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

[Enzan]

Is there any word on how the Mac version is doing?

[0x517A5D]

ATI 4870, december 2008 drivers.  XP SP3 up to date, recent install, so little cruft.

It works well with one slight issue.

I get a flickering dashed line on the very bottom line of the screen, if certain things are true:

• The game is in fullscreen mode.
• The video mode is 1680x1050, native for my monitor.
• The combination of FULLGRID's Y setting and the font size is less than 1050.
• BLACKSPACE is set to YES, so that the tiles are not stretched to fill the screen.

If any of those are changed, the issue does not occur.

Here's a link to a screenshot.  That line at the bottom rapidly flickers on and off.  I suspect it is at the frequency of G_FPS_CAP.

[Baughn]

None so far.
Toady's been having trouble with stripping the binaries sufficiently; the OS X defaults reveal more debugging information than he's comfortable with.

[Baughn]

ATI 4870, december 2008 drivers.  XP SP3 up to date, recent install, so little cruft.

It works well with one slight issue.

I get a flickering dashed line on the very bottom line of the screen, if certain things are true:

• The game is in fullscreen mode.
• The video mode is 1680x1050, native for my monitor.
• The combination of FULLGRID's Y setting and the font size is less than 1050.
• BLACKSPACE is set to YES, so that the tiles are not stretched to fill the screen.

If any of those are changed, the issue does not occur.

Here's a link to a screenshot.  That line at the bottom rapidly flickers on and off.  I suspect it is at the frequency of G_FPS_CAP.

Hum. A consequence of only calling glClear *once*, I expect, on a double-buffered screen.

The drivers are supposed to clear the buffers for us, but in your case there seems to be an off-by-one error - they forget about the last line.

[David]

dwmapi.dll and efsadu.dll were missing for me. Once I had these, it worked, but there was no increase in frame rate.

[Talvara]

after installing the win32 stuff it worked for me, havent looked at the framerate yet, it seems alright.

what I did notice tough is that this version is alot more unforgiving on bad tilesets. if you make a mistake (ie. pointing at an non existant file) it will let you know and crash.  where the old version just sucked it up and showed you the curses instead.

I prefer this better btw

edit: it seems like a fair increase of speed aswell for me. I only tried on embark.

did 2 embarks on 2 human towns. the Open GL one got me a fps of 300+
and the origional one first got me 100fps, but after a minute that droped to 30ish fps

Im running on a laptop windows Vista (dualcore 3.2 gig) (3gig Ram) ( nvidea Gforce 8600M GT (512 ram))

[bhelyer]

if you make a mistake (ie. pointing at an non existant file) it will let you know and crash.

It's not crashing if it's done on purpose, it's just exiting. :]

[Baughn]

No, it's crashing.

It's just a controlled crash. Fatal errors are never a good thing, but they're much better than *byzantine* errors, and even if the consequence would be no worse than showing the wrong tiles.. that's still showing the wrong tiles.

Presumably, if you add a timeset, you'd like DF to actually use it. 

[Appelgren]

When I first ran 40d2 it wouldn't start, warning me about "incorrect application configuration". I tried to solve it by unpacking the win32_libs.rar, but I got the same error message. When I ran Dependency Walker it couldn't find the files EFSADU.DLL and MSJAVA.DLL

[Baughn]

Neither of those are actuallly used by DF, nor does it link directly to them. msedf.dll is new, but that's actually some encrypting filesystem thing..

Side-along libraries are an utter annoyance to handle. I'm installing microxp on a spare computer right now, which should help me diagnose these problems - with any luck I'll get the same errors you do.

Oh, and 40d3 is out.

[Appelgren]

Ok, I'll try that.

[Veroule]

The latest 40d3 runs for me without anything special.  My immediate reaction is that the initial loading of it is horrible.  This is just a result of having so many different dll's to pull in.

Next I found a flaw when comparing with the current 40d for basic speed.  I set the 40d3 init information to a matching level with 40d
FPS:10000
G_FPS:10
PARTIAL_PRINT:YES:1
I didn't match grid sizes yet.  Got an FPS of 200 at the main menu with 40d3.
I then shifted the G_FPS to 50 for each.  40d produced the same FPS it had been (partial print of course), whereas 40d3 suddenly jumped to 1000 FPS.  Since this is an exact ratio match 10->50 and 200->1000 I am thinking there is an error in the math determining how often to update.  Setting the G_FPS lower should allow an increase in FPS, not the other way around.

[Baughn]

The logic is fine, but the current timer logic has a hard clamp of 20 frames per graphical update. I've suggested that that should be increased.

[IndonesiaWarMinister]

Isn't partial print broken in 40d3?

It will be fixed in 40d4 (Baughn told me in the IRC)

Is it?

[Baughn]

Broken in 40d2 and 40d3, should be fixed in 40d4.

That people get increased performance *anyway* is certainly interesting.