New MAME release

[Retrospect]

Now then guys n galls .... 

 

I've modified a layout from Barry Nelson (kept his name in the layout file as well as mine added)

 

This is a MAME bezel thingy ... place the zip file in your ARTWORK folder, select the TI994a (hopefully both US and Europe should work)

And it'll give you this rather lovely scene of an early 80s living room with shit-horrible wallpaper , a dusty not kept after TI99 and an old Hitachi television.

I've also added a CRT filter to the screen area. Note: There's no wires to/from the TI to the telly.  I'm not THAT good. 

 

ti99_4a.zip <<< put in artwork folder

 

 

[+mizapf]

Did you already experiment with the HLSL?

 

OSD Menu -> Slider controls -> step down to "Window 0 Screen 0 Effect" and press cursor right, stepping through all options. For a TV set, the setting "crt-geom" could fit nicely.

[Retrospect]

21 minutes ago, mizapf said:

Did you already experiment with the HLSL?

 

OSD Menu -> Slider controls -> step down to "Window 0 Screen 0 Effect" and press cursor right, stepping through all options. For a TV set, the setting "crt-geom" could fit nicely.

Thanks @mizapf , no , i didn't experiment with HLSL , what I did was place a .png file with CRT effect over the screen area (which had been blanked for the ti screen already)

I might have a look at that option now though.  What would be really nice now, is some way to have the screen rounded around the edges, maybe even a bit of CRT vignette too.  Phospher?

[+mizapf]

Actually, you can easier get to the option in the command line, as I just checked:

 

mame ti99_4a [...your options...] -bgfx_screen_chains crt-geom

 

So the effect I mentioned in the OSD menu is in BGFX, not in HLSL.

 

More details on https://docs.mamedev.org/advanced/index.html

[+Schmitzi]

On 4/1/2023 at 7:59 PM, mizapf said:

Thanks for asking. No, everything quiet here: https://www.ninermame.org/info/releases

During the last months I've been a bit busy with real-life things, in particular research work and the start of the new courses at our university. But I'll continue as soon as enough free time shows up.

 

 

 

Hi, at what version have the command line parameters been changed ?

Still trying to find out but no chance. I think it was around 0.172 or some later ?

thx

 

[+mizapf]

I guess you mean the ioport change. See also https://ninermame.org/info/impchange

 

[+mizapf]

Release 0.254 is here, nothing new, everything still working.

 

Nevertheless, I uploaded new TI99 builds of MAME to WHTech for Linux, Windows, and Raspberry OS. In the macOS folder I put a full package copied from https://sdlmame.lngn.net/

 

See

https://ftp.whtech.com/emulators/MAME/ti99/linux/

https://ftp.whtech.com/emulators/MAME/ti99/windows/

https://ftp.whtech.com/emulators/MAME/ti99/raspbian/

https://ftp.whtech.com/emulators/MAME/full/macos/

 

 

[.mad.]

Thanks for the new version! :)

 

can i ask why your texas only compile is so large at 114 MB?.

when i make it, it's closer to 50MB.

 

i guess it has something to do with the debug code or extra commands in the compile commandline.

just for a laugh, i opened the mame.exe in a hex editor and cut the last 45MB from it (114-70MB).

and to my surprise, it still works with all texas games tested.

the debug menu still opens with [-debug], but probably doesn't have all the features now.

 

code cut from here...

446:6090 > search text "debug_aranges"

 

PLEASE can you do a test compile without the debugger and extra stuff, and see how small you can make the mame.exe.

thanks!

[+mizapf]

Which build do you mean? (Linux, Windows, RaspberryOS, macOS)

 

The biggest builds are the macOS variants, but they are full builds with all drivers, not only TI-99. And I don't build them, due to the lack of a Mac, just copied them from the mentioned SDLMAME address.

 

As my Firefox shows, the other builds in the WHTech folders are 56 MiB, 48 MiB, and 68 MiB.

 

 

[.mad.]

sorry, i didn't realise it wasn't your build.

i guess you can't do anything about it.

 

i was talking about this ti99 only "mame0254b_ti99_win64bit.zip"

yes the ZIP file is 68MB, but when unpacked the mame.exe is 117MB.

[+9640News]

@mizap

 

Changing the subject a bit.  Back a few months ago, you indicated TIPI mouse support (not the QEMU connection), was not implemented.  Will this be implemented anytime soon?  Just inquiring.......

 

Beery

[+mizapf]

2 hours ago, .mad. said:

sorry, i didn't realise it wasn't your build.

i guess you can't do anything about it.

 

i was talking about this ti99 only "mame0254b_ti99_win64bit.zip"

yes the ZIP file is 68MB, but when unpacked the mame.exe is 117MB.

OK, you were referring to the exe file; I just looked at the archive files.

