- Blind mode tutorial
lichess.org
Donate

v2.4.55 and plans for the future. Need help!!

#1

Here we go, new features!

Possible new features

  • tap&drag shape drawing on mobile - this was implemented, tried and removed because it didn't work. I will fix it when I have the time. I have to test it directly on a mobile, since it involves multifinger gestures. When/If this will work, one will be able to tap with a finger on the board then draw with a second finger (which is how I understand it will work on the mobile app) - it will only draw green, though, so I am not sure how useful it is, and how is this gesture any different from a zoom in? I think I need to understand how it works on the mobile app. Can you guys tell me if you find this a useful feature?

Other stuff

I have exhausted the quick wins, the features I can just whip up in an hour and get on with my life. In a week I am going out of town for a week, so probably development is going to be slow. I looked into performance a bit, but I think for a significant improvement of performance you will have to wait for a rewrite (version 3).

As for the companion app, I am having writer block, to be honest. I wanted to write it in .NET MAUI, because it was multiplaform, but then the first thing I wanted to do - the login - I had to implement different code for Windows and for non-Windows. That's multiplatfuss, maybe! I hesitate to do it in Tauri, like the Local Lichess Engine and I don't want to do it for Windows only, which would be the fastest thing and I would do it in WPF. Also the system service thing I believe it's specific for each OS anyway. Three different apps doesn't feel good.

Maybe you guys can suggest something?

The companion app will be important for a multitude of reasons:

  1. it will set up the external engine stuff as a service - so you start the machine, you don't log in, and it provides a running set of local engines, like StockFish, Leela, Dragon, Maia, whatever you want, that you can access from any device once you log in
  2. I want to make it use the engines directly if you're on the same machine - it's really slow to start with the API - or find some optimization so it doesn't have to take seconds to start analysing when you move to another position
  3. it should pause the engine process and restart it when you stop/restart the engine (or it stops because of reaching the engine limit) - right now what happens is that it goes to level 27 or whatever and it stops, giving you the option to go deeper (the little + sign). When you click the + sign, it starts the engine again, with a limit of 99. So it goes again from 0 to 27 for no good reason.
  4. it would provide access to external resources that browsers refuse to go to because of "security" or CORS or whatever. A lot of APIs will become available to the extension.
  5. it will provide access to the file system and/or local databases, allowing optimal access to huge files, fast caches, full control of data. That means curating your own local database of moves, reading and writing SCIP and Chessbase files, etc.
  6. it will give you the opportunity to run really long taking processes in the background - which on the browser it's always at the whims of whatever browser you use and how many pages you open and when someone closes it.

That's just from the top of my head. It does require a lot of commitment, though. Let me know what you think. I really need to get around this blockage.

Here we go, new features! - Retry on Space flow option in Interactive Lessons - when enabled, if you have selected Sequential or Spaced Repetition and you have reached the end of a variation, then pressing Space will either 1) retry if there are more variations to play 2) move to the next chapter if not (https://siderite.dev/blog/lichess-tools---user-manual/#extendedInteractiveLessonFlow) - read position command - new /readposition command will read the position piece+position by piece+position so you can set up a board. Space pauses the read, Esc or moving from the position abandons it. (https://siderite.dev/blog/lichess-tools---user-manual/#commands) - Time controls theme - new theme will show time controls (in the format TIM+INC) on the top-right corner of minigame popups, as the ones you see when you hover a playing user or a game link (https://siderite.dev/blog/lichess-tools---user-manual/#themes) - FEN based wiki tooltip on study opening link - the Wiki pages based on FEN feature makes sure you see the Wikibooks page in no matter which moves you made to get there. But it only works in Analysis Board and not study. Now, if the current position does have a Wiki entry, it will add it as a tooltip of the Explorer opening name. (https://siderite.dev/blog/lichess-tools---user-manual/#wikiFen) Possible new features - tap&drag shape drawing on mobile - this was implemented, tried and removed because it didn't work. I will fix it when I have the time. I have to test it directly on a mobile, since it involves multifinger gestures. When/If this will work, one will be able to tap with a finger on the board then draw with a second finger (which is how I understand it will work on the mobile app) - it will only draw green, though, so I am not sure how useful it is, and how is this gesture any different from a zoom in? I think I need to understand how it works on the mobile app. Can you guys tell me if you find this a useful feature? Other stuff I have exhausted the quick wins, the features I can just whip up in an hour and get on with my life. In a week I am going out of town for a week, so probably development is going to be slow. I looked into performance a bit, but I think for a significant improvement of performance you will have to wait for a rewrite (version 3). As for the companion app, I am having writer block, to be honest. I wanted to write it in .NET MAUI, because it was multiplaform, but then the first thing I wanted to do - the login - I had to implement different code for Windows and for non-Windows. That's multiplatfuss, maybe! I hesitate to do it in Tauri, like the Local Lichess Engine and I don't want to do it for Windows only, which would be the fastest thing and I would do it in WPF. Also the system service thing I believe it's specific for each OS anyway. Three different apps doesn't feel good. Maybe you guys can suggest something? The companion app will be important for a multitude of reasons: 1. it will set up the external engine stuff as a service - so you start the machine, you don't log in, and it provides a running set of local engines, like StockFish, Leela, Dragon, Maia, whatever you want, that you can access from any device once you log in 2. I want to make it use the engines directly if you're on the same machine - it's really slow to start with the API - or find some optimization so it doesn't have to take seconds to start analysing when you move to another position 3. it should pause the engine process and restart it when you stop/restart the engine (or it stops because of reaching the engine limit) - right now what happens is that it goes to level 27 or whatever and it stops, giving you the option to go deeper (the little + sign). When you click the + sign, it starts the engine again, with a limit of 99. So it goes again from 0 to 27 for no good reason. 4. it would provide access to external resources that browsers refuse to go to because of "security" or CORS or whatever. A lot of APIs will become available to the extension. 5. it will provide access to the file system and/or local databases, allowing optimal access to huge files, fast caches, full control of data. That means curating your own local database of moves, reading and writing SCIP and Chessbase files, etc. 6. it will give you the opportunity to run really long taking processes in the background - which on the browser it's always at the whims of whatever browser you use and how many pages you open and when someone closes it. That's just from the top of my head. It does require a lot of commitment, though. Let me know what you think. I really need to get around this blockage.

This topic has been archived and can no longer be replied to.