conanbatt / openkaya Goto Github PK

The function getTimeSettings is undefined but referenced in the code.

Uncaught ReferenceError: getTimeSettings is not defined
(anonymous function):9292/users/scarsick/:274
c.event.handlejquery.tools.min.js:67
c.event.add.j.handle.o

$("#update").click(function() {
        var info = {
            info: {
                description: $('#description').val(),
                show_timestamps: $('#show_timestamps').attr('checked'),
                time_settings: getTimeSettings(), //here
                email: $('#email').val(),
                old_password: $('#old_password').val(),
                password: $('#password').val(),
            }
        };
       //...

By forking the process and using signals, we could tell the bot stuff live as it plays.

This is the game: https://beta.kaya.gs/gospeed/match/40358

Make test case or load it in the SE demo to confirm it produces an error.

Then fix it :)

Hi,

Creative Commons is for content, not for code. I would suggest to use a license for code, particularly, you might be interested in Affero GPL (which has the 'share-alike' but it is also for websites), but it is up to you.

Changing the license is less harder when is done at the beginning of a project, later it can be a painful process (to contact all contributors).

With the following SGF, the SE marks the lower left group as dead , when its obviously alive.

(;FF[4]SZ[19]KM[0.5]RU[Japanese]OT[N/A]HA[2]AB[dd][pp]PW[MoffyUK]WR[6k]PB[MoffyUK]BR[6k]C[MoffyUK[6k]: Hello
fluffm[8k]: hi
MoffyUK[6k]: Thanks for playing
fluffm[8k]: hvae fun!
MoffyUK[6k]: You too
];W[dq]WL[1778.097];B[qd]BL[1792.645];W[nq]WL[1738.757];B[pn]BL[1785.498]C[MoffyUK[6k]: I see you also found Kaya ranks a little demanding
MoffyUK[6k]: You started at 6k?
];W[pr]WL[1712.237]C[fluffm[8k]: i think i started at 7 and then won a game
MoffyUK[6k]: OK
fluffm[8k]: but im 6k on KGS
MoffyUK[6k]: OK
];B[qq]BL[1748.713]C[MoffyUK[6k]: I was 3k on KGS
fluffm[8k]: yeah i think it's a bit rougher on kaya
];W[kp]WL[1694.621]C[MoffyUK[6k]: Yes
];B[qk]BL[1716.044]C[MoffyUK[6k]: It is nice not to play a fast player - the last was ultra fast even to the bitter end - I managed to stay slow and was over 100 ahead when he escaped
fluffm[8k]: haha
];W[od]WL[1676.383]C[fluffm[8k]: that's nice
];B[pf]BL[1691.714]C[MoffyUK[6k]: Fast players try to force mistakes - they rarely win on clean merit I feel
];W[qc]WL[1652.474]C[fluffm[8k]: i agree... but i do also sometimes get impatient and play too fast
MoffyUK[6k]: Impulisve play has strangled my 22 years of Go playing - when I can stay calm and think I play differently
];B[pd]BL[1652.367];W[pc]WL[1638.161];B[co]BL[1638.929]C[fluffm[8k]: why do you think the ranks are different here than kgs?
];W[do]WL[1609.314]C[MoffyUK[6k]: A good question
MoffyUK[6k]: Not really sure
];B[cp]BL[1625.987];W[dp]WL[1596.817];B[cq]BL[1606.455];W[oe]WL[1573.926];B[fc]BL[1580.001];W[pe]WL[1557.083];B[qe]BL[1560.020];W[qf]WL[1549.429];B[qg]BL[1515.966];W[rf]WL[1542.001];B[og]BL[1508.803];W[rg]WL[1494.755];B[qh]BL[1491.615];W[dn]WL[1422.060];B[cm]BL[1480.414];W[cn]WL[1408.063];B[bn]BL[1477.098];W[dm]WL[1399.031];B[cl]BL[1473.307];W[kc]WL[1369.897]C[MoffyUK[6k]: Nice game so far
];B[dl]BL[1437.372]C[MoffyUK[6k]: I am trying to play every move with meaning - I so very often just invade impatiently
fluffm[8k]: yeah it is a nice game... although i regret the top right corner... that was silly
];W[ho]WL[1337.524]C[MoffyUK[6k]: Yes
MoffyUK[6k]: Completing the upper left shimari still requiers a second move
MoffyUK[6k]: 3rd move I mean
fluffm[8k]: huh. i didnt know that
];B[dh]BL[1380.743]C[MoffyUK[6k]: OK, let me play under it to show you
MoffyUK[6k]: If your stones had been at C16 and E17 then it is secure
];W[bd]WL[1295.452]C[fluffm[8k]: is it c15?
];B[cc]BL[1339.536];W[de]WL[1279.516]C[MoffyUK[6k]: No, shimari is C16 and E17
MoffyUK[6k]: C15 to complete yes
MoffyUK[6k]: For what you had
fluffm[8k]: ok, yeah thats what i imagined
];B[ce]BL[1295.945]C[MoffyUK[6k]: 4-4 stone mostly intended as influence
];W[cd]WL[1260.704];B[cf]BL[1278.042];W[dc]WL[1251.309];B[ed]BL[1271.407]C[MoffyUK[6k]: You get a great result here
];W[cb]WL[1243.186]C[fluffm[8k]: yeah i thought d15 might be overplay
MoffyUK[6k]: It may have been - I do not know the joseki
];B[ic]BL[1233.184];W[ke]WL[1220.429];B[ie]BL[1201.121];W[di]WL[1203.830]CMoffyUK[6k]: OK, now for some entertaiiung - but risky play
MoffyUK[6k]: Maybe
(;B[eh]C[MoffyUK[6k]: There was always the danger of compromising my lower side
fluffm[8k]: yeah. i think im ok with this result
MoffyUK[6k]: I am trying to play with correct shape
fluffm[8k]: i dont think i know what correct shape is yet
MoffyUK[6k]: Now you speed up - when you smell blood - but it is precisely when you should still read
fluffm[8k]: true :)
MoffyUK[6k]: Whoa, getting complex now
fluffm[8k]: yep... im getting more and more behind, im afraid
MoffyUK[6k]: NO problem
MoffyUK[6k]: No*
MoffyUK[6k]: I know that I am capabke of Dan level play when I can sustain concentration
MoffyUK[6k]: Have I made any moves so far that were questionable, do you think?
fluffm[8k]: i dont think so
MoffyUK[6k]: Thanks
MoffyUK[6k]: I can go through the game after if you want - there were times when you were not hard enough on me
fluffm[8k]: ok
MoffyUK[6k]: T13 perfect move
fluffm[8k]: thanks
MoffyUK[6k]: N9 also
MoffyUK[6k]: Getting too complex to track now
fluffm[8k]: alright, we're out of the woods for now
MoffyUK[6k]: I misread
MoffyUK[6k]: Well played!
fluffm[8k]: thanks!
MoffyUK[6k]: That was good stuff
fluffm[8k]: arr the lady is pressuring me
MoffyUK[6k]: OK
MoffyUK[6k]: No worries
MoffyUK[6k]: A free game after all
MoffyUK[6k]: 1 liberty short of a seki
fluffm[8k]: phew!
MoffyUK[6k]: Once again, it is the overstretch - on the right - that is my undoing
fluffm[8k]: i still dont think youll lose anyway
MoffyUK[6k]: True
MoffyUK[6k]: But the last game I played I made no obvious mistakes
fluffm[8k]: wow
MoffyUK[6k]: Very rare indeed
MoffyUK[6k]: Thanks
fluffm[8k]: thank you
fluffm[8k]: hey farful
MoffyUK[6k]: I think this better
fluffm[8k]: unfortunately i dont actually have time for a review
MoffyUK[6k]: What do you think?
MoffyUK[6k]: OK
fluffm[8k]: ive got class to go to!
MoffyUK[6k]: No worries
MoffyUK[6k]: Relaly enjoyed the game
MoffyUK[6k]: Many thanks
fluffm[8k]: but thanks again for the thoughtful game
]))

