fluid-tech IRC Logs-2013-07-03

[02:48:03 CDT(-0500)] <chris> hello guys

[02:48:11 CDT(-0500)] <chris> chris from CERTH here

[02:48:29 CDT(-0500)] <Guest93798> anyone here?

[14:15:17 CDT(-0500)] <colinclark> how's it going, nanook_?

[14:15:22 CDT(-0500)] <colinclark> anything new or cool today?

[14:16:45 CDT(-0500)] <nanook_> hey colin. not really. going to get some work done in the next few hours though

[14:19:29 CDT(-0500)] <colinclark> that's great

[14:19:57 CDT(-0500)] <colinclark> I was planning out improvements to the Flocking scheduler last night

[14:20:11 CDT(-0500)] <colinclark> I'll post them to Github so you can see where I'm planning to go with it

[14:20:21 CDT(-0500)] <colinclark> it likely won't affect your work for a while, but I'll keep you in the loop

[14:20:38 CDT(-0500)] <nanook_> sure. thats cool.

[14:53:19 CDT(-0500)] <yzen> cindyli: do you know what actual preference name we have for:

[14:53:25 CDT(-0500)] <yzen> layoutControls

[14:53:33 CDT(-0500)] <cindyli> toc

[14:53:38 CDT(-0500)] <yzen> links and inputs larger

[14:54:06 CDT(-0500)] <yzen> that's the actual schema name right ? fluid.uiOptions.toc ?

[14:54:15 CDT(-0500)] <cindyli> those are for links control panel, yzen

[14:54:40 CDT(-0500)] <cindyli> yes for your other 2 questions

[14:55:14 CDT(-0500)] <yzen> ok

[14:55:29 CDT(-0500)] <yzen> so there's no such thing as layout controls any more right

[14:56:18 CDT(-0500)] <cindyli> layout controls will only contain table of contents, yzen

[14:56:53 CDT(-0500)] <cindyli> do you think we should rename "layoutControls" to "tableOfContents"?

[14:57:11 CDT(-0500)] <yzen> cindyli: well not exactly

[14:57:24 CDT(-0500)] <cindyli> ok

[14:57:28 CDT(-0500)] <yzen> cindyli: I'm trying to figure out what's the name you are using for your top level grade name

[14:58:13 CDT(-0500)] <cindyli> do you mean the very top? that would be defined in auxiliary schema, the first line

[15:14:42 CDT(-0500)] <cindyli> michelled: i issued a pull request with more UIO renaming: https://github.com/fluid-project/infusion/pull/355. can you take a look when you have time? If it looks too big to complete before your vacation, drop me or Justin_o an email so Justin_o can have a look.

[15:37:55 CDT(-0500)] <yzen> Bosmon: if i define an argument map for the grade

[15:38:20 CDT(-0500)] <yzen> Bosmon: do i need to define an appropriate init function as well ?

[15:38:52 CDT(-0500)] <Bosmon> yzen - assuming one of the existing ones won't do

[15:39:00 CDT(-0500)] <Bosmon> But I am concerned about your desire to define an argument map

[15:39:05 CDT(-0500)] <Bosmon> What is the use case?

[15:39:22 CDT(-0500)] <Bosmon> I'm not sure that even the argument maps we have already are not a mistake

[15:39:32 CDT(-0500)] <Bosmon> But it is too late to go back on our 2007 (container, options) signature

[15:40:26 CDT(-0500)] <yzen> Bosmon: well , we are talking about the builder

[15:40:27 CDT(-0500)] <Bosmon> Remember that in the real world, noone calls functions any more

[15:40:39 CDT(-0500)] <Bosmon> So they don't care what signatures things have...

[15:41:00 CDT(-0500)] <yzen> Bosmon: as you might remember , it only works with 2 thing primary and aux schemas

[15:41:03 CDT(-0500)] <yzen> things

[15:41:04 CDT(-0500)] <yzen> args

[15:41:26 CDT(-0500)] <Bosmon> Well, surely they are just elements in an options structure

[15:41:32 CDT(-0500)] <Bosmon> Why should they be direct function arguments?

[15:42:09 CDT(-0500)] <Bosmon> It will already annoy people by having to write "null" when they don't want a primary schema.....

[15:43:58 CDT(-0500)] <Bosmon> You could just make the "UIOptions Builder" a plain function of course

[15:46:11 CDT(-0500)] <yzen> Bosmon: http://25.media.tumblr.com/tumblr_m4rrd0U3NS1r05tymo1_250.jpg

[15:47:44 CDT(-0500)] <Bosmon> yzen - didn't we have a new one of those

[15:47:57 CDT(-0500)] <yzen>

[15:48:09 CDT(-0500)] <Bosmon> Oh yes, here it is

[15:48:09 CDT(-0500)] <Bosmon> http://jonyiveredesignsthings.tumblr.com/image/52664694149

[15:48:19 CDT(-0500)] <Bosmon> The one that explains your recent enthusiasm for TypeScript

[15:49:00 CDT(-0500)] <yzen> Bosmon: what takes precedence , supplied schema or schema grades?

