Fandom

Foldit Wiki

2012 april 4: chat on scores and casp10

1,465pages on
this wiki
Add New Page
Comments0 Share


[23:50:00] <@beta_helix> As you all know too well, the puzzle load has been a bit nuts since CASP ROLL started...

[23:50:45] <TimovdL> Was your goal reached, more variation?

[23:50:52] <@beta_helix> we also have these very exciting Design puzzles (that we hope you are also excited about) but CASP 10 starts in a month and for CASP 9 we basically had to suspend all non-CASP puzzles for 3 months... and even then it was too much of a load!

[23:51:07] <@CFC> ....with the upcoming onslaught of CASP10, are you happy with a client that can cope? from your point of view?

[23:51:18] <@rav3n_pl> got 9026 in gallery puzzle :D

[23:51:47] <@beta_helix> so we have been talking about splitting Foldit into different categories, to start: CASP & design puzzles.

[23:52:32] <@beta_helix> There would be ranks/scoreboards for each category so you wouldn't have to play all the puzzles, just the ones you like.

[23:52:49] <frood66> u intend to split foldit into different sections of play?

[23:53:00] <@beta_helix> (such as if you are bored with CASP puzzles)

[23:53:18] <@rav3n_pl> hmm it can be bad idea....

[23:53:26] <@marie_s> i dont mind score and rank, just have the choice is good for me

[23:53:26] <spmm> that may put twice as much pressure on

[23:53:46] <@rav3n_pl> global rank + sub rankings, this way it can work

[23:53:47] <frood66> well u can of course

[23:53:55] <@beta_helix> jflat06, Seth, feel free to jump in here. The goal would be for you to only play the puzzles that you enjoy and NOT have to play every single Foldit puzzle.

[23:53:58] <Tlaloc> what about a overall score, and separate subscores and tables for the different categories

[23:54:18] <@auntdeen2> how would you score teams?

[23:54:21] <@rav3n_pl> lol great minds... ;]

[23:54:24] <Tlaloc> or what rav3n said

[23:54:32] <spmm> I enjoy all of them getting points on all of them is hard :)

[23:54:33] <Tlaloc> lol

[23:54:33] <@rav3n_pl> hey T

[23:54:44] <@beta_helix> Again, this is just in the early stages and we are still brainstorming different ideas, but we wanted to let you know early on what we are thinking and our intentions with this.

[23:54:46] <frood66> u currently have 166 players online - why not split it?

[23:54:55] <TheGUmmer> why split it? can you just post more puzzles, lets us pick the ones we want to play and adjust the scoring?

[23:55:01] <silverberg> I must admit, I feel compelled to play all of them

[23:55:02] <@CFC> no, no... hangon ....with devs here online, the upcoming onslaught of CASP10, are you happy with a client that can cope? from your point of view?

[23:55:37] <@rav3n_pl> scissors tool!

[23:55:43] <spmm> what about lowering the points on the design puzzles - would reduce convergence and have less effect on global points

[23:56:14] <@auntdeen2> CFC - good question... the client is still very script driven atm

[23:56:20] <@rav3n_pl> we need scissors for upcomming denovo CASP puzzles!

[23:56:34] <Chopsticks94> why scissors?

[23:56:54] <TimovdL> You can use the alignment tool as a scissor

[23:56:57] <@rav3n_pl> to make easier modelling

[23:57:01] <@Madde> I'd like a working cutpoints tool, too

[23:57:22] <@marie_s> will we have de novo or only templates?

[23:57:35] <@rav3n_pl> alingment tool is still not good for me

[23:58:21] <@beta_helix> @marie that is another question we need to address. We were hoping that the natives for the CASP ROLL puzzles would be released so we could do some analysis on those results for CASP 10.

[23:58:41] <Tlaloc> when does CASP 10 start?

[23:58:51] <@CFC> soon

[23:58:54] <spmm> May

[23:59:00] <Susume2> do you know when the casp roll natives will be released?

[23:59:06] <@rav3n_pl> I want to select multiple pieces form many tmepletes then align them and close cutopints,. but I cant do it in one thread, need to select parts separatly... not very good. Make combination of 2 aligmenet works, but more? no

[23:59:15] <@beta_helix> early May


[00:00:06] <@Madde> was this an answer to Susume's question?

[00:00:07] <frood66> beta: maybe i am being daft - did u just suggest u split players?

