That's it. (r14300) It's f***ing excellent!MJS wrote:I think he has the nightly, and that is why he has now discovered PBS, and is grateful for it.
YAPP - Yet Another PBS Patch (now in trunk!)
Moderator: OpenTTD Developers
-
- Route Supervisor
- Posts: 466
- Joined: 15 Jun 2007 09:27
- Skype: madchimiste
- Location: Berlin, Deutschland
Re: YAPP - Yet Another PBS Patch (now in trunk!)
-
- Engineer
- Posts: 111
- Joined: 08 Feb 2007 11:59
- Skype: a.locritani
- Location: Italy
Re: YAPP - Yet Another PBS Patch (now in trunk!)
Sorry for reviving this topic, but I think it's better to ask question about YAPP here instead of ask them in general, to avoid confusion to people that don't use trunk version but only releases.
anyway, my question:
I've made a simple test scenario: 3 station, 2 trains, a single line between with overtaking track. (savegame attached was saved using this build)
strange things happen
1) trains don't find a free path, even if one exists (screenshot 1 and 2) - in screenshot 2 I removed second platform on station in order to "force" a situation where one train have to leave otherwise deadlock (and they lock, doh...)
2) train pass a PBS signal from the back (it's ok) when a slightly longer path with no signal to pass from the back exists (screenshot 3)
but...
3) this behavior it's not used to overtake (train should pass signal from back to overtake the stopped train) (screenshot 4 - on following post)
by the way, without changing parameters in openttd.cfg I wasn't able to make a train overtake another train using double track. and what exactly the following parameters do?
yapf.rail_pbs_cross_penalty = 300 -- penalty for level crossing?
yapf.rail_pbs_station_penalty = 800 -- penalty for entering a (not in order list) station
yapf.rail_pbs_signal_back_penalty = 1500 -- penalty for passing a pbs signal from back
yapf.rail_doubleslip_penalty = 100 -- ?
if these questions were previously asked, please excuse me and please tell me where I can find answer: I was not able to find them
thanks
anyway, my question:
I've made a simple test scenario: 3 station, 2 trains, a single line between with overtaking track. (savegame attached was saved using this build)
strange things happen
1) trains don't find a free path, even if one exists (screenshot 1 and 2) - in screenshot 2 I removed second platform on station in order to "force" a situation where one train have to leave otherwise deadlock (and they lock, doh...)
2) train pass a PBS signal from the back (it's ok) when a slightly longer path with no signal to pass from the back exists (screenshot 3)
but...
3) this behavior it's not used to overtake (train should pass signal from back to overtake the stopped train) (screenshot 4 - on following post)
by the way, without changing parameters in openttd.cfg I wasn't able to make a train overtake another train using double track. and what exactly the following parameters do?
yapf.rail_pbs_cross_penalty = 300 -- penalty for level crossing?
yapf.rail_pbs_station_penalty = 800 -- penalty for entering a (not in order list) station
yapf.rail_pbs_signal_back_penalty = 1500 -- penalty for passing a pbs signal from back
yapf.rail_doubleslip_penalty = 100 -- ?
if these questions were previously asked, please excuse me and please tell me where I can find answer: I was not able to find them
thanks
- Attachments
-
- screenshot1: trains don't find a free path, even if one exists
- Trenhill Transport, 30th Aug 2000.png (45.31 KiB) Viewed 1924 times
-
- screenshot2: trains don't find a free path, even if one exists
- Trenhill Transport, 7th Oct 2000.png (44.36 KiB) Viewed 1854 times
-
- screenshot3: train pass a PBS signal from the back (it's ok) when a slightly longer path with no signal to pass from the back exists
- Trenhill Transport, 12th Nov 2000.png (46.93 KiB) Viewed 1851 times
Last edited by a.locritani on 18 Oct 2008 07:15, edited 1 time in total.
-
- Engineer
- Posts: 111
- Joined: 08 Feb 2007 11:59
- Skype: a.locritani
- Location: Italy
Re: YAPP - Yet Another PBS Patch (now in trunk!)
fourth screenshot (max 3 attachement for post...) and savegame
- Attachments
-
- screenshot 4: this behavior it's not used to overtake (train should pass signal from back to overtake the stopped train)
- Trenhill Transport, 23rd Nov 2000.png (46.56 KiB) Viewed 1860 times
-
- Trenhill Transport, 3rd Nov 2000.sav
- (10.49 KiB) Downloaded 170 times
Re: YAPP - Yet Another PBS Patch (now in trunk!)
