fluid-tech IRC Logs-2012-11-22

[13:06:48 CST(-0600)] <kasper> Bosmon: re your review

[13:07:10 CST(-0600)] <kasper> I'm having troubles to find a nicer solution than my current fluid.merge

[13:07:15 CST(-0600)]

[13:07:44 CST(-0600)] <Bosmon> Yes - but why doesn't this require a perfectly standard merge operation?

[13:08:16 CST(-0600)] <kasper> as opposed to the "preserve" option?

[13:08:19 CST(-0600)] <Bosmon> yes

[13:08:59 CST(-0600)] <kasper> wehn doing it without the "preserve" it would simply overwrite "outervar" and set it as the new value

[13:08:59 CST(-0600)] <kasper> as opposed to merging

[13:08:59 CST(-0600)] <Bosmon> In general, I'm increasingly recommending people to just use $.extend for their merging purposes

[13:09:02 CST(-0600)]

[13:09:19 CST(-0600)] <kasper> ok, I can give that a go

[13:09:25 CST(-0600)] <Bosmon> The semantics of fluid.merge are going to change shortly and mostly it is less likely to do what people want

[13:10:27 CST(-0600)] <kasper> haha, so you're knowingly changing it to something that is less likely to do what people want.. That seems like a rather odd decision

[13:11:12 CST(-0600)] <Bosmon> I can't help it - it needs to be improved

[13:11:31 CST(-0600)] <Bosmon> And $.extend had always been there as a more primitive and probably more performant utility

[13:14:07 CST(-0600)] <kasper> Bosmon: cool.. just substituted with $.extend and it worked

[13:14:41 CST(-0600)] <Bosmon> excellent!

[13:14:53 CST(-0600)] <Bosmon> Now just your peculiar approach to XML-encoding remains

[14:28:22 CST(-0600)] <kasper> yura1:

[14:29:07 CST(-0600)] <yura> yes kasper

[14:29:40 CST(-0600)] <kasper> I'm currently trying to get andrei working, and there seem to be an issue with application specific stuff still (in hierarchical at least)

[14:30:03 CST(-0600)] <kasper> yura: not sure if it's just me doing something wrong though

[14:30:17 CST(-0600)] <kasper> https://github.com/kaspermarkus/universal/blob/GPII-52/testData/preferences/andrei.json#L23

[14:30:33 CST(-0600)] <kasper> so first of all, did we decide that solutions should contain their ID rather than name?

[14:30:46 CST(-0600)] <yura> kasper: so they just dont show up ?

[14:33:29 CST(-0600)] <kasper> yura: yeah, he only matches on 'fluid.uiOptions.linux' and 'org.gnome.desktop.interface'

[14:33:38 CST(-0600)] <kasper> both of which I think are able to handle common prefs

[14:34:42 CST(-0600)] <yura> kasper: ok can you do a quick test ? try putting a simple 1 to 1 map in default transformation block and see if that map works and you get your app

[14:36:23 CST(-0600)] <kasper> as in the capabilitiesTransformations block of the solution entry?

[14:36:25 CST(-0600)] <kasper> yura: ^

[14:37:00 CST(-0600)] <yura> kasper: no ontologies.json file in test data

[14:40:18 CST(-0600)] <kasper> yura: I'm unsure what 1 to 1 map you want me to put there.. something like: "display.screenReader.applications.org.gnome.orca.id": "display.screenReader.applications.org.gnome.orca.id",

[14:40:26 CST(-0600)] <kasper> argh, crap.. need escapes

[14:40:30 CST(-0600)] <kasper> hang on

[14:40:43 CST(-0600)] <yura> it doesnt i think

[14:41:17 CST(-0600)] <kasper> yura: well in either case, it doesn't change anything

[14:41:39 CST(-0600)] <yura> hmm

[14:41:45 CST(-0600)] <yura> i might be misunderstanding the problem

[14:43:28 CST(-0600)] <kasper> yura: when logging in with andrei (hierarchical) the only two things i get are 'fluid.uiOptions.linux' and 'org.gnome.desktop.interface'

[14:43:48 CST(-0600)] <kasper> where as I should have got webanywhere, orca and the keyboard beep enabled

[15:36:39 CST(-0600)] <yura> hey Bosmon

[15:38:08 CST(-0600)] <Bosmon> JURA!

