Comments/Ratings for a Single Item

I have added the pieces system variable. This is the array stored in a set file. Its keys are piece labels, and its values are the filenames of image files. It basically tells which graphics to use for the pieces. I also fixed things so that you can set or get specific values of this array. This is done by following the variable name with a period and the element name. Because this will not work for labels that include periods, I have also added the pieceimg operator. This returns a specific element of the $pieces array.

The point of this is to make it even easier to support multiple sets when matching sets have not been provided. It can also be used to eliminate the need to use aliases. Here is how it may be used:

1. Set groupsets to an array of the allowed sets. Each should be the basename of a set file:

setsystem groupsets array abstract5 alfaerie-many;

2. Create an empty array:

set mypieces ();

3. For any pieces whose images already have the same label in different sets, set the corresponding elements of your array to those pieces:

set mypieces.k $pieces.k;
set mypieces.K $pieces.K;
set mypieces.q $pieces.q;
set mypieces.Q $pieces.Q;
set mypieces.r $pieces.r;
set mypieces.R $pieces.R;
set mypieces.b $pieces.b;
set mypieces.B $pieces.B;
set mypieces.n $pieces.n;
set mypieces.N $pieces.N;
set mypieces.p $pieces.p;
set mypieces.P $pieces.P;

Or just:

foreach label (k K q Q r R b B n N p P):
  setelem mypieces #label pieceimg #label;
next;

4. For each allowed set, use an if-elseif-else block or a switch-case block to set an individualized value for any pieces that are different between sets:

if == pieceset abstract5:
  set mypieces.s $pieces.s;
  set mypieces.S $pieces.S;
elseif == pieceset alfaerie-many:
  set mypieces.s pieceimg .de;
  set mypieces.S pieceimg .DE;
endif;

5. Change the value of $pieces to that of the array you created:

setsystem pieces #mypieces;

Although you could have edited $pieces directly, this has the advantage of stripping $pieces of all the pieces you will not be using in your game.

I have added the system variable groupsets and the nullary operator pieceset. My intention behind adding these is to allow a developer to support multiple piece sets even when he cannot find piece sets that use the same keys for the same pieces.

Here is how to do this:

1. Check the box for "Bypass Error Check on Piece Label".
2. In the Pre-Game code, use setsystem to set groupsets to an array of the basenames of the sets you will allow a player to select from.
3. Still within the Pre-Game code, use an if-elseif-else block or a switch-case block to check the value of pieceset against the basename of each allowed set, use the add command to change the pieces on the board to match the images you want to use from that set, and use the alias command to set appropriate aliases.
4. Within your other code, consistently do operations on the piece aliases rather than the piece labels. Note that the fairychess include file is already designed to do this, and it is the recommended include file for new games.

I have started work on updating the documentation to include previously undocumented operators. I'll update this comment as I make more progress. So far, I have gone through the nullary operators, which do not take any arguments.

I have added documentation for the previously undocumented operators: args.

I removed the following undocumented operators: debug, dump.

I removed documentation for the following non-existent operators: player.

I modified the dump command to use var_dump() instead of print_r, and I modified it to report on a single value when it is passed as an argument. The use of var_dump() adds information on the type of each value.

I added documentation for the new operators int and type.

Where can I find an editor so I can have a piece set uploaded?

Bump. I'm making a preset for Dai Shogi and help would be appreciated.

I[s] there a notation for octal, hexadecimal or binary constants in GAME code?

I have just added a new int operator that recognizes whether a number is a hexadecimal, octal, decimal, or binary literal and converts it to an integer. Hexadecimal literals begin with 0x, octal with just 0, decimal with nothing special, and binary with 0b. It checks the first characters to determine the base, then uses PHP's intval() function.

I[s] there a notation for octal, hexadecimal or binary constants in GAME code?

No. My tests show that PHP has this feature, but for some reason, it hasn't carried over to GAME Code. It looks like it has to do with handling numbers as strings. In a further test I ran, PHP would convert base 10 strings into numbers, but it would not convert other bases into numbers when they were enclosed by quotation marks. Additionally, PHP's is_numeric() function does not recognize strings in other bases as numeric.

My current automation code tabulates the capability of a move as bit flags packed into a single word. And often has to test one of these flags. Currently I use the expression & << 1 N #mode for this, where N is a constant indicating the number of the bit I want to test. This slows down the program by requiring extra shift operations at run time.

While I could build in conversion to handle this, it might be more costly than using a bit shift.

I could of course calsulate how much << 1 N is as a decimal number in every case, and write that into the code instead, but this makes the code very hard to understand, susceptible to typos (N can get quite large, like 28) and therefore hard to debug and maintain.

What I do is use constants with meaningful names. Doing this makes the code much more readable, and it reduces the chance for error by limiting the actual numeric values to some constant assignments.

An alternative would be to have a bit operator such that bit N #mode would calculate & << 1 N #mode (i.e. mask out the Nth bit).

What is #mode supposed to be?

