PDA

View Full Version : software for EASY book editing/study



gambitron
27-04-2005, 04:23 PM
i've posted this topic in THREE different chess forums INCLUDING this one and despite having well over 100 views, have yet to see a SINGLE peep out of anyone. the BIGGEST thing holding me back in chess is a tool that will let me EASILY edit my books (eg. add out of book replies) and then systematically study them. am i REALLY the only person in the world who can't believe that NO-ONE has created a program that's easier to study books with than creating trees (by hand!) on paper and then studying them line by line (wasting tons of time resetting pieces and having to TOTALLY scrap your page of theory EVERY time you need to add new theory).

am i REALLY the only one who can picture an intuitive easy to use program that takes ALL of the drudgery out of studying books and dealing with the maze of transpositions that is open ended and allows you to both SEE and jump anywhere in your tree is a NECESSITY?!!

please read my idea on the software (i'd design myself if i were a hacker) and add your two cents to it. i'll get NOWHERE with the company that's FINALLY taken me seriously if no-one else will add their two cents.

i'd give ANYONE who'd create this program $200 right now!!!! it seems so obvious and NECESSARY yet i can't get ANYONE ANYWHERE to add a thumbs up to it or ANY programmer to say "eureka!" this is perfect! for what it's worth... my constant nagging to bookup is what finally got "systematic book study" to be added, but the program works the exact OPPOSITE of what's intuitive to me.

hello everyone.

i've been trying and trying to get software creators such as chessbase and bookup to create a "systematic book study" program for years. in fact, my repeated nags are probably what got mike at bookup to add the feature, but once he did, i tried bookup and spent a WEEK creating a "scandinavian portugese" mainline only book to study from only to have it totally ruined when i opened up my own games. there were many things about bookup that seemed couterintuitive and clunky so i tried passing my ideas to conveka (chess assistant) and they have showed interest in possibly incorporating my ideas into a new feature and asked if i'd discussed my ideas anywhere else, so i'm trying to get some other people's comments and ideas to bring to them.

please feel free to mention ANY dream feature that you'd like to see in a book study tool. i'm not talking about a database utility mind you, just a tool that will let you create "mainlines to systematically study" that can read databases for new moves by the OPPOSITE side of whatever line you're studying.

these are my ideas for the "ultimate book tool"

1. i want to study whatever book or chapter (folder) that i create systematically. i want to study EVERY move in a book no matter how many moves into a tree i get. if i have branches at move 20, i want to study those. i also want to study the same moves with different move orders wherever transpositions occur. there should be several different ways to study games...
....one game at a time stepping through the entire tree one branch at a time
....totally random
....weighted by move strength
....weighted by frequency of over the board encounters BY ME, not grandmasters and so on.

2. there should be a filter where a book never repeats until EVERY line in a book or chapter has been studied. if i keep making the wrong move in one line, the software should have a weighting system that keeps drilling me on it until i get it right.

3. there should be a hint mode and a way to manually repeat. bookup locks up as soon as you don't know a reply which totally undoes learning. you lose your place having to back up and then face whatever different line it choses and can't automatically get back to the spot you messed up at. if a user gets a line right, he should not see it again until the rest of the book has been completed successfully. bookup wastes too much time repeating the same lines over and over while totally neglecting others.

4. i want a tool that lets me see my ENTIRE tree at once so that i can easily jump back and forth when, say, adding new variations out of an actual book and so that i don't get lost. bookup only shows one line at a time so you never know where you are in the "maze of theory" while chessbase buries every new line inside the main line in a confusing jumble of variations within variations.

THIS is what a 'book navigation' text window should look like in my opinion.

Scandinavian Defense Book
1. e4 d5
....2.e5 Bf5
........3.d4 e6
........3.Nf3 e6
....2.exd5 Nf6
........3.d4 Bg4
............4.f3 Bf5
...............5.Bb5+ Nbd7
...............5.c4 e6
............4.Bb5+ Nbd7
............4.Be2 Bxe2
........3.c4 c6
........3.Nc3 Nxd5
and so on, but without all of the dots which are only there in case this forum forces every new line flush left like others do. to see how this style of book creation works in action, visit nicbase online at:
http://www.newinchess.com/NICBase/
then make whatever moves you want on the board and then hit the "key" icon to see the possible next move replies.

note: i was ALREADY notating mainlines from books by HAND this way for manual study before i found nicbase online which uses the very same INTUITIVE concept. i've created entirely new books of transpositional lines that i have no theory for there in just minutes! it's so easy... just point at a move (or make it on the board) and then a list of replies in the database pops up for you to pick which one you want to make YOUR main line. the only thing i'd do different is make all of the replies you DON'T chose disappear to keep things simpler and uncluttered, but there could be a mode where the moves you didn't chose can be displayed too if you want to create a "second choice" line to study against.

