2015.09.25 Dev Chat

From Eterna Wiki

( Link back to Dev Chat archive )


On MIT Rapid Diagnostic Kits, Next / Prev Lab Buttons, Badges, & Open Source Licensing


DevChat T-15m 

DevChat T-05m

RedSpah:  I T S H A P P E N N I N G [2:52 PM]

DevChat T-01m

DevChat T-30s

DevChat T-10s

...3

...2

...1

KA-BOOOM!!!

bekeep: I feel like the rocket took off a few minutes early [2:57 PM]

Elves: good thing we all brought our helmets! :D [2:57 PM]

bekeep: And reinforced the hull [2:58 PM]

syang: a dramatic ending to the devchat wait! :) [2:58 PM]

Elves: hehe [2:58 PM]

bekeep: I'm pretty sure Rhiju will be here [2:58 PM]

bekeep: Not sure what the agenda is [2:59 PM]

Elves: huzzah! [2:59 PM]

starryjess: hi [3:00 PM]

RedSpah: Hi Elves, bekeep, syang [3:00 PM]

RedSpah: and starry [3:00 PM]

syang: hi [3:00 PM]

Elves: hi starry!! hi red! :D glad to see you [3:00 PM]

bekeep: Hi starry, red, syang! [3:01 PM]

starryjess: hi elves, red..glad to be on for a change! :P [3:01 PM]

starryjess: hi bekeep [3:01 PM]

Nando: hi everyone [3:01 PM]

Elves: :D hi syang, bekeep & Nando & everybody  [3:02 PM]

rhiju: h everyone [3:02 PM]

rhiju: *hi [3:02 PM]

Elves: hello Rhiju! [3:03 PM]

starryjess: hi Nando and rhiju [3:04 PM]

rhiju: who did the chatbot? [3:04 PM]

rhiju: the one that counts down? [3:04 PM]

jnicol: Hi everyone [3:02 PM]

jnicol: nando did [3:02 PM]

Nando: I'm afraid I'm the one to blame (again) [3:04 PM]

rhiju: hats off, nando! [3:05 PM]

rhiju: now what are we discussing? [3:05 PM]

Elves: hi John! :D  [3:05 PM]

rhiju: i know nando was requesting feedback on game improvements ...  [3:05 PM]

rhiju: and bekeep, elves, and i (among others) have been conversing about licenses [3:06 PM]

starryjess: a license to fold? O.o [3:06 PM]

rhiju: :) [3:06 PM]

rhiju: wanted to show folks one thing: [3:07 PM]

Elves: lol starry - a license for others to use our research [3:07 PM]

rhiju: check out: opendx.mit.edu [3:07 PM]

rhiju: dang it, make it link: http://opendx.mit.edu [3:07 PM]

rhiju: i met the leader of opendx last week, and it turns out they are huge fans of eterna [3:07 PM]

rhiju: but they did not know about our upcoming eterna-medicine initiative... [3:08 PM]

rhiju: but its perfect timing [3:08 PM]

Elves: oh wow! there is perfect overlap here [3:08 PM]

rhiju: they are expanding their opendx kits to amplify nucleic acids... [3:08 PM]

rhiju: the main thing they need for their kits are actually sensor molecules [3:08 PM]

Elves: !!! [3:08 PM]

rhiju: in their case, they have been using antibodies or nucleic acids to viral proteins [3:08 PM]

rhiju: and then converting the binding events into colorimetric signals [3:09 PM]

rhiju: using the same principles as pregnancy tests [3:09 PM]

rhiju: what they don't have for sensing RNA signatures is a way to sense the signature [3:09 PM]

rhiju: and that's what we can provide [3:09 PM]

rhiju: so this is potentially very exciting [3:09 PM]

Elves: = rapid diagnostics in the field? [3:09 PM]

rhiju: yes exactly, rapid diagnostics in the field -- as they have demonstrated recently for ebola [3:10 PM]

