Angband 4.2.4

[animal_waves]

X11 frontend & shift-numpad to run does not work?

It seems I can't run with the SHIFT-<numpad> when using the X11 frontend (tried with numlock enabled & disabled), is it normal?
Any advice on how to debug this (via a log/verbose argument or whatever external tool)? I'm using an Azerty keyboard, that may have something to do with it...
Weird thing is CTRL-<numpad> works for interacting with doors, traps, etc.

I already tried to change the keyboard layout of my desktop environment as well as exporting the KEYMAP variable to the angband binary with no luck.

It works fine on other variants (Sil-Q & FCB) using X11.

[edit]
Unrelated, but if someone know if it's possible for the SDL2 fronted to use system fonts instead of those fonts shipped with the game I'll be interested to know how!
[/edit]

[backwardsEric]

It seems I can't run with the SHIFT-<numpad> when using the X11 frontend (tried with numlock enabled & disabled), is it normal?
Any advice on how to debug this (via a log/verbose argument or whatever external tool)? I'm using an Azerty keyboard, that may have something to do with it...
Weird thing is CTRL-<numpad> works for interacting with doors, traps, etc.

There is the debugging command to display the last 8 keystrokes that were relayed from the frontend. To use that press ctrl-a followed by L (it's also accessible via the Enter menu: Enter -> Hidden -> Debug mode commands -> Query -> Keystroke log). Since enabling the debugging commands marks the save file, it's probably best to try that with a character that you're not planning to submit to the ladder.

Sil's and Frogcomposband's key handling in the X11 front end look fairly similar. vanilla's handling changed around version 3.0.7 and added special case handling of the keypad.

[animal_waves]

Thanks that was helpful!

It seems the combo with the CTRL modifier interpret the numpad key as as number while the combo with the SHIFT modifier interpret the same key as a direction, so that explains the bug I guess.

Cfr. attachment: I typed CTRL-6 to clear a rubble, then SHIFT-6 to run to the right.

I hope it can help!

[backwardsEric]

Thanks that was helpful!

It seems the combo with the CTRL modifier interpret the numpad key as as number while the combo with the SHIFT modifier interpret the same key as a direction, so that explains the bug I guess.

Cfr. attachment: I typed CTRL-6 to clear a rubble, then SHIFT-6 to run to the right.

I hope it can help!

There's a default keymap to run right for {SMK}[Right] but not for {SK}[Right]. It's not clear to me from history for the default keymaps why that is. You can likely trigger the default keymap by using ALT-SHIFT-6 or add a custom keymap so SHIFT-6 works to run right. Will look into putting {SK}[Right] (and the equivalents for 1, 2, 3, 4, 7, 8, 9) into the default keymaps.

[Nick]

There's a default keymap to run right for {SMK}[Right] but not for {SK}[Right]. It's not clear to me from history for the default keymaps why that is. You can likely trigger the default keymap by using ALT-SHIFT-6 or add a custom keymap so SHIFT-6 works to run right.

It may be a question of whether NumLock is on or not.

[backwardsEric]

Unrelated, but if someone know if it's possible for the SDL2 fronted to use system fonts instead of those fonts shipped with the game I'll be interested to know how!

An ugly way to do it would be to symbolically link the other fonts you want to use into lib/fonts in the Angband directories. As an example, on a Debian system with the fonts-liberation package installed, you could:

1. Change directories to lib/fonts in the directory where Angband is installed.
2. Run "ln -s /usr/share/fonts/truetype/liberation/LiberationMono-*.ttf ."
3. Now those fonts will be available the next time you start the SDL2 front end.

The caveats with that are:

1. You have to have write access to the directories where Angband installed.
2. The font files linked will have to be monospaced and in a format that the SDL2_ttf library can read.
3. There's a hardwired limit of at most 128 fonts the SDL2 front end will show in its font selector.
4. lib/customize/font-sdl2.prf references a centered dot for open floors that's present in Angband's .fon files but may not work well with other fonts. Setting up a custom font-sdl2.prf in your Angband user directory, ~/.angband/Angband, would work around that. Having "featpen floor:*:1:0x2e" (to use a plain period, referenced by its ASCII code) or "featpen floor:*:1:0xb7" (Unicode centered dot) in that font-sdl2.prf file may work.