[00:02:26] <@CFC> :) we're all at the sharp end, we play, and we feed back

[00:02:29] <Susume2> frood I think beta meant split the puzzles into groups, not split the players into groups

[00:02:49] <Tlaloc> but split the scores by types

[00:02:53] <Tlaloc> global scores

[00:03:00] <@SethCooperIRC> hello

[00:03:12] <spmm> hi Seth

[00:03:15] <Tlaloc> hi Seth

[00:03:17] <TimovdL> hi

[00:03:21] <TheGUmmer> hi seth

[00:03:25] <@CFC> hey there Seth

[00:03:38] <karstenw> split proteins into categories. so, something like: evo rank, casp rank, ED rank, Design rank. hmm

[00:03:52] <@jflat06> we arent sure about the groupings

[00:03:55] <@jflat06> but thats the general idea

[00:04:03] <spmm> two scoreboards is not going to reduce the pressue or convergence to get points

[00:04:42] <@beta_helix> it could reduce the pressure if you don't have to play every single puzzle, right?

[00:04:44] <spmm> imo reducing points for design will not discourage players who will play for rank

[00:04:49] <spmm> in the puzzles

[00:05:00] <TimovdL> I dont like it, maybe a limit on the number of scores that are counted for 1 person?

[00:05:43] <spmm> I still play beginner puzzles for no points just dont spend a loto of time on it

[00:05:48] <@CFC> With the upcoming onslaught of CASP10, are the devs currently happy with a client that can cope? Is it currently 'set' from a dev point of view?

[00:06:27] <Tlaloc> The QTTNs are the only ones that I can can score well on any more...the competitiion has become fierce in the last six months.

[00:07:21] <TimovdL> (dont like the idea of splitting into groups)

[00:07:42] <bertro> what Timo said

[00:08:14] <@mimi2> have either of you exerienced a CASP

[00:08:38] <@SethCooperIRC> CFC what do you mean?

[00:08:52] <@marie_s> have we have to choose between the 2 rebuild?

[00:09:03] <bertro> been following pretty well CASP ROLL

[00:09:16] <TimovdL> mimi i love the casp

[00:09:22] <karstenw> if you have one client, you'll need to commit yourself to one style puzzle, or you will score very poorly in all categories.

[00:09:43] <Susume2> I like the idea of puzzle categories; I don't ahve time to play every puzzle but would like to play every casp one

[00:10:07] <@SethCooperIRC> basically, we'd like to be able to post more puzzles at once

[00:10:13] <@marie_s> madde makes a scoreboard for casp anyway

[00:10:27] <Tlaloc> CASP is pretty much a report card on how well we are doing.

[00:11:03] <silverberg> hmm, might that dilute the quality of solutions?

[00:11:29] <karstenw> leave it alone with one exception--- make a rank for casp.

[00:11:33] <@marie_s> yes, if you want to play all of them, noif you play the oneyou more gift on

[00:11:37] <@SethCooperIRC> right, we'd like to do this in such a way that it doesn't

[00:11:46] <frood66> sunds like a reconfiguration to two games - no prob with that

[00:12:11] <silverberg> I was just going to type that as a self response

[00:12:12] <@beta_helix> @frood, that is exactly the genera idea: break it up into 2 games

[00:12:17] <spmm> design puzzles are so easy to copy though

[00:12:21] <frood66> not sure how it affects foldit tough

[00:12:45] <@marie_s> no need to do 2 games, in my opinion

[00:12:59] <frood66> guess u have to ask what u want beta

[00:13:06] <spmm> agree marie

[00:13:23] <Tlaloc> I'd say that separate categories scores would do what you want.

[00:13:41] <@SethCooperIRC> it does seem like the way things are set up now, people want to play all the puzzles

[00:13:41] <@jflat06> not separate games, just separate score categories, yes.

[00:14:08] <spmm> so everone will just keep copying

[00:14:09] <Schleicher> maybe different puzzle menu

[00:14:17] <Tlaloc> but would you still have a global category?

[00:14:20] <frood66> but with just an average of what? 150 gamers - split.....

[00:14:30] <@jflat06> we're interested in *some* grouping of puzzles into different types, and there would be a scoreboard for each type

[00:14:33] <@marie_s> spmm i dont see you point against dsign, they are very useful

[00:14:44] <@CFC> SethCooperIRC - I just wonder, given the frantic activity we'll experience when CASP10 goes live, do we rely on the current client?

[00:14:44] <@auntdeen2> you (devs) need to realize.. that despite computer limitations, all of us DO want to play all the puzzles

[00:14:47] <frood66> 75 each - oh boy

[00:15:01] <@marie_s> no auntdeen not all

[00:15:13] <spmm> I really like design marie I would still play them

[00:15:20] <@auntdeen2> lkay, marie - most :-)

[00:15:31] <mat747too> i agree auntdeen

[00:15:35] <@SethCooperIRC> CFC I'm not sure what you mean, is there some change you're suggesting?

[00:15:37] <@jflat06> CFC, there are issues with the current client, and likely we will try to resolve some of them before CASP, but some will not get resolved.

