Template talk:Game of Go position

From Wikipedia, the free encyclopedia

Contents

[edit] Manual

Well I am in the process of creating a fully documented manual for both games and inclusion in articles and documentation. Not completed yet, but here it is: Wikipedia:Sandbox/Game of Go/Manual. -- Natalinasmpf 21:46, 8 May 2005 (UTC)


[edit] Example Use - A demonstration on how it all comes together

[edit] Single Letter Invocation

Entering "b" renders a black seed. Nothing renders an empty square, while "x" renders an empty star point. "w" renders a white seed. The number at the lower right corner controls the size. It really should stay at 30 though, otherwise it begins to fragment because of the coordinates. Smaller sizes will have to use no coordinates for now, which is available at Template:Game of Go Position NC.


{| style="margin:1em;"
| style="border: solid thin; padding: 2px;" |
{{Game of Go Position|=

 19 | | | | | | | | | | | | | | | | | | | |=
 18 | | | | | | | | | | | | | | | | | | | |=
 17 | | | |b| | | | | | | | | | | | | | | |=
 16 | | |a|w| | | | | |x| | | | | |x| | | |=
 15 | | |b|w| | | | | | | | | | | | | | | |=
 14 | | | | | | | | | | | | | | | | | | | |=
 13 | | | | | | | | | |w| | | | | | | | | |=
 12 | | | | | | | |w|b| | | | | | | | | | |=
 11 | | | | | | | | |b|w| | | | | | | | | |=
 10 | | | |x| | | | | |x| | | | | |x| | | |=
  9 | | | | | | | | | | | | | | | | | | | |=
  8 | | | |w|b| | | | | | | | | | | | | | |=
  7 | | | | |w|b| | | | | | | | | | | | | |=
  6 | | | | |b| | | | | | | | | | | | | | |=
  5 | | | | | | | | | | | | | | | | | | | |=
  4 | | | |x| | | | | |x| | | | | |x| | | |=
  3 | | | | | | | | | | | | | | | | | | | |=
  2 | | | | | | | | | | | | | | | | | | | |=
  1 | | | | | | | | | | | | | | | | | | | |=
     a b c d e f g h j k l m n o p q r s t

|30}}

|-
|}





19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
a b c d e f g h j k l m n o p q r s t


[edit] Two Letter Invocation

See the manual for instructions. This is only a demonstration of the template's two letter capacity.


{| style="margin:1em;"
| style="border: solid thin; padding: 2px;" |
{{Game of Go Position|=

 19 |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |=
 18 |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |=
 17 |  |  |b1|  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |=
 16 |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |=
 15 |  |  |  |w2|  |  |  |  |  |xl|  |  |  |  |  |xl|  |  |  |=
 14 |  |b3|  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |=
 13 |  |  |  |w4|  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |=
 12 |  |  |  |  |  |  |  |bl|  |  |  |  |  |  |  |  |  |  |  |=
 11 |  |  |  |  |  |wl|  |  |  |  |  |  |  |  |  |  |  |  |  |=
 10 |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |=
  9 |  |  |  |xl|  |  |  |  |  |xl|  |  |  |  |  |xl|  |  |  |=
  8 |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |=
  7 |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |=
  6 |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |=
  5 |  |  |  |  |wl|  |  |  |  |  |  |  |  |  |  |  |  |  |  |=
  4 |  |  |  |xl|  |bl|wl|  |  |xl|  |  |  |  |  |xl|  |  |  |=
  3 |  |  |  |bl|  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |=
  2 |  |  |  |  |  |bl|  |  |  |  |  |  |  |  |  |  |  |  |  |=
  1 |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |=
     a  b  c  d  e  f  g  h  j  k  l  m  n  o  p  q  r  s  t

|30}}
|-
|}


19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
a b c d e f g h j k l m n o p q r s t


[edit] Help!

The only problem, is I can't align the last row. Help! -- Natalinasmpf 01:21, 8 May 2005 (UTC)

Partially resolved the issue by removing borders, but I would really like them. -- Natalinasmpf 01:21, 8 May 2005 (UTC)

You are trying to get the borders on the outside of the numbers? As in the whole table? Dalf | Talk 01:23, 8 May 2005 (UTC)

Not really. But that actually might be a good alternative, how would I do that? What happened was when I put it inside, the last row became detached because of the parameter lengths of the last row (ended up making a newline, instead of being continuous). -- Natalinasmpf 01:27, 8 May 2005 (UTC)
Have a look at [1] and see what I did (which is essentially steal the example off of Template talk:Chess position. Dalf | Talk 01:29, 8 May 2005 (UTC)
I suppose you can either incluse teh whole thign in a table like I have done as part of the template and add a few options for border or noboarder and caption. Or you can let people either add the "frame" and caption themselves or not. Either way its pretty cool. Dalf | Talk 01:32, 8 May 2005 (UTC)

[edit] Size problem

If you make it much smaller than 30 (which is really too big for most articles) you start to get gaps because of the width of the digets. You might instead of doing pixel size do some sort of small/normal/large format and remove the numbers where needed. Or perhaps we can change the font size. I tried it at 15 and 10 and both looked strange. Dalf | Talk 01:36, 8 May 2005 (UTC)

Apparently its just because of the co-ordinates. Anyway, this should only be a problem on the larger boards, I will create a 13x13 board, and maybe a 9x9, or a 5x5 for example positions. Hmm, all the boards needed lol, because it doesn't have a fixed position like chess. But the examples won't have co-ordinates, after that. Here's our smaller board, size 15: -- Natalinasmpf 02:52, 8 May 2005 (UTC)






I actually think instead of (or in addition to) the other sized boards we should think about doing one that is a quadrant of the board with un-terminated liberties where its cut. This is a format that I have seen used in Go gooks rather a lot and also in some of the go articles here. This way you could for example create examples of various joseki without having the whole board like here:

Image:Joseki1.png

Also a middle of the board view might be usefull like here:

Image:Go_adjacent_stones.png

The the problem is going to be finding the minimum number of templates to do the job and keeping it simple and not confusing. The KISS principle works here aswell. I think the names for them will be the most important part. Dalf | Talk 04:30, 8 May 2005 (UTC)

Okay, I have created Template: Game of Go Position 13x13 and Template: Game of Go Position 9x9, they don't have cut liberties however, they are whole boards. But would this be sufficient? -- Natalinasmpf 05:13, 8 May 2005 (UTC)

I think we can do it with the pictures that you have already. I will give it a go in a secound (after I think about a good name for it. Also I am looking through some go books and we might want to add images that include numbered stones. 1-9 for each white and black will be enough for most examples I think unless someone decides to write an article about any well known games. The books I have also have instances where a location is occupied by an 'a' or 'b' or 'c' to indicate possible next moves when they want to discuss the benifits of each. Dalf | Talk 05:26, 8 May 2005 (UTC)
Oh I went ahead and created Template :Game of Go Position 9x9 Quarter and Template :Game of Go Position 9x9 Middle for boards based on context, where it would be needed (ie. analysing openings near the end of the board or ones in the middle). Probably won't be used as much, but you could probably agree it would be useful. -- Natalinasmpf 06:29, 8 May 2005 (UTC)
Oh, thanks for helping me making the corrections. I was wondering why it seemed to suspiciously easy to transfer one type of board to the other. Oh, I'm trying to devise a general manual for using all of the boards, ie. education about the principles, what does which, how it works, the different edit sets for the boards, etc. but that would take some good explanations, so if you could help me in that area...part of it is already in here but I plan to move it onto a template page or something similar, or at least something where there is an easy single refeence to. Now all I have to do is create some formats for them. (Format being what you easily post to invoke the template.) -- Natalinasmpf 06:13, 8 May 2005 (UTC)

Taken from your own sandbox, here they are, I guess. Oh by the way aren't star points present in 13x13 boards as well? -- Natalinasmpf 06:48, 8 May 2005 (UTC)



Analyse the centre!