[15:41:37 CST(-0600)] <yura> Bosmon: do we really need a default ontology entry? it feels like it adds a lot of complexity

[15:43:00 CST(-0600)] <yura> meaning it looks like we would need to specify a straight value map between all existing preferences

[15:43:08 CST(-0600)] <Bosmon> yura - it isn't the "default" ontology but the "registry" one

[15:43:12 CST(-0600)] <yura> yes

[15:43:17 CST(-0600)] <yura> not flat one

[15:43:21 CST(-0600)] <Bosmon> ?

[15:43:22 CST(-0600)] <yura> the one used by the match maker

[15:43:41 CST(-0600)] <Bosmon> The one used by the matchmaker should be the A4A ontology

[15:43:47 CST(-0600)] <Bosmon> Is this what you are calling the "default" ontology?

[15:44:21 CST(-0600)] <yura> Bosmon: the one currently used by he match maker

[15:45:39 CST(-0600)] <Bosmon> yura - I guess we will need to circle round

[15:45:56 CST(-0600)] <Bosmon> There should be only TWO ontologies in the system

[15:46:01 CST(-0600)] <yura> yes

[15:46:05 CST(-0600)] <Bosmon> i) the one that encodes the registry terms, ii) the A4A ontology

[15:46:12 CST(-0600)] <Bosmon> Are you saying there is a 3rd one?

[15:46:52 CST(-0600)] <yura> no , i m talking about the one called default in the system: https://github.com/GPII/universal/blob/master/testData/ontologies/default.json

[15:48:13 CST(-0600)] <Bosmon> I can't say I can recognise what that is

[15:48:24 CST(-0600)] <Bosmon> And should clearly have been much more careful during code review....

[15:48:29 CST(-0600)] <Bosmon> Can you describe what this ontology is?

[15:48:48 CST(-0600)] <yura> it's nothing

[15:48:50 CST(-0600)] <Bosmon> Somehow this document appears to be transforming the A4A ontology onto itself

[15:48:54 CST(-0600)] <Bosmon> But I'm not sure what role it has

[15:48:55 CST(-0600)] <yura> yes

[15:49:20 CST(-0600)] <yura> it has no role other than ontology documentation

[15:49:59 CST(-0600)] <Bosmon> There's a problem with these transformation documents, in that they don't specify the PAIR of ontologies at either end of the transformation

[15:50:10 CST(-0600)] <Bosmon> I think if this was clearer, it would be easier to see why the "default" document shouldn't exist

[15:51:41 CST(-0600)] <yura> Bosmon: i guess my original point is that we dont really need this document

[15:52:28 CST(-0600)] <Bosmon> yura - my follow-up point is that if the role of these documents was clearly specified, we would know how to make this unnecessary document go away

[15:53:07 CST(-0600)] <Bosmon> So yes - your original point is completely correct

[15:54:54 CST(-0600)] <colinclark> As I understand it, this bit of transformation is there in order to support preferences documents that have already been converted to the AccessForAll 2.0 hierarchical representation

[15:54:55 CST(-0600)] <Bosmon> To the extent that we have a thing called "the default ontology", this should simply consist of a label, recording which of the 2 real ontologies we have is the default one (if we need that for some purpose - I'm not convinced we do)

[15:55:12 CST(-0600)] <Bosmon> We shouldn't have a thing called "the default ontology" which is specified through a transformation

[15:55:34 CST(-0600)] <colinclark> Which, in a sense, was perhaps what obscured us from noticing the fact that our scheme for supporting flat preferences wasn't quite yet ready

[15:56:54 CST(-0600)] <colinclark> I guess I see two issues here, if I'm not totally confused myself:

[15:57:35 CST(-0600)] <colinclark> 1. These ontology transformations should be labelled based on a. the source ontology and b. the target ontology

[15:58:03 CST(-0600)] <colinclark> this notion of "default" is really hiding the fact that what that transformation represents is really the "AccessForAll 2.0 to itself" ontology transformation

[15:58:39 CST(-0600)] <colinclark> 2. The registry format-flat preferences-is itself an ontology

[15:59:11 CST(-0600)] <colinclark> maybe 3. If we have any notion of some ontology being "default," it should presumably be the registry format. But then the question would be, "in what cases do we need a default?"

[15:59:21 CST(-0600)] <colinclark> yura, Bosmon: Am I confused?