[00:15:42] <spmm> but would not feel the need to spend hours trying to get .0000001 points for rank

[00:15:53] <karstenw> evo/solo/casp. more categories than that comes with other drawbacks. bbl

[00:16:43] <@mimi2> in some ways because you want more divergence for CASP it need

[00:16:48] <@auntdeen2> jflat - which will not be resolved?

[00:16:54] <frood66> why are we discussing these small point when such a suggestion has been made?

[00:16:56] <@mimi2> a different way of ranking

[00:16:59] <@jflat06> (in time for CASP)

[00:17:00] <@rav3n_pl> oh it can be categorized more: qttn, denovo, mutable etc

[00:18:07] <@SethCooperIRC> this may be a silly question, but why do you want to play all the puzzles?

[00:18:09] <spmm> you could have lower points for design and give bonus points

[00:18:29] <@SethCooperIRC> to get enough points to keep up, or wanting to see them, or something like that?

[00:18:29] <@rav3n_pl> i want all for global rank :D

[00:18:44] <@CFC> seth - because that's how we used to play

[00:18:49] <@auntdeen2> Seth - because we are all OCD (and you wouldn't have a game if we weren't ;-P )

[00:18:50] <frood66> if this line is followed - foldit will fold - there will and can only be new games

[00:19:19] <TimovdL> I just play some for global rank, still have to skip some

[00:19:22] <itskimo> ore puzzles not less :)

[00:19:26] <@TGIRC> @Jflat, Beta and Seth ... if we do this split, what are the implications for chat? Separate rooms or each category or everything still feeding into Global? I am asking from a moderator perspective

[00:19:33] <mbinfield> If the amount of clustering in design puzzles is excessive, could the devs modify the score so that it includes a factor that increases your score based on how different your design is from other peoples' solutions?

[00:19:34] <frood66> following on - u will have to start new games

[00:20:06] <Tlaloc> Because I think that each puzzle contributes to the understanding of protein folding in general...which can lead to good things for humanity :)

[00:20:15] <@rav3n_pl> for hat number of players mergr fo all puzzle chat rooms can be good idea

[00:20:18] <frood66> the new games will have to be seperate

[00:20:31] <@marie_s> I dont see in what on more casp roll or qttn help me understanding anything

[00:20:36] <@SethCooperIRC> Chat could stay pretty much as it is

[00:20:43] <itskimo> one ring to rule...

[00:20:47] <frood66> how?

[00:20:48] <@TGIRC> Cool! Thanks

[00:21:27] <frood66> seth: the new games would have different motives

[00:21:30] <spmm> and becasue I always like to try every puzzle and do as well as I can because as tlaloc says even if it is not that good it may help

[00:21:45] <@CFC> ....hummph

[00:21:46] <itskimo> yep

[00:22:01] <TheGUmmer> @spmm i agree

[00:22:16] <@CFC> ...Seth? You do understand it used to be more fun?

[00:22:32] <mat747too> seth - you force us to play all the puzzles when you give puzzles points

[00:22:40] <@SethCooperIRC> CFC, how did it used to be more fun?

[00:22:45] <@CFC> ...when we had control?

[00:22:57] <@SethCooperIRC> what kind of control?

[00:23:06] <TheGUmmer> how about so no script puzzles

[00:23:10] <TheGUmmer> some

[00:23:17] <Tlaloc> You actually have more control now...the scripts have parameters.

[00:23:19] <@CFC> Foldit v Rosseta

[00:23:39] <Tlaloc> I think no script puzzles would be good.

[00:23:47] <@marie_s> you can win only by scripts

[00:24:01] <@auntdeen2> currently, yes, marie

[00:24:04] <spmm> are we straying off topic

[00:24:09] <Tlaloc> I learned a lot about proteins when hand folding in the early days.

[00:24:17] <@auntdeen2> not the way it used to bbe - which fostered more diversity

[00:24:18] <frood66> turning the clock back is not a sensible option that i can see

[00:24:22] <@thomirc8> spmm, no, this is central to the discussion

[00:24:28] <@rav3n_pl> win only by scripts? no way

[00:24:44] <mat747too> no all players have the time to play without scripts

[00:25:04] <@SethCooperIRC> well, this is mostly theoretical at this point

[00:25:12] <spmm> I'd like no script puzzles but not in CASP

[00:25:34] <TheGUmmer> no as a class by itself

[00:25:44] <Tlaloc> No, certainly not in CASP, where any way you can get ahead is good.