{| style="margin:1em;"
| style="border: solid thin; padding: 2px;" |
{{Template:Game of Go Position 9x9 Middle|<!--
 1 --> | | | | | | | | |<!--
 2 --> | |b| | | | | | |<!--
 3 --> | | | | | | | | |<!--
 4 --> | | | | | | | | |<!--
 5 --> | | | |b| | | | |<!--
 6 --> | | | |w|w| |b| |<!--
 7 --> | | | | | |w|b| |<!--
 8 --> | | | | | | | |b|<!--
 9 --> | | | | | | | |b|20}}
 <!-- a b c d e f g h j  -->
|-
| style="text-align:center" | Analyse the centre!
|}

See Empty triangle for an example of this in an article. Dalf | Talk 07:20, 9 May 2005 (UTC)



Analyse the corner!


{| style="margin:1em;"
| style="border: solid thin; padding: 2px;" |
{{Template:Game of Go Position 9x9 Quarter|<!--
 1 --> | | | | | | | | |<!--
 2 --> | |b| | | | | | |<!--
 3 --> | | | | | | | | |<!--
 4 --> | | | | |x| | | |<!--
 5 --> | | | |b| | | | |<!--
 6 --> | | | |w|w| |b| |<!--
 7 --> | | | | | |w|b| |<!--
 8 --> | | | | | | | |b|<!--
 9 --> | | | | | | | |b|30}}
 <!-- a b c d e f g h j  -->
|-
| style="text-align:center" | Analyse the corner!
|}

Wow you are fast ...... perhaps ignore my comments below. This is probbly enough and if people want to do somethign diffrent they can do it manualy.Dalf | Talk 06:57, 8 May 2005 (UTC)

[edit] Other views

I am looking through some go books and it seems that depending on the example just about any view might be useful. Half the board or just one side+corner as I just made below. I wonder if there is a way with the templates to create a sort of intermediate situation where it is still partly manual but more flexible. Perhaps a template that creates a single row of variable length and optionally terminated with walls and corners. I am new to templates so I am going to have to do some reading but that way people can do things like the one below or just a stretch of wall or just a corner etc.

This also solves a size problem for cases where the example only needs a small amount of board space. Basically I think we should try and find a solution for partial boards that only uses one or two templates but can do any X by Y segment of a 19x19 board. Dalf | Talk 06:55, 8 May 2005 (UTC)

[edit] Additional images

I noticed that you (User:Natalinasmpf) created these images. First off wow! What program did you use to make them? Was it hard to get them to line up? Second, I mentioned this above but we were focused on layout and sections of the board etc. I think if we had images for the following people could basically anything you see in a Go book:

  • Both colors of stones with the numbers 1 through 9 or higher in them as in the above image to indicate sequence of moves.
  • Both colors of stones with a circle in them as in the image above. This is frequently used in books or online Go to indicate the last move or the next move or a stone that was there at the start. In some books a triangle is used instead of a circle.
  • The letters a, b, c, etc. In the books I am looking at there are no 'liberty' lines coming away from the letters, and they are used to represent possible next movies in a given position or simply to indicate an important point for analysis of the position. Dalf
I actually used gimp lol, and none of its advanced features (like special highlight, none of that), so you could duplicate the efforts in paint (if you're a Windows user) as well. Most of the technique I used in this was actually something I was taught at Primary 6 (three years ago) concerning tessallations where I had to line up geometrical figures on a board and make them match up, (rotational symmetry was involved as well)...so the critical thing here is that I made it from the center (its 135x135), so its thus at coordinate 67,67 where I started out with one dot, then I had four lines of 67 pixels in length branch out to the border from the centre. Then I thickened each line in proportion. This creates your empty position. To create corner or side positions, I simply eliminated one "arm", and rotated it 90 degrees for every position until I had all four corners, all four sides, etc. Then I proceeded to create an even circle of equal radius, then filled it with white or black, etc then aligned it with the lines, which wasn't that hard as its not antialiased so the distinction for the middle was quite clear. I actually made a smaller circle, before, and was about to discard it for being too small when I realised it could make a good star point. So yes, that mathematical unit was actually quite useful. Well now that you know the method, you can try it in paint as well. The circle diameter, I believe was 89x89, so you can duplicate my efforts for the numbers, etc. I'm about to use ImageMagick to antialiase the images, you might want to as well. -- Natalinasmpf 22:23, 8 May 2005 (UTC)
I am not sure I can specify exact pixels with paint (or at least I cant figure it out) but I seem to be doing ok eyeballing it. After I upload the images if they are suboptimal I am sure someon will fix them :) Dalf | Talk 22:51, 8 May 2005 (UTC)
Look at the bottom of the paint program - the coordinates are listed there, so you can see the coordinates of where your cursor is hovering above. -- Natalinasmpf 22:55, 8 May 2005 (UTC)
I am going to make some of these images. Because of the number of images I do not think we are going to be able to get away with one letter for all of them. But I am going to use things like 'o' and 'O' for white with a circle and black with a circle and so on. Right now I am looking at doing:
  • Stones in each color and for each diffrent board position with numbers in them. These will have two letter designations for hte template probbly w1sd, b1sd, w1md, and so on.
  • Stones in each color with a circle, and X and a Triangle in them. These will be designiated o, O, t,T.
  • Empty points with a,b,c,...,f. Which I am thinking we can denote with A B C D E and F.
  • Empty points with a large X in them. Denoted X and not x.

my artictis abilities are limited but I already have a few of these done so I will upload them later today. (I just saw your post here .... I am using PAINT since my unix box with gimp on it is not up at the moment.)Dalf | Talk 22:41, 8 May 2005 (UTC)

That would require a massive shifting in the template though. We could just add a legend and key. we have about 62 different combinations we can add, and only 3 of these taken up. -- Natalinasmpf 23:00, 8 May 2005 (UTC)
I dont understand what your saying here? This shoudl not require changing the template at all. You can pass multi character values to templates. It wont LOOK as pretty in the un-edited mode but it will work perfectly. People do the chess tempalte all on one line frequently. But just to make sure let me go ahead and upload some of hte images and demonstrate. Dalf | Talk 23:18, 8 May 2005 (UTC)
Yeah, I can, but I want some prettiness in unedited mode, yes. Perhaps I will reupload all the filenames to suit two letter systems again. But that would require changing everything that invokes a one letter system. Hmm, you can go ahead and upload it, but I think we won't be using it for awhile (not today, at least.) Also, I prefer the letter and number first, not the other way round. That way it looks a bit clearer.

