New Features, Changes and Fixes

@Tepin That looks correct. Your king sits on top of a gold, so it moves like a gold and thus cannot move to 59.

ops you right
but still, i had to manually resign. I thought it was automatic

@Tepin You had pawn takes horse as a legal move. Your pawn was on top of your king, so it had the moves of a king.

ohithere - Thanks for the feedback. I certainly intent to focus less on variants, I agree that the main focus should go into the 'main game'... Right now my focus will go into puzzles and mobile app, and obviously bug fixes and quality of life improvements.
Puzzle storm is ready, only reason I have not enabled it yet is because we don't have enough puzzles to make sure that you won't be getting repeated puzzles on your nth run. Puzzle storm starts easy and slowly gets harder, so it needs a sufficient amount of puzzles in each difficulty. So first I need to create a lot more puzzles, and that's what I plan to do now. I understand puzzles are very important feature, regardless of puzzle storm, so I will try my best.
I'm going to look into 'current game' and probably rework it a little, I think it's not exactly bugged, but rather picky about the games it displays.
About the analysis - I find the in-browser YaneuraOu relatively fast, obviously it's not close to chess engines, but on my machine I get around 240k nodes/s (old thinkpad), which I think is fine. It spawns 32 threads, so it's multithreaded.
I don't remember talking about implementing external engine. I know lichess has done so - github.com/lichess-org/lila/pull/10867 and it would be nice feature to have as well, but I think this would be a very niche thing, that only few users would find useful, but I can be convinced of the opposite.
The websocket error is certainly a serious one, I will look into it. Although not being able to reproduce it hinders my ability to fix it. How often does it happen?

Thanks, Wanderer! Regarding the multithreaded engine, I only see one core being used on my machines (across browsers and either on macos or windows) at 706k nodes/sec, even if I have 9 CPUs enabled for evaluation. On this same machine, if I use the native yaneuraou configured with "only" 5 threads I get over 5 million nodes/sec. When you enable multiple threads in lishogi analysis do you see more than 1 core being used on your machine? That would be amazing if I'm missing something on this one.

Regarding the websocket error, yeah, I understand how difficult it could be to resolve, especially if you can't reproduce and few (or no?) other people are reporting the problem. It seems to happen perhaps 1-2 times per week, almost always on Brave Browser on either MacOS x64 or ARM on 4 different networks, with or without VPNs (I've tried a bunch of things). I have seen it once or twice in Chrome, if I'm not mistaken, but definitely not often ... however whenever the problem starts happening on Brave I switch to Chrome or Safari and I'm good there until it starts working again in Brave the next day or so. I'll switch my default browser to Chrome for now and see if maybe it stays away.

Thanks again for your thoughtful and helpful response...great news on the puzzle front!

So only notice that tourneys' trophies are too small (silver and bronze) but probably I fix that myself.

My mistake that I downloaded source SVGs in Pychess repo but Pychess has different interface than Lichess, Lidraughts or our project)

Thanks a lot Lishogi to add all this features ! Thank you for your work!

Reconnecting