Concept Phase Oct09-May12

[Alberth]

I want to make a new GUI system, but mine will not be 40.000 lines of code long. And if it does. The lines will be userfriendly coded

Could you elaborate on this? I don't understand your phrase 'userfriendly coded'. What does that mean?
How is our code not?

To be honest. I didnt even have a close look at the GUI code. I just looked trough all the source code files of ottd and i dont feel very comfortable with it. I dont want to discuss about that. I have no arguments to provide. So whatever you will reply, i will say: You are right.

I don't think 'my' system is the best one, it is not even original, I just stole bits and pieces from other GUI systems, and put them together in a way that I think will work for OpenTTD
I am just very curious about what you mean and why you think it is not good (ie what you will do differently). I'd like to know what it is that bothers you, as I may be able to change it.

[Dante123]

To fasten the process of generating new graphics for the game, I will build in a 3D-renderer that generates 2D-graphics of the provided models. So that your 3D-knowledge will be very useful.

the is already a program called GTKradiant for 3d building. this was written for games like Quake 3 and enemy territory. maybe could be usefull for you for some base or so?
i believe its open source, and mostly abandoned in development.
but its a very easy to use program, and could be of some use to you (to prevent having to make some from scratch)

[Dante123]

Tracks will be smoother, so that will look more realistic. To manage that in code, there will be finished modules like a switch. You build complete track modules instead of one-tile-tracks like in ottd.

Trains will not have one-tile-long vehicles anymore. All engines and waggons will have a length of two tiles!

saw a strange thing in 1 of your drafts. i pointed out in the image. i presume you keep the signal building the same (or the biggest part of it) from ottd? at least the normal sig, entering sig, in-between sig and exiting signals ?

[smallfly]

saw a strange thing in 1 of your drafts. i pointed out in the image. i presume you keep the signal building the same (or the biggest part of it) from ottd? at least the normal sig, entering sig, in-between sig and exiting signals ?

I plan to make the setting of a switch visible to the user, so that you can see in which direction a train goes when passing it. Because the switches are set to straight-ahead (instead of turning) by default, two trains can easily pass each other. No signals necessary at this point.

[smallfly]

I am just very curious about what you mean and why you think it is not good (ie what you will do differently). I'd like to know what it is that bothers you, as I may be able to change it.

I will realize the GUI complety in object-oriented (class based) code. (Please dont say, you also did it that way ... )

[Alberth]

I am just very curious about what you mean and why you think it is not good (ie what you will do differently). I'd like to know what it is that bothers you, as I may be able to change it.

I will realize the GUI complety in object-oriented (class based) code. (Please dont say, you also did it that way ... )

Ok, I won't.

All windows do not derive from the Window class, and all nested widgets are not in a class hierarchy starting with the NWidgetBase class. The WindowDesc class does not contain general window construction information.


I am anxiously awaiting your design. Please contact me again in a year or two when you have a full implementation. (Assuming you don't work on other parts of the game, otherwise I'll have to wait longer.)

[Dante123]

saw a strange thing in 1 of your drafts. i pointed out in the image. i presume you keep the signal building the same (or the biggest part of it) from ottd? at least the normal sig, entering sig, in-between sig and exiting signals ?

I plan to make the setting of a switch visible to the user, so that you can see in which direction a train goes when passing it. Because the switches are set to straight-ahead (instead of turning) by default, two trains can easily pass each other. No signals necessary at this point.

mm but what if, 1 train is going straight, and the other need to drive off the main track, by crossing the road ? this is a pretty realistic situation.
like i drafted in the attached image. at what point will you need to point out the waiting points for the trains so the 1 wont block the other.
with switched, you mean the get automatic set ? not like the player has to set them manual ?

[smallfly]

with switched, you mean the get automatic set ? not like the player has to set them manual ?

I dont get what you mean, perhaps my english is too bad. But if you wanted to ask if players have to set switches manually, I answer "oh my god, no! the switches are set automatically like in ottd; you just see in which direction the next train passing will go to"

[smallfly]

I added three new sketches to the first posts of this thread. Have a look

[Sarvesh50]

Do u need testers?
I have to say your project is like a dream but if this is a one man project it will take a real long time for the release right?

But the only thing i can say for now is

[KillerMapper]

Looks really nice!

[smallfly]

Do u need testers?

That would be nice But temporarily the project is in a "pre-concept" phase. The only thing needed: ideas & wishes

I have to say your project is like a dream

A dream, I'm dreaming for many years now. And I will not stop making efforts to get such a game produced. Even if will take many years...

but if this is a one man project it will take a real long time for the release right?

It wont be a one man project. I can manage the concept phase alone and even the coding of the game itself. But I will not make the 3D-models of trains, buildings etc. I will also need someone explaining how the network model of ottd functions, what pitfalls to keep in mind, etc. Im very glad to have found the right programming language for this project as well as a portable graphics library (C/C++ & SDL). I tried to program the game with Java and C# but didnt get far (because of performance issues). I also need someone telling me which other libraries to use.

Im looking for libraries that are portable (windows, mac, linux) and useful for the game. Which libraries are useful? Which ones save developing time? Im sure I do not even know all kinds of libraries. Which ones exist? I know graphics and physics libraries and some more. Which ones do you know?