[00:26:02] <@marie_s> i win 523 ony by scripts

[00:26:15] <TimovdL> That would be a good separate category (only hand)

[00:26:54] <@rav3n_pl> lol, put scripts you ahve used marie somehwew ;]

[00:26:57] <mbinfield> I dunno.. there's certainly value in knowing how to do manual folding, but it seems a bit like an old computer engineer complaining about how people understood computers better when you had to wire a logic circuit by hand

[00:27:13] <frood66> ok - so we have 160 ish players on line working on 5 puzzles. The suggestion is 80 players on CASP and 80 playing about - how exactly does that work?

[00:27:32] <jflat06IRC> so many of you know about the pi helix problem that has caused problems in the symmetry puzzles that we have posted

[00:27:37] <TheGUmmer> from the about page: We?re collecting data to find out if humans' pattern-recognition and puzzle-solving abilities make them more efficient than existing computer programs at pattern-folding tasks" scripting doesn't help this

[00:28:10] <jflat06IRC> the issue arises because the score function of the game is able to be exploited to gain points incorrect with pi helices.

[00:28:25] <jflat06IRC> in order to fix it, we need to change the score function.

[00:28:26] <Tlaloc> Not so, GUmmer...knowing which script to run at which point is a human thing.

[00:28:39] <jflat06IRC> what this means is that scores before and after the score change will be different.

[00:28:41] <@TGIRC> I don't thiink you are looking at any "hard" partitioning Frood ... Players will just play the puzzles they want to play ... it will reflect in seperate scoreboards but no one is "locked in" to any one group of puzzles

[00:28:46] <@Madde> I'd like more de-novos (where scripts are useless when your manual start wasn't good enough)

[00:28:55] <BitSpawnIRC> !!! no scripts? and after that? we can foldit with a pencil and paper :D

[00:29:12] <jflat06IRC> in order to do this without messing up any puzzles in progress, we need to have a period of time where no puzzles are up

[00:29:14] <@auntdeen2> how would you know which script to run if you werent inspired by your hand work to see something new?

[00:29:17] <BitSpawnIRC> I write 2 or 3 scripts for EACH puzzle, scripts are necessary

[00:29:23] <frood66> tgirc: now explain the scoring :)

[00:29:33] <itskimo> no scripts puzzles willl only be won by a few players

[00:29:58] <spmm> how long do you need jflat?

[00:30:16] <jflat06IRC> our current plan is to bring all puzzles down sometime next week, where we will post the score function fix, and then we will repost those puzzles with the ability to load in solutions

[00:30:25] <Tlaloc> so you'd let all the puzzles expire, then release a new client with the new scoreboard.

[00:30:29] <@rav3n_pl> jeff, just announce thet all scores woll change up or down and do it

[00:30:37] <@rav3n_pl> we not need pause

[00:30:50] <@rav3n_pl> just do this update after some puzzle close

[00:30:51] <TimovdL> Second that

[00:30:52] <infjamc> Raven: the problem is that all scores will change

[00:30:57] <Tlaloc> Just set the expiration date for all the puzzles to the same time.

[00:31:02] <jflat06IRC> the problem is that some people may keep an older client around, and that client might give them an advantage because they will score differently

[00:31:05] <@rav3n_pl> all scores in ACTIVE puzzles

[00:31:11] <infjamc> If one's score drops after the score function update, the rankings could be off

[00:31:18] <jflat06IRC> tlaloc, people usually dont like a bunch of puzzles ending all at once

[00:31:29] <@rav3n_pl> but scores are "confirmed" on server side?

[00:31:34] <Tlaloc> But for this, you'd really need to.

[00:31:36] <jflat06IRC> which is why we were planning on just taking them down and then reposting

[00:31:46] <@rav3n_pl> just update leadrboard via server side not client side

[00:31:49] <spmm> jflat wil lyou do this in main and deveprev at the same time?

[00:32:10] <@rav3n_pl> and post WARNING on front page that update is NEED to get proper scores

[00:32:19] <@rav3n_pl> when noone will complain

[00:32:35] <@mimi2> I think a clean break and restart sounds sensible

[00:32:41] <jflat06IRC> right spmm

[00:32:42] <@rav3n_pl> nooooooooooo

[00:32:57] <TimovdL> Maybe a trick to block scores of old clients?

[00:33:20] <@CFC> Aaaach....

[00:33:42] <Tlaloc> Nobody else, WOW, for example, lets you play with an old client.

[00:34:00] <TheGUmmer> start fresh, set everyone to zero :)

[00:34:08] <jflat06IRC> our experience is that people dont like being forced to update

[00:34:14] <@rav3n_pl> rofl @gummer

[00:34:17] <@TGIRC> There is no function to "lock out" a puzzle from an older client? Kind of a co-requisite to a puzzle that the client be at a certain level?