Elves: :D :D :D [3:10 PM]

rhiju: also, *cheap* diagnostics [3:10 PM]

Elves: this is very very cool!  [3:10 PM]

rhiju: we were planning to test eterna molecules in devices here at stanford, but there was going to have to be some technology development to get to a clinically relevant device --  [3:10 PM]

Elves: do they have any tests ready to deploy or is this still proof of concept? i see some pages not fully finished on the site [3:10 PM]

rhiju: you can get free sample kits now! [3:11 PM]

Elves: omg! [3:11 PM]

rhiju: they have been focusing on deploying kits with partners in, e.g., parts of africa (for ebola) -- and in nicaragua (for TB!!) [3:11 PM]

Elves: ahhhh!!! this is awesome news [3:11 PM]

rhiju: for the TB, they are focusing on kits to test adherence to antibiotic protocols after TB has been idenfitied ... they do not have a kit for detection of TB [3:12 PM]

rhiju: which is what we will be working on starting in january [3:12 PM]

rhiju: ok, i've been blabbing a lot [3:12 PM]

Nando: those are great news [3:13 PM]

rhiju: yea! [3:13 PM]

rhiju: why are their two bekeep users logged in to chat? [3:13 PM]

Elves: two separate windows / tabs? [3:13 PM]

Nando: ben has different IDs on prod and dev [3:13 PM]

Elves: ohhh [3:14 PM]

rhiju: ahhh [3:14 PM]

Nando: the DB hasn't been synchronized in a long time [3:14 PM]

hoglahoo: ehhh [3:14 PM]

Elves: hehhhh :D [3:14 PM]

rhiju: dang, i can't clone myself. [3:14 PM]

Nando: so, I'll make a short process of the Down,Up thing: nobody said anything, so I'm assuming everyone is happy with hidden keyboard shortcuts [3:15 PM]

Nando: that's a good thing too, less work for me! :P [3:15 PM]

Elves: is that the next / prev feature?  [3:15 PM]

Nando: yes elves [3:16 PM]

Elves: were we going to get buttons in the UI for that? [3:16 PM]

bekeep: haha - didn't even realize that myself [3:16 PM]

rhiju: nando, i bet we'll come up with something for you to do for the up/down. and @bekeep, you'll have to show how to use in the tutorial sequence. :) [3:16 PM]

Elves: i think there was one suggestion to put arrows on either side of the lab title or something? [3:17 PM]

rhiju: i meant puzzle progression [3:17 PM]

rhiju: has anyone out there got the eterna100 badge? [3:17 PM]

Elves: and to apply to lab archives as well [3:17 PM]

bekeep: Yes - at the end of next week, I'll send out a link for the puzzle progression for folks to walk through [3:17 PM]

bekeep: (won't be all of the puzzles and all of the info, but about half of them will be there) [3:18 PM]

Elves: awesome! [3:18 PM]

Elves: *high five* to all devs for the recent and future updates :) [3:18 PM]

bekeep: (btw, Nando deserves a TON of credit for making updates to the tutorial creator, which has allowed us to move quickly on the puzzle progression) [3:19 PM]

rhiju: thanks for the praise -- more coming from jnicol, nando, and bekeep... [3:19 PM]

rhiju: yea,  i can attest that the new puzzle progression is really beautiful.  [3:20 PM]

bekeep: Relatedly, there will also be a documentation update on the tutorial creator coming soon [3:20 PM]

Elves: i have no eterna100 badge, alas! I think I'm only up to 55 [3:20 PM]

Nando: I'm at 90% of the Eterna10... [3:20 PM]

rhiju: lol [3:21 PM]

rhiju: how do the new data look? [3:21 PM]

rhiju: any surprises with 'R97' designs? [3:21 PM]

Nando: it seems not... [3:23 PM]

JR:  I am 2 away from Eterna 100, but then you may move the cheese, I am glad you are together on this, I have seen one too many  bankruptcies due to poor direction by one person and I would not want this to happen to this site.  [3:23 PM]

rhiju: good luck on final 2, jr! [3:25 PM]