Ie 1w and 1b doesn't look as good as w1 and and b1. -- Natalinasmpf 00:08, 9 May 2005 (UTC)

I have all the numbers through 9 and the letters a and b done. I only have uploaded through the number 3 and I am going out to dinner but will upload the rest this evening. Also I shoudl have named the letter a ones with a uppercase A instead of lower since the b havse to be uppercase as not to conflict with the black stone images. We might think about moving the a images later but for now it works. If we decide not to move them then I will put c, d, e, and f all lowercase so that B is an exception with a good reason. I will work on the stones with Triangles in them tonight as well. Dalf | Talk 01:43, 9 May 2005 (UTC)

When you're done, can you help me complete the section on the the two letter (and one letter and explain the difference if you can) system in the manual, as well as update it for the Weiqi-image page? Keep the original naming convention and call it as the original version, but mention there's a second naming convention as well. And on the subject, you got my other post regarding things like "wl" and "xl", right? -- Natalinasmpf 02:55, 9 May 2005 (UTC)

[edit] Eureka!

Can't believe myself. The solution is so simple. We simply have two ways of citing the template, where both a two letter and one letter system will work by simply uploading duplicates of the one letter images with two letter ones. The one letter images still remain, and the one letter invocations and two letter invocations can both be used.

The only thing that this really requires changing are the documentation manuals. -- Natalinasmpf 00:17, 9 May 2005 (UTC)

I already tried this with redirects on the letter a but the re-direct did not work. Uploading the file to each name needed DOES work but I asked abou the redirect have a look:

Wikipedia talk:Uploading images#Image_redirects


Uh reminder, we want an "Weiqi-aacb.PNG", for conformity to a two letter system. Unless you want it twice? There also doesn't seem to be any lines on that image? How would that line up? Yeah, basically I'm going to try uploading the one letter images as two letter now. Begin! -- Natalinasmpf 01:15, 9 May 2005 (UTC)
Oh can I reserve the image names "Weiqi-*lmn.PNG" for duplicates of existing images...the point is aesthetics...."x1" looks more "square" (this suited to the board) then "xx". Where mn is whether its a side, corner and such, of course. Ie. this is for the star points and basically anything that uses a one letter system, so it easily fits in with the two letter citation.
I don't know if I like that. You are correct that it will make the templates look good in the EDITOR but they will look the same in the article and more importantly doing the two letters for all of them makes the template MUCH harder to use. Now you have to type more characters, or worse if you leave a leading or trailing space in the requirment people will leave them out all the time and wonder why it does not work. I just used these images (but not the templates) to add a diagram to an article, go have a look it is Avalanche opening. I am also going to write a small CGI program that I can run on my webpage that will allow people to put in dimisions say from a to f and 1 to 10 with stones at ..... and it will generate the longer version. Dalf | Talk 04:17, 9 May 2005 (UTC)

Do post the source code when you're done, I might end up trying to implement some sort of script (if the wikisoftware allows it) for some flexible board making. ;-) The requirements won't require a trailing space - the two character version will be "xl", "wl", "bl" etc which looks more aesthetic in terms of alignment. There won't be a trailing or leading whitespace requirement. Plus, if they used "x" or "xl", it doesn't matter, just that the "x1" will be there as an option for ease in alignment for two letter invocations. After all, that's how it went in the chess template. -- Natalinasmpf 05:07, 9 May 2005 (UTC)

Sure thing, though I am going to chesck if this is poissible using subst and templates for individual lines etc. If it is possible using the limited "programability" of MediaWIki templates then I will do it that way. As to the two char issue, there are going to be three char ones aswell. If anyone writes an article about a well known game they will probbly want to use stones above 10. I have seen them up in the hundreds in some books (though I am not going to make that many unless I script the whole thign with GIMP and Perl including the upload. But, it is likley that you will see them above 10 (infact I might do a few more myself though not before I get the non-number ones all done. Dalf | Talk 05:18, 9 May 2005 (UTC)
Finished uploading the duplicate images - you can use two letter invocations and have a perfectly aligned editor board now. :D Found to my pleasnt surprise I don't have to upload a duplicate of the empty position images, since more than one whitespace is ignored. -- Natalinasmpf 00:24, 10 May 2005 (UTC)
I wish Image redirects worked ..... would make this all much faster. Dalf | Talk 00:48, 10 May 2005 (UTC)
Okay, trying to create a two letter invocation board to demonstrate. -- Natalinasmpf 01:41, 10 May 2005 (UTC)

[edit] Help with Triangle

So the combination of my limitations and MS Paint's limitations have finally defeated me. I cant seem to get a triangle inside these so that it fits and looks nice. This is pretty commonly used in Go diagrams so it would be nice if we could get it. Can you do any one of the stones with an equilateral triangle in it and upload it (use the uppercase T). From there I can cut and paste etc and get it into the others (I knew I should have installed the GIMP). Dalf | Talk 05:53, 9 May 2005 (UTC)

[edit] Template Management

[edit] List of Templates

A list of templates for management of changes, correlations or modifications.

[edit] Bugs

Record bugs (for any template here).

Discovered a bug, not really with the template, but with the way we invoke it. There should be only be one space between the -- and the next "|" character. The problem is that for now the template ignores the extra whitespace, but see " x" or " b" as "_x" and "_b", so, this should be fixed. -- Natalinasmpf 22:23, 8 May 2005 (UTC)

[edit] Error in template

Go diagrams never use the letter I. I removed it. Gkhan 23:20, May 8, 2005 (UTC)

Why doesn't it? If so, can you do this for all the other templates, as well as the examples, etc. (agh, I realise maintaining this is a pain.) Also concerning the one letter system, I made that mistake of not discarding the second letter format used in the chess template, should have kept it. -- Natalinasmpf 00:02, 9 May 2005 (UTC)

I honestly have no idea. But it is indeed so, download any Go-SGF reader you want to verify. And I will remove from the other templates, I just forgot :P Gkhan 00:19, May 9, 2005 (UTC)
this is because in MANY fonts the letter i and l and the number 1 all look the same. Dalf | Talk 00:32, 9 May 2005 (UTC)

Oh, also do this for the rest of the pages that use the template (the manual, etc.), if you can hunt them down. -- Natalinasmpf 01:21, 9 May 2005 (UTC)

[edit] No comment tags needed

I would use dummy parameters between rows, as Sverdrup suggested, demonstrated at User:Sverdrup/test4. It requires renumbering the parameters in the template.

