way orders are set up and handled. Right now it's one semi-static order list for the whole train and a sorting yard needs dynamic order lists per locomotive and each wagon.
I always know what kind of game I want to play, but it's overwhelming to set up, if even possible. It's tons of packages, forums, idk.. so I just end up playing vanilla, with maybe some rules among the players. But I wish there was a comprehensive interface to allow me to set all the stuff that I want.
From the top of my head (haven't played in over a year) it would be: cargo destination for people, cargo prices reflect proximity of other productiosn, mono/maglev only has passenger cabs, regular rail still gets upgrades, no planes I guess, "automatic" public transport within cities - no need to build stops, just buses driving around (transferring people to train stations). And probably more I can't think of now.
I’m the same. I once suggested on the Transport Tycoon forums that there be a mechanism for curated sets of NewGRFs that work well together to be selected from the main page and was resolutely told that it’s up to the user to put together their own sets as they like.
EDIT: it looks like something like this is now in v16 as “collections”! I wonder if these can be shared.
If you're happy with not doing cargo and can tolerate an also slightly weird pax distribution, I highly recommend [NIMBY Rails](https://store.steampowered.com/app/1134710/NIMBY_Rails/).
Thanks for recommending this! It’s looks exactly to my taste, but I hadn’t heard of it.
Anyone have any favorite ai suggestions? I am never sure which ones are too easy, too hard or simply crash all the time.
Simutrans?
Half joking because if you are really into TTD The Simutrans interface feels terible(It looks like TTD but does not feel like TTD, this is often why photoshop people hate the gimp so much) but...
I started with simutrans, and have a hard time with ttd. Their systems disgust me in a way that they probably should not but it is that same bugbear, it looks like simutrans but does not feel like simutrans.
If anyone does try it out, beware pak128. The paks are more like mods that change the goods and price ratios rather than the mere graphics updates that their name would suggest. As someone who plays the hell out of pak64 I can't make a functioning network in pak128. I almost suspect it is broken. The price ratios just never work out. pak192comic however, is easy, fun and looks good.
https://www.simutrans.com it is also free on steam, probably the easiest way to play.
If I had to describe the difference, it is that simutrans tries a little harder with it's systems, it take a little more care to make a functioning network in simutrans. Factories have specific destinations and storage sizes. for example iron ore to steel mill will jam full if you fail to deliver the steel to the car factory and no steel will be made if you fail to deliver the coal it needs. So freight network revenue is fragile until the entire chain is complete. Passenger service is tricky because the passengers want to go somewhere(exactwhere is not specified, you have to treat it statistically). If your network does not go there they won't take your trains. So the ideal is to have service everywhere, but you you have to start small and small pax networks make no money, thus why it is tricky. The money is made on the fast long distance routes, and you take a loss on the slow urban routes just to make sure you have their destination/origin covered.
If you're starting out, I really recommend the Master Hellish youtube channel, here's one introductory video: https://www.youtube.com/watch?v=P_pGUdMjWVs
It's actually not difficult to setup at all. The game just work out of the box without any mod. However there's a learning curve. Find a tutorial somewhere online and follow it thru then you're golden.
It took me maybe an hour to learn the first time I played it. It's worth it! :D
Tho after finishing the game a few times it gets old.
That’s not what they said. The game does work without mods, but the economy is weirdly nonsensical (AFAIK you can push unnecessary amounts of cargo over long distances, and the game rewards you anyway) and I personally find it kind of pointless, compared to something like Transport Fever 2.
You're right. I misread. I guess I need coffee.
Yeah cargodist's indeed broken. You could milk plenty amount of money by just setting up two airports far away.
Maybe try Simutrans. It's got actual economy simulation. I tried it and I didn't like it tho. I couldn't make a profit in that game and kept losing. :(
> the economy is weirdly nonsensical, you can push unnecessary amounts of cargo over long distances, and the game rewards you anyway
Wait... same like in real life you mean?
No. A coal plant will happily pay millions of dollars for coal hauled from the far side of the map, despite there being a coal mine right next door. Not like real life at all.
This would be awesome for OpenMW, too. Morrowind is kind of dated. But I like the original style. So a curated easy install of improvement mods as a service would be great.
I made version 12.2 my programming hobby project. First time I tackled it years ago was to have it run at real time instead of the default fast pace (even with the day length patch). Worked great including a feature to really schedule routes and have working yards for vehicles to wait. But when I wanted to change internal mechanics to add other features I hit the spaghetti code wall several times (and the code after 12.2 is just more complex with every version bump).
Took a break from it for a while and picked it up again, this time by not doing anything much feature-wise but pulling the code apart and de-C++-ing it. Bit by bit building a categorised code tree with single functions per file, deconstructing classes to plain functions and data structs, with the goal to convert from C++ to clear C. And then one day consolidate code to have for instance the same diagonal road options as current train tracks.
Might be pointless and take a long time to finish, but it's comforting to turn something overly complex into simple, small pieces and also create a unique code formatting style to visualise elements of a line of code for easier reading. Another fun part is reading about data design and coming up with a custom data container to replace the std::vector uses.
How fast do the trains and buses run on "real time" speed?
I suppose a 15 meter bus going 50km/h travels roughly its own length (13.40 meters to be exact) every second, so it takes 3 seconds to pass a grid square - checking out a YouTube video, it seems that's already the speed right now!
