I agree it's not obvious, but how would it be fixed?eshield wrote:I believe this is not that hard to check if cargo is accepted or not on unload.
(I cannot think of anything that could help the player in this situation)
Moderator: OpenTTD Developers
I agree it's not obvious, but how would it be fixed?eshield wrote:I believe this is not that hard to check if cargo is accepted or not on unload.
I found an Apple-specific bug in my code which looks like it would plausibly cause issues similar to this.strabist wrote:The attached savegame has quite a few strange issues:The issues above can be reproduced simply by starting a new game, immediately saving and then loading the saved file. Games saved in previous versions of vanilla OpenTTD and JGRPP load fine.
- It loads without any financial history, but retains delivered cargo and performance rating graphs
- It loads with a balance of $0 and loans disabled
- Vehicle count is erroneously displayed as 0 in the detailed performance rating
- No income is generated from cargo delivery (the cargo payment rates graph also shows a flat $0 for all cargos)
- Airplane running costs, infrastructure costs and interest (despite having no loan) are still deducted from the balance
I am currently running jgrpp-0.25.0 (compiled with clang-6.0.0) on MacOS 10.13.4. I have been compiling new versions roughly on a weekly basis for quite a while, and this is the first time I am experiencing this issue. Please let me know if you need any other information.
Thanks for your great work!
That is not related to your problem though. There is already a setting "Unload if accepted" which is default setting for new order anyway.eshield wrote:I believe this is not that hard to check if cargo is accepted or not on unload.nihues wrote:You need a second (different) station to load the new coal industry and unload on that station, otherwise it will be a mess to handle...
BTW, one of the golden rules when you are reporting bug to modification of original software, is that if it exist in the original version as well, then it should be reported to the original developer, not the developer of the modification. This is the case here, as this "feature" exist since original Transport Tycoon.eshield wrote:Thanks. "No loading" works really well! But, this "feature" kinda is not that obvious for user.andythenorth wrote:Don't use 'unload and take cargo'. Use 'no loading'.
Really? That's pretty common in git, with cherry-picks and the like. Wouldn't expect a git client to break on that.ino wrote:my GUI Git Client (GitKraken) doesn't like it when you time-travel commit (i.e. the parent commit is later than current commit).
Yeah I am not sure what is happening either, but basically: https://imgur.com/temDhhQpeter1138 wrote:Really? That's pretty common in git, with cherry-picks and the like. Wouldn't expect a git client to break on that.ino wrote:my GUI Git Client (GitKraken) doesn't like it when you time-travel commit (i.e. the parent commit is later than current commit).
Yeah I currently have set mine to give up. I'm really happy with the departure boards, they're awesome, but they indeed do not cope with conditional orders well. Would love to see them updated to cope better, but I do understand that they cannot know where the trains are going until they depart from a station and decide where they're going.ino wrote:There's option for conditional order with departure board: Giving up (default), assume taken, and assume not taken.
And in theory, no, you cannot really realistically fix departure board with conditional order. This is especially true with current departure board implementation: the implementation is sub-optimal and is hard to do many changes (my scheduled dispatch was (improperly) hacked in, for example). I might get around re-writing departure board sometimes though, as this also bugged me a lot.
vrn wrote:Unfortunately, the game can't know in advance whether conditional orders will be taken or not, which also messes with CargoDist for example. In your case, I would split your ICE's in two groups, one going to Duisburg and the other to Oberhausen-Bonn. I'm by no means an expert, but I think it should be possible to schedule the trains in a way that the arrivals of the first and the second alernate.
The two savegames which you posted previously which use load through with full load any cargo and refit seem to behave fine here.McZapkie wrote:Version 0.25.0 still cannot handle refit+full load+load trough.
Separate load trough+full load or refit+load trough works fine, except minor issue with full load+trough load with mixed trains - sometimes it stuck for a while with "cargo reserved to load" - probably because this cargo is waiting for cars which are still outside of the station.
Generally trough load is a very nice feature, not just eyecandy one, it is also useful to delay train load/unload in case of daylength (industry have latency between unload and production, train without delay would unload too fast and not catch an output). Of course delay can be forced by timetable, but unload trough is more convenient to set.
