Lasker vs Fehmers, 1933

Iwhat happened with Lasker vs G A Fehmers, 1933 was that the version in ACB had a couple of extra moves at the end. Normally, I check each game by comparing the final position with a copy of the <CG> database which I download periodically. Since the final positions didn't match, I tagged it as 'nnot in <CG>."

This is a typically lazy way to operate, but more thorough ways are also more time intensive. Believe me, I spend enormous amounts of ime verifying or correcting information already.

By the way, the game also appeared in <Cess Review> for 1934. That version was identical to the <CB> version, and the game was added to the proper collection.

If I ever get around to submitting the games -- and there are about 2500 at present -- I will check more thoroughly before sending a game in.

Soerabaijasch handelsblad, 18-Nov-1933 has

16...h6 17. Rf4 Bd7 18. Bxd4 1-0

[Event "Ventnor City"]
[Site "Ventnor City, NJ USA"]
[Date "1939.07.09"]
[EventDate "1939.07.08"]
[Round "2"]
[Result "1/2-1/2"]
[White "McCormick, Edgar Thomas"]
[Black "Burdge, Harold"]
[ECO ""]
[WhiteElo "?"]
[BlackElo "?"]
[Source "365Chess"]

1.d4 Nf6 2.c4 e6 3.Nc3 d5 4.Nf3 c6 5.e3 Nbd7 6.Bd3 Be7 7.0-0 0-0 8.b3 Qc7 9.Bb2 dxc4 10.bxc4 e5 11.Qc2 exd4 12.exd4 Re8 13.Rfe1 Nf8 14.Ne5 a6 15.Ne4 Be6 16.Ng5 Rad8 17.Bf5 Bb4 18.Red1 Bxf5 19.Qxf5 Re7 20.a3 Bd6 21.c5 Bxe5 22.dxe5 �-�

===

½

<MannBee> I could fix it myself, and if there is no other option I will. Given the volume of pgn submissions, however, I would rather let people know about how to clean their own pgns of unwanted code before submitting.

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

<PB and J> I recommend this method to "clean" all pgns before submission:

It's a quick, easy and accurate method to ensure no rogue coding gets in to your submissions. It should work with any chess engine.

1. Copy and past your score into a chess engine.

2. Cope and paste the pgn from the engine to a Windows note pad. Note pad will not carry unwanted code, unlike Microsoft word or word pad. So be sure to use Note pad as your "go to" carrier.

3. Copy and paste the pgn from the note pad into the Pgn upload utility.

I do this for every game I submit and I no longer have any problems, but I sure used to have problems. I was driving <Annie K> crazy.

Anyways we did some experiments together and the protocol I published above there has not failed one time ever since. I use an ancient engine <Shredder Mark III>, and it automatically repairs any notation errors, syntax errors and unwanted code that may be present in the score.

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

If you don't want to use that method, you could also just post your pgn here in your own forum, and you would catch some coding errors. You can see that the "html 1/2-1/2" code is present here in the example at hand.

So in this case, if you manually typed in the "1/2-1/2" at the end of the pgn here in your forum, then re- posted it, the pgn would come out clean. You would see the "1/2-1/2" true, not filtered through html. code (�-�). I just practiced this method here in your forum a minute ago, and it works.

The � character is actually the UTF-8 encoding for the replacement character, as shown by copying and pasting it into a python session:

� = (the 3-character string 0xEF 0xBF 0xBD in UTF-8 land)

https://www.google.com/search?q=%EF...

Most likely, the result was "mojibaked" by a text editor while trying to handle the ½ character.

* * * * *

The ½ character can be represented various ways in HTML land:

<
&frac12; (½)

&#xbd; (½)

&#189; (½)
>

I like to just use 1/2 and avoid all this trouble. Likewise, I use "=" in xtabs as well, for portability, and also because I think the table reads better.

* * * * *

Think this post was long? Some people make an entire blog post on the subject:

https://codingrigour.wordpress.com/...

* * * * *

I'm a little out of practice at submitting games, but i think the major problem was that I submitted the game before noon. I am not a morning person.

Actually, the problem turns out to be a blessing in disguise since I had reversed the players' colors. I have just resubmitted, with corrected colors. Darn mornings.

<PB on Rye>

Thanks for further information. It should be ok now so I'll go look at it.

More importantly, how are your rent collections going on Ventnor Avenue?

