uPortal IRC Logs-2014-01-13

[10:35:39 CST(-0600)] <apetro__> meta-fragments for the win

[10:59:29 CST(-0600)] <jwennmacher> apetro__, that's a bit of a confusing statement out of context.

[10:59:56 CST(-0600)] <jwennmacher> BTW, why do you come up as apetro_ and apetro__? Can't get enough of us that you want to hear what we say twice?

[11:01:44 CST(-0600)] <apetro_> at the moment, apetro_ is my laptop and apetro_ is my desktop

[11:02:17 CST(-0600)] <apetro_> they're in general configured to try apetro[*] or so until they find an available username

[11:02:47 CST(-0600)] <apetro_> apetro[]* ?

[11:02:50 CST(-0600)] <apetro__> anyway

[11:14:42 CST(-0600)] <apetro__> so, speaking of underscores. Naming metafragments with IDs with initial underscores?

[11:40:10 CST(-0600)] <tlev> +1 to underscores for metafragments

[11:43:28 CST(-0600)] <apetro__> well, I'm convinced.

[11:43:57 CST(-0600)] <apetro__> configuring in dev to log the layout and subsequent transformations of, incidentally, is very helpful.

[11:50:08 CST(-0600)] <apetro__> so, I'm eyeball deep in layout processing today. Kind of fun. What are y'all working on?

[11:50:19 CST(-0600)] <apetro__> @lfuller , I haven't seen an InfusionSoft portlet yet...

[11:51:43 CST(-0600)] <tlev> Trying to deploy 4.1 to our predevelopment environment

[11:52:17 CST(-0600)] <lfuller> lol

[11:52:47 CST(-0600)] <lfuller> well... please bear in mind that I signed a non compete and purposely avoided any portal/portlet related work for 2 years.

[11:53:32 CST(-0600)] <lfuller> probably, unnecessary.. but there were so many other issues to deal with, I just didn't see a need to go out of my way.

[11:53:48 CST(-0600)] <apetro__> sure. Best to cross those things out of employment agreements.

[11:54:29 CST(-0600)] <apetro__> so, at the risk of also running up against your nondisclosure , roughly what are you working on these days?

[11:54:33 CST(-0600)] <lfuller> Haven't had any customers request an integration with their portal yet.

[11:55:12 CST(-0600)] <apetro__> I've often thought that what InfusionSoft desperately needs is login via SAML, incidentally.

[11:55:45 CST(-0600)] <lfuller> Well... when I was first hired infusionsoft was in a position a lot of start ups find themselves in. Tons of code generated for that next whiz bang feature, with a whole lot of technical debt accumulating.

[11:56:24 CST(-0600)] <lfuller> Infusionsoft has a CAS integration they've gotten a good deal value from.

[11:57:17 CST(-0600)] <lfuller> So... a good deal of training engineers. Putting decent review processes in place to ensure new hires were good quality adds. Lots of refactoring.

[11:57:46 CST(-0600)] <lfuller> all the while the functional teams are cranking on new features.

[11:59:08 CST(-0600)] <lfuller> Unfortunately.. much of my time is spent on technical debt, rather than working on fun development.

[11:59:44 CST(-0600)] <lfuller> Initially.. I got a heck of a lot better at reading heap and thread dumps. Also have gotten way, way better at chasing down weird concurrency issues.

[12:09:31 CST(-0600)] <jwennmacher> apetro: what do you mean by a meta-fragment?

[12:09:45 CST(-0600)] <jwennmacher> This is for the Favorites email you sent out earlier.

[12:17:52 CST(-0600)] <apetro__> yes, the immediate use of a meta-fragment I'm working on is storing Favorited portlets into a _favorites fragment

[12:18:56 CST(-0600)] <drewwills> apetro__ happy Monday – just read and replied to your email

[12:20:04 CST(-0600)] <drewwills> is the favorites 1 portlet that stores the list of the user's portlets, or a fragment that stores the list?

[12:22:43 CST(-0600)] <apetro__> well, there are both things

[12:22:52 CST(-0600)] <apetro__> there's the _favorites DLM managed fragment

[12:23:06 CST(-0600)] <apetro__> which is a regular old Fragment except filtered out on traditional layout display

[12:23:27 CST(-0600)] <apetro__> then there's a Favorites portlet which renders the content of the _favorites fragment as the listing of the user's favorites

[12:23:37 CST(-0600)] <drewwills> and a Favorites portlet places somewhere else on a more-normal fragment?

[12:23:46 CST(-0600)] <apetro__> and which allows users to remove nodes from that fragment, add nodes to that fragment, etc.

[12:24:06 CST(-0600)] <drewwills> How does the Favorites portlet access the layout?