I tend to re-arrange my commit graph a bit before pushing, I can't see anything obviously wrong with the commits which appear to be referenced in your screenshot though.ino wrote:Yeah I am not sure what is happening either, but basically: https://imgur.com/temDhhQpeter1138 wrote:Really? That's pretty common in git, with cherry-picks and the like. Wouldn't expect a git client to break on that.ino wrote:my GUI Git Client (GitKraken) doesn't like it when you time-travel commit (i.e. the parent commit is later than current commit).
Conditional orders are evaluated at the time which the train reaches them. In the general case it's not possible to predict which path the train will later take.Xaxa wrote:I think I keep pushing conditional ordering beyond its limits and it's not working properly for me (and it's messing with the departure boards).
In the example all of the IC 28X and 29X trains should enter Bonn Vaysplatz. However, only 50% should continue onwards to Duisburg Moers, while the other 50% should go to Oberhausen-Bonn. However, with the current way I conditionally ordered the schedule, all trains go via Duisburg Moers every time.
It also messes up the departure boards, as it notifies that all trains that enter Dusseldorf go via Duisburg Moers and end at Oberhausen-Bonn, which is an impossible route. Meanwhile, the Oberhausen-Bonn departure boards show ghost trains to Dusseldorf before disappearing off the board altogether (see attachment 2).
In theory, could I make the conditional orders work this way with the departure boards showing the correct routes? And make it so that 285 goes to A, 286 goes to B, 287 goes to A, 288 goes to B, etc etc. (Instead of all trains going to A first, then B the second time they come around.)
The yellow commit (at the begining of white arrow I drawn is the parent of the teal commit which the arrow pointed to. It took me a while to realise that lol.JGR wrote:I tend to re-arrange my commit graph a bit before pushing, I can't see anything obviously wrong with the commits which appear to be referenced in your screenshot though.ino wrote:Yeah I am not sure what is happening either, but basically: https://imgur.com/temDhhQpeter1138 wrote: Really? That's pretty common in git, with cherry-picks and the like. Wouldn't expect a git client to break on that.
gitk renders the DAG fine here.
Yeah that's kinda bugged me too, and if you use scheduled dispatch (which I assume you do), I do have a solution in mind, just that I have no time to implement it yet (it's kinda huge).Xaxa wrote: I originally had them as two separate groups, but the problem is that I would have to schedule them every 60 minutes instead of 30 minutes with two different routes. As two separate groups where they depart 30 minutes from one another, they will take up 2 platforms at the terminus stations, rather than one if they're scheduled every 30 minutes. (When scheduled every 60 minutes in two groups they would have rerturn times of 40+ minutes, while in one group with two routes, they have turning times of around 15~ minutes which is a lot better.)
Thanks. Upgraded to 0.25.1 now and saving/loading seems to work fine!JGR wrote: I found an Apple-specific bug in my code which looks like it would plausibly cause issues similar to this.
I've updated the branch with this and some other changes (commit fd4d9591).
Can you update and then try saving/loading again?
Thanks for testing.
The lower 32 bits of all 64 bit quantities was written to savegames as 0 in that version on Apple/OSX.strabist wrote:Thanks. Upgraded to 0.25.1 now and saving/loading seems to work fine!JGR wrote: I found an Apple-specific bug in my code which looks like it would plausibly cause issues similar to this.
I've updated the branch with this and some other changes (commit fd4d9591).
Can you update and then try saving/loading again?
Thanks for testing.
Side note: Games saved on buggy versions still do not work properly, with most of the errors intact and a new one in tow (see attached picture)
It would be nice if you attach a save game or crash dump here.LSky wrote:Since 0.25.1 my game crashes when I hit 'Find missing content online', worked fine in 0.25.0
This is fixed and will be in the next release. Thanks for the report.LSky wrote:Since 0.25.1 my game crashes when I hit 'Find missing content online', worked fine in 0.25.0
Code: Select all
small_font = Arial
medium_font = Tahoma Bold
large_font = Arial
mono_font = Courier New
Users browsing this forum: Google Adsense [Bot] and 16 guests