The International Simutrans Forum

 

Author Topic: Current development projects: coding help welcome  (Read 28530 times)

0 Members and 1 Guest are viewing this topic.

Offline AndrewTraviss

  • *
  • Posts: 14
  • Languages: EN
Re: Current development projects: coding help welcome
« Reply #35 on: June 13, 2016, 11:20:52 AM »
Alright, I have some ideas based on what you wrote. I think I will create a thread to discuss the overall design of maintenance, rather than focusing on just one or two of these features, as it seems maintenance is being completely overhauled as a concept between all of the outstanding items.

Should this discussion happen in the Development board, or on the main Experimental Discussion board?

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: Current development projects: coding help welcome
« Reply #36 on: June 13, 2016, 01:26:31 PM »
Sorry for the delay in replying: if you are discussing the merits of possible new features, rather than just the mechanics of implementation, may I suggest that the thread be posted in the discussion board rather than the development board?

I shall look forward to reading it in any event.

Offline AndrewTraviss

  • *
  • Posts: 14
  • Languages: EN
Re: Current development projects: coding help welcome
« Reply #37 on: June 17, 2016, 03:58:54 AM »
Now that I'm actually digging into the code, I suspect that all of the German is going to be more of an obstacle for me than C++

Offline Rollmaterial fi

  • Devotee
  • *
  • Posts: 567
  • Languages: EN, FR, DE, FI, SE
Re: Current development projects: coding help welcome
« Reply #38 on: June 17, 2016, 11:06:40 AM »
Now that I'm actually digging into the code, I suspect that all of the German is going to be more of an obstacle for me than C++

This thread should help you with that.

Offline AndrewTraviss

  • *
  • Posts: 14
  • Languages: EN
Re: Current development projects: coding help welcome
« Reply #39 on: June 17, 2016, 03:53:22 PM »
This thread should help you with that.

Oh, excellent. Thanks!

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: Current development projects: coding help welcome
« Reply #40 on: June 17, 2016, 08:12:52 PM »
I have tried to translate some things in comments where possible. If you get stuck with anything, let me know.

Offline NoMorePacers

  • *
  • Posts: 183
  • Languages: EN
Re: Current development projects: coding help welcome
« Reply #41 on: June 19, 2016, 06:34:12 PM »
Would the liquid insolvency for network games only thing require only a bit of tinkering with simuconf.tab, or is it wider-scale?

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: Current development projects: coding help welcome
« Reply #42 on: June 19, 2016, 09:29:15 PM »
It would require changes to the code: if it were only changes to the simuconf.tab file, it would have been done long ago.

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: Current development projects: coding help welcome
« Reply #43 on: November 08, 2016, 09:01:54 AM »
I have updated this to add reference to a possible future feature, waiting time limits based on individual stops, and revenue from stop extension buildings.

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: Current development projects: coding help welcome
« Reply #44 on: November 14, 2016, 02:04:54 PM »
I have added two further items to the list, both relatively minor features affecting trams and road vehicles.

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: Current development projects: coding help welcome
« Reply #45 on: December 10, 2017, 04:07:44 PM »
I have updated this list to acknowledge the completion of the low bridges feature from earlier in the year, to move completed projects to another post as space was running low in the original post and to add a new project, industry road connexions.

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: Current development projects: coding help welcome
« Reply #46 on: December 18, 2017, 11:42:53 PM »
I have updated this to add several features discussed and considered recently, as well as to update the text of the proposed car parking feature to fit in with recent developments in passenger generation and classes.

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: Current development projects: coding help welcome
« Reply #47 on: December 21, 2017, 11:12:31 PM »
I have updated this to add fractional power and tractive effort values.

Offline freddyhayward au

  • *
  • Posts: 32
  • Languages: EN
Re: Current development projects: coding help welcome
« Reply #48 on: August 12, 2019, 11:28:26 PM »
I understand that current development is focused around bug fixing and other more pressing issues (and I lack the skills to contribute code), but I wanted to share my thoughts on private car behavior. The key limitation as I see it is that only one vehicle per direction can occupy a 125m stretch of road at any given time. Currently, ludicrous situations arise from this such as small towns becoming gridlocked. cars also slow down large maps significantly, but disabling them results in congestion dropping to 0, reducing public transport usage. Regardless of CPU performance, it would be impossible to realistically simulate private car traffic on a one-car-per-tile basis. I see three possible approaches to this problem, each of increasing complexity. All, I believe, would produce far more realistic results than the current.

1) retain physical car simulation but represent multiple simulated cars as one physical car. For example, at 5 simulated cars per physical car, an 8-tile (1km) stretch of road would have a capacity of 40 private cars per direction at a given time rather than 8.

2) abandon physical car simulation. calculate traffic, routes & congestion per town. adjust per-town speed limits based on congestion. A particularly congested city might limit the speed of all buses within its borders to 35 km/h.

3) abandon physical car simulation. calculate traffic, routes & congestion per tile. adjust per-tile speed limits for player road vehicles: for example, a bus travelling on a congested section of a city street might be limited to 25km/h.

Offline ACarlotti

  • *
  • Posts: 474
Re: Current development projects: coding help welcome
« Reply #49 on: August 13, 2019, 01:27:02 AM »
There is another option that would partly help - allowing multiple vehicles to occupy a single tile while travelling in a single lane.
I think the graphics scale is approximately 30m per tile*, compared to a default physics scale of 125m per tile. So effectively we are simulating modern cars as if they are 16m long instead of 4m long. If we allow multiple vehicles in a tile, then a stationary queue would have approximately 1/3 of the capacity it would have in real life (allowing for a bit more space between cars in real life). The impact only flowing traffic would be less - maybe 50%-80% of the real capacity for a city road, and perhaps >90% of the real capacity for motorways.

Offline freddyhayward au

  • *
  • Posts: 32
  • Languages: EN
Re: Current development projects: coding help welcome
« Reply #50 on: August 13, 2019, 03:05:00 AM »
Another advantage of that approach would be its application to player vehicles. In reality, an advantage of buses (and trams, though these are handled as rail vehicles) is that multiple vehicles from different routes can board at a single stop simultaneously. This is currently impossible on two-way road tiles. As someone familiar with the game internally, how feasible would your option be? OpenTTD has this feature:

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: Current development projects: coding help welcome
« Reply #51 on: August 13, 2019, 10:05:22 AM »
I agree with A. Carlotti - this would be a preferable solution and also assist with player vehicle interactions with each other. However, this would be a substantial and complex coding project interacting with some code that has so far remained untouched from Standard and there are many higher priorities for my time at present, and I am hampered in what I can do by needing to upgrade my computer, but the parts for doing so being not available until next month.

If anyone else would like to implement the feature of road vehicles being able to occupy fragments of a tile (tiles are divided into 1/16ths for some purposes and vehicle lengths can be specified in their .dat files as these 16ths), so there are at least some fragments of code in the existing game that can in principle be extended to support this.

Offline ACarlotti

  • *
  • Posts: 474
Re: Current development projects: coding help welcome
« Reply #52 on: August 13, 2019, 02:25:00 PM »
As someone familiar with the game internally, how feasible would your option be?
It would be a substantial undertaking, which should possibly be tied in with changes to how junctions and lanes work. I think there has already been suggestion of reworking some of this stuff in Standard, but it is definitely a long-term ambition rather than and easy patch.