I calculated it long ago but I can't find the original notes. It can be extrapolated from the default 74 ticks a day (and the distance represented by 1 tile) to ((33 ticks per cpu second) * 86400) in my patch, and I think it came about a tile size of around 8 metres or so. Too long ago to be sure, but I found a time speedup of 4x compared to realtime was a good compromise. It made transport by slow vehicles viable to keep factories running, or a lot less were needed to keep up deliveries. The newer setting for minutes per tick made realtime possible eventually. I had it set to 11, I think.
It was also quite easy to limit production (or passengers) in the code from 06:00 to 20:00 a day for a little extra realism with either emptying a station during the night or have no traffic servicing buildings. Even weekends are possible. Warehouses would be a nice addition.
Folks should also checkout OpenLoco which is a full reimplementation of Chris Sawyers Locomotion: https://openloco.io
Deploying and building OpenTTD if my favourite exercise when i am trying to learn new CI tool or orchestrator or packaging format. Simple enough, yet fun.
Drive-by questions for any OpenTTD aficionados: Can anyone recommend a how-to to get the original MIDI music files working on the Android release?
The music is uncommonly good in the original releases.
Sorry if it's a meta post, or a faq, or if I'm just missing some HN lore here, but I finally have to ask : why is it that, everytime there's a new release for OpenTTD (and 0ad, for that matters), it makes it to the front page ?
I can kinda get why there are often news about Factorio, for example (it's not subtly related to programming), and there is nothing wrong with open source projects getting visibily even when they're not "yet another web framework" or "the last open weight LLM wrapper".
But why those games in particular ? Are the authors regulars of HN ? Is there an interesting forking history / backstory ?
Any way, kudos for the release, of course :)
I speak only for myself, but: I can't think of any other computer game that I've spent time on, and that I look forward to playing again, over such a long span of time.
While I don't play it often, I have been playing it for over 20 years.
It's impressive (again, at least to me) for its longevity if nothing else. Just seeing that it continues to progress makes me happy in dumb ways.
Yeah same here.
It has the competitiveness of games like Civ, The Settlers and so on. But with the world building aspects of games like Sim City.
There aren’t many other games that strike this balance as well.
I first started playing the original DOS game in the 90s and I honestly think OpenTTD looks and plays just as well now a the original did 30+ years ago. The level of detail in the pixel art is ideal for the type of game it is.
OpenTTD and 0ad are open-source games that have been under development for many years, which tickles the hacker sensibility. I think these two games fit into the HN mindset much more than Factorio, which is a newer, proprietary game.
Because Transport Tycoon was back then a HUGE success, and it was developed by Chris Sawer, a some-how game-dev-celebrity because of his former successes
I think it's just that it's a cool game that's popular with HNers. Same with 0ad and other software like GnuCash. Always interesting to see what new features they have introduced and even just read how others play the game.
Hell, just read the article on a new feature added in this release: https://www.openttd.org/news/2026/06/25/backwards-driving
Totally a self-inflicted pain with a ton of ways to solve it!
For some people it's the game. Like I played the original (not even TTD, just TT) and this is still a very fun game. In a nutshell it's an optimization sandbox since the beginning, with more and more ways to invent and customize solutions for the troubles you make yourself.
About the history of the project you better hear from the horse's mouth: https://www.openttd.org/news/2024/03/06/happy-birthday
The flip, I think, is realistic enough. It could be just slower and the speed of turnaround reflect whether the loco needs to turn around. But the slow reverse that can be solved with two locos is a nice upgrade too.
Wonder if the Yogscast Jingle Jam will use v16 in this year's stream. Might stop Duncan crashing so many trains if they're all driving slower backwards.
hah I played for an hour this new beta. It's fun. I haven't played for years.
I couldn't workout the train signals though, there are 4 of them and they have a digit as well with them, not sure what's going on there. I'm probably not the target audience :)
Sorry to be the one to say RTFM but … https://wiki.openttd.org/en/Manual/Building%20signals
Yeap. That's exactly what I meant by not the target audience.
For most signalling you want the path signals. They come in 2 flavours: both kinds only do the path in one direction, one kind is non-blocking to trains running the opposite way, the other kind is blocking, one way only.
This combo lets you set up an upline and downline with a terminal station: you have one way only signals on the line, and non-blocking signals on the station entrances pointing into the station, and a crossing between. Trains will get to the end of the upline, and wait to get a path to the end of the station. Then they'll turn around and wait to get a path to the downline.
>This combo lets you set up an upline and downline with a terminal station:
That actually is nothing to do with path based signals, and could be done quite simply with purely vanilla TTD signals.
What path-based signaling allows is for multiple trains to make non-conflicting moves through a single block at the same time, which basic signals wouldn't.
That's not quite true. The original signals would let a train into the cross over block when no station lines were available because they didn't have that pass through. Depending on the layout it would either get stuck or end up heading in the wrong direction. Pre-signals could solve it but they weren't in the original.
You just put a signal at both ends of the passing loop. Just the way they’re signaled in real life - one signal to authorize a train into the loop, one to authorize movements out of the loop.
Path based signals (nor pre-signals for that matter) didn’t exist in plain original TTD and constructing such track layouts was easy and straightforward.
Separating locomotives from trains next?
And then, finally, hump yards?
> Separating locomotives from trains next? Someone wrote a patch for version 8 or 9 years ago which enabled coupling and decoupling, although one needed to set it up in a specific order to avoid failing couplings or train crashes. It's a shame the devs never picked it up.
> And then, finally, hump yards? I doubt a hump yard will ever be a feature because as far as I see it, it would need a complete overhaul of the way orders are set up and handled. Right now it's one semi-static order list for the whole train and a sorting yard needs dynamic order lists per locomotive and each wagon.
cool hopefully this stupid crash bug is fixed now!