Alternatively (here shown for chess):

{{chess position|=

 |rd|nd|bd|qd|kd|bd|nd|rd|=8
 |pd|pd|pd|pd|pd|pd|pd|pd|=7
 |  |  |  |  |  |  |  |  |=6
 |  |  |  |  |  |  |  |  |=5
 |  |  |  |  |  |  |  |  |=4
 |  |  |  |  |  |  |  |  |=3
 |pl|pl|pl|pl|pl|pl|pl|pl|=2
 |rl|nl|bl|ql|kl|bl|nl|rl|=1

|30}}

gives


8
7
6
5
4
3
2
1
a b c d e f g h

(where the numbers 1-8 are just an example, not needed) because the "=" assigns a value to an unused parameter (its name is the empty string), which has no effect.--Patrick 15:10, 19 May 2005 (UTC)

Renumbering the template? All 362? *dies* (my other concern was what about the image size parameter - the board looks so big for chess, never mind what could happen with Go). -- Natalinasmpf 19:01, 19 May 2005 (UTC)

Nah, you don't need to renumber the template. That's the point with the "=" signs. --jacobolus (t) 19:48, 19 May 2005 (UTC)
Actually, I think this is enough cleaner that it'd be worth it to update the "manual" to reflect the version with the "=" signs. --jacobolus (t) 20:40, 19 May 2005 (UTC)
{{Game of Go Position|=

 19 | | | | | | | | | | | | | | | | | | | |=
 18 | | | | | | | | | | | | | | | | | | | |=
 17 | | | |b9| | | | | | | | | | | | | | | |=
 16 | | |a|w| | | | | |x| | | | | |x| | | |=
 15 | | |b|w| | | | | | | | | | | | | | | |=
 14 | | | | | | | | | | | | | | | | | | | |=
 13 | | | | | | | | | |w| | | | | | | | | |=
 12 | | | | | | | |w|b| | | | | | | | | | |=
 11 | | | | | | | | |b|w| | | | | | | | | |=
 10 | | | |x| | | | | |x| | | | | |x| | | |=
  9 | | | | | | | | | | | | | | | | | | | |=
  8 | | | |w|b| | | | | | | | | | | | | | |=
  7 | | | | |w|b| | | | | | | | | | | | | |=
  6 | | | | |b| | | | | | | | | | | | | | |=
  5 | | | | | | | | | | | | | | | | | | | |=
  4 | | | |x| | | | | |x| | | | | |x| | | |=
  3 | | | | | | | | | | | | | | | | | | | |=
  2 | | | | | | | | | | | | | | | | | | | |=
  1 | | | | | | | | | | | | | | | | | | | |=
     a b c d e f g h j k l m n o p q r s t

|20}}

gives:


19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
a b c d e f g h j k l m n o p q r s t



Ok - go ahead - but do it for all versions, and fix the wording about "comment tags" so that it gets modified for use with the equals sign. There are also many pages, such as Template:Weiqi-image and Template talk:Game of Go Position (this page was actually an informal manual) which promote the previous method as a way to invoke the template which has to be updated. I will have to update most of these myself (including my own page) to correspond with the new measures, and perhaps also for the two letter invocation boards as well. -- Natalinasmpf 20:54, 19 May 2005 (UTC)

[edit] Number and Kifu Problem

By using this template, it will be very easy and convenient for us to record a Kifu (棋谱). However, to make a Kifu, we not only need to add "w" and "b", but also need to add numbers like b27, w34, b235, etc. And the current template does not support such format. I found that when a number bigger than 10 is added, the go board will be split. This causes difficulty in recording Kifu. Can we work out a solution? Your help will be greatly appreciated. --Neo-Jay 21:02, 27 January 2006 (UTC)

Ultimately, when the Mediawiki software upgrades, I might think of a proper mechanism for replacement (I'm currently contemplating programming a template that will use template:qif...it will take a lot of work to develop a 3-digit template using the current template software (note I how I've already overextended the original Template:chess position concept!) I think currently having to split it into new diagrams and new captions would be better - we can tackle stuff 9 at a time; would this be troublesome? Elle vécut heureuse à jamais (Be eudaimonic!) 21:56, 27 January 2006 (UTC)

Many thanks for your reply. Best regards for your work. --Neo-Jay 22:13, 27 January 2006 (UTC)

[edit] Four-Quadrant Coordinate System

I create a 4-Quadrant coordinate system and just modified the template to reflect this version. I think it is much better than the current 1-Quadrant coordinate system. The new one divids the Go board into four parts (North-East, North-West, South-East, and South-West), not simply one. It is easier to remember and, especially, compatible to the traditional Go number system (3-3, 4-4, etc). For instance, under this 4-Quadrant system, readers will be much easier to understand the "Double 4" fuseki (see an example at Takemiya Masaki). This 4-Quadrant coordinat system can be our Wikipedian (i.e., Wikipedist) version of Go Board Coordinate System. We can promote it to the world and replace the current various coordinat sytems for the game of Go. What do you think of it? Your opinions are greatly appreciated. Thank you very much. --Neo-Jay 21:24, 26 January 2006 (UTC)

It looks good, although I'm not sure about the conflict between notation systems (especially since the transclusion context (ie. the standard way to copy and paste a board template) might have to be modified? Most of my games involve alphanumerical notation...is there a standard way to notify this? Could we perhaps offer a range of options?

I was thinking of using template:qif, even... (that way, alphanumeric would be invoked by default but it could defer to quarter). Good suggestion though; is this a standard practice in your geographic location? Perhaps it just deviates? The 4-4, 3-3, thing has never been a huge problem for me using 19x19 notation...I just think of it as the amount of areas away from the board.

Meanwhile, you can help me finish the images that aren't converted in as seen in Template:Weiqi-image. There are two nightmares I have to deal with in addition - uploading this to commons, and uploading this to http://games.wikicities.com (which has a Go area but as I don't participate often and being one of the initiators of the templates apparently it has stagnated a bit *gulp*) ... then there's the whole issue of having 350+ different images for Template:Xiangqi position (possibly tens of thousands if we want arrows and diagrams)...well, that's just my rant anyway. My current priorities are to finish that, but I often don't have time...what is needed is just to open those PNG images (for the numbers) in paint and add and remove the liberty lines as needed (by eyedropping the yellow blackground and filling the lines with them). Elle vécut heureuse à jamais (Be eudaimonic!) 00:41, 27 January 2006 (UTC)

Many thanks for your comments. I must admit that this 4-Quadrant system has never been used in any geographic location before. I create it and think that it is much better than the curent existing various systems. However, if you think that it is important for us to adopt the mainstream version, please feel free to change the 4-quadrant system back to the former alphanumerical system. Thanks. --Neo-Jay 01:23, 27 January 2006 (UTC)
And sorry that I do not understand how to convert the images. What sofeware do I need to have? What is liberty line? How to "eyedrop" and "fill the lines"? --Neo-Jay 01:26, 27 January 2006 (UTC)

One of key weaknesses of the alphanumberical system is the incoherence of the use of alphabet. Some use A-S, while some omit the letter "i" and use A-T. To omit "i" is becasuse it is worried that "i" may be confused with the number "1". However, the letter "l" can also very likely be confused with the number "1". Accordingly, when using the alphanumberical system, the user must be clear what kind of alphabetic system is used. --Neo-Jay 01:55, 27 January 2006 (UTC)

