Website founded by
Milan Velimirović
in 2006

2:18 UTC
ISC 2020



Remember me

Forgot your
Click here!
to create your account if you don't already have one.

Rating lists


MatPlus.Net Forum General Comments on the Codex
You can only view this page!
(1) Posted by Siegfried Hornecker [Tuesday, Apr 3, 2007 03:07]; edited by Siegfried Hornecker [07-04-03]

Comments on the Codex

I took a look at the codex again and wrote into this posting what I think.

Part I

Article 2 - Sentence 2 is partially wrong! There's often no need to form the stipulation in words. Today, often common symbols are used, like #7 or h#9 etc. I think this should be added to it.
PS: Also see Article 21, e. (not my comment to it)

Article 7 - The FIDE rules don't apply to compositions since they are not made for it. Article 1 of the FIDE rules imo clearly states that they can't apply to compositions at all. See here.

Article 9,2 - A study is also considered cooked if one can win instead of draw, even if there is no other way to draw. This should be added here or in Article 11.

Article 10,2,c - I don't think this is used by anyone. It is really pointless to punish a winning study because one has a method to repeat moves. In a draw study, there could be a point at this.

Article 13,2 - I don't think in a helpplay problem, a promotion dual in the final move should be tolerated. If at all, then it should be only in few cases but not generally.

Article 14 - I think even illegal positions should be accepted but then any retro-play may not apply to it, e.g. they can't use retrospective rights. I don't know if promotion also should be considered at this, e.g. if it should be possible to promote to all pieces or only those that are on the board in the initial position. Also, this rule was ignored sometimes. I've seen an oudot task with nine black pawns some time ago, for example.

Article 16,2 - Those expressely stipulations ain't necessary, in my opinion. They don't give anything and even take the fun from solving.

Article 18 - It's possible to exclude or change this for a certain problem by stipulation, e.g. the same piece instead of the same kind of pieces.

Article 21,e - See my comment to Article 2

Article 27,b - These tournaments shouldn't be considered for official solving ratings due to the ability to cheat with literature and/or electronics.

Annex 2,1,e - This isn't necessary anymore. Instead, other possibilities of contact can be given or even online formulars.

Annex 2,1,h - Sadly, most of the time this isn't given.

Annex 2,2,b,ii - This is ignored most of the time. I really appreciate to get the awards of tourneys where I participated but most of the time they are not sent to me.

Annex 2,3,g - I think, this should be defined as 1 or 2 years since that time should be enough for a judge. It is up to the judge to not make too many tourneys at the same time. Yes, there have been soviet tourney where this were some decades!

Annex 2,4 - The address shouldn't be necessary if it is a pure online tourney. Otherwise, it is (to send a printed copy of the award). I often ignore footnote 37 and it doesn't matter to the most, by the way.

Annex 2,5,b - I think, this time should be six months plus one month for later corrections (if allowed) that appear in the next form of the award. Also see Annex 2,5,c.

Annex 2,6,b - The composition should be published within 2 years, in my opinion. Also, the solution should be published within one year after the composition.

Annex 2,6,d - This should be three months, I think.

Annex 2,6,e - 12 months should be enough for this. 18 months is too long. A good director should ever send information about receipt, if possible, anyway.

Part II - I can't aay anything about this since I'm not really a solver.

Your opinions are welcome!

A general discussion on organization of tourneys etc. was started at Andrei Selivanov's website but I don't think it was frugal, sadly.

(Read Only)pid=582
(2) Posted by Hans Gruber [Saturday, Jun 16, 2007 13:16]

Hi Siegfried,

thanks a lot for these comments. This reminds me that still some work is to be done with the Codex (I am afraid I am the chairman of the respective PCCC subcommittee).

Hans Gruber
(Read Only)pid=894
(3) Posted by Michael McDowell [Sunday, Jun 17, 2007 00:18]

Your view on Article 13,2 is far too severe in my opinion, as it would exclude all helpmates ending in a long-range mate by promotion. Are you telling me I should consign the following problem to the bin? I seem to remember it missed being included in the FIDE Album by half a point!

