Sangband 1.0.1's new thread

[hitchi]

So am i.I did recompile nss, to no effect.But S isn't alone now; NPP also segfaults the same way, while angband isn't affected.I'm googling for it.

Found a workaround on Comment 7 of this page:

https://bugzilla.redhat.com/show_bug.cgi?id=580709

Weird stuff, pppd, wich the bug report was about, was also segfaulting here, but i didn't notice because i run it from a cron job and it doesn't segfault that way?!
I'll try to find a patch for glibc that really fixes the issue and submit a bug report to Lunar(i'm a tad confused).
Meanwhile S is running in it's full glory!

[Bostock]

2-weaponing feature request

2-weaponing is great (and it's the best-balanced 2-weaponing I've ever seen). But it's annoying from an interface perspective, especially when you usually wield a shield in the second hand: first you spend keypresses unwielding the shield (quite possibly also making it fall on the floor), then wielding the second weapon and ensuring it goes to the 2nd hand.

Could there perhaps be an inscription that would make the inscribed weapon automatically wield into the 2nd hand if valid, unwielding the shield in the process if needed, and charging the appropriate turn cost? (But perhaps never forcing the shield onto the floor - while fair and logical, it's a little mean and annoying.)

[Bostock]

Use-ID Bug

If an item is destroyed during use-ID (not just consumed, but boom), you get the message that e.g. "it seems that you are using (nothing)." This is only really a problem in practice for staffs of chaos use-ID'd from the floor, but... well, just call me a perfectionist. :-)

[camlost]

I used to run sangband(sdl) in the console(linux console) framebuffer; and switch to X11 via CRT+ALT+f7.This works for any SDL apps running in the framebuffer but there *must* be a way to switch from a full screen sangband(running in X11) to windowed mode.

You can run S in windowed mode. Check the display options (=!a) for a windowed mode. You will need to quit and restart for them to take effect.

You might also be able to use keyboard commands to switch between windows, I'm not sure.

[hitchi]

Ok so I compiled 1.0.1.169 from source with make -fMakefile.sdl (.std with X11 is incompilable anyway), fired it up and -- how do one switch to other windows/desktops? This is FreeBSD (but think Linux+X11+GNOME, should behave the same).

I used to run sangband(sdl) in the console(linux console) framebuffer; and switch to X11 via CRT+ALT+f7.This works for any SDL apps running in the framebuffer but there *must* be a way to switch from a full screen sangband(running in X11) to windowed mode.

[pav]

Ok so I compiled 1.0.1.169 from source with make -fMakefile.sdl (.std with X11 is incompilable anyway), fired it up and -- how do one switch to other windows/desktops? This is FreeBSD (but think Linux+X11+GNOME, should behave the same).

[hitchi]

It's a pretty fundamental error - I'm surprised it hasn't broken any other apps.

So am i.I did recompile nss, to no effect.But S isn't alone now; NPP also segfaults the same way, while angband isn't affected.I'm googling for it.

[Magnate]

Did anyone know that you can loot traps in S? I mean, when you disarm them, you get items? It was news to me last night, but it requires a very high disarm skill. I've made it more accessible in SVN.

I didn't know that, but I shall rest more easily now that I know my rogues have a reliable supply of lead shots...

[Magnate]

Note:"mv /lib/libnss_files.so.2 some_place_else" allows the game to start(of course it also breaks my system...)

Ouch. I'm not familiar with Lunar Linux, but it sounds like a Gentoo-type system where you compile all the packages from source? If that's the case, have you tried recompiling libnss? It's a pretty fundamental error - I'm surprised it hasn't broken any other apps.

[NotMorgoth]

I've just rediscovered another one of my pet hates about Sangband:

When you get the level feeling 'Premonitions of death appall you, this place is murderous,' it causes fear, which if it happens at the wrong time, ie in the middle of a fight, can in itself have potentially fatal consequences.

I think that if there are going to be level feelings they should be of some help and not make it more likely that you will die.



On a related point, I also think the precognition messages you get from high perception could also be made more helpful.

Maybe they should be less common, but when you do get them they should always be accurate, otherwise they are worthless.

For example, when you get the message 'you sense a horde of monsters,' which I assume is meant to indicate a pit or something similar, if you know it is not certain to be correct, all it really tells you is that there might be a horde of monsters on the level. But since that is true for any level, it's not really telling you anything at all.

[hitchi]

Recompiling glibc didn't work, this is what i'm getting with "gdb -ex r -ex back -ex q"

[Thread debugging using libthread_db enabled]
[New Thread 0xb75feb70 (LWP 27058)]
[New Thread 0xb6dfdb70 (LWP 27059)]

Program received signal SIGSEGV, Segmentation fault.
0xb77363f7 in internal_getent () from /lib/libnss_files.so.2
#0 0xb77363f7 in internal_getent () from /lib/libnss_files.so.2
#1 0xb77362fe in internal_setent () from /lib/libnss_files.so.2
#2 0xbfffe548 in ?? ()
Backtrace stopped: previous frame inner to this frame (corrupt stack?)
A debugging session is active.

Note:"mv /lib/libnss_files.so.2 some_place_else" allows the game to start(of course it also breaks my system...)

[camlost]

Did anyone know that you can loot traps in S? I mean, when you disarm them, you get items? It was news to me last night, but it requires a very high disarm skill. I've made it more accessible in SVN.

[camlost]

Thanks for the tip camlost; i used gdb to debug it and found out it's glibc acting up.

I will try recompiling glibc, thanks again.

It's possible that the fault is in S's code. If recompiling glibc doesn't fix it and you can give me a full stack trace, I'll look into the problem. You can email it to me at middendorfproject@gmail.com

[hitchi]

Thanks for the tip camlost; i used gdb to debug it and found out it's glibc acting up.

Program received signal SIGSEGV, Segmentation fault.
0xb77363f7 in internal_getent () from /lib/libnss_files.so.2

I will try recompiling glibc, thanks again.

[camlost]

I'm using the latest stable SDL(1.2.14); SDL_image(1.2.10); SDL_ttf(2.0.10); SDL_mixer(1.2.11); SDL_sound(1.0.3) .

I compile on Linux (Ubuntu 10.04) using (according to package manager):
* libsdl1.2-dev version 1.2.14-4ubuntu1.1
* libsdl-image1.2 version 1.2.10-1
* libsdl-ttf2.0-0 version 2.0.9-1build1
* libsdl-mixer1.2-dev version 1.2.8-6build1
* libsdl-sound1.2 version 1.0.3-3build1

I am not getting a segfault on run.

As for guesses, I don't have any. I didn't do the rewrite to SDL (assuming that is a problem). Perhaps some debugger output would suggest a place to start looking into the problem?