jasig-ssp IRC Logs-2013-05-01

[10:49:30 CDT(-0500)] <JasonElwood> Jim- can you take a look at SSP-733 and let me know what has been completed please? I know the external tables have been altered but I don't believe the model has been changed

[11:28:40 CDT(-0500)] <js70> k

[11:43:54 CDT(-0500)] <js70> the models are in place for external_faculty_course_roster and external_faculty_course mentioned in the ticket, I know that I created moel support where section_number and section_code were added (just what I do). However, there is no support for section_code in the Early Alert API that I can see.

[11:44:28 CDT(-0500)] <js70> Apologize for the delay, getting ducks in a row for the upcoming iteration.

[12:02:50 CDT(-0500)] <dmccallum54> so, the interesting thing about the EA API is that it doesn't really have a notion of a course identifier at all

[12:02:57 CDT(-0500)] <dmccallum54> (section, for that matter)

[12:03:36 CDT(-0500)] <dmccallum54> it's got courseName and courseTitle and it pretty much takes the client's word for it that those are meaningful and correct

[12:03:56 CDT(-0500)] <dmccallum54> historically, the EA portlet happens to stick formatted_course into courseName

[12:04:39 CDT(-0500)] <dmccallum54> but if i were coming to this fresh there's no way that's what i'd expect

[12:06:22 CDT(-0500)] <TonyUnicon> i am trying to suggest to tyler that he change his irc nickname to tpain

[12:06:28 CDT(-0500)] <TonyUnicon> negotiations are not going well

[12:06:52 CDT(-0500)] <dmccallum54> but, the EA service does assume that this convention is actually part of the API and has lookups against external data that assume courseName is formatted_course

[12:08:22 CDT(-0500)] <dmccallum54> which is a problem, as discussed a few times on list, b/c we can end up with multiple external data records, e.g. when the same instructor is teaching the same course in multiple terms or the same course multiple times in the same term

[12:09:41 CDT(-0500)] <dmccallum54> in any event… that's a long way of saying that it might be possible to get away with just limiting the changes to the EA portlet and the roster lookup API to support section codes, but we probably want to make corresponding changes to the EA APi as well

[13:02:40 CDT(-0500)] <js70> hey jason:

[13:02:45 CDT(-0500)] <js70> https://wiki.jasig.org/display/SSP/SSP+v2.0+MAP+Tool+Phase+2-+Email+Plan

[13:03:12 CDT(-0500)] <js70> can you give me some guidance on why radio vs check box, how you want this to work?

[13:04:20 CDT(-0500)] <js70> Also is time expected outside class just 2x the credit hours. ie how to calculate.

[13:05:33 CDT(-0500)] <JasonElwood> what specifically is your question about radio buttons? each of the radio buttons are a pair, either/or scenario

[13:05:43 CDT(-0500)] <JasonElwood> time outside is 2 X credit hours, yes

[13:07:03 CDT(-0500)] <js70> the specific question is why. you have check boxes for some and radio buttons for others. To the since they are pairs and the context is obvious, why have radio buttons. .ie

[13:07:04 CDT(-0500)] <JasonElwood> I just updated the mock-up to show the pairs better

[13:07:16 CDT(-0500)] <js70> instead of With Course Description

[13:07:21 CDT(-0500)] <JasonElwood> so you can't select both

[13:07:23 CDT(-0500)] <js70> without Course Description

[13:07:34 CDT(-0500)] <js70> you can't with a check box either:

[13:08:11 CDT(-0500)] <js70> just checkbox Wtih Course Description. Unchecked means the Course Description will not be shown.

[13:08:26 CDT(-0500)] <js70> it would significantly reduce ui clutter in this case.

[13:08:55 CDT(-0500)] <JasonElwood> that would be fine. so default is without, check to add

[13:09:16 CDT(-0500)] <js70> What ever you prefer.

[13:09:43 CDT(-0500)] <js70> I would keep the Print Map with Options vs Print Map in Matrix Format as that is not a with or without.

[13:09:47 CDT(-0500)] <js70> just different

[13:09:55 CDT(-0500)] <JasonElwood> right. first one is a radio

[13:11:44 CDT(-0500)] <js70> Also, is there any reason we aren't supporting , delimited addressing for emails. So we just need a To: and a CC: ?

[13:13:06 CDT(-0500)] <JasonElwood> check out the mock-up now

[13:13:37 CDT(-0500)] <js70> That's what I'm talking about!:^)

[13:13:56 CDT(-0500)] <JasonElwood> cool. thanks for the improvements

[13:28:26 CDT(-0500)] <JasonElwood> for the sample data, you guys need external_course, external_course_term, external_course_tag, external_course_program, external_program?

[13:29:25 CDT(-0500)] <TonyUnicon> yes

[13:48:52 CDT(-0500)] <dmccallum54> for templates i believe we're going to need external_department and external_division as well

[13:49:52 CDT(-0500)] <js70> tru dat

[14:08:06 CDT(-0500)] <TonyUnicon> yes

[14:08:08 CDT(-0500)] <TonyUnicon> good point

[14:08:17 CDT(-0500)] <TonyUnicon> ill throw 'er on the pile

[14:08:39 CDT(-0500)] <TonyUnicon> almost done with the external course term stuff

[14:08:46 CDT(-0500)] <TonyUnicon> I forgot how much I had done for it during phase 1

[14:12:04 CDT(-0500)] <TonyUnicon> Dan how do you feel about this hybrid approach, slightly different than what we talked about so

[14:12:11 CDT(-0500)] <TonyUnicon> the validate course term call would look like

[14:14:00 CDT(-0500)] <TonyUnicon> "/1/reference/course/

[14:15:10 CDT(-0500)] <TonyUnicon> response would be

[14:15:16 CDT(-0500)]

[14:15:29 CDT(-0500)] <TonyUnicon> or does that just spit in the face of rest?