But the only thing i can say for now is

Looks really nice!

Nice to read, that you obviously like the idea.

[KillerMapper]

Ok course

I would like to beta test too btw.



Just to be sure, your game will be like OTTD aka managing trucks/buses/trains/... for transporting goods, passengers...?

Hope that your passenger transport will be better than in OTTD (especially with train orders)

[smallfly]

Just to be sure, your game will be like OTTD aka managing trucks/buses/trains/... for transporting goods, passengers...?

Yeah. The principles will be the same.

Hope that your passenger transport will be better than in OTTD (especially with train orders)

It will not be comparable to OTTD. In P1 every citizen will have an identity: A citizen has a flat, a job, a car. Every citizen knows exactly where he wants to go to and uses the fastest/cheapest way to work/shopping/vacationing etc. Together with all those nice features like timetabling, shared orders, etc. you wont miss anything in that game

(Its easy philosophize as long as there is no real game but only a concept )

[Dante123]

before i'm to late, i also want to be beta tester

btw, not only for the game, but also when your 3d drawing prog is ready, i think i could provide you with some good feedback maybe

[lawton27]

Just to be sure, your game will be like OTTD aka managing trucks/buses/trains/... for transporting goods, passengers...?

Yeah. The principles will be the same.

Hope that your passenger transport will be better than in OTTD (especially with train orders)

It will not be comparable to OTTD. In P1 every citizen will have an identity: A citizen has a flat, a job, a car. Every citizen knows exactly where he wants to go to and uses the fastest/cheapest way to work/shopping/vacationing etc. Together with all those nice features like timetabling, shared orders, etc. you wont miss anything in that game

(Its easy philosophize as long as there is no real game but only a concept )

If every individual person is done separately, surely that would require allot of processing power to be done in a suitable about of time.

[smallfly]

If every individual person is done separately, surely that would require allot of processing power to be done in a suitable about of time.

I know that. Of course, you wont be able to calculate millions of citizen. (at least not now, in the year 2009; but perhaps in 2020 when the game is finished ) The block based building system (one block contains one person) will lead to a less number of citizen than in ottd, so that the need of processing power decreases.

For the first version of the game I'm planning maxima like: (which will be more than enough because of the higher complexity of building rail systems, road systems, airports, flight routes, harbours, ship routes ...)

500 planes (unsigned short int)
1.000 trains (unsigned short int)
5.000 trucks (unsigned short int)
50.000 citizens (unsigned short int)

citizens will have
- location (unsigned short int 2 Bytes, unsigned short int 2 Bytes)
- flat id (unsigned short int 2 Bytes)
- job id (unsigned short int 2 Bytes)
- additional info bits like "on the way to work", "on the way to shopping" etc. (2 Byte)
all together 10 Bytes per citizen => total: 500.000 Bytes => ca. 500 kB

Is that processable for a real time game?

[Alberth]

Sounds somewhat simcity-ish.
Maybe read up on their algorithms to see how they do it?

Also, I assume you know the global OpenTTD code/data structure, and its weak points? If not, read the discussions on speeding up the game (usually multi-core discussions). It also gives you an idea on how far the OpenTTD game scales. If you want more, you'll need a very different code/data structure than OpenTTD has.


At the end of the day, there is of course only one way to find out if it is feasible, namely implement it.
You may want to make a version first that tests such things without any graphics, etc.

[Zuu]

As for libraries you have to explore a bit what is out there.

As a complement to SDL there are SDL_Image and SDL_TTF and many more. SDL_Image is a quick way to load compressed images (SDL itself can only load BMP files). SDL_TTF lets you draw text with TTF fonts so that you do not need to resort to bitmap fonts.

For helpers etc. you might have your set of helpers already from other projects. If not, there is boost which I myself do not use anymore, but some people find it useful. The quality of some libraries is quite low but there might be some with better quality worth using too.

I don't want to pack you on all my choices, but if you have a specific need then I'd give you some more tips. Also keep in mind that you always have the choice to implement things yourself which is sometimes a better choice because of the higher level of control.

[smallfly]

I don't want to pack you on all my choices, but if you have a specific need then I'd give you some more tips. Also keep in mind that you always have the choice to implement things yourself which is sometimes a better choice because of the higher level of control.

I programmed websites (inclusive cms) many years and always wanted to program everything on my own. In the last years I started to use finished modules like a wysiwyg text area based editor, file management systems etc.

I will try to use libraries for all cases that are not performance relevant (loading, saving, editoring, settings etc.) Which helper library do you prefer (it using one) (I think you mean a collection of common functions needed when realizing a game?)

Sounds somewhat simcity-ish.

Well, I also like the principles of simcity (but not the way how it was realized). P1 will be a mixture of a logistics and city building game, but the latter will be automated by default, so that you normally only deal with the logistics part.

Also, I assume you know the global OpenTTD code/data structure, and its weak points? If not, read the discussions on speeding up the game (usually multi-core discussions). It also gives you an idea on how far the OpenTTD game scales. If you want more, you'll need a very different code/data structure than OpenTTD has.


At the end of the day, there is of course only one way to find out if it is feasible, namely implement it.
You may want to make a version first that tests such things without any graphics, etc.

Thanks for the hints