Page 1 of 1
| Loader V5.16 breaking sims which worked perfectly before the loader update 14/07/2021 at 19:39 #140574 | |
58050
2659 posts |
Whilst continuing testing of the forth coming Birmingham New Street summer 1990 TT this evening it became clear that a number of things that worked perfectly before the loader update now don't. Please note that the sim era we are using is the 1976-1991 era. 1.Trains destined for Soho DED (5G52) when you signal them from NS312 thru to reverse behind NS319 worked perfectly on tyhe old loader, but now using loader V5.16 you can't set the route from into the REV SDG. That said when the tyrain actually passes NS312 & is routed in towards SB1 that signal clears but the track remains greyed out. So a quick clipping the points is necessary on all those movements. You can set the route from NS319 to Soho Depot OK & works no problem. Trains coming out of Soho Depot into NS319 REV SDG. also work OK. 2. 1A39 which is a booked WTT service of empty NPCCS from Camden CS/Wembley CS which is designed to turn the formation round runs via Stechford - Aston - Perry Barr South Jn - Perry Barr West Jn - Soho East Jn - Soho South Jn & recesses in Monument Lane loop for approx. 10 mins to allow other trains to pass. When the route between NS301 & NS245 are cancelled to signal 1A39 out of Monument Lane loop the route is set & NS295 clears without any interface from the user. 3. Other issues pretty much dealing with GPS & LOS boards have also been spotted in so much as LOS boards showing proceed aspects. I've attached some screenshots so you can see what we've found so far.   
Post has attachments. Log in to view them. Last edited: 14/07/2021 at 20:35 by 58050 Reason: corrected typo Log in to reply |
| Loader V5.16 breaking sims which worked perfectly before the loader update 14/07/2021 at 19:45 #140576 | |
|
58050
2659 posts |
This last image is at Birmingham International. This shot is something regarding the 120 seconds timeout. Updated photo of the Birmingham International station issue. Post has attachments. Log in to view them. Last edited: 14/07/2021 at 20:07 by 58050 Reason: None given Log in to reply |
| Loader V5.16 breaking sims which worked perfectly before the loader update 14/07/2021 at 20:01 #140577 | |
Steamer
3985 posts 
|
Thanks- Mantis 34162 raised. Not been able to reproduce problem 2, but I think it's another symptom of the general bug affecting shunt routes.
"Don't stress/ relax/ let life roll off your backs./ Except for death and paying taxes/ everything in life.../ is only for now." (Avenue Q) Log in to reply |
| Loader V5.16 breaking sims which worked perfectly before the loader update 14/07/2021 at 20:08 #140579 | |
|
58050
2659 posts |
Ignore that screenshot on my 2nd post. You need to look at the file ending in 0322 to see the right image.
Log in to reply |
| Loader V5.16 breaking sims which worked perfectly before the loader update 14/07/2021 at 20:09 #140580 | |
GeoffM
6376 posts
|
It seems to be where the track lists one exit signal but the route has a different exit signal. This was a recent change to handle preset shunts better so we'll be looking at that.
SimSig Boss Log in to reply |
| Some preset shunt issues 19/07/2021 at 21:57 #140659 | |
KymriskaDraken
963 posts |
Post has been automatically hidden due to a number of complaints.
Log in to reply |
| Some preset shunt issues 20/07/2021 at 08:05 #140660 | |
clive
2789 posts |
For those wondering what happened, there was a problem with present shunts that's been around for a while but hadn't been spotted until work was being done on a new sim. In particular, if I recall correctly there were no less than three different and inconsistent ways of determining whether a shunt was preset, so in principle you could have the same signal be preset or not depending on irrelevant details when the route was set. I refactored the code to make the logic the same in all places and also provide a way for developers to override it in unusual cases. This code was tested in a number of sims. It turns out, though, that some developers had been using a technique for route setting in some circumstances which I wasn't aware of, though it was perfectly legitimate. As Geoff says, this involves making the route exit signal *not* being the signal at the end of the route. This made the signal at the end of the route appear to the new logic to be a preset (after all, a preset shunt is normally a shunt signal on the route that isn't the route exit signal). That is what triggered the problem. As soon as we realized what the problem was, the fix was simple and has been made in 5.16.1, which is now available. Log in to reply The following users said thank you: Albert, Hap, 58050, Meld, Steamer, KymriskaDraken, phil1044, flabberdacks, Mikhail |