[14:16:39 CDT(-0500)] <TonyUnicon> two request params are fine too, just wondered if you thought this was slightly better dmccallum54

[14:17:13 CDT(-0500)] <dmccallum54> it's a bit spitty

[14:17:28 CDT(-0500)] <dmccallum54> it could be justified for performance reasons

[14:18:12 CDT(-0500)] <TonyUnicon> better /1/reference/course/validateTerm?courseCode=MATH101?termCode=FA15

[14:18:20 CDT(-0500)] <TonyUnicon> ?

[14:19:12 CDT(-0500)] <TonyUnicon> how would it impact performance? it would not change the underlying query that gets executed

[14:21:15 CDT(-0500)] <dmccallum54> so

[14:23:22 CDT(-0500)]

[14:24:09 CDT(-0500)]

[14:24:25 CDT(-0500)] <dmccallum54> the performance comment was w/r/t how much data is potentially sent over the wire

[14:24:26 CDT(-0500)] <TonyUnicon> ok

[14:25:03 CDT(-0500)] <dmccallum54> i.e. you're just sending a boolean back, which is the representatin of the ./validateTerm resource. as opposed to a big course struct, which is the representation of members of the "root" /course collection

[14:25:25 CDT(-0500)] <TonyUnicon> yup, eventually ill stop asking dumb rest questions

[14:25:30 CDT(-0500)] <dmccallum54> not dumb

[14:25:56 CDT(-0500)] <dmccallum54> it's just a shift away from SOAPy/RPC-y operation-oriented design to resource oriented design

[14:26:05 CDT(-0500)] <TonyUnicon> yeah

[14:27:46 CDT(-0500)] <dmccallum54> on the /1/reference… do we have any good conventions for indicating "this is external data" in the resource design?

[14:28:04 CDT(-0500)] <dmccallum54> transcript is external, but that looks just like any other operational resource

[14:29:47 CDT(-0500)] <dmccallum54> same for "instruction"

[14:30:01 CDT(-0500)] <TonyUnicon> we can introduce one, doesn't look like any convention exists

[14:30:19 CDT(-0500)] <TonyUnicon> look at ExternalStudentRecordsController

[14:30:29 CDT(-0500)] <TonyUnicon> its just

[14:30:30 CDT(-0500)]

[14:30:40 CDT(-0500)] <dmccallum54> yeah

[14:30:45 CDT(-0500)] <TonyUnicon> for term

[14:30:46 CDT(-0500)] <TonyUnicon> @RequestMapping("/1/reference/term")

[14:31:28 CDT(-0500)] <dmccallum54> well, i guess there's the "convention"

[14:31:51 CDT(-0500)] <TonyUnicon> what is?

[14:31:55 CDT(-0500)] <TonyUnicon> that there is none?

[14:33:23 CDT(-0500)] <js70> yeah, when I added most of those calls I saw the convention as data for person / person, reports, data for anything else reference. Can't verify consistency.

[14:33:24 CDT(-0500)] <dmccallum54> external terms are /1/reference. so are external courses

[14:34:02 CDT(-0500)]

[14:34:18 CDT(-0500)] <TonyUnicon> reference data could be external

[14:34:21 CDT(-0500)] <TonyUnicon> internal

[14:34:22 CDT(-0500)] <TonyUnicon> rateher

[14:34:24 CDT(-0500)] <TonyUnicon> rather

[14:34:42 CDT(-0500)] <js70> exactly.

[14:35:19 CDT(-0500)] <dmccallum54> ok. well. the /1/reference convention is there for direct access to external data outside the context of an operational entity. guess we should stick with it. and there's been talk about exposing admin screens for read/write management of external data.

[14:35:41 CDT(-0500)] <dmccallum54> so it's maybe kinda the same thing as what we've thought of as reference data to date

[14:36:37 CDT(-0500)] <TonyUnicon> ahh

[14:37:28 CDT(-0500)] <TonyUnicon> all this url talk is making me hungry

[18:20:32 CDT(-0500)] <dmccallum54> SSP 2.0.0-b1 and SSP-Platform 1.2.0-b1 have been released

[18:20:37 CDT(-0500)] <dmccallum54> woohoo

[18:20:55 CDT(-0500)] <TylerBain> yay!

[18:21:16 CDT(-0500)] <TylerBain> Do we get a cupcake from a competing team now?

[18:21:42 CDT(-0500)] <dmccallum54> it's pointless to compete with us

[18:21:57 CDT(-0500)] <dmccallum54> i think everyone's given up by now

[18:22:33 CDT(-0500)] <TylerBain> good, good

[18:23:34 CDT(-0500)] <js701> personally, I like gelato :^)

[18:29:53 CDT(-0500)] <TylerBain> that's great... still havn't gotten YOUR treat for your promotion, perhaps there's an idea there....

[18:34:58 CDT(-0500)] <js701> whoa, the burn it hurts.

[18:35:25 CDT(-0500)] <TylerBain> All I'm saying is that SOMEONE brought in gelato, and SOMEONE ELSE brought in cupcakes....