 

I built the Windows version, yes. I included the drivers ti99_4x, ti99_4p, ti99_8, geneve, ti99_2, evmbug, cortex, cc40, tm990189, and tutor. Also, I let the tools build (TOOLS=1), i.e. chdman, imgtool etc. The tools usually pull a lot of code into the exe.

 

[+mizapf]

@9640NewsIt is on my list, but not planned for a special time. My next-in-line for MAME would be the block DMA of SCSI, as soon as time allows. I'm currently spending some time on TIImageTool 3.

[+9640News]

11 minutes ago, mizapf said:

@9640NewsIt is on my list, but not planned for a special time. My next-in-line for MAME would be the block DMA of SCSI, as soon as time allows. I'm currently spending some time on TIImageTool 3.

OK, thanks for the feedback.

[.mad.]

Here is a lite version of the MAME exe 0.254 ti99 (114MB - 55MB)

https://mega.nz/file/Pgw0gYCA#-DRY5oJCsCej-oyMxNDfNoUHAlEf6Q7RFmkN7MLKVKE

 

please can you test it.

i can't see anything wrong with it, but have not tried every option.

 

if you give it the OK, i will include it with the next GB release. 😉

[+mizapf]

So how did you build it?

 

Please note that "debug build" and "-debug" are NOT the same things.

A debug build contains symbols in the executable and has no optimizations so that we can effectively debug MAME itself. This bloats the executable code and also leads to heavy performance penalties. I'd guess it is 3-4 times slower than the standard build (which means it can easily drop below the 100% emulation time, making sound choppy and the emulation slower than the real thing).

 

The debugger that is launched with -debug is a build-in feature of MAME and is intended to debug the emulated system (not MAME itself). There is no build option to exclude that feature from building as far as I know.

 

Clipping off parts of the executable does not sound good to me. I'd not recommend it, even if the thing is still running.

 

After all, I also thought about dropping those TI-99 specific MAME builds and returning to the full build, because the effect of excluding the other drivers is really minimal. It should not really bother your computer if there is a lot of executable code that is never run; it is just static memory never paged into execution. So it should not really take away available runtime memory or otherwise decelerate the execution. The only advantage is that building is much faster, and the executable is loaded more quickly.

[ralphb]

I've recently noticed that the copy protection of the MG Explorer disk is too good for MAME.  The disk contents certainly look a little off:

 

cassiopeia ~ > xdm99.py explorer.dsk 
 EXPLORER :    145 used  30 free   43 KB  1S/1D 40T  9 S/T
----------------------------------------------------------------------------
 MGEX22Z45    2  PROGRAM        256 B           P                        
BASAMPLE      2  PROGRAM        236 B           P                        
DIAGNOSE     60  DIS/FIX 80   15104 B  177 recs P                        
EXP           6  DIS/FIX 80    1280 B   15 recs P                        
LOAD          2  PROGRAM         86 B                                    
MEMEX-TEST    9  DIS/FIX 80    2048 B   24 recs P                        
MEXP          6  DIS/FIX 80    1280 B   13 recs P                        
MXBEXP       23  DIS/FIX 80    5632 B   65 recs P                        
PREASSEM      2  DIS/FIX 80     256 B    3 recs P                        
XBEXP        24  DIS/FIX 80    5888 B   68 recs P                        
XBMEMFULL     2  DIS/VAR 163    153 B    2 recs P                        
XBMEMOFF      3  PROGRAM        304 B                                    
XBSAMPLE      2  PROGRAM        252 B           P                        
Warning: Sector count does not match disk geometry
Warning: Sector count is not multiple of 8

 

MAME 237 reports this error (either controller):

 

Fatal error: Device 5.25" double density floppy drive load failed: Unable to identify the image format

 

Not sure if it's worth analyzing, as "unprotected" versions are available.  Just wanted to let you know that the emulation is not 100%.

explorer.dsk

[+mizapf]

DSK images cannot be 100%, as they lack all data outside of sector contents. In that case you have to use HFE. I know that I got it working once; you can use the search function in the forum.

 

By the way, what is a 44 KiB DSK file? This is certainly not supported, just because of its size. It must be 90K, 180K, 360K, or 720K.

 

Edit: This was the thread:

 

Interestingly, I cannot load Diskassembler, but the other ones (Advanced Diagnostics, Explorer). I found the Diskassembler image in a folder "notwork", even though that video proves the opposite. 5 years is quite enough time to start forgetting things, as it seems.

[+mizapf]

I just found out that on March 27 this year, someone removed the FM resync from the WD emulation, which was my fix for the Diskassembler in 2018. 😞

I left a note on Github and asked for discussion.

 

Edit: In fact, it seems to be broken for a longer time, but no one noticed. DiskAssembler did load, that's for sure.

