9.5 works fine, 9.8 never displays any window

[wlindley]

On (32-bit) Ubuntu 10.10, experimental 9.5 works perfectly.  Downloaded the 9.8 binary and config, and tried with Pak128.Britain-Ex-0.8 ... in the terminal we get all the way to

Code Select Expand

Show banner ...
Running world, pause=0, fast forward=0 ...

but no window ever displays.  I even went into the 9.5 directory, renamed the 'config' directory and copied 9.8's config directory into config/ there, and ran ../sim-exp-9.8/simutrans-exp-2011-0706-linux-x86 ... to eliminate any missing files; same result.  It chews CPU time but never displays any graphics.  Again, 9.5 works perfectly.

Anything else I need to do, or try, or look for? 

p.s., I tried but failed to redirect all the messages that are always displayed:

Code Select Expand

Warning: display_set_base_image_offset(): image=4921 was already moved!
Warning: display_set_base_image_offset(): image=4922 was already moved!


but haven't figured how to make the redirection work yet, any hints on the magic incantation?  And why do all versions of simutrans display thousands of warnings like that? 

p.p.s., 2011-0706 looks like an ISO date but that would be July 6, which is in the future, can we please use proper ISO YYYY-MM-DD dates?

[jamespetts]

WLindley,

thank you for the report, and sorry that you are having problems. Where did you download this Linux binary? There was a problem with the normal automatically compiled binaries recently, so DeMangler compiled some, but he was not able to test anything than the no graphics server build. Try this binary and see whether it works for you.

[wlindley]

Ah, that works!  Vielen gracias!  I mean, muchas Danken...

p.s., It was http://www.box.net/shared/qri3ao5m0m from the thread at http://forum.simutrans.com/index.php?topic=7404.0 which does not. 

[deMangler]

p.p.s., 2011-0706 looks like an ISO date but that would be July 6, which is in the future, can we please use proper ISO YYYY-MM-DD dates?

My fault that. That is a format I have personally been using for a long time for various date related things. I never realised how similar it looks to ISO 8601 before.
Thanks for pointing it out. I will be more careful

dM

[Erik]

Try this binary and see whether it works for you.

It workes for me.
It solves the bug I had reported. http://forum.simutrans.com/index.php?topic=7444.0

But now there is a other bug.
The game played well at first. But sometimes it suddenly disappeared.
I have run it with gdb again.

Here is the given error output:

Code Select Expand


Program received signal SIGFPE, Arithmetic exception.
0xf7dd3a8a in __udivdi3 () from /usr/lib32/libgcc_s.so.1
(gdb) n
Single stepping until exit from function __udivdi3,
which has no line number information.
[Thread 0xf3289b70 (LWP 27351) exited]
[Thread 0xf2a88b70 (LWP 27352) exited]
(gdb)


At the moment I can reproduce the error quiet easily.
EDIT: It seems it happens when I have the screen open of a specific train on a specific moment.

If there is some more information you need.
Just ask.

[jamespetts]

Erik,

thank you for your report. Can you upload a saved game with instructions on how to reproduce the crash?

[Erik]

Of course

The .sve seems to big for a attachment.
So here a link where I have stored it elsewhere.
erik1 exp error.sve

Just klick on the train who just reach the station.
When the train go's back and is half way the slope down while you have the window of that train open.
The game will crash at more or less 11:54.

[jamespetts]

Erik,

I can't work out what version of pak128 that you're using for this: can you give me a link to your pakset so that I can reproduce this?

James.

[Erik]

O, sorry.
Totally forgotten.
I use a few add-ons

pak128 v1.4.6
My add-ons

[jamespetts]

Erik,

I still can't get this to work with those three addons plus Pak128 1.4.6 - are you sure that you only used the three add-ons?

[Erik]

Yes I didn't send all the add-ons.
But only the few that were in use.

Here the complete bunch of add-ons I use.
This time I have checked it first. (I don't like it from myself that I give such much trouble  )

[jamespetts]

Erik,

thank you for that. I have now had a chance to test it: although it is not clear what is causing the problem, I can reproduce it on the 9.8 released binary, but not my latest development code. The problem would appear, therefore, to be solved. Hopefully, the 9.9 release will work for you.

Thank you for the report :-)