5. of course i want to chose which "windows" are displayed, where they're placed and how big the are the way you can edit fritz's layout. bookup doesn't let you use a fullscreen board which is beyond annoying. the chubby bishop looking pawns that bookup uses are terrible. pawns should look like pawns as in USCF pieces.

6. i want to use my books to study my own games to see who is the first player to leave book. if i do, i'd like to study all the moves from that branch until i get it right. by study... i mean actually play gainst the book. if my opponent introduces a new move, i'd like to create a new branch and search for OTHER lines that transpose to the new position. i really want to incorporate "out of book" transpositions into my book and use an engine to pick replies to create a main line for whatever opening i'm playing.

7. i don't want to sort databases by date, or player etc. i want to sort them by tree. any new out of book replies should be added to the proper spot in the tree. eg. (looking at the scandinavian tree in comment #4) when someone plays 2.Nf3, the move should be added before, after or in between moves 2.e5 and 2.exd5 depending on how i want to organize the tree. i'd prefer to have the program automatically rank variations by whatever moves are most common in my amateur (rarely grandmaster book following) games. there should be flexibility in sorting moves based on what criteria the user choses.

8. books (or chapters) would be organized by color so you could have one book for the moves you'd play in in a given variation as white and an entirely DIFFERENT book if you also play that variation as black. then, the "bad replies" for one side won't pollute the other side's mainline as happens in bookup which simply merges EVERYTHING into one huge book. of course you could create as many books as you'd like if you want to play more than one varation in a given position. basically eack book and chapter that you create is like a paperback book, only you can add new moves and transpositions to it and chose your new replies with a chess engine all in one interface. opening any new database and browsing it will have NO effect on your mainline books unless you either "OK" a database move in "book building mode" or chose an "open mode" which is more like a database as i understand databases.

9. there should be a flexible color coding feature that lets you chose different colors for theory in your trees so that you know where moves are coming from. eg. one author's moves could be green, another's red, any new theory generated by a chess engine blue and your own games dark grey. then you'd know who suggested (or played) any move in your book, and could add (or search for) notations for that move to add to that book. i'd chose to have a seperate notation window to keep things uncluttered in your tree window.

basically, all i want is to create opening books, skim databases for new theory to add to books and to use a tool that automates the study of trees. the way everyone's clunky database etc. programs work, it's actually EASIER for me to transcribe an indented tree (like in comment #4) and then reset chess pieces from move 1 line by line down my tree! that's annoying! someone should have created a tool that totally automates piece resetting and allow me to click on ANY point in a tree and study from there or add to the theory. that would speed study up 50 times. nobody has done it yet. if i were a hacker. i'd create a program myself.

these ideas are just crude rough drafts. there should be flexibility in a variety of "modes" that let the individual chose the features they want and work their own way, but this is the way that i'd like to to work creating "locked" books quickly and studying them quickly without having to reset pieces.

all comments are appreciated as i'd really like to get an intuitive and flexible book creation tool so i can finally study openings on a regular basis. thanks in advance to anyone who's taken the time to just read this far and extra thanks to everyone that adds their two cents to my ideas. maybe you too will have the tool of your dreams i hope.

Spiny Norman
27-04-2005, 06:40 PM
I think one of the impediments might be the sheer complexity of what you're asking. It sounds simple, but in reality the amount of data storage and computational power needed might be too much.

I suppose that one might partly get around that by having some kinds of checksum that could be created per position per game. So if you had 100 games with 40 moves per game, then your database would need 100x40x40 records (40,000). That just to identify the unique positions created.

If you then went on to try to characterise the move orders needed to reach those positions and put them into a navigable tree structure, well, my brain ain't capable enough to work out the best way to program it (e.g. how to accommodate transpositions, etc - seems horrendous to my mind). I can see the value in the functionality you propose, but it would be a major programming effort to create that functionality and encapsulate it in an easy-to-use human interface.

My best guess is that you're probably looking at somewhere between 6 - 24 months of programming/debugging/testing effort. At best (6mths) that's around a $30,000 investment. At worst (24mths) you're talking upwards of $100,000 ... :eek:

Maybe I underestimate today's programming tools? :hmm:

Davidflude
02-05-2005, 11:31 AM
I have over the last few weeks been analysing something very new.

The way that I have been doing it is

1) set up a new database in Chessbase in Chessbase format.

2) set up the start position from which I wish to start analysing as a game on this database.

3) import this game into Shredder and start analysing. Once I know what the various variations are I save each as different games back into Chessbase.

4) I then generate a Chessbase tree structure for this database.

5) I can then step through this tree effortlessly forwards and backwards.

This is not perfect. It does however allow me to analyse new lines on an on-going basis and add new material.

Having separate games is much easier to manipulate than combining all the lines into a separate game. Of course from time to time I will combine games and do a repertoire print (for those unfamiliar with this it looks like the prints in Encyclopedia or MCO or Nunn's Chess Openings).

I gather that Chessbase 9 has better features than Chessbase 8. (Read MIG's
colunms on Chess Cafe or Email him with your Queries.