[00:34:18] <mbinfield> I think some people are confused between puzzle score and points, maybe.. jflat is talking about the scoring of proteins

[00:34:19] <itskimo> level feild

[00:34:24] <frood66> no - another idea - u cannot play unless u are holding a whole cheese

[00:34:30] <genestar> hi guys, ii dont know what to do

[00:34:31] <@auntdeen2> you are really going to change dev & main at the same moment o.o

[00:34:45] <@rav3n_pl> just rosetta scoring libs

[00:34:56] <spmm> how long will the change take?

[00:34:59] <Tlaloc> I think that this sort of thing is okay to force people to update.

[00:36:15] <jflat06IRC> we cant always avoid posting to main

[00:37:14] <Tlaloc> If you expired all the puzzles at the same moment, updated the client in main, forced using the new client, and posted new puzzles, everyone would be on the same footing.

[00:37:26] <@jflat06IRC> right tlaloc

[00:37:45] <infjamc> Would the autosaves be copied over, though?

[00:37:48] <@auntdeen2> jflat - what exactly is the new scoring?

[00:38:02] <infjamc> If not, then a warning needs to be given so that people would remember to make manual saves on every puzzle

[00:38:04] <@jflat06IRC> it changes some hydrogen bonding parameters

[00:38:23] <@jflat06IRC> pi helices score well because they're using long range bonds as opposed to short range

[00:38:39] <@jflat06IRC> so it makes the bonds formed by pi helices be short range, like normal helices

[00:38:42] <@auntdeen2> this is all from the pi helix bug?

[00:38:52] <@jflat06IRC> this fix, yes.

[00:38:58] <harp> and how long would it be down for?

[00:38:58] <@CFC> ...I wonder now, just exactly what the devs what decide what the differences are now, between FoldClub and Rosetta@haome

[00:39:01] <@auntdeen2> thank you

[00:39:16] <dldahlen> can you develope a mark amrk band system instead of dragfrom point to point? it is very hard to drag through the middle of puzzle

[00:39:28] <infjamc> By the way, is the pi helix bug an issue in Rosetta 3.x as well?

[00:39:31] <Tlaloc> Is rosetta using the new scoring?

[00:39:41] <@jflat06IRC> rosetta never finds pi helices because rosetta isn't crazy enough to be able to form them in the first place

[00:39:48] <TimovdL> How about posting an update, and give people 24 hours to update without stopping the puzzles, but after that only the updated scoring counts?

[00:39:50] <@mimi2> will you take the opportunity

[00:40:10] <@auntdeen2> didahlen - sounds like a good feedback to post

[00:40:12] <spmm> just do it guys we can survivie for one day

[00:40:21] <@mimi2> to make any other scoring changes

[00:40:41] <@rav3n_pl> hahaha

[00:40:41] <itskimo> yep

[00:40:55] <infjamc> @spmm: plus, if all else fails, there's always the option of loading the expired old version of the puzzle to extract the manual save

[00:40:59] <harp> true spmm

[00:40:59] <silverberg> yes, do it

[00:41:19] <itskimo> gogogogo

[00:41:32] <@rav3n_pl> yes, close and repost puzzles configured to force updated clien may work

[00:42:04] <@jflat06IRC> right

[00:42:06] <@rav3n_pl> ofc possibility to load old saves

[00:42:06] <@jflat06IRC> that's the plan

[00:42:06] <Tlaloc> Extend the existing puzzles so they all expire at the time of the longest puzzle now.

[00:42:11] <@jflat06IRC> right rav

[00:42:35] <Tlaloc> then post 5 new puzzles with the new scoring.

[00:42:37] <@rav3n_pl> no way T! puzzle shoudl NOT end at same time!

[00:42:49] <Tlaloc> along with the new client

[00:42:55] <frood66> and what amazing advantage will this give exactly?

[00:42:56] <@TGIRC> No matter what you do Jeff, it will probably annoy someone ... I doubt it will "kill" anyone and even if it does, Bummer! It is for the good of the game and the advancement of the science. Points can be regained over time.

[00:43:19] <itskimo> yep

[00:43:41] <spmm> agree rav - jflats original suggestion is fine it obviously needs to be done to fix the PI helix scoring

[00:43:50] <@jflat06IRC> the bottom line is we HAVE to update the scoring function to match rosetta.

[00:43:55] <@jflat06IRC> there are two ways of doing this

[00:44:00] <spmm> will you keep us up to date in global via IRC?

[00:44:03] <Tlaloc> It's all temporary. A few days of furor that everyone will forget in a few weeks.

[00:44:04] <@jflat06IRC> either letting all puzzles expire on the same date

[00:44:05] <frood66> not a fast respose i notice

[00:44:18] <BletchleyPark> why not roll out a new client and keep the puzzles ?

