Hey,
I have my first test result to add to the WIKI folks started...and am not sure how to add it. Could someone message/email me if there is already a way/format established for doing so? If not I guess it is worth thinking about it.
The result is not very exciting (tried no adding to history if beta - alpha <= 1 ... it worked terribly) but I thought I would use it to learn how to contribute to the site.
-Sam
First result contribution
Moderators: hgm, Rebel, chrisw
-
- Posts: 1922
- Joined: Thu Mar 09, 2006 12:51 am
- Location: Earth
Re: First result contribution
I was replying to a post by Jaap Weidemann, but it seems it disappeared.
Anyways, the wiki in question is http://chessprogramming.wikispaces.com/. It is still undergoing a lot of progress, but it seems its only from a select few. Everyone is encouraged to come and contribute! There are still a lot of holes, mostly in advanced search/evaluation theory.
Anyways, there is not really an established format for test results yet. Probably the cleanest way would be to create a new page for each idea's test results, linked to from the main page.
It's interesting that it's taken so long for test results to pop up, as that was the intended purpose of the wiki in the first place...
Anyways, the wiki in question is http://chessprogramming.wikispaces.com/. It is still undergoing a lot of progress, but it seems its only from a select few. Everyone is encouraged to come and contribute! There are still a lot of holes, mostly in advanced search/evaluation theory.
Anyways, there is not really an established format for test results yet. Probably the cleanest way would be to create a new page for each idea's test results, linked to from the main page.
It's interesting that it's taken so long for test results to pop up, as that was the intended purpose of the wiki in the first place...
-
- Posts: 1494
- Joined: Thu Mar 30, 2006 2:08 pm
Re: First result contribution
Since I hope lots more people will contribute to the wiki, here is how to do it:BubbaTough wrote:Hey,
I have my first test result to add to the WIKI folks started...and am not sure how to add it. Could someone message/email me if there is already a way/format established for doing so? If not I guess it is worth thinking about it.
The result is not very exciting (tried no adding to history if beta - alpha <= 1 ... it worked terribly) but I thought I would use it to learn how to contribute to the site.
-Sam
1. go to http://chessprogramming.wikispaces.com/
2. If you have not already joined the space, click on "Join This Space" to the left. You have to join to contribute..this just to prevent robots from placing ads on the wiki. If you have joined already, then use "Sign In" at the upper right.
3. Just browse the wiki, and when you think of someting to correct or add, click on the "Edit This Page" link at the top, make yohur changes, then click "Save".
Mark
-
- Posts: 62
- Joined: Mon Aug 14, 2006 3:47 am
- Location: Stellenbosch, South Africa
Re: First result contribution
Sorry Zach,
I found it and thus deleted my post. Thanks for your swift reply!
Jaap
I found it and thus deleted my post. Thanks for your swift reply!
Jaap
-
- Posts: 1922
- Joined: Thu Mar 09, 2006 12:51 am
- Location: Earth
Re: First result contribution
I'd like to note that many people are still contributing, many of them newer members.
Of note is Gerd Isenberg, who as of now has amassed a whopping 403 page edits. This compared to the total number of edits for the wiki, 876...
JOIN THE WIKI!!!!
Of note is Gerd Isenberg, who as of now has amassed a whopping 403 page edits. This compared to the total number of edits for the wiki, 876...
JOIN THE WIKI!!!!
-
- Posts: 62
- Joined: Mon Aug 14, 2006 3:47 am
- Location: Stellenbosch, South Africa
Re: First result contribution
Done!Zach Wegner wrote:JOIN THE WIKI!!!!
Jaap
-
- Posts: 2250
- Joined: Wed Mar 08, 2006 8:47 pm
- Location: Hattingen, Germany
Re: First result contribution
Hi Zach,Zach Wegner wrote:I'd like to note that many people are still contributing, many of them newer members.
Of note is Gerd Isenberg, who as of now has amassed a whopping 403 page edits. This compared to the total number of edits for the wiki, 876...
JOIN THE WIKI!!!!
OK, I am quite obsessed to put all my otherwise archived stuff in the wiki.
Fortunately I had actually some rather quite weeks in office - but that will change soon again. It is more work than I initially thought - it has some addiction potential though.
Of course the number of edits counts every minor change.
The spell-checker is great - but still one needs to overwork all my grammatical errors. Actually, while working on pawn-pattern, I think it is better to split pages even more.
Cheers,
Gerd
-
- Posts: 1494
- Joined: Thu Mar 30, 2006 2:08 pm
Re: First result contribution
Sam,
A little more help. You should navigate through the site and see where the test results would best fit. History killers should effect move ordering, so I would go here:
http://chessprogramming.wikispaces.com/Move+Ordering
you will find a history heuristic link, which unfortunately does not have any content yet. So you should click that link, edit it and describe what the history heuristic does to move ordering, then add a link there to "History Heuristic Test Results". You can put your results on that page.
How bad was it (updating history on null window nodes)? I would expect it to be bad, since most of the game tree is null-window nodes, so the histoyr counters would proabbly not get enough info to be useful. Maybe using a bigger increent on non-null window nodes make sense though, to give them more weight. So many ideas, so little time to test them!
Mark
A little more help. You should navigate through the site and see where the test results would best fit. History killers should effect move ordering, so I would go here:
http://chessprogramming.wikispaces.com/Move+Ordering
you will find a history heuristic link, which unfortunately does not have any content yet. So you should click that link, edit it and describe what the history heuristic does to move ordering, then add a link there to "History Heuristic Test Results". You can put your results on that page.
How bad was it (updating history on null window nodes)? I would expect it to be bad, since most of the game tree is null-window nodes, so the histoyr counters would proabbly not get enough info to be useful. Maybe using a bigger increent on non-null window nodes make sense though, to give them more weight. So many ideas, so little time to test them!
Mark