Current master post-4.2.2

[whartung]

The floor is also harder than most orcs ;-)

But the tangential energies of floor impact combined with the integral flex ratios of the arrow material should greatly reduce breakage.

No doubt there should be some friction model based on the floor space tile type to better model the drag coefficient of the arrow (no doubt related to the fletching material, you know, in for a penny and all) to better place the arrow down the hall as it goes sliding after contact.

I leave it as an exercise for others to discern whether there should be any skipping.

[MattB]

The floor is also harder than most orcs ;-)

[will_asher]

Walls are harder than orcs, so they break arrows, too.

But if it runs out of range and lands on the floor it shouldn't break.

[fph]

Walls are harder than orcs, so they break arrows, too.

[will_asher]

I just noticed that in the description of arrows it says "35% chance of breaking on impact" but IIRC, the code has a chance of the arrows breaking even if they don't hit anything. Which is it supposed to be? on impact or on firing?

[Nick]

You didn't, Fizzix did

Well, fizzix did all the work of designing the new method, but I made the tiny change that made it actually appear in the game...

[PowerWyrm]

Neat, I forgot I did that

You didn't, Fizzix did

[Nick]

IIRC that's a new feature. Looking at the code, the summon function now has a "call" parameter which uses a new method of summoning ("calls out" a monster that is present on the level) and this parameter is set to "true" 25% of the time randomly.

Neat, I forgot I did that

[MattB]

IIRC that's a new feature. Looking at the code, the summon function now has a "call" parameter which uses a new method of summoning ("calls out" a monster that is present on the level) and this parameter is set to "true" 25% of the time randomly.

Oh. Fair enough then.
That might call an end to my traditional level-ending of standing on the stairs with a staff of summoning trying to call up some cheap low-level uniques.

[PowerWyrm]

Using this code everywhere panel dimensions (in particular "screen_wid / 2" occurences) are used fixes the problem:

int screen_hgt, screen_wid;
    int panel_wid, panel_hgt;

    screen_hgt = p->screen_rows / p->tile_hgt;
    screen_wid = p->screen_cols / p->tile_wid;

    panel_wid = screen_wid / 2;
    panel_hgt = screen_hgt / 2;

    /* Paranoia */
    if (panel_wid < 1) panel_wid = 1;
    if (panel_hgt < 1) panel_hgt = 1;

The only remaining issue with this is that the @ isn't visible if panel_hgt = 1 (it's displayed one line below).

[PowerWyrm]

If you try to use the "l"ook command when using 16x16 tile size and you end up with a panel of size 0, the command will also loop endlessly.

[PowerWyrm]

I just TO'd the Balrog of Moria, then summoned him on the same level using a staff of summoning.
Should that happen?
Feels weird to me...

IIRC that's a new feature. Looking at the code, the summon function now has a "call" parameter which uses a new method of summoning ("calls out" a monster that is present on the level) and this parameter is set to "true" 25% of the time randomly.

[MattB]

I just TO'd the Balrog of Moria, then summoned him on the same level using a staff of summoning.
Should that happen?
Feels weird to me...

[PowerWyrm]

Also the sector displayed makes absolutely no sense...

It should start at [0,0] if you're in the northwesternmost panel and should be increased by 1 each time you move east or south from there. See MAngband code:

/* Start at current panel */
	y2 = p_ptr->panel_row;
	x2 = p_ptr->panel_col;

/* Apply the motion */
	y2 += ddy[dir];
	x2 += ddx[dir];

/* Prepare to ask which way to look */
	sprintf(out_val,
	        "Map sector [%d,%d], which is%s your sector.  Direction?",
	        y2, x2, tmp_val);

/* Set the panel location */
	p_ptr->panel_row = y2;
	p_ptr->panel_col = x2;

[PowerWyrm]

- open client
- set tile size to 16x16
- use "l"ook command
- crash

In do_cmd_locate():

int panel_hgt = (int)(PANEL_SIZE / tile_height);
int panel_wid = (int)(PANEL_SIZE / tile_width);

strnfmt(out_val, sizeof(out_val),
		        "Map sector [%d,%d], which is%s your sector.  Direction?",
		        (y2 / panel_hgt), (x2 / panel_wid), tmp_val);

But PANEL_SIZE = 11, so if you use tile size greater than 11 the values for panel_hgt and panel_wid will be 0 and will provoke a division by zero error in the following line...