Comments/Ratings for a Single Item

Hello,

Aurelian Florea here. I can't login in. For some reason I keep getting the allow cookies message even if aI allow cookies for all websites. Am I still doing something wrong, or is it a problem with the website?

I temporarily disabled cookies to see if they are required for logging in, and they are. Although your data is stored in a session, the browser uses a session cookie to know which session is yours. If you're using Firefox, there is an extension called Web Developer, which will let you, among other things, look at your cookies. It will also let you disable or enable cookies.

Hello again,

Aurelian Florea here.

I do have cookies enabled but I seem to get 2 errors when attempting a login with firefox.

CSS error expected declaration but not found '*' skipped to next declaration.

Javascript error: TypeError:tribar is null

Only with the add-on I can see those.

Doest that help us?

Aurelian, I haven't been able to duplicate your error messages in Waterfox, which I normally use with Windows, or with Firefox, which I ran in Linux Mint. So I'm thinking there might be some incompatibility with one of your extensions. I think you can run Firefox in a safe mode that doesn't run any extensions. Try that and see if you can log in. If that works, run it again normally and disable extensions one at a time until you isolate the one causing the problem.

Login does not work with internet explorer either nor with mozilla with disabled addons or extensions. I'm really sorry to bother you like that Fergus, but it seems we have a problem. Could be something related to the provider or Romania (the country where I am in)? I do not know!

Hello,

Even if from a newly created different user, this is still the same Aurelian Florea (the one with Enep and Apothecary).

I managed to sign in with a newly created user, but I still cannot sign in with my main user catugo. What does this mean Fergus? Any idea?

I don't know what the cause could be, but like H. G. Muller, your last login was on December 16, 2016. I reset your catugo password and emailed the new password to the address you have on record for that userid.

This is Fergus Duniho, signed into Aurelian Florea's original account with a new password just to test that it works. It does.

I'm back too!

Thanks, Fergus!

help! i never recieved the email! my email should be listed under ancalagon, but if not, it is [email protected].

I can't login with my old account (user ID is MThomas).

I tried to change the password, and I got the email with the link to the page to enter the keyword and new PW, but after entering these it says: "failure to decrypt email adress".

I tried two times, and I'm sure I entered the correct keyword.

What should I do?

I just tried changing my own password as if I had lost it, and it worked for me. So the script appears to be working. If you tried it two times, you might have mixed something up. Each time you request to change your password, it creates a new keyword and a new link for you, and only the latest ones will work. To make sure you are using the right ones, keep the page with your new keyword in your browser, and compare its unique id to the one you get in your email. When you get an email with a matching unique id, you need to use the link in that email to change your password. When I checked your database entry, the privateKey field was not set, which means that you will have to start over.

Hang on. I'm going to change the script to better inform you of obsolete links.

I have now updated the script. If you use an obsolete link, it will inform you of this, and it will tell you how to find the correct link. If you have the correct link, it will tell you the unique id associated with your keyword, so that you are using the correct one.

Thank you, it worked now :)

Joining the Chess Variant Pages is now easier than before. Just fill in your information and get your new password by email. There are no longer multiple stages to go through before you're registered.

If you make a mistake, it is also now easier to delete your account. If you have not yet signed in, you can delete your new account so long as you're still at the same IP address you created it with. If you have already signed in once, you can delete it so long as you have not left any footprint on the site with it.

Hello, long time lurker of this great site! 

I created an account for this years ago as akinfantryman.  I must have written down the email wrong, because I don't believe I ever got an email with the inital sign in information, and I have tried several times to reset the password over the years.  (it recognizes there is an akinfantryman account, but I can't access it.)  I just created this sudo-account, mostly to post this message....I would rather have my usual handle akinfantryman.  Is there any way to recover that account?

Thank you!

hi. 1st thing that should hit u when u arrive on this site is a live game started and picked randomly from all existing chess varients with the user vs a general AI opponent that can play any of the games. This feature is strangely absent. Why?

You want a programmer to make it for you?

Chess V plays some of them including one of mine but you should realize (I think you had punily intended to hint you are yourself a programmer) that is not an easy task to take as such a general AI is rather hard to build. But maybe we should try.

You may download ChessV here : http://www.chessv.org/

Good luck !

I wouldn't say there is anything strange about its absence. You are the first person who has ever said we should have this, and no such AI exists. At best, we have programs and scripts that have been created to play some games. Even Zillions-of-Games, which has more games available for it than any other program of its kind, can only play those games that have been programmed for it. There is no AI that can read rules written in English and play the game against someone. What this site is the home of is Game Courier, which lets you play Chess variants with other people by correspondence. It is as versatile as Zillions-of-Games with respect to the games that can be played with it, but it is only for correspondence games, not for playing against a computer.

Actually equiping the Interactive Diagram with an AI is one of the (alas many) things on my to-do list. As configuring the diagram for the more common variants is rather straightforward (and a wizard exists to make that even simpler), this would be able to support a large fraction of the variants described here.

Furthermore, the input format for the diagram (currently piece names and their moves in Betza notation) could be easily extended to handle some of the more 'exotic' rules that are common (zoning of the board, limitations to what can capture what, special promotion rules). Currently the diagram can only implement such rules by allowing them to be specified by additional (optional) JavaScript routines on the web page that contains it, which is a bit cumbersome.

I am not sure challenging visitors on the home page by an AI in a random variant is a good idea, though. I do think it would be a good idea to have the possibility in articles that describe a variant to play the latter against an AI. The AI does not have to be strong; in fact it would do more harm than good if it was. It would most likely be used by people that just read the rules for the first time, and it would be rather discouraging for them if they get crushed without a chance. So I am aiming for something that does not entirely play like an idiot, but can be beaten by a beginner who puts in some effort.

A good example of what I am after is the Peace Chess article: this contains a diagram that acts as a simple computer opponent. Eventually I want all articles I authored to contain such a feature. As most of those already contain Interactive Diagrams, this should happen automatically once I equip the script for the Diagram with an AI. That would also bring to life the Diagrams I posted in the Comments section of many articles not authored by me.

For other variants it still would require configuring the Diagram, and putting it on the page, though. Like Fergus says, there is no way yet to compile the English description to a formal rules set on which the AI could act. It is unlikely there will ever be one: the English descriptions are too inhomogeneous for that.

HG,

How have you build and implemented the AI in peace chess?

Just a fixed-depth tree search (iteratively deepened) where every move (rather than every turn) counts as a ply, and unfinished release chains are evaluated as a loss. As I really have no idea how this game should be played, there hardly is any evaluation. I just give a penalty for embracing pairs close to your own King, which is extra high when the enemy piece in the pair is actually aimed at your King.

So it basically moves randomly, except that it is pretty clever in detecting possible King captures, and immediately finishes you off when it sees one.

As it is not possible to give away material in this variant, this doesn't lead to completely crazy play.

All right, thanks, but the initial question is pretty good :)!