Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

Faster than 40d, no problems

42 (26.1%)

Faster than 40d, problems

72 (44.7%)

No slower than 40d, no problems

14 (8.7%)

No slower than 40d, problems

16 (9.9%)

Slower than 40d, no problems

2 (1.2%)

Slower than 40d, problems

3 (1.9%)

Doesn't work (please explain)

12 (7.5%)

Total Members Voted: 160

[Corona688]

C) providing functionality for an area where a library is broken does not amount to rewriting that library

Er, what functionality that is not provided?  SDL does provide full keyboard translation functionality.  D11 just didn't use it.  Think we all got scared by the big bad 'unicode' word, picturing some insane overcomplex parsing system.  But all it ultimately means is that SDL can tell you the exact letter a combination of keystrokes means no matter what language and arrangement your keyboard is.

Hmm.  When you guys start allowing binding to international keys, how do you plan to save the beyond-ascii characters in the config file?  Or is this solved already?  Only got a US keyboard myself, so I wouldn't know.

[Baughn]

Input issues aside, I've gotten a few more optimizations and bugfixes into the rendering code. You wouldn't think it'd be possible.. heh, but adding -mfpmath=sse hardly counts as my doing, does it?

It's in 40d13-hed, as usual. And I think that's it, no more bugs. I hope.

[SimRobert2001]

question for you guys: Why can't i mine? I try to hit the enter key, but no selector comes up, os i can't select any area to mine.

in addition, i can't move up or down z levels at all.

i'm also using D13

[0x517A5D]

I've got a couple of things.  .40d13 on Windows.  ATI Radeon 4870, Feb 2009 drivers.  Standard American querty keyboard.

KEY_HOLD_MS doesn't seem to have any effect.

KEY_REPEAT_MS seems to control the initial delay, like KEY_HOLD_MS is supposed to.

Nothing seems to control the actual held-key repeat rate that KEY_REPEAT_MS is supposed to control.  It seems to use the hardcoded (init.h) 150ms value.


GRID:xx:yy controls both the windowed grid size and the fullscreen grid size.  FULLGRID:xx:yy seems to do nothing.  This matters when the fullscreen font is different than the windowed font.  I use the curses_640x300.png (8x12) for fullscreen, and a doubled version of that font (16x24) for windowed.  When I switch to fullscreen mode, the font changes properly, but the screen only shows the same amount that windowed mode would show.  I can then use the mousewheel to zoom, which does increase the number of columns and lines displayed, eventually enough to fill the screen.  If I do zoom, then switch back to windowed mode, the window is zoomed out quite far.


In fullscreen mode, my flashing bottom line bug occurs in .40d10 through .40d13.  It does not occur in .40d9.  I do believe this to be a driver issue.  But it would be nice if you would find a different way to blank each frame in fullscreen mode.  Maybe draw a black quad that's extremely zoomed in on?


I will add these reports to the github tracker.

[Baughn]

GRID and FULLGRID are both about to be deprecated, as they can perfectly well be automatically calculated. There may be an initial zoom setting as a replacement.

Thus, their not doing very much anymore is quite deliberate.

[Baughn]

In fullscreen mode, my flashing bottom line bug occurs in .40d10 through .40d13.  It does not occur in .40d9.  I do believe this to be a driver issue.  But it would be nice if you would find a different way to blank each frame in fullscreen mode.  Maybe draw a black quad that's extremely zoomed in on?

Can you check whether this happens in ACCUM_BUFFER or FRAMEBUFFER mode? Test both, please.

[Random832]

C) providing functionality for an area where a library is broken does not amount to rewriting that library

Er, what functionality that is not provided?  SDL does provide full keyboard translation functionality.

Um, I thought someone actually SAID that it fails to translate for non-US keyboards - i.e. if you press the key next to tab, it always comes back with the keycode for "q" even though on a french keyboard it is "a".

[Veroule]

KEY_HOLD_MS doesn't seem to have any effect.

KEY_REPEAT_MS seems to control the initial delay, like KEY_HOLD_MS is supposed to.

Nothing seems to control the actual held-key repeat rate that KEY_REPEAT_MS is supposed to control.  It seems to use the hardcoded (init.h) 150ms value.