[+mizapf]

Even worse, another MAME dev removed the write support in the HFE format support a year ago. https://github.com/mamedev/mame/commit/66c0222c690fb6fb0e093fad37d42a68fea66a2f

 

Not even noticing me, guess I'm expected to check github and the forums all the time. 😞

 

Edit: OK, I got the info that the save code was stateful (i.e. depended on the previous load) which is a bad idea. Well, then, I'm going to rewrite that. But just clipping out without telling the file owner (and I'm clearly mentioned in the header)...

[+arcadeshopper]

fyi 

cat error.log 
[:ioport:peb] AMA/B/C address prefix set to 70000
[:ioport:peb] Slot 2 = :ioport:peb:slot2:samsmem
[:ioport:peb] Slot 3 = :ioport:peb:slot3:speech
[:ioport:peb] Slot 4 = :ioport:peb:slot4:tirs232
[:ioport:peb] Slot 5 = :ioport:peb:slot5:tipi
[:ioport:peb] Slot 6 = empty
[:ioport:peb] Slot 7 = empty
[:ioport:peb] Slot 8 = :ioport:peb:slot8:hfdc
Soft reset
[:ioport:peb:slot4:tirs232:uart0] resetting UART
[:ioport:peb:slot4:tirs232] (RS232/1) No serial output attached
[:ioport:peb:slot4:tirs232:uart1] resetting UART
[:ioport:peb:slot4:tirs232] (RS232/2) No serial output attached
[:ioport:peb:slot4:tirs232] (RS232/1) No serial output attached
[:ioport:peb:slot4:tirs232] (RS232/2) No serial output attached
[:ioport:peb:slot5:tipi] Trying to connect to Raspberry Pi at tipipz2:9901/tipi
[:ioport:peb:slot8:hfdc] FD connector 1 with 5.25" double density floppy drive
[:ioport:peb:slot8:hfdc] FD connector 2 with 5.25" double density floppy drive
[:ioport:peb:slot8:hfdc] FD connector 3 has no floppy attached
[:ioport:peb:slot8:hfdc] FD connector 4 has no floppy attached
[:ioport:peb:slot8:hfdc] HD connector 1 has no drive attached
[:ioport:peb:slot8:hfdc] HD connector 2 has no drive attached
[:ioport:peb:slot8:hfdc] HD connector 3 has no drive attached
[:maincpu] ** RESET triggered
[:ioport:peb:slot5:tipi] Connection established
[:ioport:peb:slot5:tipi] TIPI server offers ASYNC communication
[:ioport:peb:slot5:tipi] Closing connection
[:ioport:peb:slot5:tipi] Trying to connect to Raspberry Pi at tipipz2:9901/tipi
[:ioport:peb:slot5:tipi] Connection established
[:ioport:peb:slot5:tipi] TIPI server offers ASYNC communication
[:ioport:peb:slot4:tirs232:uart0] resetting UART
[:ioport:peb:slot4:tirs232] (RS232/1) No serial output attached
[:ioport:peb:slot4:tirs232:uart1] resetting UART
[:ioport:peb:slot4:tirs232] (RS232/2) No serial output attached
[:ioport:peb:slot5:tipi] Got error: End of file (2)
[:ioport:peb:slot5:tipi] Stopping TIPI
 

This is latest code checked out from git and compiled yesterday. Sill getting the eof and no connection

Distributor ID:    Ubuntu
Description:    Ubuntu 22.04.2 LTS
Release:    22.04
Codename:    jammy

 

tipi is a actual pi set up and works fine in js99er 

 

Greg

[+mizapf]

This is the already known EOF error (next to last line); I don't have a solution yet, but I will check as soon as possible.

[+9640News]

On 6/10/2022 at 6:30 PM, jedimatt42 said:

Did you do this part: emulation installation · jedimatt42/tipi Wiki (github.com)

 

specifically the following part:

 

 

@mizapf

 

I went looking for this information on ninerpedia thinking I might find the information, but ended up finding this note in this topic pointing back to Matt's comments.  As a suggestion, you might want to consider adding these notes or at least a reference to @jedimatt42's site, this needs to be done.  I now have Version 3.xx working under MAME.

Beery

[+mizapf]

I don't have a section about TIPI on Ninermame yet, but I'll keep a note.

 

By the way, I set up Ninermame (www.ninermame.org) to keep everything related to MAME a while ago, so you better watch that site for future updates on that topic, not Ninerpedia.

[jenorton]

Hi:

 

Sorry if this isn't the best place for this, but, there are quite a few topics.

 

Anyway, I want to have 3 or 4 floppies, but, I can't seem to find a way to do it from the command-line.  All I find are -flop1 and -flop2. -flop3 doesn't seem to be recognized.

 

Thanks!