[15:49:09 CDT(-0500)] <Bosmon> yzen - supplied schema takes preference

[15:49:15 CDT(-0500)] <yzen> ok

[15:49:30 CDT(-0500)] <Bosmon> And overrides schema grades for any preferences that it covers

[15:49:40 CDT(-0500)] <Bosmon> The rest are filled in from the schema grades

[15:59:02 CDT(-0500)] <yzen> Bosmon: can we use fund in place of funcName?

[15:59:18 CDT(-0500)] <Bosmon> yzen - right now they are interchangeable, yes, although that is a bit unfortunate

[15:59:31 CDT(-0500)] <yzen> even in invokers ?

[15:59:37 CDT(-0500)] <Bosmon> They really should have a different implementation, so you should be careful to try to use the right one

[16:00:26 CDT(-0500)] <Bosmon> For example, using an IoC expression for funcName which evaluates to a string should then look it up as a global name....

[16:00:36 CDT(-0500)] <Bosmon> If you find a problem, please produce a JIRA

[16:00:45 CDT(-0500)] <Bosmon> Yes, even in invokers

[16:01:08 CDT(-0500)] <Bosmon> We've had "direct function" invokers for a while

[16:01:14 CDT(-0500)] <Bosmon> Since I did the Uploader update, I believe

[16:02:53 CDT(-0500)] <Bosmon> You can even write "this-ist" blocks in invokers too

[16:02:55 CDT(-0500)] <Bosmon> https://github.com/fluid-project/infusion/blob/master/src/webapp/components/uploader/js/HTML5UploaderSupport.js#L311-L316

[16:06:15 CDT(-0500)] <Bosmon> yzen: https://github.com/amb26/infusion/blob/FLUID-4035/src/webapp/demos/keyboard-a11y/js/five-star.js#L118-L121

[16:06:37 CDT(-0500)] <yzen>

[16:06:39 CDT(-0500)] <yzen> nice

[16:08:58 CDT(-0500)] <yzen> Bosmon: is it possible to write an indexer that returns the matching grade and that's it

[16:10:01 CDT(-0500)] <Bosmon> yzen - I guess you could have that effect by just making a "constant return" from your indexing function

[16:10:10 CDT(-0500)] <Bosmon> That is, having it return "index" whatever defaults it is given

[16:10:18 CDT(-0500)] <Bosmon> But I'm not sure whether that has the effect you're asking for

[16:10:21 CDT(-0500)] <Bosmon> Could you give an example?

[16:10:55 CDT(-0500)] <yzen> actually, sorry when i index, i only want the final component names that's all

[16:11:13 CDT(-0500)] <Bosmon> yzen - the result of the index is always the component names

[16:11:16 CDT(-0500)] <Bosmon> Could you explain more?

[16:12:48 CDT(-0500)] <yzen> constant works

[16:13:33 CDT(-0500)] <colinclark> you really are a man of few words, yzen

[16:14:51 CDT(-0500)] <yzen> Bosmon, colinclark well I see how the result value is a list of grades, which is what i need, but I am trying to generate an index with a sensible key that points to that list

[16:15:11 CDT(-0500)] <yzen> Bosmon: the sensible key would be something like "schemaGrades"

[16:16:17 CDT(-0500)] <yzen> Bosmon: afaik the return of the indexer function is expected to be an array

[16:16:29 CDT(-0500)] <yzen> or object

[16:17:10 CDT(-0500)] <Bosmon> yzen - but why do you want to find ALL of them in a list?

[16:17:15 CDT(-0500)] <Bosmon> Surely you only ever want to find one of them at a time

[16:17:43 CDT(-0500)] <Bosmon> yzen - the indexer function returns a string or list of strings

[16:18:08 CDT(-0500)] <yzen> Bosmon: right

[16:18:34 CDT(-0500)] <Bosmon> yzen - why do you want them all?

[16:19:30 CDT(-0500)] <yzen> Bosmon: i need all schema grades to be able to build a complete schema, they i m doing it is by supplying all grades that are schema grades as grades of the final built schema component/grade

[16:19:44 CDT(-0500)] <Bosmon> yzen - what do you need a complete schema for?