I know I am a little late to the game here, and I know this is on some level a display and layout issue but I think Wikipedias polocy about no original research applies here. Unless this is a coordinate system that is not only in use but accepted in general then our implication that it is correct (by using it) and especially our promoting it is not appropate. Personally I think there are a lot of shortcommings to the systems that are activly used but wikipedia policy is pretty clear on this. I think we shoudl revert. Dalf | Talk 06:06, 27 January 2006 (UTC)
I disagree. The 4-quadrant system is not an article in Wikipedia, but a technical solution we use to display go board. Many, if not all, technical solution of Wiki including this template itself are Original. If we establish an article to introduce Go board coordinate system, I absolutely agree that we should not promote this 4-quadrant system but should introduce the existing main versions. However, what we are doing is to find a way to display Go position and play Go in The sandbox. I do not think that use of this coordinate system in the template violates wiki's rules. And we should note that even the previouse coordinate system in this template also omits the letter "i". It is not a mainstream practice. Frankly speaking, there is no mainstream coordinate system at all now. --Neo-Jay 06:29, 27 January 2006 (UTC)
Once the template is used in an article it becomes part of the article and is an implicit clame of correctness, it is also not just how we display the board but what we are displaying that matters. Have a look at the maps used in articles, they are all subject to the same policies. Further in as much as users will want to use wikipedia to learn about Go, the incompatability with all published content and all the content found on the web is problematic. If you check the Go litrature and the online servers there are one or two standards that they all use. If you look at Go board manufacturers or offical tournaments they have standards too (though the boards dont frequently use lables at all). Dalf | Talk 06:57, 27 January 2006 (UTC)

I just added the template to explain the Go board position. This shows the great advantage of the 4-quadrant system. It so clearly identifis the Hoshi (星) with (4, 4), Komoku (小目) with (4, 3) and (3, 4), Mokuhazushi (目外) with (3, 5) and (5, 3), and Sansan (三三) with (3, 3). It may not be a big deal for a high-level player, but it is perfect for a beginner. --Neo-Jay 06:42, 27 January 2006 (UTC)

I am only mostly of the opnion that this is Original Research (which is why I did not revert) so I posted a question here [2] I think I make my objects a little more clear there so go have a look. Dalf | Talk 07:09, 27 January 2006 (UTC)
Many thanks for your considerations. It's so nice of you. OK. I understand your position. And I checked many main Go websites, including American Go Association, Japan Go Association, Sensei's Library, Go Base, etc. I does not find any of them uses coordinate system. Can it prove that the "mainstream" and "correct" dispaly of a Go board is no number and no alphabet at all? If so, to aviod users' taking our alphanumeric system as "an implicit clame of correctness", we must not use any coordinate system. Correct? --Neo-Jay 07:28, 27 January 2006 (UTC)
Hey, I am editing in good faith here I coudl have just reverted the changes rather than discussing them. I suspect that the American Go Association does record games on paper or on computers as well as the others. The online ones use a file format that uses a coordante system similar to the one we had. KGS uses one almost exactly like the one we had. Dalf | Talk 08:42, 27 January 2006 (UTC)
Thanks. I know that you are in good faith. I apprecaite it very much. You said: "My consern is that if I buy a Go book or a report of a recent tourny I will be given the games written in some system that will be diffrent from this one." [3] I do not think that you can buy such a book. For almost every Go book discusses Go position like 3-3, 4-4, 3-4, etc. No book really introduces Hoshi as d4, p4 (or q4?), j10 (or k10?, l10)... and Komoku as c4, e3, r16, p17... and Mokuhazushi as ..... Actually, such alphanumerical system is also not used in the report of Go games. Have you heard anyone talking about Go game with such a confusing system? I checked the KGS website and have not found the webpage using a coordinate system. Instead on its Introduction page, there is no coordinate syste at all. Why? Just for the alphanumerical system is extremly confusing. So, if we really do concern not buying a confusing book, we must not use any current coordinat system at all. Thank you. --Neo-Jay 09:14, 27 January 2006 (UTC)
Ok, I will drop the point, though I suspect that someone else will be bold like you (though it may be months and months) and change it back so you should be ready with the same argument then. Dalf | Talk 09:30, 27 January 2006 (UTC)
Thank you for your discussion. You are very nice. I kowtow to you. (OrZ) --Neo-Jay 09:38, 27 January 2006 (UTC)

I also found a coordinate system that omits letter "J" [4], not "i" (which the pervious Wiki coordinate sytem omits), and not "l". So, which system is exactly correct? I think the only correct solution, as I put above, is No Coordinate System At All. Thanks. --Neo-Jay 08:06, 27 January 2006 (UTC)

Another argument is this. Wikipedia is an encyclopedia. But Wikipedia CAN have its own Rules of Layout, even different from those of other encyclopedias. Then the coordinate system for Go board can also been seen as a layout rule in Wikipedia for Go position. Why cannot it be original? --Neo-Jay 08:30, 27 January 2006 (UTC)

My consern is that it is not a presentation issue but a content issue. Layout rules are pure presentation, ways to convey information. I understand your position that the position of the stones is the only information being presented, and you may be correct. I am not 100% comfortalbe with it (yet), though I do think your system might be easier. Would be intresting to see a game notated in it, and that is where my problem comes in. The coordinate systems are the basis of the recording systems used to record games, if we use a diffrent one then we add a barrier for other editors to contribute or use wikipedia allongside other sources. The other systems as you ahve pointed out are far from standard but they are a lot closer to eachother than this system is to any of them. I wonder why no one has ever used a system like this before? Dalf | Talk 08:38, 27 January 2006 (UTC)

My concern is the current invocation/transclusion system is based around alphanumerics, and other than a need for a total overhaul it would also make existing citations of the template difficult to follow. Perhaps a separate template? (If only there was an article for standard Go notation - like algebraic notation, or something). Elle vécut heureuse à jamais (Be eudaimonic!) 22:00, 27 January 2006 (UTC)

Thanks for your concern. I do not see many articles cite the alphanumerical system. On the contrary, many articles use such 3-3, 3-4, 4-4, system, which can be better displayed by this 4-quadrant system. See e.g., Go terms, Fujisawa Hosai, Takemiya Masaki, Go opening Theory in "Go Strategy and tactics, Basic joseki in "Joseki"... Especially see the article of Avalanche joseki. It originally introduced the joseki by the alphanumerical system. But it was changed to the symmetrical notation system at 03:53, 19 May 2005 by Dalf because Dalf concerns that "this joseki can be played in any corner." This again showes the advantage of the 4-quadrant system. It DOES conform to the customary citation system for Go. To record a whole game, usually the alphanumerical system is also not used. See the article of Kifu. We can find that the Kifu is customarily recorded in a grid diagram. To display the Go in an alphanumerical system is a Western idea. And it "did not catch on, ... While a typical piece of chess literature is in algebraic notation punctuated by occasional diagrams, go literature mostly consists of diagrams with a sequence of plays marked, and prose commentary" (See Kifu). Given this situation, I argue that use of any alphanumerical system for Go board position will violate the tradition and custom of Kifu. Then choosing a proper coordinate system for Go position should be an issue of technical solution or layout, and subject to the full discretion of Wikipedia. And among all the options, the 4-quadrant system is, in my view, still the best one. Thank you so much for your consideration. --Neo-Jay 22:43, 27 January 2006 (UTC)
Why is the 1-quadrant alphanumerical system universally adopted for Chess, but not for Go? Two key differences between these two games matter. Firstly, Chess is a moving game, while the Go stones do not move once they are placed on board. Accordingly, it is easier for Go Kifu to be simply recorded in even a single grid diagram, which is unimagable for chess. For a long time, there has been no strong incentive to promote Go people to develop a mature solution for the coordinate system. Go players simply talk about Go by 3-3, 4-4, 5-4, 3-9, etc. Secondly, there are only 8 numbers for every side of a chess board, while 19 for Go. To put the whole Go board into one quadrant will make it very difficult for players to remember. So go players traditionally conceptualize the Go board into four equal parts. Thus the 1-quadrant alphanumerical system totally disorder go players' customary perception. That is why although Western people tried to apply chess' alphanumerical system to Go, they fail. Now the 4-quadrant system, which can also be seen as an alphanumerical system, provides a perfect solution to "alphanumerize" the Kifu. --Neo-Jay 23:39, 27 January 2006 (UTC)
I still think it should be an option. The quadrant system sounds good for noting concepts, while the a1, b2, etc. system seemed okay for notating games...Elle vécut heureuse à jamais (Be eudaimonic!) 02:22, 28 January 2006 (UTC)
Thank you for your comments. Yes, the a1, b2, etc. system is indeed Okay for notating game. But the 4-quadrant system is also Okay for doing that. And we should note that neither of them is the mainstream solution for Kifu, which is typically recorded in grid diagrams without any coordinate system. If we really want to have a coordinat system, we DO have discretion to choose a better one. Thanks and Happy New Year. --Neo-Jay 02:38, 28 January 2006 (UTC)