This is indeed strange. Please provide a savegame so that I can analyze your patch/advanced settings, specifically your pathfinder penalties. EDIT: Ah, it seems you did provide one. It wasn't there 10 minutes ago, was it? Anyway, I'm taking a look at the savegame now.
No, as far as I know, it is the penalty for finding a path over a track piece which is reserved by another train. This penalty causes trains to try to find routes around occupied/reserved track, as long as the detour is not too far.a.locritani wrote:yapf.rail_pbs_cross_penalty = 300 -- penalty for level crossing?
It's the penalty for reserving a piece of track which belongs to a station platform, including those of stations in the order list. This penalty doesn't matter to the pathfinder, because all possible platforms cause an equal penalty, so the pathfinder effectively ignores the penalty. I'm not sure if this means that trains will have a tendency to prefer shorter platforms in non-uniform stations, i.e. whether they generate a smaller penalty.a.locritani wrote:yapf.rail_pbs_station_penalty = 800 -- penalty for entering a (not in order list) station
That is correct.a.locritani wrote:yapf.rail_pbs_signal_back_penalty = 1500 -- penalty for passing a pbs signal from back
For an explanation of what a double slip is, please click here.a.locritani wrote:yapf.rail_doubleslip_penalty = 100 -- ?
The best place to find such stuff should be the YAPP wiki article. However, it does not contain the above information. I guess I should expand the article to explain these patch settings.a.locritani wrote:if these questions were previously asked, please excuse me and please tell me where I can find answer: I was not able to find them
-
- Engineer
- Posts: 111
- Joined: 08 Feb 2007 11:59
- Skype: a.locritani
- Location: Italy
Re: YAPP - Yet Another PBS Patch (now in trunk!)
Yeah, I've added it later... I don't know if the savegame will load, because I've not made it with a nigthly or a release version, but with a custom build (you can find link to this version in my first post)Tekky wrote:EDIT: Ah, it seems you did provide one. It wasn't there 10 minutes ago, was it? Anyway, I'm taking a look at the savegame now.
oh, thanks! just what I needed.Tekky wrote: For an explanation of what a double slip is, please click here.
I searched the wiki before asking, obviouslyTekky wrote: The best place to find such stuff should be the YAPP wiki article. However, it does not contain the above information. I guess I should expand the article to explain these patch settings.

Re: YAPP - Yet Another PBS Patch (now in trunk!)
If I rebuild your screenshot 1 in plain trunk, the trains drive just like you expect. Please check the penalties using the in-game console as the values in openttd.cfg don't affect savegames. If the penalties are correct it is possible that the custom build is broken.a.locritani wrote: 1) trains don't find a free path, even if one exists (screenshot 1 and 2)
rail_pbs_cross_penalty -> penalty for crossing (using) an already reserved tile.a.locritani wrote: yapf.rail_pbs_cross_penalty = 300 -- penalty for level crossing?
yapf.rail_pbs_station_penalty = 800 -- penalty for entering a (not in order list) station
rail_pbsstation_penalty -> the same but specific for station tiles.
-- Michael Lutz
-- Michael Lutz
-
- Engineer
- Posts: 111
- Joined: 08 Feb 2007 11:59
- Skype: a.locritani
- Location: Italy
Re: YAPP - Yet Another PBS Patch (now in trunk!)
You're right, it works in trunk! in custom build penalties are the same that in openttd.cfg. I'll put a question on topic where I took the custom build.Michi_cc wrote:If I rebuild your screenshot 1 in plain trunk, the trains drive just like you expect. Please check the penalties using the in-game console as the values in openttd.cfg don't affect savegames. If the penalties are correct it is possible that the custom build is broken.
Problem 2 and problem 3 still happen in trunk, anyway
-
- Engineer
- Posts: 111
- Joined: 08 Feb 2007 11:59
- Skype: a.locritani
- Location: Italy
Re: YAPP - Yet Another PBS Patch (now in trunk!)