[16:20:03 CDT(-0500)] <Bosmon> Surely you only need to complete the schema up to what was requested from the auxiliary schema

[16:20:32 CDT(-0500)] <Bosmon> If a schema grade wasn't requested either in the primary schema, or indirectly from the auxiliary schema, it shouldn't appear

[16:20:32 CDT(-0500)] <yzen> Bosmon: ok well this was one of the things we were thinking about

[16:21:04 CDT(-0500)] <yzen> how can a schema grade be requested in the primary schema if it might constitute a primary schema?

[16:21:38 CDT(-0500)] <yzen> and when you say primary schema what exactly do you mean

[16:21:53 CDT(-0500)] <Bosmon> Well, to be more precise, the primary schema influences which schema grades are NOT requested

[16:22:01 CDT(-0500)] <Bosmon> Given the schema has already chosen to override them

[16:22:08 CDT(-0500)] <Bosmon> But in fact, they will end up as schema grades anyway

[16:22:26 CDT(-0500)] <Bosmon> Since the system will have to synthesize something for the other grades to refer to

[16:48:26 CDT(-0500)] <colinclark> did we lose yzen for the day, or did the problem get solved?

[16:50:04 CDT(-0500)] <Bosmon> We lost him for a while

[16:50:09 CDT(-0500)] <Bosmon> He said he'd try to carry on the conversation later

[16:58:38 CDT(-0500)] <colinclark> cool, ok

[22:17:15 CDT(-0500)] <yzen> Bosmon: ayt?

[22:17:30 CDT(-0500)] <Bosmon> JORA!

[22:17:54 CDT(-0500)] <yzen> shall we carry on the conversation ?

[22:18:08 CDT(-0500)] <Bosmon> sure

[22:18:37 CDT(-0500)] <yzen> so this is the workflow that i head in mind :

[22:19:11 CDT(-0500)] <yzen> we might or might not have schema grades on the page ..

[22:20:27 CDT(-0500)] <yzen> or rather lets start from the other side

[22:20:52 CDT(-0500)] <yzen> the result of the primary schema builder would be a combination of 1 or more schema grades

[22:20:59 CDT(-0500)] <yzen> or rather an array

[22:22:05 CDT(-0500)] <yzen> that array would be applied as a dynamic grade to another component, the one that needs to reference things from the schema

[22:22:37 CDT(-0500)] <yzen> Bosmon: you mentioned that it makes sense to only limit them to the things that are declared in aux schema

[22:23:32 CDT(-0500)] <Bosmon> yzen - yes

[22:24:08 CDT(-0500)] <yzen> Bosmon: if there is no aux schema do we use all of primary schema ?

[22:24:50 CDT(-0500)] <Bosmon> Yes, although we don't plan to handle that case now

[22:25:05 CDT(-0500)] <Bosmon> It requires some kind of support for the "default auxiliary schema"

[22:33:08 CDT(-0500)] <yzen> ok so then the process i m thinking of is

[22:35:01 CDT(-0500)] <yzen> primary schema grade is aware of something aux schema built e.g. all top level grade names + optional supplied schema => index of only necessary schema grades + applied supplied schema if necessary => a list of schema grades that will be added to a component which will result in a complete required schema

[22:41:45 CDT(-0500)] <Bosmon> yzen - could you explain what you mean by "primary schema grade is aware of something aux schema built" ?

[22:42:08 CDT(-0500)] <yzen> sorry i mean primary schema builder

[22:42:33 CDT(-0500)] <Bosmon> And can you explain what you mean by "is aware of" and also "something" : P

[22:44:11 CDT(-0500)] <yzen> right so we want to filter out schema grades that are irrelevant to what's in aux schema

[22:44:14 CDT(-0500)] <yzen> that's what i mean

[22:44:24 CDT(-0500)] <yzen> i m not sure at what stage this filtering should happen at

[22:50:21 CDT(-0500)] <Bosmon> yzen - it will be easiest just to use the grade indexer in the standard way

[22:50:39 CDT(-0500)] <Bosmon> That is, i) find all the "preference keys" from the aux schema, and then ii) look them up in the standard schema index

[22:50:57 CDT(-0500)] <Bosmon> It will take the indexer the same amount of time to compile the index in any case, and you may as well make use of the work it does

[22:51:10 CDT(-0500)] <yzen> that makes sense

[22:51:15 CDT(-0500)] <Bosmon> Otherwise you will just be doing lots of silly expensive linear scans through the total grade list

[23:34:43 CDT(-0500)] <Bosmon> Ah yzen

[23:34:47 CDT(-0500)] <Bosmon> This article is amazing: http://n.exts.ch/2013/07/syncing_couchdb

[23:35:45 CDT(-0500)] <Bosmon> "Don't go back to throwing away data just because it looks old." : P