Elves: Omei, Meechl, and Eli posted spreadsheets of the data in the thread here for R97 http://www.eternagame.org/web/news/6153537/ [3:25 PM]

bekeep: Us neither. : ) [3:25 PM]

rhiju: we are beginning work on an in-game viewer of these data [3:26 PM]

Elves: :D that is awesome!! [3:27 PM]

rhiju: we'll post links & questions to getsat as we start to prototype. but if there are suggestions for how to display the array-based fluorescence data, please fire away now. (or later in getsat) [3:27 PM]

Elves: we had been hoping simply for a button to download a spreadsheet, and I know LFP6 was working on a script to pull the data for each lab. an in game viewer would be even better. [3:27 PM]

Elves: very cool! [3:27 PM]

Elves: For the badges, would it be easy to make them clickable so that they launch to their description / position in the roadmap?  [3:28 PM]

rhiju: yea, an in-game viewer will be important, particularly as we expect to bring more and more folks into the lab with the new puzzle progression and 'mission videos'.  [3:28 PM]

Elves: i.e. from one's profile [3:28 PM]

Elves: we also were hoping for additional data visualization options in-game, like the dot plot and melt plot graphs.  [3:29 PM]

rhiju: @elves, for badges-->roadmap, that sounds like a good idea, but can you clarify a little?  [3:29 PM]

rhiju: why would auser want to go into roadmap from a badge they've already obtained? [3:29 PM]

Elves: if the viewer could incorporate a few ways to toggle the visualization of the data then people who don't have bandwidth to make or interpret raw data could look at patterns in the data viz [3:29 PM]

Elves: yes [3:29 PM]

jnicol: for the badges, once completed, the link is no longer available in the roadmap [3:27 PM]

Elves: from my profile, or anyone's http://prntscr.com/8khrv5 [3:30 PM]

Elves: from my roadmap http://prntscr.com/8khs55 [3:30 PM]

Elves: so the point is that when i see a new badge in my profile, clicking it and going to the full description tells me what it is [3:31 PM]

rhiju: ah, OK! [3:31 PM]

Elves: and if i see a badge i don't have in someone elses profile, clicking it and going to my roadmap tells me if i have it and what it is [3:31 PM]

Nando: no clicks, just a mouse-over [3:31 PM]

jnicol: and not all badges are diaplyed in the roadmap [3:29 PM]

Elves: Nando good point, though i'd like both personally [3:31 PM]

Elves: jnicol that's good to know, I didn't know that [3:32 PM]

jnicol: ie, eternacon, beta tester badge [3:29 PM]

Elves: ahh [3:32 PM]

Elves: perhaps roadmap could be updated to have a special badges container [3:32 PM]

jnicol: and even some top 50, 25 player badges only display in the profile [3:30 PM]

Elves: maybe an image tooltip mouseover with the full description of badge? [3:33 PM]

rhiju: any comments on license? I know JR had some good points over in getsat [3:35 PM]

rhiju: https://getsatisfaction.com/eternagame/topics/-official-dev-chat-followup-input-on-a-new-license-for-eterna-content [3:35 PM]

rhiju: one of the key questions there is who is the license for? [3:36 PM]

rhiju: eterna has several groups of stakeholders, and potentially more in the future. [3:36 PM]

rhiju: the primary group of stakeholders at the moment are players. [3:37 PM]

JR:  you need to deliver if you spend the money on a new licence, that is your bet. [3:37 PM]

JR: or wait and see, that would be the second bet.  [3:37 PM]

Elves: i was wondering if we can codify a ceiling on what partners can charge for any medicines, sensors, etc. that we produce. or maybe some kind of built in charity program where if they charge for medicine, for... [3:38 PM]

...each one that is sold, one is given for free?