the flow now is:

1 - load SGF file
2 - gen ai move..

3- Wait until new sgf is received from server
-> if move, go back to 1

It could be:

1 - load SGF (for handicap)

2 - gen ai move (and add it to a local tree)
3 - Wait until new move is received from server
-> if move,
a. add it to a local tree
b .go back to 2

• If there is a syncronicity issue, or bot restart, go back to step 1

The estimator has a lot of comments and lots of spaces. The code is pretty large, so a minified version is key.
Lets keep a production version of it on the same folder, and make the tests run both javascript files (just to make sure compression is not breaking anything).

The visual estimator gives decent results for the games shown, but once put on the server, all games will be under heavy estimation and bring about lots of unthinkable scenarios.

We should make a test not only testing a lot more sgfs, but one that will make it easy to add another sgf so when it fails, we can easily expand the test cases.

This can be done by having one large HTML test file with an absurd number of sgfs and run the estimator with them. Proabbly better not to have the visual representation of it for performance issues, unless it explodes or marks something really wrong.

We could run the estimator on finished games and compare the estimator result to the real result, so we can also spring out cases where the estimator fails.

This is not really a bug but more of a concern about what will be handled by the client-side code. I have looked quickly through the code and I've seen some code that gives the impression that the client-side code will be responsible, in some cases, of informing the server of a loss or a win. Giving that kind of power to the client-side code has security flaws, and it wouldn't be hard to exploit them if there's no server-side validation...

Could we have a bit more information about client/server responsibilities and security mechanisms that will be put in place?

Thanks!

Time controls are pretty hardcoded on the client. Server will start controlling the times for bot soon and bots should be able to configure themselves on the different time systems.

https://beta.kaya.gs/gospeed/5839 coord G3 should be black point.

https://beta.kaya.gs/gospeed/match/36706 misses 4 points!

At the moment kayabot just throws everything to stderr (or stdout). When there happens something unexpected, it might produces something interesting.

But if there is tons of status updates and you haven’t redirected everything to file for post processing (not fun if it generates much output), those status updates will consume whole buffer of terminal, making it impossible to see an error that has happened, like, a week ago or so.

My suggestion for the problem: make it possible to log out errors differently. For example, allow redirection of debug messages to separate file and other errors to another. Or even allow redirection of debug messages to a file (/dev/null for people that aren’t interested :P) and throw everything else (unexpected) to console, so person maintaining the bot can look at the bot if it has done something funny.