fluid-work IRC Logs-2012-05-01

[11:26:09 CDT(-0500)] <anastasiac> cindyli, yura, can we check-in after lunch?

[11:26:28 CDT(-0500)] <cindyli> sure, anastasiac

[11:49:56 CDT(-0500)] <anastasiac> jameswy, question about access to the IDI site: anyone can create an account for themselves? What default privileges would a new account have?

[11:51:58 CDT(-0500)] <jameswy> anastasiac: Good question.

[11:52:07 CDT(-0500)] <anastasiac>

[13:43:55 CDT(-0500)] <Bosmon> Hi, cindyli - did you have your CHECK-IN?

[13:44:14 CDT(-0500)] <cindyli> what is that, Bosmon?

[13:44:38 CDT(-0500)] <Bosmon> Apparently you were going to have one "after lunch"....

[13:45:13 CDT(-0500)] <anastasiac> Bosmon, we did check-in, now cindyli and I are pairing

[13:45:28 CDT(-0500)] <Bosmon> How is it going

[13:45:37 CDT(-0500)] <Bosmon> I am just chatting with colinclark about our transformer architecture

[13:45:48 CDT(-0500)] <colinclark> yes, it's true

[13:45:59 CDT(-0500)] <Bosmon> I think the GPII "demultiplexValue" transformer and the A4A "valueMapper" transformer are really the same thing

[13:46:14 CDT(-0500)] <Bosmon> But need some work to be put into a form where they are "obviously invertible"

[13:46:38 CDT(-0500)] <Bosmon> And even further work to be put into a form where they are "actually invertible" : P

[13:46:39 CDT(-0500)] <anastasiac> Bosmon, the transformation architecture doesn't seem to handle arrays very well

[13:47:16 CDT(-0500)] <Bosmon> I am working on a rewrite in my branch here: https://github.com/amb26/infusion/blob/FLUID-4695/src/webapp/framework/core/js/ModelTransformations.js

[13:47:22 CDT(-0500)] <anastasiac> we want to be able to create an array in the output model, but the dot notation is not working as we'd like

[13:47:25 CDT(-0500)] <Bosmon> What problems are you having with arrays, anastasiac?

[13:47:29 CDT(-0500)] <Bosmon> Ah yes

[13:47:34 CDT(-0500)] <Bosmon> I am aware of that issue

[13:47:41 CDT(-0500)] <Bosmon> I will probably be working on that today too

[13:47:48 CDT(-0500)] <cindyli> super

[13:47:58 CDT(-0500)] <anastasiac> so should we hold off until you're done?

[13:48:03 CDT(-0500)] <colinclark> cindyli and anastasiac: You'll want to keep a close eye on Antranig's branch

[13:48:06 CDT(-0500)] <Bosmon> Although right now I can only think of a solution where there was ALREADY an array in the input model

[13:48:15 CDT(-0500)] <colinclark> it's a pretty substantial and pretty nice refactoring of the transformation framework

[13:48:16 CDT(-0500)] <Bosmon> Can you give me an example of a thing you are trying to make work?

[13:48:22 CDT(-0500)] <anastasiac> well, that's the problem: there isn't

[13:48:32 CDT(-0500)] <Bosmon> Ok

[13:48:38 CDT(-0500)] <Bosmon> That's relatively problematic then

[13:48:40 CDT(-0500)] <anastasiac> source model has "captions: true, transcripts: true"

[13:48:51 CDT(-0500)] <Bosmon> And could only really be attacked by producing a .... SCHEMMA.... describing the output model

[13:48:53 CDT(-0500)] <anastasiac> output needs an array with one entry for captions, one for transcripts

[13:49:01 CDT(-0500)] <Bosmon> I guess I will beef up the system a bit so it can deal with that

[13:49:35 CDT(-0500)] <anastasiac> we were trying an approach where we do an initial pass at transformation where we create an empty array

[13:49:47 CDT(-0500)] <Bosmon> https://github.com/amb26/infusion/blob/FLUID-4695/src/webapp/framework/core/js/ModelTransformations.js#L99

[13:49:53 CDT(-0500)] <Bosmon> The comment on this line refers to this problem

[13:50:02 CDT(-0500)] <Bosmon> "default model structure" is created implicitly by the action of fluid.set

[13:50:15 CDT(-0500)] <Bosmon> We have exactly the same issue in CSpace which yura suffered for a while

[13:50:49 CDT(-0500)] <Bosmon> Unfortunately this piece of core framework ("trundlers") is due to be abolished soon, so we will just have to hack a simple and temporary solution (at the framework level) in for now

[13:50:58 CDT(-0500)] <Bosmon> But hopefully the standard for the SCHEMMA can be reasonably stable

[13:51:08 CDT(-0500)] <anastasiac> so Bosmon, it sounds like cindyli and I should hold off on this, and wait for your branch rewrite?