I did some test and I think the problem is the length of track before the signal: if I create a line
|----6----| |----6----|
====== ======
SS==========SS==========SS
|-------10-------| |--------10------|
where S is a station tile and = is the track (junction are not shown)
it works, but If I reduce the length of overtaking track to 3 then lock happens, also in trunk
it's strange and I don't know if this is a penalty-related problem
|----6----| |----6----|
====== ======
SS==========SS==========SS
|-------10-------| |--------10------|
where S is a station tile and = is the track (junction are not shown)
it works, but If I reduce the length of overtaking track to 3 then lock happens, also in trunk
it's strange and I don't know if this is a penalty-related problem
Re: YAPP - Yet Another PBS Patch (now in trunk!)
Code tags should help here.a.locritani wrote:I did some test and I think the problem is the length of track before the signal: if I create a line
Code: Select all
|--6-| |--6-| ====== ====== SS==========SS==========SS |---10---| |---10---|
where S is a station tile and = is the track (junction are not shown)
it works, but If I reduce the length of overtaking track to 3 then lock happens, also in trunk
it's strange and I don't know if this is a penalty-related problem
Melt with the Shadows,
Embrace your destiny...
Embrace your destiny...
-
- Chairman
- Posts: 776
- Joined: 20 Jan 2007 12:08
- Location: Germany
Re: YAPP - Yet Another PBS Patch (now in trunk!)
I managed to backport a.locritani's savegame from ITiM to clean cargodest. Maybe it helps to find the problem.
Please note that the last merged trunk revision of cargodest was 14417
Please note that the last merged trunk revision of cargodest was 14417
- Attachments
-
- Trenhill Transport, 4. Nov 2000.sav
- Savegame for cargodest h527e0a444916
- (172.51 KiB) Downloaded 172 times
"The bigger the island of our knowledge, the longer the shore of our ignorance" - John A. Wheeler, Physicist, 1911-2008
Re: YAPP - Yet Another PBS Patch (now in trunk!)
Hello, I love YAPP but I think this has a bug. Look at my screenshot.
I want to make Semi-directional station. that is: train can come into both side, but i has to go out in one direction. Of coarse I don`t want to use depots or loops.(In practical passenger service rarely uses loops.)
|Station|--------<In<--
--->In>------------└-------┘------->Out>---
So i made an design, but it has a bug. Train goes to reserved space. Although in the second screnshot, tain 12 is waiting, but suddenly it starts to moving and it crashes.
I`m currently using Yet another PBS patch (v8.2) (in the Russian OpenTTD Community patchpack (1.3.0))
I want to make Semi-directional station. that is: train can come into both side, but i has to go out in one direction. Of coarse I don`t want to use depots or loops.(In practical passenger service rarely uses loops.)
|Station|--------<In<--
--->In>------------└-------┘------->Out>---
So i made an design, but it has a bug. Train goes to reserved space. Although in the second screnshot, tain 12 is waiting, but suddenly it starts to moving and it crashes.
I`m currently using Yet another PBS patch (v8.2) (in the Russian OpenTTD Community patchpack (1.3.0))
Re: YAPP - Yet Another PBS Patch (now in trunk!)
Don't combine PBS and pre-signals. (Though it probably is still a bug)
Put a one-way pbs where your pre-signal entrance is, put normal pbs signals facing away from the station where your pre-exits are, and kill the one-way combo pre-signals. That should sort it.
Put a one-way pbs where your pre-signal entrance is, put normal pbs signals facing away from the station where your pre-exits are, and kill the one-way combo pre-signals. That should sort it.
Melt with the Shadows,
Embrace your destiny...
Embrace your destiny...
Re: YAPP - Yet Another PBS Patch (now in trunk!)
This is not a bug but just a small timing issue that can happen when you use a two-way station (trains can enter from both ways) with presignals. As Thief^ said use pbs-signals only. You can replace you leftmost pre-signal with a one-way pbs signal, that way trains won't leave in that direction. Replace all exit-signals with normal pbs signals.
Re: YAPP - Yet Another PBS Patch (now in trunk!)