[15:59:23 CST(-0600)] <colinclark>

[15:59:32 CST(-0600)] <Bosmon> colinclark - I believe that you are not confused at all

[15:59:42 CST(-0600)] <colinclark> what a rarity

[15:59:51 CST(-0600)] <colinclark> yura: Do you think I'm confused?

[16:06:32 CST(-0600)] <yura> so Bosmon, the document that makes sense : https://github.com/GPII/universal/blob/master/testData/ontologies/A4A-2.0.json what should it be called

[16:07:23 CST(-0600)] <Bosmon> The name of the document should list the 2 ontologies

[16:07:51 CST(-0600)] <Bosmon> This should probably also appear inside it somehow... but I guess you couldn't find it without the naming convention, so that isn't so important

[16:08:59 CST(-0600)] <Bosmon> I guess we need to pick some kind of delimeter character...

[16:09:10 CST(-0600)] <yura> "/"

[16:09:15 CST(-0600)] <Bosmon> ha

[16:09:19 CST(-0600)] <Bosmon> Very funny

[16:09:20 CST(-0600)] <Bosmon> ok

[16:09:30 CST(-0600)] <yura> really ?

[16:09:55 CST(-0600)] <Bosmon> Well sure, why not just use a further level of subdirectories

[16:10:01 CST(-0600)] <yura> so it would be "to/from"

[16:10:07 CST(-0600)] <yura> ya that's what i thought

[16:10:15 CST(-0600)] <yura> so as colinclark was mentioning

[16:10:29 CST(-0600)] <yura> A4A-2.0/flat was it ?

[16:10:39 CST(-0600)] <yura> oh he s gone

[16:10:42 CST(-0600)] <yura> does that make sense ?

[16:10:53 CST(-0600)] <yura> Bosmon: ^

[16:10:53 CST(-0600)] <Bosmon> We should probably call it "registry"

[16:10:57 CST(-0600)] <yura> ok

[16:10:57 CST(-0600)] <yura> so

[16:10:59 CST(-0600)] <Bosmon> Which will be a bit more descriptive

[16:11:05 CST(-0600)] <yura> "registry/flat"

[16:11:06 CST(-0600)] <yura> ?

[16:11:10 CST(-0600)] <Bosmon> no, the other one

[16:11:15 CST(-0600)] <Bosmon> Mr. Poirot : P

[16:11:20 CST(-0600)] <Bosmon> A4A-2.0/registry : P

[16:11:30 CST(-0600)] <yura> riight

[16:35:20 CST(-0600)] <kasper> Bosmon: ... got a question about my annoying &amp; workaround in the code

[16:35:34 CST(-0600)] <Bosmon> KASPARD!

[16:35:42 CST(-0600)] <kasper> ... so what I currently have will mess up &lt; &gt; etc

[16:35:51 CST(-0600)] <kasper> making the &amp;lt;

[16:35:53 CST(-0600)] <kasper> etc

[16:35:58 CST(-0600)] <kasper> which isn't very nice

[16:36:06 CST(-0600)] <Bosmon> It sounds like it's just worth ditching that worthless library entirely

[16:36:46 CST(-0600)] <kasper> Bosmon: yeah - BUT a temporary solution is better than none at all atm.. especially considering two of the implementers going to the hackathon are using XML file

[16:36:48 CST(-0600)] <kasper> s*

[16:38:11 CST(-0600)] <Bosmon> Just looking into its code

[16:38:12 CST(-0600)] <Bosmon> https://github.com/GPII/universal/blob/master/node_modules/xml-mapping/node_modules/xml-writer/lib/xml-writer.js#L161-L165

[16:39:12 CST(-0600)] <Bosmon> It does look horribly broken

[16:39:46 CST(-0600)] <Bosmon> Although it does look like it's at least ATTEMPTING to escape the characters

[16:40:19 CST(-0600)] <kasper> yeah, but I checked (removed the 4 lines from our code) and it caused one of the tests to fail

[16:40:36 CST(-0600)] <kasper> I think perhaps the issue is that they only take it into consideration in attributes and not in values

[16:41:30 CST(-0600)] <kasper> ie: <a param="foo&bar">boo&far</a> ---> <a param="foo&amp;bar">boo&far</a>

[16:41:40 CST(-0600)] <Bosmon> Very strange

[16:41:44 CST(-0600)] <kasper> well.. that came out pretty nonsensy

[16:42:00 CST(-0600)] <kasper> but yeah, guess you get my point

[16:42:09 CST(-0600)] <Bosmon> It does look as if it is attempting to escape the text