[13:51:10 CDT(-0500)] <Bosmon> I imagine it will look a lot like the format we have for mergePolicies

[13:51:22 CDT(-0500)] <Bosmon> anastasiac I think so, yes, unless it really blocks your workflow

[13:51:39 CDT(-0500)] <Bosmon> I really expect to have something moderately workable by the end of today.... since we are in quite a hurry with GPII at the other end too

[13:51:53 CDT(-0500)] <anastasiac> no, that should be fine - I can just pick this up again when your branch is ready

[13:52:41 CDT(-0500)]

[13:53:49 CDT(-0500)] <Bosmon> I think we have generally come to the conclusion that "recursive schemas" although they may have some advantages for interchange, should be cursed at the implementation level

[13:54:06 CDT(-0500)] <Bosmon> The performance characteristics of using them in CSpace ended up to be just too bad

[13:55:12 CDT(-0500)] <Bosmon> So I think we will frequently be going with the "flat schema" approach, like the one we use in mergePolicies

[13:55:33 CDT(-0500)] <Bosmon> Our GPII partners love flat things, in any case

[13:55:54 CDT(-0500)] <anastasiac> I'm starting to love them, myself

[13:56:06 CDT(-0500)] <Bosmon> Like.... pizzas? : P

[13:56:12 CDT(-0500)] <anastasiac> mmmm

[13:56:34 CDT(-0500)] <anastasiac> beaver tails

[13:57:11 CDT(-0500)] <colinclark> that's a very canadian reference

[13:57:24 CDT(-0500)] <anastasiac> indeed

[13:57:46 CDT(-0500)] <Bosmon> Or perhaps the PANCAKE TORTOISE

[13:57:47 CDT(-0500)] <Bosmon> http://www.reptilesweb.com/images/stories/datso_pictures/med_PancakeTortoise.jpg

[13:58:03 CDT(-0500)] <anastasiac> oo, flat. interesting

[13:58:10 CDT(-0500)] <travis_84_______> Not to interrupt...

[13:58:23 CDT(-0500)] <Bosmon> Hi there, travis_84_______

[13:58:27 CDT(-0500)] <colinclark> travis_84_______: what, you don't like tortoises?

[13:58:38 CDT(-0500)] <colinclark> Bosmon: I guess your tortoise interest has largely been eclipsed by cats

[13:58:40 CDT(-0500)] <travis_84_______> haha

[13:58:42 CDT(-0500)] <colinclark> I can't imagine they get along well

[13:58:47 CDT(-0500)] <travis_84_______> had one as a kid

[13:58:52 CDT(-0500)] <colinclark> cool

[13:59:06 CDT(-0500)] <colinclark> I remember when I first met Bosmon, he would travel around to see different types of tortoises

[13:59:08 CDT(-0500)] <Bosmon> colinclark - it generally has, yes... I remember that time I spent in the Amsterdam Zoo in 2007, when it eventually dawned on me that tortoises don't really do anything.......

[13:59:14 CDT(-0500)] <colinclark> lol

[13:59:43 CDT(-0500)] <Bosmon> I had a nap next to the GIANT TORTON they had there, and it didn't really do anything for me..

[13:59:49 CDT(-0500)] <Bosmon> As opposed to having a NAPP with a CATTT

[14:00:45 CDT(-0500)] <travis_84_______> so you think...fear the quiet animals.

[14:01:00 CDT(-0500)] <travis_84_______> They plotting

[14:01:09 CDT(-0500)] <colinclark> lol

[14:01:11 CDT(-0500)] <colinclark>

[14:02:52 CDT(-0500)] <travis_84_______> So I found an indepth guide in setting up PhoneGap to develop http://mobile.tutsplus.com/tutorials/phonegap/phonegap-development/

[14:03:17 CDT(-0500)] <travis_84_______> it;s not looking good

[14:04:14 CDT(-0500)] <travis_84_______> You have to have a different OS to still dev for the different platforms???

[14:04:46 CDT(-0500)] <Bosmon> travis_84_______ - how do you read that?

[14:04:51 CDT(-0500)] <travis_84_______> it's not as all inclusive or so it appears

[14:05:47 CDT(-0500)] <Bosmon> Ah, I guess the point that it says that it requires an "intel-based Mac in order to do iPhone development"

[14:05:53 CDT(-0500)] <Bosmon> Does that match with your experience, colinclark?

