[Patch] Improved Timetable Management [V2.31tr SVN15778]

[rbn2903]

click set time, backspace it, ok, then hit set time again and the automatic value is set

I'm not sure, I completely understand your problem (trying to reproduce) but with this I think can help you If you add a vehicle to an already started headway, then you open the headway window and CTRL+Click on Set intervall which then proposes a new, shorter time that fits. So a new headway can easily be started.

[PhilSophus]

When adding a vehicle to a timetable which already has headway set, if you then try to set the headway you have to first clear it (click set time, backspace it, ok, then hit set time again and the automatic value is set). If you don't you get the rather unhelpful "timetable cannot be started" or something to that effect (can't check the actual message on this PC)

As Rainer said, <CTRL>-click proposes a new fitting interval (without CTRL it uses the previous value and proposes an automatic one only if none was set before).

It would be handy if the headway could be a little more automatic (eg the "automatic" button being fully auto, if more trains get added to the shared order list, it just shortens the headway accordingly) etc.

Reworking of manual vs. automatic modes is already on my todo list. This includes finer-grained control on what should be done automatic and what not and the possibility to change each gap length separately in manual mode. However, I can't tell when I'll implement it.

(timetables seem to get a bit pissy with daylength, I don't know if you've fixed this as I haven't played with both ITiM and Daylength)

What does happen with daylength? I'm not sure it would work out-of-the-box (especially when using days as the timetable unit, there might be conversion errors, but virtual time converts directly to ticks), but on the other hand, I don't have reports of it not working, either. That either means no one tried, or there are no substantial problems.

[audigex]

With timetables: it's more a conversion problem. The auto-fill works correctly, but you (or rather, I, there may well be a way) can't set the times yourself. I'll have another look later with some more info.

With the first part, the CTRL+click is definately the solution I was looking for. Would it be possible for this to be a visible, rather than hidden feature though? Obviously it's no longer a problem for me, but unless people read the entire thread (wishful thinking) they wouldn't know.

With regards to more automation, sounds excellent just what you have at the moment is great, but to be able to click a "completely automatic" style button would be great. Especially for headway, but if there's any possibility of a more auto- autofill, that'd be grand. Might get complicated with different speed trains though.

Edit: Poking through the settings a bit more, I've discovered the "wall clock" aspects of the timetabling. Again, no idea how well if at all this integrates with daylength. Would be superb if it did though

[PhilSophus]

With the first part, the CTRL+click is definately the solution I was looking for. Would it be possible for this to be a visible, rather than hidden feature though? Obviously it's no longer a problem for me, but unless people read the entire thread (wishful thinking) they wouldn't know.

It's in the tooltip That is not too much text, I hope

Especially for headway, but if there's any possibility of a more auto- autofill, that'd be grand. Might get complicated with different speed trains though.

Well this won't come until I implement kind of timetable stop watches, which is on my todo list, but very much at the bottom, i.e. probably far ahead.

[audigex]

If it's on the to do list, I'm happy; everyone has priorities.

Again, great work with the patch - I'll keep playing with it for a while at least, I doubt I'll like playing without headway anyway

[PhilSophus]

I put that post from cargodest thread here for reference, as I think it would be more appropriate to continue discussion here:

Possible bug? Here's how I broke it...

1) Create train x between stations A and B
2) Start train x (no idea whether it has to complete any runs beforehand)
3) Create train y with shared orders with x
4) Delete train x's orders, or the train itself

Result: Stations A and B with "n passengers to A/B, no route found". Also at any stations with no direct route to A, but a route to one of B, get the same result "n passengers to A, no route found". This message remains for as long as I've been willing to sit and watch. The only solution I've found is to delete the shared orders on train y, then re-create the orders again manually. I'm assuming that if you then remove train y after creating train z, the same problem occurs.

I'm not sure if it's totally reproducable but I've definately seen it at least twice.

Patches used:
ITiM
Passenger Reduction (currently off)
Cargodest

Can supply a GRFlist if required.

I'm having trouble replicating it at the moment (although I wasn't using any ITiM features at the time). I'll keep trying.

