Bug: Unable to start game with pak.64eperimental and sim exp. 10.10

[jk271]

Hello, I have compillied myself latest simutrans experimental from 10.10 branch. I am not able to start game and play with pak.64experimental 0.2.
Edit:
Program crashes during pak files loading process loading of demo.sve.

Pak64(not experimental), pak.german-experimental and pak128.britain work well.
I am using OS: 32-bit Linux.

gdb backtrace with some lines of text output from program:
Calculating textures
...done
FATAL ERROR: loadsave_t::rdwr_str()string longer (2057) than allowed size (256)

Program received signal SIGABRT, Aborted.0xb7fe2424 in __kernel_vsyscall ()
(gdb) bt#0  0xb7fe2424 in __kernel_vsyscall ()
#1  0xb7c90781 in *__GI_raise (sig=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
#2  0xb7c93bb2 in *__GI_abort () at abort.c:92
#3  0x0826b8f1 in log_t::fatal (this=0x83ceed0, who=0x829adc8 "loadsave_t::rdwr_str()",     format=0x829b000 "string longer (%i) than allowed size (%i)") at utils/log.cc:257
#4  0x0809c603 in loadsave_t::rdwr_str (this=0xbfffd900, s=0xbfffceb8 "\330\316\377\277\214]ͷ", size=256)    at dataobj/loadsave.cc:732
#5  0x08273ed9 in vehikel_t::rdwr_from_convoi (this=0x8c026d0, file=0xbfffd900) at vehicle/simvehikel.cc:2150
#6  0x082780a2 in waggon_t (this=0x8c026d0, welt=0x88a76c8, file=0xbfffd900, is_first=false, is_last=false)    at vehicle/simvehikel.cc:3076
#7  0x081e92cd in convoi_t::rdwr (this=0x9392018, file=0xbfffd900) at simconvoi.cc:2835
#8  0x081ec055 in convoi_t (this=0x9392018, wl=0x88a76c8, file=0xbfffd900) at simconvoi.cc:223
#9  0x082604fe in karte_t::laden (this=0x88a76c8, file=0xbfffd900) at simworld.cc:5089
#10 0x082619f6 in karte_t::laden (this=0x88a76c8, filename=0x8d76ebc "/home/pokus/simutrans/pak.64-experimental/demo.sve")    at simworld.cc:4718
#11 0x0821746e in simu_main (argc=1, argv=0xbffff534) at simmain.cc:1022
#12 0x0829217e in main (argc=1, argv=0xbffff534) at simsys_s.cc:682

I made git bisect.First bad version is f9907e45 (FIX: Error in the recording the last stop position of ships)

[ras52]

I wonder whether this may be the same bug as the one I'm currently seeing with pak128.Britain-Experimental on the 10.x branch.  It may not be the same, and the manifestation is certainly different, but it seems quite possible to me.

Edit:  Are you easily able to try building pak64.experimental from source using the current makeobj from 10.x?  I don't think it's likely, but it perhaps possible that the pak64.experimental binaries are incompatible with simutrans 10.x.  (For what it's worth, I can duplicate this, and I think it's quite possibly the same as the bug I mentioned above.)

[Carl]

This bug doesn't seem to arise in Windows -- I can launch pak64.experimental with the latest binaries, even though many of the pak files were generated using makeobj for 10.x.

[jk271]

I am sorry for misleading information. Loading paksets works ok. Loading of demo.sve caused failure.

I used latest working version of simutrans-experimental which was simutrans-exp-2012-02-03-6d03bfe-dbg. I saved default game and overwrote demo.sve by saved game. Now it works also with latest self-compiled versions.

File demo.sve on my computer was probably corrupted. I do not know why. It is weird for me. Thank you for your assistance and sorry for time used for this topic.

[Carl]

Thanks for the report. I'll check the demo.sve with the latest version to see if there's a problem.