[14:06:06 CDT(-0500)] <colinclark> Most definitely

[14:06:20 CDT(-0500)] <colinclark> Apple offers a very restrictive platform

[14:06:48 CDT(-0500)] <colinclark> To even touch their development tools for iOS, you need to:

[14:06:48 CDT(-0500)] <colinclark> 1. have access to a Mac

[14:06:55 CDT(-0500)] <colinclark> 2. pay the fee to join their development program

[14:06:58 CDT(-0500)] <colinclark> etc.

[14:07:17 CDT(-0500)] <travis_84_______> I have found article saying Apple is blocking apps from Phone

[14:07:28 CDT(-0500)] <travis_84_______> Gap*

[14:07:29 CDT(-0500)] <colinclark> There are certainly other platforms that you can do development and testing on for free

[14:07:32 CDT(-0500)] <Bosmon> Ok.... travis_84_______... I think if you don't have access to such a thing, it is perfectly fine for this project to say that it is "portable in theory"

[14:07:40 CDT(-0500)] <Bosmon> And for you to just work on an Android version

[14:07:49 CDT(-0500)] <colinclark> and there is enough of a network of mobile developers here to help test your phone gap code elsewhere

[14:07:59 CDT(-0500)] <colinclark> Bosmon: I don't think it will be just "portable in theory"

[14:08:08 CDT(-0500)] <Bosmon> I heard reports of problems with phonegap apps in the past, but more recently I hear that Apple is mostly fine with them in their app store so long as they meet their other requirements

[14:08:17 CDT(-0500)] <colinclark> travis_84_______ just needs to hook up with a buddy who can compile an iOS app and test it for him

[14:08:24 CDT(-0500)] <Bosmon> ok

[14:08:38 CDT(-0500)] <colinclark> If I remember correctly, justin_o, jameswy, and I are all registered Apple iOS developers with access to the toolchain

[14:09:13 CDT(-0500)] <colinclark> So while you probably won't be able to verify that it works on platforms you don't have, travis_84_______, you shouldn't have any trouble developing cross-platform code with PhoneGap

[14:09:16 CDT(-0500)] <Bosmon> But I think it will require a little more than just compiling an existing app, since I imagine travis_84_______ might have to do some work on PhoneGap itself

[14:09:27 CDT(-0500)] <colinclark> Ah, aha

[14:09:30 CDT(-0500)] <Bosmon> Part of the plan is to work on a workable shim for some of the HTML5 Speech APIs

[14:09:42 CDT(-0500)] <colinclark> Yeah, that'll involve some extra buddy work

[14:09:53 CDT(-0500)] <colinclark> I think one of us would probably still be willing to help out with some testing and tweaking

[14:09:56 CDT(-0500)] <Bosmon> Cool

[14:09:57 CDT(-0500)] <colinclark> if we can get the code close

[14:10:03 CDT(-0500)] <colinclark> You want speech input? Or speech output?

[14:10:34 CDT(-0500)] <colinclark> travis_84_______: The whole "Apple blocking PhoneGap" thing is history

[14:10:37 CDT(-0500)] <travis_84_______> maybe both

[14:10:52 CDT(-0500)] <travis_84_______> but priority is input

[14:11:06 CDT(-0500)] <colinclark> It was quite early on in the life of PhoneGap, when developers had to do a lot of customization to make it work. And lazy developers weren't bothering, and thus Apple was rejecting poorly-formed PhoneGap apps

[14:11:20 CDT(-0500)] <travis_84_______> sorry slow typing today

[14:11:22 CDT(-0500)] <colinclark> they have quite clearly stated their support for PhoneGap over the past several years

[14:11:26 CDT(-0500)] <colinclark> no worries, travis_84_______

[14:11:28 CDT(-0500)] <travis_84_______> hands are stiff

[14:11:43 CDT(-0500)] <colinclark> For output, you could probably use Speak.js as a temporary shim

[14:11:48 CDT(-0500)] <colinclark> It doesn't sound very good, but it does work

[14:12:05 CDT(-0500)] <colinclark> for input, I think one of us can probably help compile and test your modifications to PhoneGap

[14:14:18 CDT(-0500)] <travis_84_______> ok that sounds good

[14:14:40 CDT(-0500)] <travis_84_______> we still have yet to really get that far

[14:15:10 CDT(-0500)] <travis_84_______> I am working on the UI/UX approach atm

[14:16:32 CDT(-0500)] <travis_84_______> do I want to start the program with a choice dialog for asking if the user needs speech or touch, or should it be detected? stuff like that

[14:17:53 CDT(-0500)] <travis_84_______> if it is detected I still can't assume they are going to use a touch device for touch and not speech input