4 Road 2 way station issues
i have a 4 road by-directional station
with the signals set as in the first example (all PBS) trains will only go to the shortest road avalable to them the next train will wait for a path untill the first has gone,
this means that a maximum of of the 4 roads are used and when a train reverses at the station only one
in the second example 2 way normal signals have been placed inside the PBS ones,
this allows trains to use all roads avalable but results in deadlocks and infrequent crashes!
The third method i tried was using presignal blocks around the PBS exits similar to method 2 (in this case the entry signals are one way presignal enntries not PBS one ways)
this seems to work most of the time but can sometimes deadlock due to bad timing
there must be a way to impliment such a station with YAPP can anyone tell me how
i have a 4 road by-directional station
with the signals set as in the first example (all PBS) trains will only go to the shortest road avalable to them the next train will wait for a path untill the first has gone,
this means that a maximum of of the 4 roads are used and when a train reverses at the station only one
in the second example 2 way normal signals have been placed inside the PBS ones,
this allows trains to use all roads avalable but results in deadlocks and infrequent crashes!

The third method i tried was using presignal blocks around the PBS exits similar to method 2 (in this case the entry signals are one way presignal enntries not PBS one ways)
this seems to work most of the time but can sometimes deadlock due to bad timing
there must be a way to impliment such a station with YAPP can anyone tell me how
- Attachments
-
- Meningworth Transport, 15th Jul 1930.png
- PBS signals only
- (45.19 KiB) Downloaded 117 times
-
- Meningworth Transport, 30th Jul 1930.png
- PBS with 2 way non PBS inside
- (55.83 KiB) Downloaded 87 times
-
- Meningworth Transport, 20th Nov 1931.png
- PBS with Presignals inside
- (58.32 KiB) Downloaded 94 times
Re: YAPP - Yet Another PBS Patch (now in trunk!)
Clear out your openttd.ini's yapp/pbs sections (or delete it entirely). That layout should work, but the old pbs/yapp pathfinder penalties from the older pbs can cause trouble.
Melt with the Shadows,
Embrace your destiny...
Embrace your destiny...
Re: YAPP - Yet Another PBS Patch (now in trunk!)
No, it won`t work. The train comes from the reverse direction can`t go to the station, because they can`t find the safe position.Yexo wrote:This is not a bug but just a small timing issue that can happen when you use a two-way station (trains can enter from both ways) with presignals. As Thief^ said use pbs-signals only. You can replace you leftmost pre-signal with a one-way pbs signal, that way trains won't leave in that direction. Replace all exit-signals with normal pbs signals.
Re: YAPP - Yet Another PBS Patch (now in trunk!)
This is the layout I was trying to describe:
Re: YAPP - Yet Another PBS Patch (now in trunk!)
Yexo: I don't like the third signal from the top in your layout, as trains waiting in front of that signal will be blocking the junction. I suggest you remove that signal.
By the way, what NewGRF are you using? I really like the graphics of the track.
By the way, what NewGRF are you using? I really like the graphics of the track.
Re: YAPP - Yet Another PBS Patch (now in trunk!)
That signals is indeed questionable. I usually build it there as it improved throughput of the station and I design the exit track so trains should never have to wait in front of that signal. However, if there is a chance trains could stop in front of that station it should indeed be removed.Tekky wrote:Yexo: I don't like the third signal from the top in your layout, as trains waiting in front of that signal will be blocking the junction. I suggest you remove that signal.
By the way, what NewGRF are you using? I really like the graphics of the track.
The tracks are from YARRS.
Re: YAPP - Yet Another PBS Patch (now in trunk!)
If you want to improve throughput on the exit line, I suggest that you create a second exit line for at least one train length and merge the two exit lines afterwards. This creates a lot more throughput than an additional signal.Yexo wrote:That signals is indeed questionable. I usually build it there as it improved throughput of the station and I design the exit track so trains should never have to wait in front of that signal. However, if there is a chance trains could stop in front of that station it should indeed be removed.Tekky wrote:Yexo: I don't like the third signal from the top in your layout, as trains waiting in front of that signal will be blocking the junction. I suggest you remove that signal.
Thanks for the information!Yexo wrote:The tracks are from YARRS.
Who is online
Users browsing this forum: No registered users and 2 guests