Although I’m a big fan of Google products, and although you folks can certainly come up with a better color scheme than my ludicrous tries, I see one major problem with the current Google Code PGN reader.
The depth of PGN annotations supported by this version of the reader is unacceptably small. I had to edit these very brief annotations to get under the maximum…
If a PGN reader can handle this, that would be more than fine IMO!
I agree that is not one of the strengths of Google’s offering, but then again, it is free. You get what you pay for. Since you didn’t identify the source of the other example that you prefer, it’s hard to draw comparisons.
PS: Your preferred example really isn’t a pgn reader. It is a static document. Also, I do believe you can use Forsyth–Edwards Notation (FEN) for positional setup and the PGN4WEB software (same as Google?) will handle that just fine. Of course if you are trying to format your text in a forum post or blog similar to the PDF sample, that would be awkward. But its really a different media too. You have less control of the browser being used to format the content.
I meant to say that if one could present the entire contents of said static document in a given PGN reader, then that reader would be adequate.
Some context: in the USCF Issues Forum, this browser had been proposed as ideal for the USCF website.
I sent an email to ask, and Paolo Casaschi was kind enough to reply:
Yet another correction: pgn4web is hosted on code.google.com, but otherwise has no association with Google.
I agree it would be cool if one could feed a static document into a chess reader and have it be smart enough to pick off the chess notation and replay the game. That would require quite a bit of “intelligence” in the software to know where regular human language ended and the specialized chess language began (or resumed). That level of intelligence would likely be cost prohibitive (but not impossible) to develop.
PGN and FEN was a structured language created to allow a computer to properly parse the chess notation imbedded in a document and replay the game. It was created at a time when the computer processing capabilities were more limited. It set a standard, and much like html or xml it will probably remain the standard for quite some time. Unfortunately, PGN and FEN is a compromise between man and machine and isn’t as user friendly or flexible as a human brain.
I have been impressed with many of the advances made by computer science over the years. Perhaps someone will solve this problem and create a better chess reader that is more “intuitive” to use. It probably will happen, its just a question of time.
I have tried to use OCR reading software designed to create a textual FEN from an optical scan/photo of a chess position diagram from a book or magazine.
But the error rate during translation was awful.
Maybe it would help if one exact standard font for the piece icons in a chess diagram could be anointed by FIDE. Then OCR developers would have an easier or at least more plausible task.
But I presume FIDE would mess up the task, like choosing a font that is not free to the public etc.
The free (with permission) “Chess Alpha” font has been such a boon to the growing number of self-published chess authors that its inventor, Eric Bentzen (EnPassant.dk), deserves an honorary award from the USCF or similar. For instance, Ray Cheng used Chess Alpha in his popular book Practical Chess Exercises.
Into Bing.com or Google.com, search for (without the quotes) “OCR chess”.
That would be interesting. The best I’ve been able to come up with is a spreadsheet that lets one enter a diagram as one sees it (top to bottom) and it then generates a FEN string.
That is, enter:
r bqkbnr
pppp ppp
n
P
B P
N
PPPP PPP
RNBQK R
(with spaces on all empty squares) and it will spit out:
r1bqkbnr/pppp1ppp/2n5/4P3/2B1P3/5N2/PPPP1PPP/RNBQK2R w - - 0 1
All the other FEN parameters are selected in the entry screen as well. (And I messed up… it’s black’s move in the above.)
I wonder how much of a challenge it would be to take that diagram entry and translate it to a diagram (using fonts.)
(Or enter a FEN string and likewise have it generate the diagram.)
Still, OCR would be sweeter.
Instead of all empty squares, I had the option of using a dot to represent an empty square so there would be a hint as to where the squares are. From Winboard, output from “copy position to clipboard” would be pasted here to create a diagram using the Chess Case font. The “dot” or text diagram can be used to make minor changes. The FEN line above also works.
It’s output was used to create a diagram in a lesson plan.
I use Babeschess as a PGN reader.
I didn’t even mean that, just meant that the reader should be able to handle similarly detailed annotations. (And it can.)