[edit] Breaking pages

This modification to the template breaks articles already using the template. If you want to use another coordinate system, please create another template, but don't mess with a template which is already in use.

I am going to revert the template to the classic type - the one we started to use. --Panairjdde 23:05, 28 January 2006 (UTC)

Can you identify which specific articles are broken?? I have not found one. On the contrary, your reversion breaks many articles like Go terms, Fujisawa Hosai, Takemiya Masaki, Go opening Theory in "Go Strategy and tactics, Basic joseki in "Joseki", etc. ... As I said, it is the so-called classic coordinates that is inconsistent with the notation system used in many articles. Wiki should be a unified system and find a best solution. And if your reason is right, then those templates which have been used in articles shall not be modified anymore. This violates the key ideal of Wiki: a living system to which every user contributes everyday. Your reason will make Wiki a static and even dead system. Please provide your claimed broken articles EXCEPT the sandbox games. And let's find whether we can fix such problems. Otherwise, I will revert it again. --Neo-Jay 03:47, 29 January 2006 (UTC)
You took a template with a corrdinate system, used it in articles adopting another coordinate system, and then, once you discovered they were different, you changed the template. Why? Why not using a template suitable for your needs, instead of changing another? And what's the problem with sandboxes? If, for resons I do not understand, you need a different coordinate system from the one used, for example, in SGF files, why don't you make your template, instead of changing the one already existing? I understand you support the coordinate system used in China, but please, bear with wath is used in the rest of the world. --Panairjdde 21:42, 29 January 2006 (UTC)
I think that I have made my position very clearly. Wiki's Go articles need a UNIFIED solution, not SPLIT into DIFFERENT systems. And I have also put it very clearly, your argument will make Wiki become a STATIC AND EVEN DEAD system, not a LIVING one. We are seriously discussing which solution is the BEST coordinates. And my new system DOES NOT break any article. And you have not yet answerred my key question: WHAT SPECIFIC ARTICLES DO YOU THINK ARE BROKEN by my system?? If you DO NOT UNDERSTAND why we need a different coordinate system from the one used, TRY to take some time to read the relevant discussion here. And if you still just keep saying LET THE OLD TEMPLATE ALONE, and cannot provide SUBSTANTIAL reason why my edit is BAD, then I will revert it again to my system. --Neo-Jay 22:00, 29 January 2006 (UTC)
  1. stay calm
  2. the article broken is the sandbox game
  3. if the old template was really not consistent with the articles you cited, it was the fault of those using the template in the wrong way
  4. your coordinate system needs three variables to indicate one point on a plan, and this is wrong
  5. you can follow the josekis on a single quadrant of the goban, if you need to stick with the 4-4 convention
  6. when you don't want/can't use a kifu to record your games, the 1-19/a-t system is used. example: SGF.
--Panairjdde 22:18, 29 January 2006 (UTC)
  1. Thanks. I do and I will.
  2. This means that you admit that no article is broken, but just sandbox game. These two concepts are extremely different.
  3. Those users have not used the Template in the wrong way. The template itself is a wrong way. Those users just follow the traditional and customary way to talk about Go. And if we agree that we do need a unified coordinate system, the 1-19/a-t coordinates will inevitably cause confusion.
  4. My system needs two, not three, variables: horizontal and vertical. This is same as the 1-19/a-t system. For example, the point t19 is (probably, depends whether you omit "i", "j", "l" or not) N1E1 in my system, which is much more visual. Although it looks long, it is not wrong. I do not understand why you call it as "wrong".
  5. A single quadrant of the goban also needs a coordinate system, to which the 1-19/a-t obviously cannot provide a good solution, but my system can.
  6. When I don't want/can't use a kifu to record my games, the 4-quadrant coordinates can also be used, and, more importantly, it is a better solution.
Thanks for your serious discussion. I hope we can continue. --Neo-Jay 23:07, 29 January 2006 (UTC)
I just want to collaborate with you to build a better Wikipedia.
As far as I know, they are both on Wikipedia. But let's talk about "good coordinate systems".
The template has a 1-19/a-t system, you (they) use it with 4-quadrants coordinates and the template is wrong? Why did not you (they) use a different template in first place?
Because you need to indicate 4 variables [N/S][1-9][E/W][1-9]. This can have a meaning in a spherical coordinate system, such as in latitude/longitude, but in a two dimensional space you need only two coordinates!
Why dont' you use the interval 1-10/a-k for a quadrant?
I frankly don't see how it is better to write S3W4 instead of 3D.
Same here. Best regards.--Panairjdde 10:13, 30 January 2006 (UTC)
You said: "Why did not you (they) use a different template in first place?" Let's clarify what "In the first Place" means. I have make it clear once and once again. I make it here for the Nth time: Go people have used the four-part notation system to talk about go game for THOUSANDS OF YEARS! Just since the 20th century (or the end of the 19th century), some Western people have tried to apply Chess' coordinates to Go. And they FAIL (see Kifu, and also the "Go West" section in "History of Go"). The current custom of notation system is still 4-4, 3-3, 4-3 system, and the Kifu is still recorded in grid diagrams without any coordinates. Those people who have promoted the chess-style coordinates just think of Go board in the same way as they think of Chess. Obviously they have not really understood the characteristics of Go. And we can clearly see what is at the FIRST PLACE, and what is at the SECOND. Actually, it is the 1-19/a-t that is the 2nd place and plays as an invader who breaks the custom and causes inconvenience. That's why I call it as a Wrong Way.
You said: "you need to indicate 4 variables [N/S][1-9][E/W][1-9]..." You had said that I need THREE variables, now you say it is FOUR. And I notice that you again think it is THREE below (and call it as "insane"). Can you clearly count the insane number firstly? But let's just talk about this 4 (?) variables. I have to remind you that the North-South-East-West system had been used long before human beings agree the earth is round. Such system has absolutely been used in two dimensional space. To identify a center point and describe other points with N-S-E-W is much more clearly than the two (?) coordinates. My system takes Tengen as the center point and marks it as (0,0), which perfectly reflects the meaning of Tengen (Center or Origin of heaven), and describes the four corners as North-East, North-West, South-East, and South-West. This is so easy to understand once you understand the map of your city. Have you seen anyone talks about a city map with two (?) coordinates? That's why I argue that my system is easier to remember and more clear than the 1-19/a-t system.
Let's talk more about the advantage of the 4-quadrant system. Please see the "Two-dimensional coordinate system" in "Cartesian coordinate system". You will find that the 4-quadrant system is also used for two dementional space in mathematics. I explictly admit that my idea exactly comes from this two dimensional cartesian coordinate system. My original though was to simply put the Go board in such cordinates and to mark every point like (-1, 3), (9, -9), etc. I do not know how you count the varibles in this system. Maybe two (if you think "+" and "-" are integrant parts of the numbers). Then you may not call it insane. Or three or four (if you think "+" and "-" is, or are, also varible(s)). Then it is still insane. But anyway I make two improvements to this system: 1) use N-S-E-W to replace the "+" and "-"; 2) reverse the order of numbers. I call them as "improvements" becasuse I think that 1) North-South-West-East is more visual than "+" & "-" and does not have sequence requirement, 2) to upgrade number from side to center is consistent with go players' custom. Do you still think it as insane?
You said: "Why dont' you use the interval 1-10/a-k for a quadrant?" The 1-10/a-k can only rightly describe the lower-left corner, while my system can perfectly describe all the four corners. Which one is better?
You frankly don't see how it is better to write S3W4 instead of 3D. I once again ask you to think of all the Four Corners, not just one. Probably you may find that I am not so "INSANE".
Thanks for your comments. I think that you are serious although you called my system insane. But, please DON'T insert your comments into mine again. We are discussing about "Breaking Pages". Please don't break other ones' words first. Thank you. --Neo-Jay 21:36, 30 January 2006 (UTC)


