NAND Patch

Wolf01 · Post by **Wolf01** » 24 Jul 2006 16:51

Owen wrote:-> deifne: It was temporary; I'm moving it to a patch option

i already moved it, just dl my patch and use it

meowsqueak · Post by **meowsqueak** » 25 Jul 2006 10:04

Oh, sweet sweet goodness... geek++

DredFurst · Post by **DredFurst** » 02 Aug 2006 00:12

1 statement:

OpenTTD Does electronics simulation

meowsqueak · Post by **meowsqueak** » 02 Aug 2006 01:14

DredFurst wrote:1 statement:

OpenTTD Does electronics simulation

Not really, it does asynchronous logic simulation, and with some considerable track-laying could also do synchronous logic. You might also think of the number of trains passing a point in unit time as some sort of 'current'. It does not model electronics because there's no way to represent complex impedances (if you want to consider restriction of train flow as resistive, there's still no way I can see to model inductive or capacitive impedances, as there's no way to integrate or differentiate the current flow), amongst other things.

There's also no clear way to model potential difference, except with the speed of a train, maybe...

I think the analogy starts to break down.

The useful thing about this stuff is it lets you create much more complex systems, where one part of track can influence the operation of another. The NAND function is the basic building block of this. A lot of digital logic can now be implemented with tracks and trains, but it remains to be seen what sort of useful transport-related control structures people can come up with.

Exciting patch.

iNVERTED · Post by **iNVERTED** » 02 Aug 2006 10:55

I appreciate your work and all... but surely programmable signals would be far better than any number of extra pre-signals - they would remove the need for laying un-used track, so we can make things more compact

edit: regardless of that, I just found a place in my game where a NAND would be very, very useful.

Rens2Sea · Post by **Rens2Sea** » 02 Aug 2006 18:04

Got nothing to do with this weird patch, but it's also with a little train that regulates the traffic.

Edit: fixed and added a funny maglev one, both 2 trains left - 2 trains right etc.

If you move the two-way signal from the maglev choohoo, you can change the amount of trains per branch at a time, atm it's 2-2, but can be set to 4-1 or something

skidd13 · Post by **skidd13** » 26 Oct 2006 19:08

The patch is awesome!!!! I updated it against trunk with some minor changes.

EDIT: Added Savegame with the logic gates (OR,AND,NAND,NOR,XOR,XNOR).

ddream · Post by **ddream** » 27 Nov 2006 22:18

I've updated this patch to MiniIN 7267 and made graphic for nand signals, this graphic is necessary to run patched game. It is my first approach to OTTD sourcecode, so I could made something wrong.

mamuf · Post by **mamuf** » 07 Jul 2007 01:28

I was playing around with Situation Switcher (found the idea at OpenTTDCoop) and made few pretty good and functionable variations. But I still knew something like NEG/NAND presignal would be handy to make it really perfect.

Then I tried to make my own NEG presignal but I could hardly understand the OTTD code around signals, although it actually worked. The main problem was the looping of these negations and the game has always freezed so I stopped trying for that time.

And then I found this thread after I saw a mention about NAND gates in the thread about programmable signals. Does anybody still work on this patch? And there is no activity in the other thread either

.

Wolf01 · Post by **Wolf01** » 07 Jul 2007 08:38

i know that owens come out again some days ago with this idea, and he was thinking about making all signals become red when looping

skidd13 · Post by **skidd13** » 07 Jul 2007 13:53

I rewrote the nand patch. Technicaly it works, but I have the feeling that something is wrong with it.

TODO:
- Savegame conversion to be save with the semaphores
- New sprites for the nand signal

mamuf · Post by **mamuf** » 07 Jul 2007 13:59

skidd13 wrote:I rewrote the nand patch. Technicaly it works, but I have the feeling that something is wrong with it.

TODO:
- Savegame conversion to be save with the semaphores
- New sprites for the nand signal

I am just working on it, too. I used a patch made from the 7267 MiniIN and merged it with the latest trunk manually but it has some issues, too. I think it is somewhere around signal bits because the game thinks I have looped NANDs and the ctrl-driven signal editing doesn't work well. (It is very similar to the first problems I've encountered while I was working on my own NEG signals.)

