PicoChess v3.3 Bookworm Images [12 Mar 2025]

[RandyR]

Hi Everyone,

Hopefully I have solved the issues with the previous images when using the retro machines. See the previous thread for problem details. Here are the links. Let me know if you still have issues.

Enjoy.

Randy


64-bit Bookworm Lite (4.4 GB)

64-bit Bookworm Desktop (5.6 GB)

[Francois Vannier]

Hi Randy,

Thanks.

Good news :

> Working as good as before + tutor OK : AC Prodigy, Etek AVRMax, FIDChesster, MMIV, NovagRobot...

Bad news :

> Existing retro Novag Sapphire2 regressed : OK before, now stuck at engine setup. Artwork On or Off does not help

> Trying the new engines : I just tested  the 2 Novag & the Saitek: none of them is working. They're stuck at engine setup

If the only issues were the new engines not working, we'd cope with that, I guess. The most annoying issue is previous engine(s) no longer working like Novag Sapphire2. I stopped the tests due to this issue.

Kr,

Francois

[Randy Reade]

Hi Francois,

It looks like my fix didn't work.

Henri sent me a PM and I'm looking at it now.

Sorry for the frustration.

Randy

[Samtec]

for me, it worked nice. thanks for your time and work. 

[RandyR]

Thanks for the feedback. It's a fairly intermittent issue. I can switch retro engines 50 times with no issue then try again later and it will hang at engine setup using the same retro machine selection. It's now appears related to updating the RPiOS files. Using an up-to date OS with the picochess folder from an old stable image has the same issue. That's unfortunate as it will be hard to maintain going forward. Perhaps PicoChess 4 will be a solution. :^)

I'll try to put together an image that has updated PicoChess files with the older OS and see how that goes.

Stay tuned...

[Johan Sjöblom]

Hard work and I wish you the best of luck. I can test the new image mame engine setup with picochess 4 when Im back home from travelling end of next week. Maybe I could find some additional info on what is going on?, but I have a feeling that you might find some potential root cause soon.

--
You received this message because you are subscribed to the Google Groups "PicoChess" group.
To unsubscribe from this group and stop receiving emails from it, send an email to picochess+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/picochess/82f194f8-1cea-4dba-a71c-6157ac1378ffn%40googlegroups.com.

[Johan Sjöblom]

Just a thought, if the problem seems to be that mame is always working with new/updated  Pi OS,... maybe it (mame/mess) should be compiled from source on a new OS setup? It's not likely that a new OS setup would disable some function in the old chess library, but that day may come, so I will continue to test and stabilise the python picochess 4.0... 

BR, Johan

[RandyR]

I usually cross-compile MAME on my PC to save compile time but the one on my recent images was compiled on an up-to-date RPi4b. There's a similar issue when using x86 which, unfortunately, sent me down the wrong path when I thought it was the same as on the Pi (the symptoms were the same - hangs at 'Engine Setup'). But the x86 issue is MAME keeping a thread open after PicoChess sends the quit command and this interfered with running a new retro machine. On the Pi, it was a loss of communication between PicoChess and MAME (via the chessengine plugin) when starting the new retro machine.

I was unable to determine what was interfering (intermittently) on the Pi. And on the x86 PC I don't know why there is a hanging thread after sending the quit command (or closing MAME with the mouse) but if you don't use the chessengine plugin there's no threads leftover after closing.

Randy

[Johan Sjöblom]

Good info. I am still on the learning path, and here on the shore of Madeira my brain is obviously not in full use :-)

I do however remember that the existing 3.x picochess has a quite "aggressive" quit function in uci/UciEngine class. If I remember correctly it first sends quit uci message and then kills the mame. Maybe there should be a "sleep wait" for 0,5-1sec before any kill command is used? If the kill command is processed faster than the quit message the communication could hang? Just a guess on my side.

Sending a quit uci message should in most cases be enough?

--

You received this message because you are subscribed to the Google Groups "PicoChess" group.
To unsubscribe from this group and stop receiving emails from it, send an email to picochess+...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/picochess/8ceaa5b1-ffd0-4ec6-be6c-ec061658be53n%40googlegroups.com.

[Randy Reade]

Hi Johan,

The kill command for mess was needed to prevent a long shutdown delay when shutting down the Pi via the PicoChess menu, if I recall correctly. But, I'm sure you'll get to testing that portion of your code eventually. But, in the meantime you should put PicoChess out of your mind and enjoy your southern getaway. 😁

To view this discussion visit https://groups.google.com/d/msgid/picochess/CAEk%2BOSTQzVbv1RdzOQRZ05KL1d4HdURvzbfLjtdyaGLXh6663g%40mail.gmail.com.