[16:42:17 CST(-0600)] <Bosmon> But for some reason only if there are any attributes

[16:42:52 CST(-0600)] <kasper> stupid XML parser

[16:42:54 CST(-0600)] <kasper>

[16:43:49 CST(-0600)] <kasper> I guess the question is whether to leave it in or remove it completely and make it a requirement that you cannot have &s in the XML doc

[16:44:07 CST(-0600)] <kasper> both are pretty horrible options

[16:44:25 CST(-0600)] <kasper> not sure which is worst

[16:47:12 CST(-0600)] <Bosmon> Don't be ridiculous : P

[16:47:13 CST(-0600)] <Bosmon> We will fix it

[16:47:39 CST(-0600)] <Bosmon> To start with, make your own fork of the project and open up its test cases

[16:47:41 CST(-0600)] <Bosmon> https://github.com/GPII/universal/blob/master/node_modules/xml-mapping/node_modules/xml-writer/test/misc.js

[16:47:46 CST(-0600)] <Bosmon> And add one of your own to these

[16:47:51 CST(-0600)] <Bosmon> And then see why the damn thing is broken

[16:53:11 CST(-0600)] <kasper> I hate you antranig

[16:54:12 CST(-0600)] <kasper>

[16:58:46 CST(-0600)] <Bosmon> Well, you hate CATTTs too

[16:58:51 CST(-0600)] <Bosmon> I know it doesn't count for anything : P

[17:05:12 CST(-0600)] <kasper> Bosmon: looks like they added quite a few tests since we started using it .. https://github.com/lindory-project/node-xml-writer/blob/master/test/misc.js

[17:05:18 CST(-0600)] <Bosmon> ok

[17:05:24 CST(-0600)] <Bosmon> Is it possible they even fixed the bugg?

[17:05:26 CST(-0600)] <kasper> including loads of &amp; tests – probably found the bug

[17:05:30 CST(-0600)] <Bosmon> Awesome

[17:05:35 CST(-0600)] <kasper> at least the tests runs

[17:05:38 CST(-0600)] <Bosmon> Better just do an update then....

[17:05:41 CST(-0600)] <Bosmon> I'm sure you hate me less now : p

[17:05:42 CST(-0600)] <kasper> now I dont hate you quite as much

[17:05:45 CST(-0600)] <kasper> hahahahaha

[17:06:07 CST(-0600)] <Bosmon> We may as well reap SOME of the benefits of being "social coders"

[17:06:11 CST(-0600)] <Bosmon> Rather than merely being scumbags

[17:06:20 CST(-0600)] <kasper> haha

[17:06:47 CST(-0600)] <Bosmon> Once again I am listening to the works of DENMARK'S FOREMOST COMPOSER.......

[17:07:34 CST(-0600)] <kasper> is it pleasing you?

[17:07:47 CST(-0600)] <Bosmon> yes

[17:07:53 CST(-0600)] <kasper> perhaps I should actually try listening to him as well

[17:08:32 CST(-0600)] <Bosmon> It is a great one to destroy the neighbours with

[17:09:08 CST(-0600)] <kasper> haha, especially 10 past midnight

[17:26:33 CST(-0600)] <kasper> Bosmon: ok, updated pull request

[17:27:08 CST(-0600)] <Bosmon> I guess a REPOSITORY is a bit like a TREASURY

[17:27:15 CST(-0600)] <Bosmon> So it makes sense that as a KASPARD you would be in charge of one

[17:27:27 CST(-0600)] <Bosmon> GASPARD DE LA HUBB!

[17:27:33 CST(-0600)] <kasper> haha

[17:27:42 CST(-0600)] <kasper> I like that

[17:27:55 CST(-0600)] <kasper> makes me sound important

[17:28:10 CST(-0600)] <Bosmon> We will have to vote you to an official GPII title

[17:41:08 CST(-0600)] <kasper> perhaps we can use the Implementers workshop setting up the facilities and preparing the inauguration

[17:46:12 CST(-0600)] <Bosmon> I think a GASPARD needs some kind of CROWN too

[17:46:29 CST(-0600)] <Bosmon> That represents all the RICHES that he holds within his REPOSITORY : P

[17:52:18 CST(-0600)] <kasper> haha, it's gonna be AWESOME!