uPortal IRC Logs-2012-10-09

Owned by Apereo Infrastructure
Last updated: Oct 09, 2012
6 min read

[10:58:34 CDT(-0500)] <athena> EricDalquist: you get a chance to look at dmccallum54's SSP branches?

[10:58:46 CDT(-0500)] <athena> they look ok to me, though i don't imagine there's much to really look for here

[10:59:03 CDT(-0500)] <EricDalquist> yeah I was going to after I got the 4.0.7 release rolling

[10:59:07 CDT(-0500)] <athena> awesome (smile)

[10:59:11 CDT(-0500)] <EricDalquist> working on that right now

[10:59:18 CDT(-0500)] <EricDalquist> and then we'll get the SSP branches moved

[10:59:20 CDT(-0500)] <EricDalquist> should be easy

[10:59:21 CDT(-0500)] <athena> figured - saw lots of email

[10:59:30 CDT(-0500)] <athena> cool, will be good to have that project public

[10:59:33 CDT(-0500)] <EricDalquist> I think it is just adding his repo as a remote

[10:59:37 CDT(-0500)] <EricDalquist> pulling down the branches

[10:59:43 CDT(-0500)] <EricDalquist> then pushing the branches to the uPortal repo

[10:59:45 CDT(-0500)] <athena> dmccallum54: is there any documentation on how to add SSP to an existing uportal installation?

[11:00:24 CDT(-0500)] <dmccallum54> no

[11:00:56 CDT(-0500)] <dmccallum54> the docs that do exist are focused on either running SSP-Platform or a developer standalone mode: https://wiki.jasig.org/display/SSP/SSP+Development+Environment+Setup

[11:01:17 CDT(-0500)] <dmccallum54> my todo list has an entry for updating those docs to at least reflect the new repo/project names

[11:01:45 CDT(-0500)] <dmccallum54> but i don't have a task for any new docs

[11:02:03 CDT(-0500)] <athena> well, maybe can wait until the first person asks

[11:02:05 CDT(-0500)] <dmccallum54> can open a jira for the 'add to existing uportal' use case if you think it will be helpful

[11:02:15 CDT(-0500)] <athena> but i'd at least hope that some existing uportal adopters would be interested in adopting SSP as well

[11:02:42 CDT(-0500)] <athena> so far with umobile we've had a lot more interest with existing jasig participants

[11:02:43 CDT(-0500)] <athena> you guys probably have a better sense of your target audience than i do though

[11:02:48 CDT(-0500)] <dmccallum54> well

[11:05:05 CDT(-0500)] <dmccallum54> imho something like 'you can drop this (or this a collection of smaller, simpler, SSP-related portlets) into your existing uP' is the more compelling message, as compared to 'in order to run this at all you need uP'

[11:06:04 CDT(-0500)] <dmccallum54> so i hope we can get to a point where your doc is an absolute must-have

[11:17:23 CDT(-0500)] <EricDalquist> also I think the live nightly builds should be more stable again

[11:17:36 CDT(-0500)] <EricDalquist> I added logic to the quickstart to wait for tomcat/hsql to start and stop

[11:31:12 CDT(-0500)] <athena> yay

[11:35:13 CDT(-0500)] <drewwills> EricDalquist have you seen the proposed change for PERSONDIR-66? https://github.com/Jasig/person-directory/pull/2

[11:35:22 CDT(-0500)] <EricDalquist> I saw the pull

[11:35:27 CDT(-0500)] <EricDalquist> but haven't had time to digest it yet

[11:35:41 CDT(-0500)] <EricDalquist> I'm getting caught up after last week off, I'll see if I can look tomorrow morning

[11:37:25 CDT(-0500)] <drewwills> that's fine... it would be cool, on one level, to get it into 4.0.7... but not critical – I can bump the properties version in the uP pom of just one portal, once a person-dir release it cut

[11:37:44 CDT(-0500)] <EricDalquist> too late for 4.0.7 now (smile)

[11:38:06 CDT(-0500)] <drewwills> ah ok, i was on pto

[11:38:17 CDT(-0500)] <EricDalquist> well I just pushed to the tag about 5 minutes ago