I there a notation for octal, hexadecimal or binary constants in GAME code? My current automation code tabulates the capability of a move as bit flags packed into a single word. And often has to test one of these flags. Currently I use the expression & << 1 N #mode for this, where N is a constant indicating the number of the bit I want to test. This slows down the program by requiring extra shift operations at run time. I could of course calsulate how much << 1 N is as a decimal number in every case, and write that into the code instead, but this makes the code very hard to understand, susceptible to typos (N can get quite large, like 28) and therefore hard to debug and maintain. Written as a hexadecimal it would be much more obvious and less error prone, as even for N=28 it would just read 10000000. An alternative would be to have a bit operator such that bit N #mode would calculate & << 1 N #mode (i.e. mask out the Nth bit).

Where can I find an editor so I can have a piece set uploaded?

Thanks again Fergus.

Can you delete 18. resign move from the database?

I think this is causing a problem after a won command.

Instead of doing that, which would have fixed only one game, I did a more global fix. In the won, resign, and lost commands, I replaced $movenum, which is recorded in the log, with $mline[$mln]->movenum, which is generated by the program and indicates the number of the current move.

Fergus,

The problems with the "won" command are back.

This game is again showing the wrong winner:

https://www.chessvariants.com/play/pbm/play.php?game=Latrunculi+XXI&log=j_carrillo_vii-cvgameroom-2020-281-060

Can you delete 18. resign move from the database?

I think this is causing a problem after a won command.

= = =

Source code:

Thanks Fergus!

There seems to be something weird with the "won" command.

I have now fixed this. See the comment below.

The correct game was this one:

https://www.chessvariants.com/play/pbm/play.php?game=Petteia+-+Dmytro+Variation&log=j_carrillo_vii-halfer-2020-283-953

Seems to be fine now. Thanks.

It was showing that your opponent had won, yet you had entered the won command. So, I fixed that. It now makes use of the $players variable if available. This gets set when an invitation is accepted, and it keeps a permanent record of who moved first, which makes it easier to tell which player entered the command.

Fergus,

I must be going crazy... :-(

Now this game:

https://www.chessvariants.com/play/pbm/play.php?game=Petteia+-+Dmytro+Variation&log=j_carrillo_vii-halfer-2020-283-953

(the game whose moves I couldn't see before) is showing that my opponent won the game. This game is another one where I ended the game with the "won" command.

I issued the "won" command as white, but the log shows that my opponent (black) won.

There seems to be something weird with the "won" command.

Below the page source section with the problem:

Apologies Fergus, gave you the wrong link.

The correct game was this one:

https://www.chessvariants.com/play/pbm/play.php?game=Petteia+-+Dmytro+Variation&log=j_carrillo_vii-halfer-2020-283-953

But I just checked again and now the log is there. It was weird. last time I checked there were only two lines in the log, White to play and Jose Carrillo won the game.

And I saw it as well in the page source.

Seems to be fine now. Thanks.

And thanks for correcting the winner in the other comment I submited.

Cheers!

So I could write something like

def BadZone == 9 rank #destination =origin =destination =locustsqr =dropsqr =droppedpiece;

to get a function called with 5 parameters, and only testing if the second is on rank 9?

First, a matter of vocabulary. Function definitions have parameters, but function calls have arguments. An argument is a value passed to one of the parameters of a function. This will test whether the fourth argument from the right is a coordinate on rank 9, which would be the tenth rank. As long as you include five arguments in your function call, it will be the second argument from the left. If you had only four arguments, it would test the first one from the left. If you had three or less, that parameter would not be assigned a value. I know it's a bit counter-intuitive, but it's what comes from evaluating functions in reverse order, as Game Courier does.

Also, something went wrong with this game, which also just ended with the "won" command:

https://www.chessvariants.com/play/pbm/play.php?game=Petteia+-+Dmytro+Variation&log=j_carrillo_vii-halfer-2020-285-568

Are you sure you gave me the link to the correct game? This one doesn't include the won command in it.

Now it shows that my opponent (who resigned) won the game.

That's now been corrected.

Fergus,

Also, something went wrong with this game, which also just ended with the "won" command:

https://www.chessvariants.com/play/pbm/play.php?game=Petteia+-+Dmytro+Variation&log=j_carrillo_vii-halfer-2020-285-568

I'm trying to view the moves to replay the game, and now I can't see any of the moves.

It just show the starting board without any of the moves played.

Thanks in advance.

Thanks Fergus.

Now that the "won" command is working (and because by opponent had resigned to the game, while the won command didn't work) there was a strange behavior that affected the result for this game:

https://www.chessvariants.com/play/pbm/play.php?game=Latrunculi+XXI&log=j_carrillo_vii-cvgameroom-2020-281-060

Now it shows that my opponent (who resigned) won the game.

Can you please correct the result for this game? Thanks again.

To summarize, I fixed it.

Fergus,

I'm not sure I understood your response.

Does the won command don't work anymore when entered as a move?

How are we supposed to claim a win when we checkmate in a preset that does not enforce the rules?

It would make more sense for the "won" command to work as a move than the "lost" command. For losing a game we have the alternative to "resign" command, but what is the alternative when we win by checkmate and we should claim a win immediately, rather than ask the opponent to resign a checkmated position?