[Naranathu Bhranthan]

Books

Hi there,
I can not find any books in the object.txt.
Can anyone please tell me where they are listed?
I am mainly interested in the level I have the best chances to find those that are not available in the shop, so that info would be of help, too.
Thanks!

[DudeMan]

Hi there,
I can not find any books in the object.txt.
Can anyone please tell me where they are listed?

They can be found from class.txt. The information you seek is on a line with 'book-properties'.

[Naranathu Bhranthan]

Found the ranger-book properties with the druid. Thanks a lot!

[Chud]

I think I have possibly found a bug with the order actions are taken in the dev release I'm playing -- 4.2.4-99-g3b1884c5c on Windows

I realize that this is not the most current, but I did skim the notes in later updates and didn't see anything that looked related.

It seems that sometimes when actions are done for a turn, and the game is sitting waiting for my input, I will enter an action, and then the monsters make their moves, and only after they are done my action then executes.

I've noticed what I thought were odd behaviors before that could be the same thing, but I found an example that's more clear.

High level priest fighting 3 huorns. My glyph of warding breaks, I get hit a few times, and then it's my turn. I cast glyph of warding again. Next I get a series of messages about them casting darkness and me resisting the effect, during which "glyph of warding" is not displayed in the status line. Only once those messages are cleared does the status line update.

It might well be just a message-order glitch though, since as far as I remember they do not actually hit me in this situation. It does seem like there are times when I enter a move and the monsters get to move before my input executes though - it is hard to duplicate, but I'll grab a save if I can.

My guess is the 4.2.4 release version does the same thing; I jumped to the dev version because I thought the 4.2.4 release seemed like it was being occasionally weird like this.

[Chud]

The more I think about it the less sense this makes - sorry. I'm sure the order of who moves when is governed by a fairly outer-level loop, and I imagine there probably can't really be any way to get it out of sync. Hmm...

[smbhax]

^c ("special - break") seems to have a default keymap to Q ("Kill character & quit").

Is that supposed to be there? Did it always do that? In the past few weeks I started noticing that I was getting "Are you sure you want to kill your character?" when I would accidentally hit ^c instead of the key right next to it, ^x, when I wanted to save. Seems like a dangerous arrangement. But I don't *think* I somehow set that keymap myself; found it in a fresh install even after deleting all the old junk from my Windows\User\[username]\AppData\Local\VirtualStore\ProgramFiles\angband folder. Besides in the Windows standard version, I also get it in my MSYS2 SDL2 version.

And if I clear that keymap:
- pressing ^c just pops up the usual unmapped key "press ? for help" text--so does "special - break" just mean keymap Q?
- if I ^x (save and quit), then run the game again, ^c is back to being keymapped to Q!

Is any of that supposed to be happening?

[Nick]

^c ("special - break") seems to have a default keymap to Q ("Kill character & quit").

Back in the old days before Microsoft started using Ctrl-C for copy, it was the key combination to stop a running program. Rather than actually allowing that to happen, Angband catches the key combination and asks the player if they're trying to kill the character.

[gglibertine]

So I've been playing a half-troll warrior recently to keep myself distracted, and thinking about the issue with light weapons giving more blows and thus being more damaging than heavy weapons.

This is probably a dumb question, but would it be possible to change the way damage and blows are calculated such that DEX contributes more to light weapons, and STR contributes more to heavy ones? Or does it already work that way and I'm just clueless?

(I haven't looked at the code because I probably wouldn't understand it; I don't math good and my rudimentary coding skills are decades out of date.)

[Estie]

It's not so much that dex doesn't contribute enough to #blows, it's more that str contributes too much. Why does high str allow you to attack faster with a dagger ?

Long ago I made a proposal to let #blows be determined by dex alone and have str put a cap on that number dependent on weapon weight. You could adjust the numbers such that the damage output would be similar to the status quo; the difference would be that the level 1 half troll would achieve his best result with something like a longsword (~2 blows) as opposed to a dagger (same ~2 blows) or a lance (~1 blow). As the dex goes up, so does the #blows, as the str goes up, so does the weight allowance so eventually he gets the same #blows with the lance.

At the same time the level 1 halfling warrior would get ~3 blows with the dagger and only ~1 with anything heavier, while other classes use the same cap for str but get a lower base #blows from their dex stat (eventually capping out at the current max values of 6, 5 or 4).