JR: I am not in a position to know which bet is best. [3:39 PM]

Elves: like in order to get access to our license, partners would have to agree to some reasonable terms of pricing & distribution to prevent the common shenanigans in big pharma these days [3:39 PM]

rhiju: based on discussions with pharma, etc., it seems clear that if eterna decides to go after 'bounties', there will be a different agreement for each such case, since each partner will have their own legal infrastructure.  [3:40 PM]

rhiju: we do not have to define that license now. [3:40 PM]

rhiju: *those licenses [3:40 PM]

Elves: yes, i mean for the underlying foundational license, would be our opportunity to codify anything we want applied across the board [3:40 PM]

JR: to get too far into the logistics without delivering has never been a good idea in my experience  [3:40 PM]

rhiju: @JR, yes we don't want to overoptimize early, as they say in software development [3:41 PM]

Nando: @JR: not preparing anything is a vastly more dangerous game/bet [3:41 PM]

JR: remember, Das Labs has the ability to not take the data public if they find something interesting [3:42 PM]

rhiju: @JR, what data? [3:42 PM]

rhiju: *which data? (oops, data is plural) :) [3:42 PM]

JR: the data you are protecting with the new licience [3:42 PM]

rhiju: right, but i don't think the das lab (or, here actually the greenleaf lab) would want to protect the data. [3:43 PM]

JR: That is something you know , but i do not.  [3:44 PM]

rhiju: but that's a good point. open access to all data is something to spell out in an agreement that users can see. and at least i hadn't thought about that before. [3:45 PM]

Elves: maybe it would help to clarify the purpose / goals of the foundational license. for example, is it intended to define what is our IP, who can use it for what purposes, etc.? [3:45 PM]

JR: I would put a halt to production  and write a paper, but maybe that's not how it works. [3:46 PM]

rhiju: @JR, like write a paper now? [3:46 PM]

Elves: publishing a paper on our research puts it in the public domain, which protects it from being patented by others and ourselves as well, regardless of license, if i understand correctly [3:47 PM]

rhiju: well, actually, the minute we make the sequences and data public, they are protected from patent by others. [3:47 PM]

JR: we are poking in the dark, if you find something i would think you would delay the outtput and write a paper.  [3:47 PM]

Elves: i think we have some number of months after publishing to change our minds and seek patent, then it's public domain [3:47 PM]

Elves: yes, though first to file means that public domain is strongest when published in writing in a publicly recorded document, as opposed to prior art through web publishing, but i may not understand for sure [3:48 PM]

Elves: that would be a good question for the IP contacts [3:49 PM]

rhiju: @Elves -- in talking to stanford tech licensing as well as corportate folks, it is highly unlikely that anyone would bother to patent a sequence if it was publically viewable ahead of time [3:49 PM]

Elves: that's good to know! [3:50 PM]

rhiju: my proposal is that all eterna sequences, scripts, analysis, forum posts be submitted under CC (our current license) or appropriate equivalents that allow non-commercial reuse & enforce attribution [3:50 PM]

Elves: then it's a matter of making that license compatible with any additional licenses applied on top of it [3:52 PM]

JR:  It you find something interesting, can you stop transmitting that information to the internet and write a paper per your grants? [3:53 PM]

rhiju: in terms of a way to sustain eterna financially, there are several potential options, but one involves licensing our private github-hosted software, written by developers, to for-profits. that;s how rosetta & foldit work (i'm a part of that rosettacommons consotrium who benefit from the license). in our case, the package would be eternabot. [3:53 PM]

rhiju: @JR, that's an interesting question, particulary the way you phrased it. [3:53 PM]

rhiju: eterna development is currently funded by an NIH grant [3:54 PM]

rhiju: and we presented a data management plan where we promised to release all chemical mapping data in the RMDB and in the Eterna game. [3:54 PM]

rhiju: so we are obliged to make the data public, at least at some point.  [3:54 PM]

jnicol: speaking of making synthesis data public, I just uploaded the data for R97 [3:52 PM]