[00:44:37] <@jflat06IRC> or closing all the puzzles, posting the update, and then reopening them and letting people use their old solutions

[00:44:41] <Tlaloc> The problem is that the scoreboard won't reflect things.

[00:44:54] <BletchleyPark> you reset the scores

[00:45:01] <infjamc> Actually... would it be okay to apply a "grandfather clause" of sorts?

[00:45:07] <BletchleyPark> and upon loading the old solutions you will have new value for money

[00:45:11] <@rav3n_pl> we talking only on current puzzle scores, not points gained b4!

[00:45:12] <infjamc> Namely, let the current puzzles be scored under the old score function

[00:45:28] <infjamc> ...and only apply the corrected score function to newer puzzles

[00:45:47] <Tlaloc> eek...that makes for ugly stuff in the code.

[00:45:50] <@TGIRC> It would be a relatively easy fix to tag the solution witht he client version. no recovered solutions from N - 2 or the like

[00:45:53] <@rav3n_pl> client cant use 2 scoring libaries at same time

[00:46:06] <@mimi2> if there are no pi helics in the current puzzles then will there actually be a difference

[00:46:12] <spmm> surely score change will only happen for people with pi helixes?

[00:46:25] <infjamc> What I'm saying is that how much of a difference would it make if the existing puzzles are mis-scored under the old function?

[00:46:34] <@jflat06IRC> not exclusively, but the differences should bey very small if you arent making pi helices

[00:46:34] <TheGUmmer> mmmmm pi

[00:46:37] <@CFC> did happen, and done, mbin

[00:46:51] <infjamc> If the answer is "not that much," then we might as well keep the existing puzzles open

[00:47:18] <itskimo> developers choice

[00:47:27] <Tlaloc> What if you just reset all the scores on the server and required the new client

[00:47:38] <BletchleyPark> exactly

[00:47:41] <@rav3n_pl> in some u-turn bonding it can affect score too (if i understand problem correctly)

[00:47:42] <itskimo> yep

[00:47:48] <spmm> why make it more complicated keep it as simple as possible and I vot for not closing all a t the same time

[00:47:50] <Tlaloc> as soon as a new score posted to the server, you'd get your new rank

[00:48:03] <BletchleyPark> correct

[00:48:34] <itskimo> what is important the scores or the science?

[00:48:53] <@TGIRC> The Science chat was earlier today MBinfield

[00:49:11] <BletchleyPark> scores is what keeps us competing

[00:49:40] <itskimo> compete of cooperate

[00:49:56] <@jflat06IRC> it might be that the score differences are so small (assuming you didn't make pi helices) that we wont even need to close the puzzles

[00:50:10] <@TGIRC> ... and new scores do not preclude the possibility of competing ... just on a new (more accurate) footing ... no biggie

[00:50:20] <itskimo> good

[00:50:23] <@jflat06IRC> if that actually ends up being the case, then it'll just be an update

[00:50:56] <Tlaloc> just do what you need to do, jeff...at most it has wierd scores for a week.

[00:50:59] <@auntdeen2> can I go back to an earlier discussion with a thought - then I need afk

[00:51:01] <@auntdeen2> @beta... rather than splitting the game into 2... Timo's suggestion much earlier about having X number of puzzles count per player, mixed with mbinfield's thought of adding points for diversity might be a better way to go… that could possibly satisfy everyone

[00:51:03] <@betahelix> we just wanted to warn you ahead of time!

[00:51:12] <@rav3n_pl> jeff, you have all best solutions on server, just load them to old and new scoring and see difference, it it will be ~5pts or so do it w/o any closing.reposting ;]

[00:51:21] <itskimo> Keep it simple

[00:51:41] <@betahelix> @auntdeen, that is why we wanted to bring this up to all of you early on, to get your feedback and suggestions so that we can bring them to the Foldit meetings.

[00:52:08] <@SethCooperIRC> yes, it seems like the simplest thing would to just be update without having to close the puzzles

[00:52:23] <itskimo> go for it

[00:52:28] <@auntdeen2> beta - that might be a harder system for you, but might satisfy the most of us for different reasons

[00:52:30] <@SethCooperIRC> if all the update does is penalize this one rare case of hydrogen bonds

[00:52:44] <Tlaloc> Just post in the opening message that scores will change.

[00:52:55] <@SethCooperIRC> then people would still potentially get credit if they had already made them in the currently open puzzles

[00:52:57] <@betahelix> so you have been warned: nobody try any pi-helices this coming week! :-P

[00:53:04] <@SethCooperIRC> but, they wouldn't get credit in the future

[00:53:12] <itskimo> no stress just fun :)

[00:53:14] <Tlaloc> lol

[00:53:23] <mottiger> then i'm glad i don't know what pi-helices are :D

[00:53:24] <mat747too> beta - what about a non global points category

[00:53:26] <infjamc> beta: could you ban my account for a week to get rid of the temptation? (j/k :-)

[00:53:30] <spmm> but the old ones will not be rescored? just the active puzzles?

