I'll enter the <CG standard castling symbols> or whatever it's called

Kashdan - Isaacs was already in the db.

<cg.com> castling characters need to be the capitol letter O.

So <O-O> will never cause a game to be rejected.

<o-o> don't use lower case letter o.

As does the PGN standard...

https://en.wikipedia.org/wiki/Castl...

Curious, the wiki page claims 0-0 and 0-0-0 (with number 0) is used for algebraic and descriptive notations.

I like to denote the number 0 with a stroke to distinguish it from the letter O, likewise some fonts add a dot in the middle of the number zero to do the same. But to tell the truth, most chess books use fonts where it's hard to tell the difference (maybe just by the width).

* * * * *

Just curious - what's a "canceled game"?

I suppose I might find it in a previous post on this forum, but it would be nice for the PGN to have some indication.

To add a bit more about copy/pasting pgns.

If you copy paste pgns from:

rusbase
ozbase
britbase

there will be no issues with castling syntax.

Isn't a "canceled game" a game which was played, but later disallowed from official scoring in a tournament?

For example if a player withdraws, but had played several games before withdrawing, the aribiter may cancel those game results.

The record of the game having been played can be preserved, but should then have "canceled" listed in the header to make it clear the result did not count in the final scoring of the event?

One more thing about pgn errors-

I can only report about <Shredder Classic III> since it is the only chess engine I ever use.

If I copy paste a pgn into this engine, if there is a problem with syntax, one of two things will happen.

1. The incorrect syntax items will show up in the engine move list as "brown colored" and you will not be able to play through the moves on the engine.

2. The engine will sometimes auto-correct syntax errors in the move list.

Whether I am adding in moves manually, or copy-pasting from a database, I always put every pgn into my chess engine, so I never have problems getting a game published because of syntax errors in the move list.

As you both know, I've submitted a few games to <CG> in my time. In fact, it's more like a few thousand games. This problem with "0-0" vs. "O-O" has not come up before.

Here is my usual process. I download or hand enter game scores into my home copy of ChessBase, and check them against what sources I may happen to have. After this, I copy the game score into a NotePad file which contains a schematic of the PGN form, fill in the informnation, and copy from that itno the submission page on <CG>.

(Yes, I do every game individually. I don't feel comfortable submitting huge bunches of unchecked scores. Accuracy is very important to me.)

This process represents castling as "0-0" (numbers) rather than "O-O" (capital letters). I've been doing this for around fifteen years, and I don't recall any problems with the castling move (unless you guys have been Extra Nice and keep correcting my scores without telling me!).

There are still two games left. I'm going to resubmit them using my usual method. If they are rejected, then I'll have to make some changes to get the repsreentation right. But unless this is something instituted in the last few days, I can't believe it's the problem.

Game Collection: US Championship (1946). Games of Lewis J. Isaacs

<Phony Benoni>

I could not possibly agree more with you about using <Cancelled game> as opposed to <+>.

Like you, over the years I have uploaded a ridiculous quantity of games, but it was fairly recently that I finally figured out what <+> meant.

===================

Just one more bit of information on <castling syntax> that I remembered since my last post.

It was <Annie K> who figured out what was wrong with one of my uploads- I had used the numeral syntax <0-0>.

I had been using that syntax before, but not encountered this problem, which was- and remains- puzzling to me.

Still, <Annie> insisted that the best practice for me was to use the capital letter <O-O> syntax for all future uploads, and that is what I began to do. Two years later now and I have not had a single upload rejected due to the <O-O> syntax.

As usual with cg.com mysteries, there always seems to be another layer eh? I think much of this is down to <Daniel> having written all the code and fashioned it in a way that if it did not work- he could fix it, I presume by adding more code? At any rate, the new webmaster and techs have made a heroic effort to sort out Byzantine coding problems over the last few years.

<Phony Benoni>

As I just reported, I also always used the numerical <0-0> syntax on uploads, and I did not get a rejection until two years ago when <Annie> figured it out.

So I took her advice, switching to <O-O> syntax and there have been no further problems.

It strikes me that your plan to upload more <0-0> syntax castles has a flaw- it's possible that it does not trigger a rejection, but this gives you no useful information if the rejections are intermittent, as they seem to be. In other words, it is not an effective experimental design.

Much simpler, and preferable, would be using the <O-O> castling syntax in each and every case, since there is no evidence that it has ever caused a rejection.

I see that Lahde's biography of Kashdan (game 156, p, 119) has <36...Nc3>, citing "L'Echiquier", 1930, p. 1071. I don't have the original of that publication to check.

However, "Weiner Schach-zeitung", November 1930, p. 327, gives <36...Lb2-c3!>.

It looks to me like <36...Bc3> was the original move, and 36...Nc3 a later error. Pending further test results and FDA approval, of course.

Game 15, Tuesday, February 7, 1899: Hymes 0 Janowski

<A contemporary copy of the score would be welcome.>