Even if you had disabled all parts of ITiM, the part I suspect (the order list rewrite) can not be disabled. It would definitely help me much, if you were able to reproduce a state where it always happens, as I need to watch it happening in the debugger to find out what's wrong. I've also had cargo with stale destination from time to time, but never could pin down when it appeared, let alone under which circumstances.

[audigex]

I'll keep trying. It was definately something to do with replacing the train (manually). I just can't remember the order in which I did things. I'm certain I made the second train before removing the first, but I can't remember whether I started it, whether I removed the firsts orders, whether I'd already sent it to the depot, whether there were any other trains running that route (although I'm fairly confident not)

There's just so many variables... I'll keep playing though.

Edit: It keeps happening, but I can't reproduce it when I want it to. It's definately when adding trains to a route then removing the original ones, however, because it happened three times while doing that as I upgraded from steam to diesel. All I can suggest is to make three or four routes on a map running shared orders, then replace the trains on the routes one at a time. I just can't make it happen

[rbn2903]

Hi Philsophus,

I can't help it, but I always get the weirdest ideas to make this patch even better If a train is autofilling a timetable but somehow has to stop at a junction due to another train blocking its way, is it possible, not to record this time? Or is this impossible to do, because the train then has to accelerate again which takes longer than driving constantly?

Another thing. After a certain engine becomes available, i usually update to this (lets say: DBSet XL BR116 (max. speed 120km/h) to BR101 (max. speed 220km/h)). Would a be possible to shorten the single timetable durations with the factor 120/220 (= 6/11; of course only travelling, not waiting times), so that you don't have to autofill the timetable again?

I assume that people new to this patch could be overwhelmed by all the features it brings But I still hope it will finally make it into trunk, but there are already quite good competitors like cargodest and freeform map edges

Greets,
Rainer

P.S.: I'd love to post my current ITiM-savegame so that anyone could see, how I use this but it has lots of NewGRFs :/

P.P.S.: Did you update your first post when you last updated to v2.30?

[PhilSophus]

I can't help it, but I always get the weirdest ideas to make this patch even better If a train is autofilling a timetable but somehow has to stop at a junction due to another train blocking its way, is it possible, not to record this time? Or is this impossible to do, because the train then has to accelerate again which takes longer than driving constantly?

Another thing. After a certain engine becomes available, i usually update to this (lets say: DBSet XL BR116 (max. speed 120km/h) to BR101 (max. speed 220km/h)). Would a be possible to shorten the single timetable durations with the factor 120/220 (= 6/11; of course only travelling, not waiting times), so that you don't have to autofill the timetable again?

I don't intend to change the autofill itself (well I already did with the non-destructive mode ) and I don't even see how you could reasonably distinguish between normal and abnormal held-ups.

As for the second suggestion. Given that autofill is non-destructive I don't see a real need for this.

P.P.S.: Did you update your first post when you last updated to v2.30?

Umm, yes, at least it had a history entry on 2.30 when I just checked Anything wrong with it?


Just a short notice: Progress in ITiM might slow-down a lot in future as my spare time might reduce significantly in future. So, don't expect so much new features either from my already long todo list or newly suggested.
However, I intend to keep ITiM updated to current trunk/cd and to fix bugs (if you provide me with means to reproduce them and find out whether they are ITiM bugs - looking at audigex -).

[rbn2903]

As for the second suggestion. Given that autofill is non-destructive I don't see a real need for this.

Ok then, was just an idea.

[...] at least it had a history entry on 2.30 when I just checked [...]

Actually I meant the sections before for example: 6. Non-Destructive Auto-Fill has become kind of obsolete And C. Patch Settings should contain "Show notice when vehicle has finished timetable autofill: Off, On"

Just a short notice: Progress in ITiM might slow-down a lot in future as my spare time might reduce significantly in future. So, don't expect so much new features either from my already long todo list or newly suggested.