[14:18:50 CDT(-0500)] <Bosmon> travis_84_______ - I guess ultimately user choice would have to be part of the loop, you can't necessarily expect to always detect what their context is

[14:19:03 CDT(-0500)] <Bosmon> But making a UX plan for just one modality will already be quite taxing enough

[14:20:50 CDT(-0500)] <travis_84_______> true, but I am also trying to make it that if I dev this initially for a single model of use, to be able to add-on other models easily later

[14:21:42 CDT(-0500)] <Bosmon> travis_84_______ - cool

[15:44:04 CDT(-0500)] <cindyli> anastasiac: i pushed the change for institution list template to https://github.com/cindyli/idi-theme/tree/IDI-8. I didn't go very far with the actual institution page, no update there.

[15:44:31 CDT(-0500)] <anastasiac> cindyli, thanks for the push - no problem, I'll finish it up

[15:44:42 CDT(-0500)] <cindyli> thanks, anastasiac

[16:17:28 CDT(-0500)] <colinclark> hey Bosmon

[16:17:36 CDT(-0500)] <colinclark> avtar has a question about the GPII preferences server and CouchDB

[16:17:43 CDT(-0500)] <colinclark>

[16:17:50 CDT(-0500)] <colinclark> I'm in a crazy GPII call at the moment

[16:17:55 CDT(-0500)] <colinclark> and won't be very useful to him

[16:17:58 CDT(-0500)] <avtar> hi hi

[16:18:05 CDT(-0500)] <Bosmon> Evening, AVTAR

[16:18:29 CDT(-0500)] <avtar> Bosmon: i was wondering if you currently do anything specific to set up couchdb

[16:18:41 CDT(-0500)] <Bosmon> Nothing specific that I'm aware of

[16:18:49 CDT(-0500)] <Bosmon> What kind of things were you wondering about?

[16:19:48 CDT(-0500)] <avtar> step 1: install the couchdb package

[16:19:59 CDT(-0500)] <avtar> step 2: wave hands about

[16:20:12 CDT(-0500)] <avtar> step 3: start using it as shown here https://github.com/GPII/universal/blob/master/gpii/node_modules/preferencesServer/src/preferencesServer.js

[16:20:20 CDT(-0500)] <avtar> it's step 2 that i'm curious about

[16:20:34 CDT(-0500)] <Bosmon> hahaha

[16:21:07 CDT(-0500)] <Bosmon> I believe just having anything running on localhost on port 5984 is as much as is necessary

[16:21:22 CDT(-0500)] <Bosmon> Although it might need to have its database created first

[16:21:51 CDT(-0500)] <Bosmon> To be clear, I don't think we actually have a preferences server yet which does anything

[16:21:56 CDT(-0500)] <Bosmon> That code you pointed to is just a mock : p

[16:24:18 CDT(-0500)] <Bosmon> Well, almost just a mock

[16:24:34 CDT(-0500)] <Bosmon> Looks like there is a version which expects there to be a db called "user"

[16:24:59 CDT(-0500)] <Bosmon> Sorry, a db with a document called user

[16:25:44 CDT(-0500)] <Bosmon> The only bit which necessarily commits us to anything couchlike is this URL: url: "http://localhost:5984/%db/user/%token"

[16:25:55 CDT(-0500)] <avtar> sorry, one sec

[16:32:59 CDT(-0500)] <avtar> back

[16:43:29 CDT(-0500)] <avtar> so looks like i just create this one database and document, then we're good to go

[16:47:03 CDT(-0500)] <Bosmon> Yup, it's a pretty basic system : P

[16:51:43 CDT(-0500)] <avtar> ok, thanks for helping out!

[17:11:54 CDT(-0500)] <colinclark> avtar and Bosmon: Catching up on your discussion...

[17:12:30 CDT(-0500)] <colinclark> If it helps any, I heard yura on the phone to Kasper today saying, "so, how's it going with your profile preparation work? Because Avtar wants to deploy the preferences server, and it might be helpful to have some preferences to actually store in the database."

[17:16:31 CDT(-0500)] <colinclark> So an empty database shouldn't be a surprise at this point

[17:17:24 CDT(-0500)] <avtar> ahh

[17:17:25 CDT(-0500)] <avtar> ok

[17:20:16 CDT(-0500)] <avtar> i'll put up an instance as-is and then hand over the keys to kasper and/or yura

[17:20:37 CDT(-0500)] <avtar> the sun finally came out

[17:20:43 CDT(-0500)] <avtar> just as i finished typing that

[19:46:40 CDT(-0500)] <travis_84> installing the android sdk and packages take forever...