Recent Posts

That came out great! 

Hi arc11,

1.- The input_mister backend requires SDL for its scancode definitions. So the actual fix for this will involve moving INPUT_MISTER into the build scripts so the dependency is properly handled, instead of moving it into input_mister which basically allows the build by breaking the module. I'll fix this when I have a chance.

2.- I'll add your fix for jammasd/rawinput.

3.- Font rendering is fine for everyone, at least with the suggested use of uismall font. I'll tell you the same I did in 2025: please provide the exact configuration that makes the fonts look bad (destkop res, ini, screen-shot, etc. -hopefully before 2027 ) anything that allows me figuring out the issue in case there is an edge case we didn't cover. The change you suggest breaks an implemented feature (integer scaled ui fonts).

Hello

I recently downloaded and compiled the mame 0.287 (windows build with msys2 without SDL )
The compilation build is fine and the mame binary works as expected.

With GM 0.287 I have the same issues as reported « Reply #467 on: June 18, 2025, 10:05:33 am ».

The compilation is broken.

Code: [Select]

Compiling src/osd/modules/input/input_common.cpp...
D:/Mame0287/src/osd/modules/input/input_common.cpp:34:10: fatal error: SDL2/SDL.h: No such file or directory
   34 | #include <SDL2/SDL.h>
      |          ^~~~~~~~~~~~
compilation terminated.
To fix it I did :

Code: [Select]

The "#define MISTER_INPUT 1"  line must be moved from input_common.h to input_mister.cpp (before the first #include)
The input of jammasd is not recognised

to fix it I did :

in src/osd/module/input/input_rawinput.cpp

Code: [Select]

      // this is for duplicate devices in a collection such as extra mouse buttons
      if (tname_basic_string.find(L"&Col01") != std::string::npos)
         return nullptr;

When I comment these 2 lines, the jammasd keyboard is correctly detected :



and the mame osd menu font size is unreadable.
to fix it :

 in the file src/frontend/mame/ui/ui.cpp
this change :

Code: [Select]

    float const one_to_one_line_height = float(raw_font_pixel_height) / target_pixel_height;
 
    // determine the scale factor
-   float scale_factor = target_font_height() * scale / one_to_one_line_height;
+   float scale_factor = std::min(target_pixel_height / float(raw_font_pixel_height) / options().font_rows(),
+                        target_pixel_width / float(raw_font_pixel_height) / options().font_rows());
 
    // if our font is small-ish, do integral scaling
    if (raw_font_pixel_height < 24)

revert this code to the standard mame code...


If it can help others....

Thanks Scott, that's actually pretty interesting to me. Tables with upper flippers! Can you give me examples of some? I want to check them our. 

*edit, you just mean tables with upper flippers in general that are scripted to allow separate control right? 

ooh I love the retro aesthetic of that! Great work, I'm sure he'll love it!

What type screen do you have in that? Is it a vertically mounted TV/monitor or something?

I was thinking along the same lines, about installation depth, and will do some experimentation.

Yeah, I was off on the completely wrong track about the slits. I see what you mean now.

How deep you install the light pipe might make the LEDs easier to see.

If you press the light pipe in so it is flush with the surface of the case, it will definitely be harder to see at a low angle.

I 100% believe you, but cannot see it in these images. Maybe a different perspective would help.

All good. I have in fact allowed for the PCB thickness in the design. There is a gap there. I'll check your other requests soon.

the wings may need to be a little larger, the holes up to 4mm diameter, and moved a little further away from the body.

Agreed.  Better 6mm too much room between the screw hole and the side of the case than 1mm too close.
- Might be good to allow enough room for a socket if anyone wants to use a bolt/hex head fastener.

I like the idea of closing the hole with a light pipe, to avoid damage from spilled sodas and such. A drawback is that the LEDs would be less visible from the side, such as when installed on the in-side a cab.

How deep you install the light pipe might make the LEDs easier to see.

If you press the light pipe in so it is flush with the surface of the case, it will definitely be harder to see at a low angle.

If you start to put in the light pipes then place the case cover face down on a hard flat surface and firmly press down in the center of the ant head logo, that should seat the light pipes so they stand proud the same amount as the ant head logo does, increasing visibility.
- If this is how you want to proceed, consider adding an outline rim on the case the same height as the logo.  That would also make it easier to 3d print face down.

Your side slits idea might further allow a compromise as it would allow the LEDs to be visible from the side (barely), while still mostly protecting from unplanned soda events.

I wasn't suggesting slits in the side of the case, just a 0.5 mm wide by 6-8 mm high slit in each LED cylinder (red lines in this pic) for greater flexibility and friction to hold the light pipe firmly in place without adhesive.




Scott

Here's a retirement present for my co-worker, who casually mentioned months ago that he loved the old Atari 2600 game Megamania.

Batocera running on an old rpi3b I had laying around. The marquee and control panel artwork was inspired from the Megamania box art. I added a Turbotwist spinner from my long defunct cocktail cabinet so he could play paddle games as well. I curated a list of mame, Atari 2600, and Atari 800 games. Overall I'm really happy with the size, aesthetic, and ergonomics. Now I need to make myself one just like it 

 

All good. I have in fact allowed for the PCB thickness in the design. There is a gap there. I'll check your other requests soon.