Nobody's gonna be mad at you for that Thats quite nice what you created so far

Greets and "Frohe Weihnachten und einen guten Rutsch ins neue Jahr!",
Rainer

[PhilSophus]

Actually I meant the sections before for example: 6. Non-Destructive Auto-Fill has become kind of obsolete And C. Patch Settings should contain "Show notice when vehicle has finished timetable autofill: Off, On"

Done. I should really make this a Wiki page, it would be much more maintainable then, and more important it would be maintainable by others

Greets and "Frohe Weihnachten und einen guten Rutsch ins neue Jahr!",

Thank you and the same back

[MJS]

Some interesting new stuff has appeared in trunk lately. Does anyone perchance have a newer build with this patch (and I myself would be most interested if cargodest is also included.)

On another note, this patch seems to work quite well. It has gotten a lot easier to handle, although setting up timetables and managing them when you replace vehicles still takes time - but at least now it's fun doing it, and that's the way it should be. Biggest advantage, at least to me, is not that the station ratings are high and stockpiles are constant, although those are nice too: I really love it that I can keep my stations a lot smaller because I know there'll never be more than one train of a certain service in the station at the same time. I'll add a screenshot to make my point clear: five lines serving an 8-platform station. (I don't count the extra platforms on the NW side of the station.)

[rbn2903]

Does anyone perchance have a newer build with this patch

I think CargoDest hasn't been updated since december and so you (an me too) would miss a lot of new introduced features. besides I didn't dare to ask. It looks so demanding, and I think Philsophus has a lot of other stuff to do

[PhilSophus]

Does anyone perchance have a newer build with this patch

I think CargoDest hasn't been updated since december and so you (an me too) would miss a lot of new introduced features. besides I didn't dare to ask. It looks so demanding, and I think Philsophus has a lot of other stuff to do

Yes, this lot of other stuff is mainly a new job

Internally, I keep ITiM updated to trunk BTW. Well, I make sure, it compiles, that is, I can't tell whether everything still works . But as you say, there has been no update to cargodest, and so neither can be to the cargodest-variant of ITiM. And since most people seem to use ITiM+cd, I didn't bother releasing trunk only.

[pshemko]

I've noticed one small bug. If a ship has to autocomplete timetable it skips the buoys in the process (i.e. the timetable remains unfilled for them). After filling it up manually with 0:01 the headway works correctly. I think the best solution would be to treat buoys the same way as depots or waypoint - without any waiting time.

[Jans]

Could some one Tell me what should I do with this:

*** OpenTTD Crash Report ***
Date: 2009-01-25 16:27:39
Build: ITiM 2.30 w/cargodest built on Dec 2 2008 12:16:53
Reason: Assertion failed at ..\src\signal.cpp:511: dir == INVALID_DIAGDIR || dir == GetRailDepotDirection(tile)
Language: german.lng
Exception C0000005 at 00442D78
Registers:
EAX: 01BA66F8 EBX: 000000FF ECX: 0012F8F8 EDX: 01BA6760
ESI: 0004720B EDI: 00000003 EBP: 0004720B ESP: 0012FB10
EIP: 00442D78 EFLAGS: 00010202

Bytes at CS:EIP:
C6 05 00 00 00 00 00 53 EB 78 0F B6 B7 5C 40 5E 00 53 6A 00 6A 00 8B C5

