Lichess Drawing limits

I wonder if LT can do anything about local persistence of drawing objects per board.

The server limit for studies, is such that, for example, one can't draw the full opposition field classes by 4 colors on all 64 squares, and have that persist upon a reload. However, the 64 squares circles can be drawn, only, not propagated.

I am not asking about changing the server limits (I must have asked already some years ago, no changes there).

But, in the same way that local PGN user grown move subtrees are made local machine persistent, could such persistence apply also to board drawings.

I can link to a public study of mine, where the server omitted about 1/3 of the objects.. and paste a snapshot of how I could have it in my solo experience of the study.. I do use empty board for chess study and use visual summaries of many things, that have some mathematical-chessy compact invariants or things to keep in mind about many other chapters.

I may use a board chapters to make a table of content in visual drawings about all unique cases that would each be having a treatment or homework in other chapters.. Or other homework on chess vision and mobility in multipath problems, and that lichess limit, gets in the way.... it might be a minuscule user case in the user base, but there are plenty such that get high profile, and attention. I am of those doing more study than performance play activities on lichess. This is not to say that others don't have a proportion of their activity in that way too. They might be content and not spend energy like me into asking with full motivation development, and imagination of how else could things be.

So for visual sake! is it possible to keep that user work at least locally (and perhaps even upon chapter PGN export, bypass the server loss-full filter (not loss-less).

Science-fiction babbling:
BTW: if LT could completely bypass all PGN steps from the server and having import export made locally, some useful things could be had, that are now tethered to the antics of study PGN handling.

LT has made a lot of useful things with user created sub-trees and managing them through PGN buffers, it could extend to the PGN export and import... a parallel local PGN extension.. PGN are portable through lichess text boxes I think. for manual sharing.. but studies can be used for solo study.. as workbooks. How remote from feasible is such idea.(this could be a separate discussion, not the priority in this thread, just that it is under the problem I am asking a band-aid work around).

Opposition field public study
https://lichess.org/study/oGkqNpmF/pbi0OZrM/

Server limit consequence, curation to keep the idea but not complete:
https://i.postimg.cc/nr0gchyw/screenshot-2024-01-29-at-15-24-17.png

Local session intent:
https://i.postimg.cc/W10FktMS/screenshot-2023-11-02-at-22-52-30.jpg

from, my comments
https://lichess.org/forum/community-blog-discussions/ublog-G4KvLwRm#2
about
https://lichess.org/@/habbis/blog/the-only-method-you-need-to-know-for-figuring-out-the-opposition/G4KvLwRm

Some chess learning stuff from lichess learning features. For those curious about the example, chess topic.
https://lichess.org/practice/pawn-endgames/opposition/A4ujYOer/eNlmabX1

I had no idea there are limits to the number of circles. I think this is just a bug. The shapes are "persisted" in the comments, with patterns that you don't see in Lichess, but you get when you save the PGN. I will look it up. It's probably impossible to do anything about it on the client side, though.

Seems to be a hard limit of 32 shapes.

32 shapes: server limit.

Are you saying that the internal data structure of the server, has a limit and that it is coincidentally, how the local client PGN export is also limiting the lichess drawing syntax on top of the PGN standard?

It could be that the internal server data structure is having the export PGN equivalent or is itself the result of PGN export mechanism?

In non-study analysis pages, lichess has recently added client side user create move trees persistence, which LT seem to be using well, thanks to you.

But studies is where the drawing communication or feature is being recordable.

And it is where, from my user experience based imagination, the new code of other analysis pages seem to have never been deployed. The ancestral study internal data structure being some sort of backbone, that maybe lichess is not wanting to tickle, their old implementation of the PGN standard.

The PGN standard, I have gathered, has had many dialects used on top, such as the drawing here, chessbase also has its own, possibly copied in other tools chewing PGN in or out. A bit like internet explorer and web programming making its own html dialects or JavaScript extensions (my tourist web tech understanding of the past, perhaps the recent web standard was not just to reduce user control of what they see, after all, the tendency with browser API for extensions, ensuring brain FOV compliance).

So, there is no client-side access to commenting? You need the pages themselves to be able to accept the same features that the other analysis pages have. Well. I asked, and now you might be having en eye if a solution were to exist. thanks for looking.

The funny thing is that if you create a new study and you add a PGN with all the shapes, it saves them correctly
Once you add anything else it limits it to 32. Since there is no API to replace a chapter with a PGN, I can't really do anything. It would have been terribly cumbersome anyway.

And if you raise an issue on Lichess, they will tell you you don't need more than 32 shapes...

thanks. Yes, this is how the technology communication rationales have evolved. I will consent to this logic.
Who needs 64... just go an play your game already. Wait, you want what? to use the board to explain some things or compress information/insights in visual form. Just go and play a game, already!

Now you are aware.

PS: how did you get to have more shapres? following the same syntax as exported ones with drawings?
you test how many? And to make sure: they do upload locally?.. that is great. I could make 2 chapters with 32 each complementary, and merge them, then upload them or share the merged PGN via text boxes.... If I understood correctly.

Well, when I saved the 32 shape chapter PGN I got something like this:

[%csl Ga8,Gb8,...,Gh5]
So I just manually added the others down to Gh1. %csl is a circle. For arrows there are two coordinates and the macro looks like this: [%cal Gd5f5]

The simplest way for you is to just draw the shapes, copy the PGN using the LiChess Tools "Copy branch as PGN" or /copypgn command, then paste it into a new chapter.

It appears you can draw more shapes in other positions and the old position shapes will remain unaffected even if more than 32. But if you change the shapes in the position where there are more than 32, you end up with 32.

Recapping:

draw the shapes,
copy the PGN (...)
then paste it into a new chapter.

So good news. improving on my clunky 2 chapters to keep one of sorts. I can directly propagate from chapter to chapter... and share the text buffer (or ideally file as lichess export does... hmm).

That last line is using import PGN via the study create chapter mechanism (?). To make this portable to incoming other readers.

Now I have a difficulty computing the last procedural statement into a static (yet, mechanistic?) picture...
(I am slow that way). Let me explicitly process through verbal stream (LOL).

It appears you can draw more shapes in other positions and the old position shapes will remain unaffected even if more than 32. But if you change the shapes in the position where there are more than 32, you end up with 32.

answered below.... faster than my abilities...

So if you have a chapter with multiple positions (moves) and you draw 10000 shapes on one position, then you export and import the PGN, then you will have 10000 shapes on that position. If you move to the next position and you draw something, there will still be 10000 shapes on the previous position. But if you move back and you draw the 10001st shape, then that will reset the shapes to 32.

In other words, there is a limit of 32 shapes per position and changing shapes on any other position will not affect others.