Angband 4.0 beta release

[passer_by]

Originally Posted by Nomad

Judging by the pattern of which doors are missing, I'm going to take a stab that the problem is with where the symbols used in room_template.txt have been semi-standardised with vault symbols, and the "place some monsters and treasures nearby" symbol, (previously !) has been changed to a number 9. It looks like the code is possibly reading 3-followed-by-9 as 39 and failing to place a door because it doesn't recognise it's a 3, or something similar.

I agree with you, this is exactly what is happening (and nice catch spotting this just from the observed behaviour).
The code that parses these rooms contains the line (gen-room.c:1913):

doorpos = atoi(t);

where t is a pointer into the room template. Due to C's confusion between char-pointers and strings, atoi will read as many successive characters as possible from the room template and try to convert them into a number.

[Nick]

Originally Posted by passer_by

I agree with you, this is exactly what is happening (and nice catch spotting this just from the observed behaviour).
The code that parses these rooms contains the line (gen-room.c:1913):

doorpos = atoi(t);

where t is a pointer into the room template. Due to C's confusion between char-pointers and strings, atoi will read as many successive characters as possible from the room template and try to convert them into a number.

Awesome, thanks both of you.

[Nick]

Update 2cc3979 fixes:

• High resist ID issues (maybe)
• Problems with recognising {worn} items
• Resting bug (but not slowness)
• Template room door problem (thanks Nomad and passer_by)
• Mana calculations, I hope (thanks passer_by)

I'm going to try and round up what is still outstanding, and maybe start a new thread.

[krazyhades]

Originally Posted by Nick

I'm going to try and round up what is still outstanding, and maybe start a new thread.

Outstanding:
1) Nick
2) other bugs maybe
3) Nick

[Zireael]

Originally Posted by krazyhades

Outstanding:
1) Nick
2) other bugs maybe
3) Nick

Agreed so much.

[wobbly]

Well I hate to be the negative one but Nick I'm disappointed. Where's this?:

http://angband.calamarain.net/view.php?date=2007-07-20

[Darin]

Version g8470954:

Every time I start a game with randarts it crashes during randart generation. I have the latest randarts.log and the Console output if this can be reproduced. If it makes any difference I am running on Mac OS X v. 10.8.5

[tumbleweed]

For me, launching angband.exe and immediately exiting afterwards reliably crashes the 4.0 betas, no matter whether it's via Ctrl+X, menu File->Exit, window controls or whatever else. No such issue in 3.5.1.

edit: The crash is an access violation (0xC0000005). Also, loading a save and then exiting immediately, or starting a new game and then exiting immediately (so basically Ctrl+N Ctrl+X) is enough to avoid the crash. (Initialization issue?)

angband-v4.0beta-187-g8470954 Windows


Also two minor cosmetic quirks:

• Both in 4.0 and 3.5.1, trying to access the help via menu Help->Contents from the title screen does nothing, despite the menu item not being grayed out.
• edit: And showing "You can't start a new game while you're still playing!" / "You can't open a new game while you're still playing!" message boxes when using the file menu during character creation, instead of making those choices possible or simply disabling the menu items seems rather odd behavior to me. </nitpick>

[Thraalbee]

Originally Posted by Nick

Update 2cc3979 fixes:

• High resist ID issues (maybe)
• Problems with recognising {worn} items
• Resting bug (but not slowness)
• Template room door problem (thanks Nomad and passer_by)
• Mana calculations, I hope (thanks passer_by)

I'm going to try and round up what is still outstanding, and maybe start a new thread.

Thanks! Mana calculations seem right. But the problem with AC in the main screen remains. When I switch armour, the AC is not updated although SP are.

[myshkin]

Originally Posted by Darin

Version g8470954:

Every time I start a game with randarts it crashes during randart generation. I have the latest randarts.log and the Console output if this can be reproduced. If it makes any difference I am running on Mac OS X v. 10.8.5

I was unable to reproduce this crash. I'd be happy to look at what debugging information you have. Also, are you starting a new game or working from an existing savefile? If the latter, please pass along the savefile as well.