[00:53:32] <@rav3n_pl> i..neeed....pi....helice... :P

[00:53:32] <@betahelix> haahahahaha

[00:53:44] <@SethCooperIRC> spmm, just active puzzles

[00:54:04] <@SethCooperIRC> we can find out the impact by running our test suite

[00:54:13] <Tlaloc> can you require the new client so everyone will be competing equally?

[00:54:24] <@jflat06IRC> we can, but only on a puzzle per puzzle basis

[00:54:31] <@CFC> ...do we have a viable, stable client for CASP10?

[00:54:39] <@SethCooperIRC> right, we couldn't require a new client for the currently open puzzles at this point

[00:54:47] <@rav3n_pl> add one more tool and we have it ;]

[00:54:47] <@jflat06IRC> CFC, you're in more of a position to answer that question than us.

[00:54:52] <@SethCooperIRC> so, in essence you'd get to pick your client

[00:54:57] <@CFC> ...hmmm

[00:55:01] <spmm> yes that is really important CFC

[00:55:02] <@SethCooperIRC> for the current puzzles

[00:55:05] <@rav3n_pl> just kill wiggle lock and bring bacl working LWS :D

[00:55:14] <@betahelix> @CFC do you mean: are we going to be posting a lot more updates before CASP starts?

[00:55:39] <TimovdL> Add a tag to the uploading of scores and you can also do it with running puzzles

[00:55:56] <Tlaloc> seems important to get this score change in before CASP and make sure it's all working for a few weeks.

[00:57:26] <@CFC> ....but we've been asking the 'stable client' question for a while...

[00:57:58] <mottiger> mine is stable

[00:58:21] <BitSpawnIRC> then I want yours mott

[00:58:39] <Schleicher> "stable" as in "not crashing"

[00:58:55] <mottiger> exactly, stable like not crashing

[00:58:56] <@betahelix> if I recall from the previous developer chat our goal was to have a "stable client" in April, so that there wouldn't be any major changes/additions to the game before CASP10 started.

[00:59:59] <@CFC> ...just, give, us, a stable client. Level playing field. It'll give your CASP10 analysis, a reference point.

[01:00:13] <itskimo> yep

[01:00:21] <mottiger> if lws is working or not doesnt make a big change

[01:00:32] <mottiger> it's just nice for the competition

[01:00:41] <@jflat06IRC> there are a lot of definitions of stable clients

[01:00:53] <@jflat06IRC> not crashing, no bugs, etc

[01:01:00] <@CFC> ...oh, jeff

[01:01:03] <spmm> lws does make a difference if puzzles are locking really ear

[01:01:03] <@jflat06IRC> not altering

[01:01:04] <@jflat06IRC> which do you mean?

[01:01:10] <spmm> early*

[01:01:17] <@thomirc8> Not crashing would be a great start

[01:01:30] <TimovdL> No hangs also

[01:01:37] <mottiger> are you all useing devprev or main?

[01:01:47] <mottiger> my main client is running for ages and does not crash

[01:02:11] <@thomirc8> mine crashes multiple times a day. Has done so since I started folding

[01:02:11] <itskimo> ditto

[01:02:11] <@jflat06IRC> i try to pay attention to crash feedbacks, or at least the ones that i can reproduce that i dont think are special cases

[01:02:17] <BletchleyParkirc> I have a remark on that

[01:02:21] <@jflat06IRC> i personally get no crashes that i dont cause myself

[01:02:34] <spmm> main has been crashing more than usualy but that may be me playing with bands

[01:02:35] <TheGUmmer> my main has been rock stable until the last CO2 and you came along

[01:02:36] <BletchleyParkirc> is the network traffic 'saved up' on reduce network traffic settings ?

[01:02:59] <BletchleyParkirc> I notice that the client communicates a lot before shutting down on those instances

[01:03:06] <BletchleyParkirc> it can take minutes

[01:03:51] <BletchleyParkirc> There also seems to be some relation to slower processing (memory moving around ?)

[01:03:53] <@marie_s> many crashed in 537

[01:03:58] <beta_helix> @TheGUmmer I don't think there has been an update to main since Monday, February 27th 2012 (correct me if I'm wrong, jflat06)

[01:04:21] <@jflat06IRC> other than devprev updates, that should be correct

[01:04:32] <BletchleyParkirc> the issue has had my interest for some time, so that date is plausible

[01:04:48] <TheGUmmer> 537 must not like my scripts then

[01:04:55] <BletchleyParkirc> you may want to check the code for reallocs

[01:05:25] <@thomirc8> BletchleyParkirc: agreed

[01:05:33] <beta_helix> 537 is pretty large and the ligands add more complexity (and probably more issues)

[01:05:57] <BletchleyParkirc> It is not just with 537, in fact that is running quite well in general.

