There are two big changes for COW coming this year.
First is gamifying the training wizard. Watch a graph of your progress memorizing your new repertoire.
Second is Epic analysis. Have Stockfish comb through your opening to show you improvements or moves that are playable but missing.
This will require a change in the database format to include bookmarks. When Epic analysis finds a bunch of lines where it has added new candidate moves, it will need to make a list of the move orders that reached the positions with those new moves. That list will be bookmarks. Bookmarks can also be used to store any line so that you can jump right to it. Updating the database forma means simultaneously updating all four apps (Windows, Mac, iOS, Android) so with me luck.
The Epic analysis feature sounds useful, but how will you handle performance and storage when Stockfish generates many candidate lines and bookmarks? It might help to add limits or filters so users don’t get overwhelmed.
xenirox wrote: Sat Apr 04, 2026 7:47 am
The Epic analysis feature sounds useful, but how will you handle performance and storage when Stockfish generates many candidate lines and bookmarks? It might help to add limits or filters so users don’t get overwhelmed.
Yes, similar to the novelty finder there will be a user set limit on how many new bookmarks the Epic analyzer will find.
For accuracy and simplicity, Epic (bookmark) is to add only The Best (Stockfish 18) candidate move at every ply (if not already present) starting at a selected End Node and working back to the first move after the Beginning Chess Position (any 1.xn like e4). Thus a subsequent Backsolve is actually correct. Will yield an ebook that appears to have been created from 'scratch' using a GM source for each ply, but at the speed of Mike's programming and your CPU.
RonT wrote: Mon Apr 20, 2026 5:03 pm
For accuracy and simplicity, Epic (bookmark) is to add only The Best (Stockfish 18) candidate move at every ply (if not already present) starting at a selected End Node and working back to the first move after the Beginning Chess Position (any 1.xn like e4). Thus a subsequent Backsolve is actually correct. Will yield an ebook that appears to have been created from 'scratch' using a GM source for each ply, but at the speed of Mike's programming and your CPU.
Yes, the Epic part is about adding new moves, the most important moves being the best ones from Stockfish. Other possibilities are adding the less-than-best moves that opponents may consider (or that you may consider) playing.
Automatically adding those moves is more interesting if you can then see where they were added. That's where bookmarks come in.