Stack trace:
01F4D870 00000001 0012FB50 01C65A08 01000003 00000000 00443028 00000001
0044DD18 0004720B 00000003 00000001 00625E78 01C65A08 01F4D818 01C65A08
0012FB64 0044DD96 01C65A08 00000001 01C65A08 00000000 004C1384 01000000
00625E78 01C65A08 0012FB90 00000000 004C17D4 00000000 01C65A08 00000000
0012FBC8 0044AEB0 00E3F4DE 00E3F4C0 000000FF 0007EFF6 005F8184 0007EFF6
00000008 00000001 0000000F 0012FBD8 00E3F4DE 00E3F4C0 0012FBD8 0042F1FF
00000000 0003FCD9 0012FBE8 004254D2 00E3F4DE 00002701 0012FC00 0042F429
00E3F4C0 00E3F4B0 7C802600 7C802600 7C80932E 00504F45 00653394 0063DDB4
0012FC18 00000000 0012FD14 0042EA97 0012FD2C 00000002 00000001 01010003
00000000 00B70000 00000000 00000000 00000000 FFFFFFFF 00000000 FFFFFFFF
0012FE38 0060029C 00000000 FFFFFFFF 00000000 00000000 00000000 7C929D8A
6C6C756E 00000100 00000000 00000004 00B70748 00000002 0012FD2C 0000FDE8
69646564 65746163 00020064 0012FCC4 7C82167A 0000001C FFFFFFFF 0000001A
6C6C756E 0000FD00 0012FD2C 00000001 00000019 0012FCF0 7C8215D3 0000FDE9
6C6C756E 00020600 FFFFFFFF 0012FD2C 00000104 00000000 00000000 FFFFFFFE
0012FF00 00456CE0 0000FDE9 00000000 0002067C FFFFFFFF 0012FD2C 00000104
00001C6C 0012FF30 004568A7 00000002 0012FE30 FFFFFFFE 00000000 5C3A4422
74736554 65706F5C 6474746E 6578652E 442D2000 00B70100 00B70640 00000080
00000002 00B70270 00B70280 0000000D 00000006 00B70640 00B77230 00B70178
00000004 00B77218 00000000 00000003 00B70178 00000010 00B70178 00B77228
00B70168 00000000 00000000 00B77228 00000020 00B77230 00000001 00B70178
00000020 00000280 00B70000 0012FBB4 00000002 0012FE7C 7C91E900 7C9201C0
FFFFFFFF 7C9201BB 00567126 00B70000 00000000 00000018 00B76ED8 0012FEEC
0012FEF8 0012FE0C 005671BC 00000018 00B76F08 00B7015C 00B70000 0012FE88