2nd HM Problem Observer 1979-80

(= 3+3 )

Helpmate in 5

1.Rh8 g4
2.Qa8 g5
3.Qh1 g6
4.Kb1 g7+
5.Ka1 gxh8Q

(Read Only)pid=910
(4) Posted by Uri Avner [Sunday, Jun 17, 2007 01:25]

McDowell's is a good example why article 13.2 should not be changed.
(Read Only)pid=913
(5) Posted by Siegfried Hornecker [Sunday, Jun 17, 2007 06:26]; edited by Siegfried Hornecker [07-06-17]

I fully agree with you. I made a mistake there.

I change it to: A promotional dual should not generally be tolerated if the theme of the helpmate is based upon promotion (then the promotion should neither be counted as one of each).
Also, it should not be accepted if there is a dual other than promoted pieces using the same line.

Does what you showed really count as a dual, By the way? I wouldn't call it a dual.
(Read Only)pid=916
(6) Posted by Michael McDowell [Monday, Jun 18, 2007 10:59]

Technically a dual, but the choice is completely irrelevant. I did see this problem criticised in print once, by an individual who presumably regards an inspiration like Loyd's Excelsior problem as worthless!
(Read Only)pid=922
(7) Posted by Hans Gruber [Monday, Jun 18, 2007 11:18]

Dear Michael,

I completely agree. The matter of "duals" in most cases is a matter of taste, esthetics, or judgmental decision. Probably it is wise not to try to restrict that in codex rules.

In your problem, much of the impression comes from the long moves - thus the "dual" is quite unavoidable. As a judge, I would not care about it.

Hans Gruber
(Read Only)pid=923
(8) Posted by Siegfried Hornecker [Tuesday, Jun 19, 2007 04:56]; edited by Siegfried Hornecker [07-06-19]

The rules of castling must be precised for chess composition. See the following problem of mine:

(= 8+5+1N )

Imitator a2

Siegfried Hornecker forum, May 30 2007
White to move and mate in three moves

1.c5 Kxg2 2.0-0 Kf3/Kh3 3.Bf4!/Kg2! mate

Is castling legal here? Does f1 count as a threatened field?

Well, this one raises the question about the legality of castling. 2.Kf1 would be illegal but before and after 2.0-0 black couldn't go to f1 with his king if the white king would be there as the imitator has no move to "south west".

PS: Note that on the mate after 2...Kh3 h4 is only guarded by Bg5, not by Rh1 since the imitator would have to go over the edge of the board.
(Read Only)pid=925
(9) Posted by Hauke Reddmann [Tuesday, Jun 19, 2007 14:10]

Well, I'd say: If Black's Xe1/f1/g1 (resp. Xe1/d1/c1) is illegal anyway
due to whatever stipulation cause (unlike due to self-check!), White may castle.
Instead of Imitator, think of e.g. grid chess - Pc2 does not control d1.

(Read Only)pid=926
(10) Posted by Siegfried Hornecker [Tuesday, Jun 19, 2007 19:19]

The thing is - black never controls f1 but 1.Kf1 would be illegal because black would control f1 then. Also, g1 is attacked by the black king but not anymore when castling is done. So does this contradict with 3.8.ii.(2).a ?
if the square on which the king stands, or the square which it must cross, or the square which it is to occupy, is attacked by one or more of the opponent`s pieces.

(Read Only)pid=929
(11) Posted by Hauke Reddmann [Wednesday, Jun 20, 2007 10:47]

You should have become a lawyer - and it's much more
lucrative than problem composing anyway :-)

(Read Only)pid=932
(12) Posted by Siegfried Hornecker [Wednesday, Jun 20, 2007 11:17]; edited by Siegfried Hornecker [07-06-20]

I'd really like to become a lawyer - like Alois Wotawa and someone else I know (but forgot who) but I can't since I'd have to study jurisdiction and it's not possible here. It would be enough
(I wanted to write but forgot what since I was interrupted. I'll add it later if I remember it)

I know a lot about the law in Germany anyway.

But I disagree. Chess composition will be lucrative soon, I hope, but I can't say at the moment why I believe so. You'll have to wait a few months.
(Read Only)pid=934
(13) Posted by Siegfried Hornecker [Saturday, Oct 20, 2007 04:29]; edited by Siegfried Hornecker [07-10-20]

Annex II, pt. 2

6 - what if the organisator only gives a computer board (like in last Schwalbe meeting)?

7d - "obvious" should be defined. Maybe the solver should point out why it is a win or draw (see Kasparian's studies).