<Phony Benoni>

On your four submissions, these three processed successfully:

John J Leary vs O Ulvestad, 1939

H R Morris vs M L Hanauer, 1939

F Reinfeld vs J W Collins, 1939

===

On the <Burdge - McCormick> game, you appear to have submitted a different score this time?

The original score you submitted is:

1.d4 Nf6 2.c4 e6 3.Nc3 d5 4.Nf3 c6 5.e3 Nbd7 6.Bd3 Be7 7.0-0 0-0 8.b3 Qc7 9.Bb2 dxc4 10.bxc4 e5 11.Qc2 exd4 12.exd4 Re8 13.Rfe1 Nf8 14.Ne5 a6 15.Ne4 Be6 16.Ng5 Rad8 17.Bf5 Bb4 18.Red1 Bxf5 19.Qxf5 Re7 20.a3 Bd6 21.c5 Bxe5 22.dxe5 1/2-1/2

but the score you re-submitted is:

1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. Qc2 d5 5. e3 O-O 6. Bd3 c5 7. Nf3 Nc6 8. a3 dxc4 9. Bxc4 cxd4 10. axb4 dxc3 11. bxc3 Qc7 12. Bd3 b6 13. Bb2 Bb7 14. O-O a6 15. Qe2 b5 16. Rfc1 e5 17. Nd2 Rfd8 18. Ne4 Qe7 19. Nc5 e4 20. Bc2 Ne5 21. Nxb7 Qxb7 22. Rd1 Nc4 23. Rxd8+ Rxd8 24. Rd1 Qc8 25. Ba1 g6 26. Rd4 Re8 27. Bb3 Ne5 28. h3 Nd3 29. Bc2 1/2-1/2

The re-submitted score was not processed, because it is a duplicate of this game:

M Bain vs T Knorr, 1939

<Henry Benoni IV, Part One>

Imo, unless the PGN result explicitly differs from the header tag result, the header tag Result should be used.

<z> I'm not sure what you mean? I agree there's a redundancy built in to the current system, but possibly some people like it that way.

I just ran some tests and I can report:

The <Pgn Upload Utility> will reject games missing a result tag, but it will not reject games that have a result tag but are missing a result at the end of the move list.

<Phony Benoni>

I am sorry I used the wrong play!

<z> <Imo, unless the PGN result explicitly differs from the header tag result, the header tag Result should be used.

(If the PGN uploader is to be permissive / fault tolerant)>

According to the tests I just ran, this is indeed the current state of affairs.

[Result "1-0"]
1.e4 resigns
1-0

The win is noted in two places.

* * * *

<z> Ok I understand now - the result is noted in two places. It's a redundancy to be sure.

Whether or not to get rid of the redundancy is another matter. I expect there would be adherents to both - status quo, or get rid of the redundancy.

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

On submitting/processing:

To submit a game, you use the <Pgn Upload Utility>.

That is the port of entry and the first gate keeper. It will reject submissions with pgn tag errors, and it will tell you what you did wrong in big red letters. That's a good thing in my opinion.

The <Pgn Upload Utility> then sends pgns that pass the gate keeping requirements to a <processing web page>. Traditionally, only admins had access to this page. Nowadays, a few of us volunteer editors also have access. We are Beta testing under the supervision of the admins. The ultimate goal is that all cg.com editors will be able to upload their own games and those of others, and to process their own correction slips and those of others.

So on the <processing web page>, there is a "receiving box" with the name of the person who submitted the pgn in the first place. I can now examine the pgn. I can also edit the pgn.

To become a published game, it still needs to be processed. I do this by clicking the "process game" button on the template.

In the vast majority of cases, there is nothing wrong with the pgns. If there is, we can either not publish it, fix it and publish it, or where possible contact the uploader and ask them to fix and re-submit.

=====

On my experiment- I sent two dummy pgns through the <Upload Utility>. One had a result marked at the end of the game score, but no result tag. This was rejected at source. It did not make it to the <processing page>.

The other had a result tag, but no result marked at the end of the game score. This did make it to the <processing page>. I attempted to process the game, but the procedure failed- the game was not published.

So you can see there's a disjunct here- if the <processing page> is not publishing games unless they have <both> result locations filled in, then the <Upload Utility> should not be passing games to the <processing page> unless they also have both result locations filled in.