[01:05:57] <Tlaloc> There are issues dealing with ligands in scripts.

[01:06:11] <@marie_s> 537 crashes when I erase a band during wiggle

[01:06:32] <Tlaloc> marie....don't do that.lol.

[01:06:34] <BletchleyParkirc> what if you just disable it Marie ?

[01:07:21] <@marie_s> I know, i just do that often, my way of folding

[01:07:38] <beta_helix> good night... this chat has hit the 2 hour mark so we should probably wrap it up...

[01:07:48] <spmm> that is what I do as well marie just delete one band and bang

[01:07:58] <TheGUmmer> freezes when I add a band when wiggling

[01:08:00] <spmm> thank

[01:08:04] <spmm> you

[01:08:11] <TimovdL> tx

[01:08:19] <TheGUmmer> thanks

[01:08:22] <@TGIRC> Beta, Jeff, Seth ... many thanks for your time and patience!

[01:08:23] <BletchleyParkirc> please make a note on checking that 'flush the communication buffer' remark

[01:08:23] <Tlaloc> thanks guys.

[01:08:30] <BletchleyParkirc> and thank you

[01:08:41] <@jflat06IRC> i can look into it BP

[01:12:04] <Jobriq> maybe foldit needs another server

[01:12:05] <@TGIRC> Define insane Tlaloc ... I've never really been clear on that ... LOL

[01:12:09] <beta_helix> @Tlaloc, the earliest this will air will be in the fall. That was my first question with them: is it after CASP 10?

[01:12:10] <BletchleyParkirc> do we need the new client for 539 ?

[01:12:14] <@mimi2> this place is always insane - but it will also get busy

[01:12:48] <TheGUmmer> insane in the membrane

[01:13:06] <Tlaloc> I would bet we'd have more new people than ever before by a lot.

[01:13:09] <@TGIRC> I would suspect not BP ... not if he is posting it now

[01:13:15] <beta_helix> @BP no, should be up soon with no update

[01:13:20] <BletchleyParkirc> thanks

[01:14:11] <itskimo> fold your membrains

[01:14:22] <Tlaloc> I hope the air conditioning in the server room is up to snuff when it airs, lol

[01:14:31] <@CFC> hmmm

[[01:14:55] <@TGIRC> @Tlaloc ... most will be "drive-bys" ... we should see an uptick in permanent numbers tho'

[01:15:06] <itskimo> bingo :)

[01:15:42] <Tlaloc> The drivebys will be insane. We were getting 3000 people on when the Nature paper posted.

[01:16:03] <Tlaloc> Some percentage will stay.

[01:16:20] <Tlaloc> The competition is so much harder than before the Nature paper was posted.

[01:16:20] <@CFC> Tlaloc... I'm curious why you would assert that?

[01:16:24] <Jobriq> dang 3000 people in this chat?

[01:16:39] <mottiger> never said it and maybe this conversation is too serious, but thanks tlaloc for all your efforts

[01:16:54] <Tlaloc> thanks, mott.

[01:17:00] <@TGIRC> I think 994 was our record for concurrent play ... we did pick up a bunch of new talent as a result of those papers

[01:17:11] <Tlaloc> assert what CFC?

[01:17:29] <Tlaloc> I have a hard time getting into the top 100 now. I used to be a top 30 player.

[01:17:44] <harp> what will 60 minutes bring?

[01:18:03] <@TGIRC> You did too good a job with your programming ... you only have yourself to blame LOL

[01:18:11] <Tlaloc> lol

[01:18:38] <Tlaloc> I do, actually.

[01:18:48] <@TGIRC> @harp ... per Beta CBS News 60 Minutes is looking for FOLDIT players in the U.S to interview for an upcoming story. We're especially interested in players with no science background, young and old, to talk about why they play. Please send a brief description of yourself and why you play FOLDIT to: journalistYourStory@hotmail.com"

[01:18:49] <Tlaloc> I give away everything I know.

[01:19:38] <@TGIRC> Makes room for new thoughts and shares the burden of remembering it all ;)

[01:20:56] <@CFC> [I missed a post in the thread. np]

[01:20:56] <Jobriq> I suppose the number of drive-by players will boost the global points the higher scorers get since the formula includes the total number of players for a puzzle

Ad blocker interference detected!


Wikia is a free-to-use site that makes money from advertising. We have a modified experience for viewers using ad blockers

Wikia is not accessible if you’ve made further modifications. Remove the custom ad blocker rule(s) and the page will load as expected.