11 - should be reworked, giving full time for 0 points sounds strange compared to 1 point with 1 minute or such (and of course, 20 points with 40 minutes or whatever a solver may need). Maybe at least 1 minute per not correctly solved problem should be given. If the time already exceeds this, nothing more has to be added.

12 - I'd like participants to get the solutions they gave back, too.

15 - Should only apply while the tourney is held. Also, the director can't see solutions???

15 - The director should not be in that jury. Instead, another neutral person should be taken there.

PS: Hans Gruber, you're still reading this or should I send per mail?


As to my problem above: It is incorrect. The solution given is incorrect but there's a not intended dualistic solution in two:

1.Rh7!! (Ia8) Kh1 (Ib8) 2.Kd1 (Ia8) or 2.Bf4 (Ia7) checkmate. :-(
(Read Only)pid=1590
(14) Posted by Kevin Begley [Saturday, Nov 10, 2007 00:48]

Re: Castling thru check in fairy conditions...

I was told some time back that to evaluate if you are castling thru check when a fairy condition is employed (such as Immitator), first make the King move onto the square in question (say Kf1 if white castles short, or Kd1 if white castles long), making the full move (that is to say, allow fairy conditions to play out: e.g., move the Immitator accordingly), and then determine if this is self-check.

If self-check, castling is not possible.
Otherwise, you are technically NOT "crossing check."

I had always assumed this was an official decision, and I operated under the assumption that everybody agreed with this.

I used this very argument in a problem I had recently composed (only a few weeks before the immitator problem appeared on here), and the editor (not going to name him here) told me my problem is not legal and belongs in the trash.

He also made a number of other claims about legalities -- some of them very shocking! -- and although he has recanted a few (when confronted with examples to the contrary), others he still claims are not 100% certain either way (he then argued that there are two types of this condition -- and no, I'm not talking about AntiCirce -- but I am yet to see an example which is specifically "his" type).

After this, I showed my problem to a few problemist friends, asking for their opinions about how to resolve castling through check.

I was delighted when a highly respected composer agreed with me, saying my problem was not only legal, but also "a perfect gem."
But, although a majority agreed the problem is legal, it troubled me that some others did not agree, and a few were honest enough to admit they were not sure.

Quite frankly, this all left me the impression that special case rules for some fairy conditions are so poorly defined that there is no certainty as to what is a legal move.

I expected things like this would have been "officially" resolved long ago.
Sadly, it seems we do not all agree on the rules for some fairy conditions.

Kevin Begley

ps: I sent the problem elsewhere (about a month ago), but have yet to hear from that editor. I expect the long delay indicates that my email probably resides in his recycle bin.
(Read Only)pid=1699
(15) Posted by Siegfried Hornecker [Wednesday, Jul 16, 2008 17:41]

Kevin, what is with your problem?

The german specialist Werner Keym proposes the following change in the newest issue of Die Schwalbe. Here's the conclusion of his article.
Some gaps in article 16 of the "Codex for Chess Composition" (1997) have been discovered recently. They can be filled by the following optimized version of article 16:
Article 16 - Dependent Moves
(1) Castling is deemed to be permissible unless it can be proved that it is not permissible (Castling convention). An en-passant capture on the first move is permissible only if it can be proved that the last move was the double step of the pawn which is to be captured (En-passant convention)
(2) Where the rights to castle and/or to capture en-passant are mutually dependent, all possible sequences must be considered according to both conventions: in this case the solution consists of several mutually exclusive parts (Partial Retrograde Analysis/PRA). In the special case where White's right to castle excludes Black's right to castle (and vice versa), an alternative is possible: the party exercising this right first is entitled to castle (Retro Strategy/RS).
(3) Other conventions should be expressly stipulated, for example if in the course of the solution an en-passant capture has to be legalised by subsequent castling (a posteriori/AP).

Note that I don't agree to the third part! I believe, AP shouldn't have to be expressly stipulated.
(Read Only)pid=2474
(16) Posted by Kevin Begley [Thursday, Jul 17, 2008 12:21]

"Kevin, what is it with your problem?"

The good news: My problem is now published -- thank heavens for Mat Plus!!
See #951 in Mat Plus #29 (Retros Section)...

I don't want to give the solution away, but watch for a Valladao Task (in 3 seperate phases of a #1).

(Read Only)pid=2478

No more posts

MatPlus.Net Forum General Comments on the Codex