Seems to be a problem in the init.txt loading.  Toady will have to fix it.

question for you guys: Why can't i mine? I try to hit the enter key, but no selector comes up, os i can't select any area to mine.

in addition, i can't move up or down z levels at all.

i'm also using D13

You are probably encountering the mouse bug.  I put a few lines in the wrong spot when merging the code for D13 off of my test files.  Turning off the mouse in the init.txt will keep it from happening.

[0x517A5D]

In fullscreen mode, my flashing bottom line bug occurs in .40d10 through .40d13.  It does not occur in .40d9.  I do believe this to be a driver issue.  But it would be nice if you would find a different way to blank each frame in fullscreen mode.  Maybe draw a black quad that's extremely zoomed in on?

Can you check whether this happens in ACCUM_BUFFER or FRAMEBUFFER mode? Test both, please.

Okay it does occur in ACCUM_BUFFER mode, but it does NOT occur in FRAME_BUFFER mode.  So that's a valid workaround.

And looking at my .40d9 config, it was in frame buffer mode.  I think it was the change in syntax that tripped me up.  Thanks.

[tdittmar]

Hi,

i tested 40d13 + head (after July 11 20:00) on my Linux system (Debian testing, 32bit, ATI X800).
There seems to be a problem with the screen update when DF is _unpaused_.
when the game is paused, dwarfs and multi-item tiles blink as expected, when the game is unpaused the screen is not updated. F12 or pausing updates the screen.
I can reproduce this behaviour with STANDARD, PARTIAL:2 ACCUM_BUFFER, FRAME_BUFFER, and VBO.

I also experience a strange behavior while starting DF: If I start it by double clicking df, it crashes while beeing in the intro. If I start df from console, it works besides the screen update problem.

Here a typical error message I get in the console:
Using OpenGL output path with buffer objects

GPU does not support non-power-of-two textures, using 512x512 catalog.
*********************************WARN_ONCE*********************************
File r300_mem.c function r300_mem_alloc line 225
Ran out of GART memory (for 1048576)!
Please consider adjusting GARTSize option.
***************************************************************************


Greetings from the sunny Provence,

Timo
 
p.s. Many thanks to Toady for making this marvelous game and also many thanks to all the others who help making a native Linux port.

[Heliomance]

I may be missing something, but I can't get it torunin the first place. I'm runnning Kubuntu Jaunty, and when I try and run the df script file, it does precisely nothing.

[tdittmar]

@Heliomance:

Have you tried running it from a console? I had the same problem double clicking the df script, but running from console worked. Did you install the head update?

Timo

[Heliomance]

Code: [Select]

tcoton@tcoton-laptop:~$ Downloads/Dwarf\ Fortress/df_linux/df
Downloads/Dwarf Fortress/df_linux/df: 5: ./dwarfort.exe: not found
Yet dwarfort.exe is right there.

EDIT: Okay, fixed that, it had the path wrong.Now I get this:

Code: [Select]

tcoton@tcoton-laptop:~$ Downloads/Dwarf\ Fortress/df_linux/df
./Downloads/Dwarf Fortress/df_linux/dwarfort.exe: error while loading shared libraries: libSDL_image-1.2.so.0: cannot open shared object file: No such file or directory
tcoton@tcoton-laptop:~$

EDIT EDIT: Fixed that too. The subtleties of Linux filepath parsing confound me. Now it seems to work.

Gah! Now it can't find data/art/curses.png, and that's not a setting in the df script file. Why does it want everything to be in the top levelof my home directory?

Gave up and copied the data folder tomy home directory. Works now, but I'd appreciate it if anyone knows how to keep the data folder in the main DF folder, and still have it work.

*twitch* *twitch* it segfaults out at worldgen. Moving everything tomy home directory fixes this,but I don't want it there! It's messy!

[bartavelle]

Gah! Now it can't find data/art/curses.png, and that's not a setting in the df script file. Why does it want everything to be in the top levelof my home directory?

It doesn't want everything in your home directory, it wants everything under the current directory. Just cd to it !

[Heliomance]

...
...
...

Why on earth didn't I spot that myself?