  1. I won't mess anymore with your "holy" comments. Let me notice, however, that you modified my comments, by removing them from where I put them.
  2. I would like to let you notice that if you want to use Chinese coordinate systems, you can do it in the Chinese WP. The fact that the four coordinate system is used in China is not a good reason (alone) to oblige people to use it on English WP.
  3. My sentence "Why did not you (they) use a different template in first place?" meant: "You used this template, which adopted the two-coordinate system, in some pages where, you say, you adopted the the four-coordinate system. You should not put the blame of this inconsistence on the template, but in the wrong use you made of it." This is no reference to Eastern/Western way of recording kifus.
  4. I repeat that it is insane to use coordinates to fix a point on a two-dimensional (and please, note the "two" in two-dimensional, it is there for some reason) and limited plane. It is useful to use N/S and W/E when (a) you have an unlimited plane or (b) when you have a spherical surface, in which is uneasy to fix a "central point". Now, (a) is the case of the Cartesian system you cited, in which the axes extend to infinite in both directions; this is not our case, since each goban has a finite number of rows and columns. (b) is the case of latitude/longitude spherical surface coordinates, in which a convention for the origin was adopted; this is not our case, since we are dealing with a finite, squared, planar surface. So the use of four coordinates in our case is "insane". At least if you are not trying to push a traditional system in this context, in which we would like to select the "best" system. Oh, and as regards city map coordinates, yes, my city map plan says something like "Mulholland drive, A-9", to indicate the streets in the index. Go figure!
  5. 1) The custom of counting from the centre can be common in your country. Do not extend this assumption to the world. 2) If using the tengen as origin of the coordinate system requires using four coordinates, let's use the corner as origin, since it requires only two coordinates!
  6. As regards the coordinates of a single quadrant. We have to define the problem here. I think we are trying to decide which system is better to define a single quadrant, in particular when we want to talk about josekis. Since there is a single quadrant, either system is good: in four-coordinates system you drop the indication N/S-W/E, to write 4-4 instead of N4W4; in two-coordinates system you put the stones in the lower-left corner. Where is the difference, when there is a single quadrant?
  7. I think that your system (not you) is insane, in the sense that you don't need four coordinates to fix a point on the goban (as I think I showed). And I do not see why we should adopt a four coordinates system that is actually made up of four two-dimensional coordinate systems, each with origin in the tengen and rotated towards each corner, as you propose.
  8. I think my comments would have been better if interleaved with yours, but I hope I made myself clear, despite this problem.--Panairjdde 11:40, 31 January 2006 (UTC)
  1. I am glad that you won't mess anymore with your humble comments. If what you noticed makes you feel uncomfortable, you should have not done the same thing to other ones first.
  2. It is not only Chinese who talk about Go by 4-4, 3-4, 3-3 etc. Those contributors adopting this notation style come from all around the world, including English speaking areas. And More Importantly, English Wiki is for All the persons who can read English, not only for those Englishmen or English native speakers. And ironically, I am afraid that even you, an Italian, don't belong to these two groups. English Wiki absolutely is also for Chinese users, just as it is also for Italian users.
  3. Those contributors, following customary way, have described the Go position in the first place. They also need a Go position template to present what they are talking about. It is an inherent demand to add a Go board diagram for them. If we agree that we need a unified solution for Go board template, and if we adopt the 1-19/a-t system, then we will face the dilemma: using it will conflict the customary notation and confuse people, not using it will casue inconvenience. Why cannot we adopt a system that conform to the customary notation system?
  4. I don't know how you define "limited" and "unlimited" plane. Let's take Italy for example. I think the territory of Italy, even including its peripheral areas, is limited. Let's see Section "Geography" at Article "Italy", it says something like "...the Adriatic Sea to the north-east, the Ionian Sea to the south-east, the Tyrrhenian Sea to the south-west and finally the Ligurian Sea to the north-west". It takes Italy as the center and describe other areas by North-South-East-West system. This is exactly the CUSTOM of human beings. And PLEASE NOTICE that such description is absolutely for a "limited" region, and has nothing to do with a spherical surface (although Italy does lie on a globe, such N-S-E-W system had been used long before human beings found the earth is a globe). Regarding your map with "Mulholland drive, A-9", I also have seen such map in the U.S. They do exisit. However, the standard changes from map to map. When you talk about it to another person, you have to have a map in hand. That is not a unified solution that Wiki needs. And such "Mulholland drive, A-9" also cannot replace the people's custom. In ordinary life, people are still talking about Rome with north, south, west, east (just count how many times these words appear in the Article Rome and how many times something like A9 appears). If we can create a Go board coordiante system that conforms to people's custom, why not?
  5. 1) As I put it, talking about Go position by 3-3, 4-4, ect is not a custom of only Chinese. It is a worldwide custom. 2) You can try to ask other Go players to adopt your so called two coordinates. You are not the first one. And I have to remind you that they have FAILED (see kifu). The reason is exactly because your system breaks Go players' custom (not only Chinese).
  6. OK. Even though we can describe josekis permenantly in the lower-left corner in the future, there are still two key problems. One, the custom is to say 4-4, 3-4, not d4, c4. Two, we still may have to introduce a part of a real game, which may happen in other three corners. If you think that there is no difference between the two coordinates at least "when there is a single quadrant", please think the cases when we have to use the other three corners.
  7. I think that I have made it clearly enough. The custom is to divide the Go board into four corners just because it is EASY, not difficult, to follow. DON'T think Go board is another chess board. Their sizes are so different that Go players have adopted a different notation system.
  8. You see. You do make yourself clear although you does not insert your ideas into mine. Many thanks. --Neo-Jay 19:41, 31 January 2006 (UTC)
Its one out of three: either I am not capable to make myself understand, or you do not understand, or you do not want to understand. Do as you please.--Panairjdde 01:53, 1 February 2006 (UTC)
I do not think content or formatting decisins should be made on the basis (or even giving consideration to) non-article space. Personally I think people playing Go or Chess on wikipedia is a waste of server resources and kinda silly since there are actual servers online which make the playing so much easier. But that is beside the point, given that there appears to be no classical convention and given that msot of the articles that need boards are discussing joseki (and therefore use the number-number convention. The new coordinate system is actually better (though I would not call the articles broken either way). We could also compromise by changing the template so that the coodinates are also provided (and omitted if not included) and then provide some blank cut and paste examples on this page with both. Dalf | Talk 22:26, 29 January 2006 (UTC)
Thank you. I absolutely agree that using this template in Sandbox game is a waste of server resources. Hope we can come to consensus to build a better coordinate system. --Neo-Jay 23:24, 29 January 2006 (UTC)
There is no need for a new coordinate system. And the one you proposed is not better. I already wrote that using three coordinates to distinguish a single point on a plane is insane: you need only two coordinates, in a two-dimension space! -Panairjdde 09:59, 30 January 2006 (UTC)