[11:38:23 CDT(-0500)] <EricDalquist> and released the staging repo

[11:39:05 CDT(-0500)] <drewwills> no worries... in that case there's somewhat less danger of bringing it into rel-4-0-patches and trying it out

[11:41:24 CDT(-0500)] <drewwills> the patch as it is takes a very conservative approach... you'd have to configire an 'unmappedUsernameAttribute' property on the DAO bean explicitly to make it behave differently

[11:42:18 CDT(-0500)] <EricDalquist> ok

[11:42:35 CDT(-0500)] <EricDalquist> I'll take a look tomorrow morning

[11:42:35 CDT(-0500)] <EricDalquist> and we can get it merged then

[11:42:53 CDT(-0500)] <drewwills> np

[11:57:12 CDT(-0500)] <EricDalquist> http://www.jasig.org/uportal/download/uportal-407

[11:57:16 CDT(-0500)] <EricDalquist> does that look good?

[12:00:08 CDT(-0500)] <athena> looks good to me

[12:32:23 CDT(-0500)] <EricDalquist> athena & dmccallum54 https://github.com/Jasig/uPortal/branches

[12:32:28 CDT(-0500)] <EricDalquist> both ssp branches should be there

[12:34:21 CDT(-0500)] <dmccallum54> thanks!

[12:34:51 CDT(-0500)] <EricDalquist> take a look and let me know if it all looks good

[12:35:04 CDT(-0500)] <EricDalquist> also we'll need to get a list of people that need push access

[12:36:01 CDT(-0500)] <dmccallum54> i can get you that. it's a short list.

[12:39:01 CDT(-0500)] <dmccallum54> branches look good to me, just clicking around.

[12:39:35 CDT(-0500)] <dmccallum54> will pull them into a private repo for our educause demo and test

[12:52:46 CDT(-0500)] <athena> w00t

[15:11:49 CDT(-0500)] <dmccallum54> EricDalquist so far so good testing the ssp-master branch

[15:12:14 CDT(-0500)] <EricDalquist> great

[15:12:42 CDT(-0500)] <dmccallum54> poking around in SourceTree, the commit graph looks like what i would expect

[15:12:56 CDT(-0500)] <EricDalquist> cool, I'll try to get those committers setup soon

[15:13:13 CDT(-0500)] <dmccallum54> np

[15:13:33 CDT(-0500)] <dmccallum54> not blocking any current work as far as I know

[15:13:46 CDT(-0500)] <dmccallum54> most everything is going on in the SSP repo

[15:14:02 CDT(-0500)] <EricDalquist> cool

[16:27:13 CDT(-0500)] <jwennmacher> Should a form be able to use a renderURL for its action or must it always use an actionURL?

[16:28:51 CDT(-0500)] <EricDalquist> it can use a render URL

[16:28:55 CDT(-0500)] <EricDalquist> but it MUST use POST

[16:29:07 CDT(-0500)] <EricDalquist> and posting to render URLs ends up with bad usability profiles

[16:29:16 CDT(-0500)] <EricDalquist> like breaking the back button

[16:29:25 CDT(-0500)] <EricDalquist> where as when you post to an action url you get a 302 to the render url

[16:29:37 CDT(-0500)] <EricDalquist> and the back button in the browser skips the action and goes to the previous render

[16:29:52 CDT(-0500)] <EricDalquist> so technically, yes you can post to a render url

[16:29:57 CDT(-0500)] <EricDalquist> functionally it isn't a great idea to

[16:33:47 CDT(-0500)] <jwennmacher> I don't fully understand ... but I'll come back to that in a bit

[16:34:02 CDT(-0500)] <EricDalquist> ok

[16:34:21 CDT(-0500)] <EricDalquist> http://en.wikipedia.org/wiki/Post/Redirect/Get

[16:34:22 CDT(-0500)] <EricDalquist> that might help

[16:40:15 CDT(-0500)] <jwennmacher> I happened to find a form using a renderURL with a parameter 'action'. When the URL was rendered on the form in the 'action=url' element it was incorrect ... sort of had 2 ampersands (rendered as "render.uP?pCt=realcalendar.n18&amp;amp;pP_action=createCalendarDefinition2"). Sort of because the encoding turned &amp; into &amp;amp; which turned the parameter 'action' into 'amp;action'. So I can see that it clearly doesn't w