[12:24:08 CST(-0600)] <apetro__> yes, the Favorites portlet is placed as a regular ol' portlet in the layout

[12:24:11 CST(-0600)] <drewwills> (just curious)

[12:24:47 CST(-0600)] <apetro__> https://github.com/Jasig/uPortal/commit/2ef6997c20a9b2d2755acb649eb6f00839d07a94

[12:25:30 CST(-0600)] <drewwills> k

[12:25:42 CST(-0600)] <drewwills> I'm glad you're working on this

[12:25:46 CST(-0600)] <drewwills> sounds like fun

[12:26:06 CST(-0600)] <drewwills> so, one suggestion I want to throw out...

[12:26:44 CST(-0600)] <drewwills> I don't believe you need a "new primitive" called Meta Fragment that includes new Java code

[12:27:08 CST(-0600)] <drewwills> i believe you can smoothly use <folder type="favorites">

[12:27:12 CST(-0600)] <drewwills> or somesuch

[12:27:16 CST(-0600)] <drewwills> within a fragment

[12:27:33 CST(-0600)] <drewwills> and whether that fragment contains other stuff or doesn't is entirely up to you

[12:27:59 CST(-0600)] <drewwills> the underscore is cool for a naming convention afai'm concerned

[12:28:45 CST(-0600)] <drewwills> example: https://github.com/Jasig/uPortal/commit/f6eee8a561641a3a53d0aa8d5e005adb16082c14

[12:29:01 CST(-0600)] <drewwills> type="emergency"

[12:32:46 CST(-0600)] <apetro__> interesting on the folder type. Will have to see what the code looks like that way.

[12:35:31 CST(-0600)] <drewwills> of courses it could really be any type you want, or even multiple types

[12:36:27 CST(-0600)] <drewwills> central suggestion is that I bet we don't need Java code that looks for fragments that start with underscores

[12:36:43 CST(-0600)] <apetro__> what's up with the .fragment-layout. files in this directory? https://github.com/Jasig/uPortal/tree/master/uportal-war/src/main/data/quickstart_entities/fragment-definition

[12:37:42 CST(-0600)] <apetro__> yeah, perhaps we need Java code that looks for a layout folder with type="favorites"

[12:38:07 CST(-0600)] <drewwills> aren't they all *fragment-layout.xml?

[12:38:31 CST(-0600)] <apetro__> and perhaps we need IUserLayoutNodeDescription to expose that "type" attribute

[12:38:40 CST(-0600)] <drewwills> ah yes, perhaps

[12:38:56 CST(-0600)] <drewwills> to date, i think only the XSL has looked at that

[12:38:57 CST(-0600)] <apetro__> no, most of them are fragment-definition files.

[12:39:10 CST(-0600)] <drewwills> oops

[12:39:20 CST(-0600)] <drewwills> yeah that's a surprise

[12:39:54 CST(-0600)] <drewwills> those should be moved

[12:40:25 CST(-0600)] <apetro__> k. noted locally, but I'm not going to jump on that, focused as I am on deliverables for this sprint.

[12:40:27 CST(-0600)] <jwennmacher> I think those files are named incorreclty

[12:40:56 CST(-0600)] <jwennmacher> they are definition files

[12:41:02 CST(-0600)] <drewwills> bleh

[12:41:04 CST(-0600)] <apetro__> aha

[12:41:20 CST(-0600)] <jwennmacher> easy fix at least

[12:41:23 CST(-0600)] <apetro__> add that to the set of import sanity checks

[13:12:22 CST(-0600)] <goldmantx1> @apetro_ IS UW going to be looking at/working on the customize drawer?

[13:33:51 CST(-0600)] <drewwills> apetro__ if you, tlev, and/or anyone at WI would like a teleconference to explore any of these topics further, I can make time for that

[13:34:57 CST(-0600)] <apetro__> @goldmantx1 I don't think UW will be using customize drawer in respondr and so won't be working on that as such, though I think we are pretty interested in our Favorites work being "compatible" for some reasonable meaning of "compatible"

[13:35:36 CST(-0600)] <apetro__> as in, don't want to be breaking things for everyone else, and do want Favorites to be reasonably adoptable elsewhere, but not looking to do a bunch of extraneous work if Favorites isn't resonating.

[13:36:46 CST(-0600)] <apetro__> @drewwills, appreciate the conference call offer. Will keep in mind. Am hoping that by doing work on feature branches (and my un-favorites is on a branch off a feature branch!) will facilitate sharing and code reviewing before things hit master

[13:37:33 CST(-0600)] <drewwills> excellent

[14:37:52 CST(-0600)] <apetro__> wrt representing the layout folder element @type attribute in Java, it looks like UserLayoutFolderDescription wants to be but is not currently representing these

[17:39:50 CST(-0600)] <drewwills> are you still there tlev?