Module information:
D:\Test\openttd.exe handle: 00400000 size: 2352640 crc: 37E970DF date: 2008-12-02 11:18:52
C:\WINDOWS\system32\ntdll.dll handle: 7C910000 size: 731648 crc: AC43D78D date: 2008-04-14 05:51:54
C:\WINDOWS\system32\kernel32.dll handle: 7C800000 size: 1063424 crc: 43AD4D1C date: 2008-04-14 05:52:14
C:\WINDOWS\system32\WINMM.dll handle: 76AF0000 size: 178176 crc: E266DD96 date: 2008-04-14 05:52:34
C:\WINDOWS\system32\ADVAPI32.dll handle: 77DA0000 size: 678400 crc: 7DCC7171 date: 2008-04-14 05:52:08
C:\WINDOWS\system32\RPCRT4.dll handle: 77E50000 size: 584704 crc: B8FA2036 date: 2008-04-14 05:52:24
C:\WINDOWS\system32\Secur32.dll handle: 77FC0000 size: 56320 crc: 75202C67 date: 2008-04-14 05:52:26
C:\WINDOWS\system32\GDI32.dll handle: 77EF0000 size: 286720 crc: 62E5D8E7 date: 2008-10-23 12:36:55
C:\WINDOWS\system32\USER32.dll handle: 7E360000 size: 580096 crc: 89D3E1AE date: 2008-04-14 05:52:32
C:\WINDOWS\system32\WS2_32.dll handle: 71A10000 size: 82432 crc: E04F0DBC date: 2008-04-14 05:52:34
C:\WINDOWS\system32\msvcrt.dll handle: 77BE0000 size: 343040 crc: 88E736C3 date: 2008-04-14 05:52:20
C:\WINDOWS\system32\WS2HELP.dll handle: 71A00000 size: 19968 crc: 495F6FD3 date: 2008-04-14 05:52:34
C:\WINDOWS\system32\SYNCOR11.DLL handle: 6BD00000 size: 40820 crc: 6D1328C4 date: 2002-11-06 18:00:38
C:\WINDOWS\system32\SHFolder.dll handle: 76730000 size: 25088 crc: 2A9ECCE3 date: 2008-04-14 05:52:26
C:\WINDOWS\system32\SHLWAPI.dll handle: 77F40000 size: 474624 crc: F8F36120 date: 2008-04-14 05:52:26
C:\WINDOWS\WinSxS\x86_Microsoft.Windows.Common-Controls_6595b64144ccf1df_6.0.2600.5512_x-ww_35d4ce83\comctl32.dll handle: 773A0000 size: 1054208 crc: FD52368A date: 2008-04-14 05:50:12
C:\WINDOWS\system32\comctl32.dll handle: 5D450000 size: 617472 crc: ACCB0A1D date: 2008-04-14 05:52:10
F:\FRITZ!DSL\sarah.dll handle: 10000000 size: 24880 crc: DAAFF75D date: 2007-09-04 09:14:44
C:\WINDOWS\system32\MSVCR71.dll handle: 7C360000 size: 348160 crc: F83AD7CD date: 2006-07-11 17:35:38
F:\FRITZ!DSL\block.dll handle: 00BB0000 size: 66864 crc: 940EC3DE date: 2007-09-04 09:14:04
C:\WINDOWS\system32\mswsock.dll handle: 719B0000 size: 247296 crc: 1323EFAD date: 2008-06-20 17:46:10
C:\WINDOWS\system32\hnetcfg.dll handle: 66710000 size: 348672 crc: 1EE21C3A date: 2008-04-14 05:52:12
C:\WINDOWS\System32\wshtcpip.dll handle: 719F0000 size: 19456 crc: 26C346BD date: 2008-04-14 05:52:34
C:\WINDOWS\System32\winrnr.dll handle: 76F70000 size: 16896 crc: 6BB1D62C date: 2008-04-14 05:52:34
C:\WINDOWS\system32\DNSAPI.dll handle: 76EE0000 size: 147968 crc: 1A76D306 date: 2008-06-20 17:46:10
C:\WINDOWS\system32\WLDAP32.dll handle: 76F20000 size: 173056 crc: C7D8F037 date: 2008-04-14 05:52:34
C:\WINDOWS\system32\iphlpapi.dll handle: 76D20000 size: 95744 crc: D7E9E8DA date: 2008-04-14 05:52:14
C:\WINDOWS\system32\rasadhlp.dll handle: 76F80000 size: 7680 crc: F4ABB355 date: 2008-04-14 05:52:24
C:\WINDOWS\system32\psapi.dll handle: 76BB0000 size: 23040 crc: 38CB6558 date: 2008-04-14 05:52:24

System information:
Windows version 5.1 2600 Service Pack 3

[PhilSophus]

Could some one Tell me what should I do with this:

Reason: Assertion failed at ..\src\signal.cpp:511: dir == INVALID_DIAGDIR || dir == GetRailDepotDirection(tile)

Does not really look like an ITiM- or cargodest-specific problem, IMHO. Neither of them does anything related to signal blocks. That is, I suspect that is a problem also present in trunk. You might want to search the bug tracker for a similar problem and if you don't find it, try to reproduce it with a clean Nightly (i.e. without patches). If you succeed in reproducing it you should open a bug report with a savegame (only from unpatched versions!) and steps needed to reproduce the problem.

[JamieLei]

Apologies if this has been mentioned before, but a great benefit to timetabling would be this patch here: http://bugs.openttd.org/task/2534

The two go hand in hand, and enable timetabling in city areas (currently trains get stuck until 100% load)

[DJ_Izumi]

Wow, it actually works. I had tried the original time tables that are in the trunk but couldn't make them work effectively. I'd make a timetable, start all my trains with the apropriate interval but they'd manage to 'clump together' eventually anyway, while all thinking they're ontime. I'm just testing this out using a money cheat to see them run but it looks good so far.

Would automatic intervall organization be an option? Where you basicly just press 'Automatic Interval', the first train automatically autofills, then generates it's own interval by dividing the total trip time by the total number of vehicles sharing the orders? Then you could construct a route, press the button, and eventually all the trains you started simultaniously will sort themselves into an evenly divided circut?

[Eddi]

that's what the "headway" button already does