[16:40:36 CDT(-0500)] <EricDalquist> oh

[16:40:40 CDT(-0500)] <EricDalquist> that is a different issue

[16:40:41 CDT(-0500)] <EricDalquist> what portlet

[16:40:47 CDT(-0500)] <athena> calendar

[16:40:48 CDT(-0500)] <EricDalquist> and what other changes are you making to it?

[16:40:51 CDT(-0500)] <jwennmacher> CalendarPortlet. I'll be fixing it

[16:40:55 CDT(-0500)] <EricDalquist> sounds like the escapeXml issue

[16:41:01 CDT(-0500)] <EricDalquist> with going from 168->286

[16:41:25 CDT(-0500)] <jwennmacher> i'm making some changes for Manhattan to support calDAV for calendaring

[16:42:17 CDT(-0500)] <jwennmacher> To clarify I was planning on fixing it by making the renderURL and actionURL. I'm not sure about the escapeXml issue you are referring to

[16:42:31 CDT(-0500)] <jwennmacher> (smile)

[16:43:41 CDT(-0500)] <EricDalquist> so the double & escaping

[16:43:49 CDT(-0500)] <athena> EricDalquist: yes, i think it's the escape issue

[16:44:01 CDT(-0500)] <EricDalquist> if the URL is used in javascript

[16:44:22 CDT(-0500)] <EricDalquist> you need to do <portlet:renderURL escapeXML='false'/>

[17:15:23 CDT(-0500)] <jwennmacher> One other thing I was curious about. I noticed a difference in url formats. Action urls have the portlet id in the url itself, but render urls have the portlet id as a query parameter. I was wondering why it is like that.

[17:15:44 CDT(-0500)] <EricDalquist> for action urls we know that the portlet is the only thing being targeted

[17:15:52 CDT(-0500)] <EricDalquist> for render urls where the state part if the url is normal

[17:16:03 CDT(-0500)] <EricDalquist> there are potentially lots of portlets on that page

[17:16:16 CDT(-0500)] <EricDalquist> so we need to specify the portlet specific bits as parameters

[17:16:28 CDT(-0500)] <EricDalquist> as you can have one render url with parameters that target multiple portlets at once

[17:18:25 CDT(-0500)] <jwennmacher> Thanks. What's an example where you'd have a render url that targets multiple portlets at once?

[17:18:38 CDT(-0500)] <EricDalquist> sure ... portlet event processing

[17:18:52 CDT(-0500)] <EricDalquist> action in portlet A sends an event

[17:18:57 CDT(-0500)] <EricDalquist> portlets B, C, and D process it

[17:19:09 CDT(-0500)] <EricDalquist> during event processing portlets can set render parameters to be used the next time they render

[17:19:16 CDT(-0500)] <EricDalquist> if they are all on that page we need the compound url

[17:19:35 CDT(-0500)] <EricDalquist> the URL part is alos supposed to be about navigational state

[17:19:40 CDT(-0500)] <EricDalquist> so when on a normal page

[17:19:50 CDT(-0500)] <EricDalquist> the portlet you are targeting isn't part of the navigation

[17:19:52 CDT(-0500)] <EricDalquist> it is just a parameter

[17:20:06 CDT(-0500)] <EricDalquist> for actions, resources, maximized, and detached urls

[17:20:11 CDT(-0500)] <EricDalquist> you will only ever target one portlet

[17:20:16 CDT(-0500)] <EricDalquist> so the portlet is part of the nav state

[17:21:09 CDT(-0500)] <EricDalquist> but I have to run home

[17:21:16 CDT(-0500)] <EricDalquist> and will be out tomorrow morning

[17:21:21 CDT(-0500)] <EricDalquist> so see everyone afte rlunch tomorrow

[17:21:56 CDT(-0500)] <jwennmacher> I know you are gone ... but thanks (smile)