[edit] After reading the article on Kifu I withdraw most my objections to the new system

It does look like most of the problems associated with consistency and methods for recording Go games will not be made worse by our adding to the mess. I do think that designing either two templates or figuring out some way to switch between the two (which would probably involve meta templates and would annoy a lot of people who are on a religious crusade against such things). Having algebraic system available does have some value for the future but I checked and this template is so rarely used that this change will not mess up anything in article space (there are some people playing games on talk pages who might be effected if they are noting the moves below the games in algebraic, but that is not important to the article space). For the most part I think that this issue is more the theory of what it right rather than being a practical problem. The Go articles on wikipedia are such a mess at this point that I don't think it really matters. Plus, as with the Avalanche joseki most articles that will need a Go board will not need the whole thing and the shapes and sizes they need will be so varied that the best solution is usually to do them custom anyway. This lets the individual editors pick. Dalf | Talk 22:37, 28 January 2006 (UTC)

Many thanks for your comments! And I think that even for those articles that do not need a Whole Go Board, the 4-quadrant coordinate system will also provide the perfect solution. If possible, we can design a template with the N-S-W-E coordinates on the four sides and users can choose to use the Whole Board or just a Corner of it. Otherwise we can establish different templates, one for the Whole board and four for the four corners. Again, all of them use the 4-quadrant system. Then when just a corner appears in the article, readers will also find a extremely friendly coordinate system. What do you think of it? Thank you. --Neo-Jay 04:27, 29 January 2006 (UTC)
I had toyed with this idea for a while. Even giving some thought to how far we can streach the limits of the (vvery limited) templating features. The truth is there are not a lot of Go articles on wikipedia and most of them need very significant work. We have three pages devoted to go terms and concepts none of which are complete and sometimes link toeachother for explination of terms that neither of them cover (circural links). Other articles that are just bad, and few that are in a state where the need for these tempaltes would make a significnat diffrence. As such I think if we want to focus our time on Go we should start with the articles and then create the templates as there i need (and by need I mean the same arrangment in more then three articles). I dont think indivudual quadrants need any lables and I think they woudl be made owrse if they had them, which is why I did not include them in the diagram on Avalanche joseki. What I think is really the problem here is that there are a number of other wiki or wiki like places on the web specifically for Go. The best contributers are all working on Sensei's Library which has better tools for doing Go markup and much mroe completness. We might investigate what the copyright status is on thier content, obviously the classic positions are not copyrightable but the commentary is what makes them useful. I am only like a 16 or 17 kyu player (if I am even that good) so there is very little that I can add and feel comfortable about.
So the short answer after sayign all that is, I think we might do better to fucus out attentions elsewhere (though doing more numberd stones for the templates might be good, to enable full Kifu. Dalf | Talk 09:16, 29 January 2006 (UTC)
I will take back a bit of the above and say that if you/we/anyone is planning on doing some articles on some joseki or tesuji doing a templates of a corner area (probbly not a full quadrant) or even a series of tempates that can make boards of diffrent sizes somethign like Template:Go board (start) Template:Go board (row) or whatever, might be a good idea. If we were to for example complete the article on Nadare joseki we woudl need at least 3 or 4 more diagrams of the same deminsion. Dalf | Talk 09:26, 29 January 2006 (UTC)
You are right. We do need to improve the articles first. I also think that the coordinate system may only play as a supplementary role since there is no concesus about the coordinate system. AND a purely algebraic notation system without any diagram is also confusing. A user must be clear what coordinate system is used ("i", "j", or "l" is omitted, or no letter is moitted at all?). So I think the best solution is to have a template with a Better coordinate system and in the meantime to provide the algebraic notation. Then even if we cannot put the numbers bigger than 10 into the board position, the algebraic notation can also clarify the series of moves. Thank you for your serious discussion. I appreciate it. --Neo-Jay 10:00, 29 January 2006 (UTC)
Why can't we do numbers larger than 10? I do not see any reason why we would not be able to. Also I think that in many cases no coordanate system is the right solution (but not always). Dalf | Talk 10:37, 29 January 2006 (UTC)
I also noticed this problem (See 3.5 "Number and Kifu Problem" in this talk page). And Natalinasmpf said that this could probably be solved "when the Mediawiki software upgrades". Such a template will absolutely be very helpful. --Neo-Jay 19:17, 29 January 2006 (UTC)
I do not think this is a problem with the mediawikisoftware it passes the argumnents the same regardles of their length. Even if it were a problem with the wiki table layout short hand we could convert to actual html tables and if necessarry (which I dont think woudl be) request changes to MediaWiki:Common.css. I will try and get this working today. Dalf | Talk 20:41, 29 January 2006 (UTC)
Perfect! Best wishes! --Neo-Jay 20:53, 29 January 2006 (UTC)
See above, it works fine though if you did a while Kifu it woudl require a lot of images and the character lengths woudl get messy. [[User:Natalinasmpf|Natalinasmpf]'s solution to all of this is to rename the images so that they are all constant width and the bug she refered to is that you cant put whitespace in the template and have it ignored. That is unlikley to be fixed as I doubt many people see it as a bug. But, so long as you dont mind the source version looking a bit messy there are easy. I will finish 10 later today (looking for an upload tool I saw a while back that makes it faster) Dalf | Talk 21:02, 29 January 2006 (UTC)
No reason other than someone needing to upload the images. What font are we using for the numbers now days? And I notice that the white stones are no longer actually white but more of a grey. Dalf | Talk 10:53, 29 January 2006 (UTC)