Nando: @JR: if they find something, why on earth wouldn't they have the right to write a paper? [3:54 PM]

JR: Does lack of licencing disallow other companies, from accessing eterna for information they are looking for? [3:54 PM]

Elves: so we won't be licensing the data, rather we are licensing the software and services that are developed? [3:55 PM]

rhiju: @elves actually just software. (can you think of any services that eterna devs or players might provide) [3:55 PM]

JR: if grant money says it's a fee for all , then Das Labs can't stop the data from getting t othe web, I would think. [3:55 PM]

Elves: i think the software is a kind of service, but yes you're right, just the software unless they want to subcontract for specific feature development [3:56 PM]

rhiju: @JR, yes that's right. of course, my lab, will's lab, etc. believe that we (as a society) learn more when data are available for lots of folks to look at. [3:57 PM]

rhiju: just to give a specific example of what might happen in the future... [3:57 PM]

JR: I am really just concerned with spending a lot od somebodies money on something that isn't research, and I feel only Rhiju would know this. [3:57 PM]

rhiju: help me understand -- are we talking about taxpayer money (which is the case now) or a for-profit corporation's money (which may or may not happen in the future, depending on our discussions now) [3:58 PM]

rhiju: for taxpayer money, our mandate is to make the data public and to write publications. which is what we are doing. quite successfully actually. [3:59 PM]

rhiju: if, e.g., novartis comes to us and asks for help to build an RNA computer to sense RNA signatures of cancer cells, that's separate. [4:00 PM]

JR: in janurary, you find something that actually benefit mankind, can you NOT post data to the web and write a paper, or does the grants insist that that data is posted anyway. [4:00 PM]

Elves: it sounds like the traditional open source situation where the core data is public, and if people want packaged GUIs and processed solutions, then we or anyone can build and sell those, while at the same time... [4:00 PM]

...anyone can access the core data and build their own platforms to serve, analyze, and otherwise use that data?

rhiju: @JR, the data must be posted at some point, typically at the time of paper publication. [4:00 PM]

JR: if you can wait till publication than should you not be covered and not need the licience? [4:01 PM]

rhiju: @Elves, yes... in principle there could be a whole economy of apps uilt on eterna data. i don't thikn that would be a bad thing.  [4:01 PM]

rhiju: @JR, yes, if we want to hold data private for some period before publishing, we have that option. Although I don't think we'd ever do that. [4:02 PM]

Elves: i like that direction, because it means that we can protect data from getting locked down, buried, or charged scalping prices by someone who owns it outright. while at the same time, we or others can make some... [4:02 PM]

...profit and fund additional research

LFP6: Just got wifi. :p Whats up all? [4:03 PM]

Elves: open source is actually not anti commercialization - in a way it promotes healthy competition in the marketplace by encouraging anyone who wants to take the core data / code, and make the best platform to... [4:03 PM]

...serve it up that is possible

Elves: hi LFP6! [4:03 PM]

LFP6: Any dev chatting left? [4:03 PM]

Elves: license discussion :D [4:04 PM]

bekeep: I think we're wrapping up soon... [4:04 PM]

rhiju: but... if devs develop a cool algorithm in eternabot based on discussions with players and what happens in january,  we can keep that private until publication.  [4:05 PM]

LFP6: License discussion all thats been going on? [4:05 PM]

rhiju: If a company reads our paper on TB and wants to do something for, e.g., diabetes, then they could do one of two things. [4:05 PM]

rhiju: they could license eternabot to use algorithms developed & tested during the TB work [4:05 PM]

rhiju: or if eternabot is not able to do it, they could give us a bounty to solve a similar/cognate problem, and then license the improved eternabot [4:06 PM]

rhiju: anyway, i got to run -- intersting discussions everyone! [4:06 PM]

rhiju: thanks for coming! [4:06 PM]

JR:  thanks [4:06 PM]

Elves: thanks Rhiju!  [4:07 PM]

bekeep: Thanks everyone! [4:09 PM]