Now I'll have a look at your version, skidd13.

mamuf · Post by **mamuf** » 07 Jul 2007 14:25

I merged my nand-update and the one by skidd13 and now I'm going to test it. The one I updated is the one which checks the NAND-loop count so I think it will be better. But there were some weird constructs so I used some pieces of code by skidd13.

mamuf · Post by **mamuf** » 07 Jul 2007 14:42

So there it is. But it still yells that some NAND signals are looped. And the signals behave weird at all. I think it is still somewhere within the signal bits.

EDIT: I fixed some little glitches and it seems to work now. The only thing I am not sure about is the value for argument 'from' on the line added to settings.cpp.

NOTE: The savegames are most likely incompatible.

skidd13 · Post by **skidd13** » 07 Jul 2007 16:13

mamuf: small hint there's an edit button, so please don't double post until it might be necessary.

I removed the loop stuff cause the warning sucks. What about a debug message (to the console) after a code preset value e.g. 20.
The loop is caused by the original pathfinder which is AFAIK still used for the signal handling. So a debug message to the console should be enough.

The savegame value in your settings.cpp is wrong. Should be 70 ATM. (Savegame bump needed).

mamuf · Post by **mamuf** » 07 Jul 2007 16:58

Yep, I already edited the last post while I had fixed version. But maybe I should've better edit the last but one

Since the NAND sets itself red while the NAND-loop reaches it's limit, the warning is really useless. I agree that if this won't be merged with trunk it doesn't need any GUI. The DEBUG call is already at line 1773 in rail_cmd.cpp. Originally it was made with printf and someone wrote here it should be DEBUG only so I changed it but I am not sure about the values of the arguments:

Code: Select all

DEBUG(misc, 0, "NOTICE: Broken off processing, found too many NAND signals\n");

So you suggest the loop check should have hardcoded limit? Maybe it still could be a patches' variable because it is used at to places in the code.

skidd13 · Post by **skidd13** » 07 Jul 2007 19:15

mamuf wrote:... but I am not sure about the values of the arguments:
Code: Select all
DEBUG(misc, 0, "NOTICE: Broken off processing, found too many NAND signals\n");
misc -> section of debug, misc seems to be OK IMO
0 -> debug_level, IIRC 0 is show message allways
"<string>" -> no need of declaration

mamuf wrote:So you suggest the loop check should have hardcoded limit? Maybe it still could be a patches' variable because it is used at to places in the code.

The multiple usage is no reason to create a patch value.

AlexW · Post by **AlexW** » 09 Jul 2007 07:31

If it is possible for someone to explain this in village idiot/simpleton/layman's terms, I would be most appreciative. I simply can't get my head around the usage of NAND, even with the picture(s) that have been posted!

mamuf · Post by **mamuf** » 09 Jul 2007 09:50

AlexW wrote:If it is possible for someone to explain this in village idiot/simpleton/layman's terms, I would be most appreciative. I simply can't get my head around the usage of NAND, even with the picture(s) that have been posted!

The question is - Do you know what logic gates are in general? Logic gates like AND, OR, NOT and ohers? If not, read the wikipedia article for example: Logic gate.

NAND is simply gate made of NOT and AND. Written in functional manner it looks like: NOT(AND(x, y)), where x, y are logic variables (values 1|0). In common electronics NAND is the simpliest gate in the meaning of number of transistors needed to build such gate. Other gates need more of them to be build. And you can build every single known logic gate of NANDs only.

Logic table for NAND:

Code: Select all

x y | NAND(x, y)
--------
0 0 | 1
0 1 | 1
1 0 | 1
1 1 | 0

DeletedUser21 · Post by **DeletedUser21** » 09 Jul 2007 10:58

I see this thread for the first time (how could I have missed this...

).

OpenTTD reaches a whole new level of geekdom!

This is awesome stuff.

Transport Tycoon Forums

NAND Patch

Who is online