[10:27:04 CDT(-0500)] <EricDalquist> well 4.0.6 is cut but found a bug in one of the last minute merges
[10:27:09 CDT(-0500)] <EricDalquist> so time for 4.0.6.SR1
[10:37:31 CDT(-0500)] <drewwills> any reason not to use string 3.1.x in a portlet? are we already doing that elsewhere?
[10:37:50 CDT(-0500)] <EricDalquist> only requirement is that you're on up4
[10:39:25 CDT(-0500)] <drewwills> roger
[10:39:52 CDT(-0500)] <drewwills> (working on the 286 port of the email-preview)
[10:40:04 CDT(-0500)] <EricDalquist> cool
[10:56:09 CDT(-0500)] <athena> yes, i'd recommend using spring 3.1
[11:01:15 CDT(-0500)] <drewwills> athena is there a good example of sending a non-200 HTTP status in case of an error in a ReaourceRequest? (and would that be appropriate)
[11:01:41 CDT(-0500)] <athena> should be pretty straightforward
[11:01:50 CDT(-0500)] <athena> just need to set the response code
[11:01:54 CDT(-0500)] <athena> like you would for a servlet
[11:02:19 CDT(-0500)] <EricDalquist> well if your forwarding to a servlet
[11:02:36 CDT(-0500)] <EricDalquist> on as plain ResourceResponse: http://portals.apache.org/pluto/portlet-2.0-apidocs/javax/portlet/ResourceResponse.html
[11:02:37 CDT(-0500)] <athena> take a look at the ResourceRequest API
[11:02:54 CDT(-0500)] <EricDalquist> note the HTTP_STATUS_CODE constant
[11:03:10 CDT(-0500)] <athena> yep
[11:03:16 CDT(-0500)] <athena> and status codes are all the same, of course
[11:06:31 CDT(-0500)] <drewwills> (should have just looked at the javadoc instead of digging through examples)
[11:18:28 CDT(-0500)] <drewwills> set it as a string? surprising
[11:18:46 CDT(-0500)] <EricDalquist> yeah, just an API limitation
[11:18:57 CDT(-0500)] <EricDalquist> but it has been tested
[11:19:04 CDT(-0500)] <EricDalquist> you can actually set any HTTP header via setProperty
[11:19:37 CDT(-0500)] <drewwills> gotcha
[11:53:18 CDT(-0500)] <EricDalquist> so any thoughts on the slf4j/logback stuff in here?
[11:53:24 CDT(-0500)] <EricDalquist> or on moving master to 4.1?
[11:59:23 CDT(-0500)] <EricDalquist> blarg
[11:59:30 CDT(-0500)] <EricDalquist> just about to publish 4.0.6.SR1
[11:59:34 CDT(-0500)] <EricDalquist> and sonatype goes down 
[12:12:20 CDT(-0500)] <drewwills> I don't know a ton about slf4j, though what I do know I like
[12:12:39 CDT(-0500)] <EricDalquist> the big difference is there is no dynamic binding
[12:12:47 CDT(-0500)] <drewwills> and we're using logback in ssp... so I have some exposure to the config
[12:12:56 CDT(-0500)] <EricDalquist> you include a library that implements the slf4j api
[12:12:58 CDT(-0500)] <EricDalquist> and it uses that
[12:13:16 CDT(-0500)] <EricDalquist> if there is no slf4j binding or multiple bindings it spits out very informative info to system.out
[12:13:45 CDT(-0500)] <EricDalquist> also ... they template based logging shorthand is nice
[12:14:00 CDT(-0500)] <EricDalquist> log.warn("foo {} bar {}", obj1, obj2)
[12:14:11 CDT(-0500)] <EricDalquist> only does a toString on the args if warn is enabled
[12:15:04 CDT(-0500)] <drewwills> yeah that token replacement in the logging messages is the part I know and like
[12:15:39 CDT(-0500)] <EricDalquist> logback also has much nicer stack trace output
[12:15:49 CDT(-0500)] <EricDalquist> each line includes the library & version number the class was loaded from
[12:16:14 CDT(-0500)] <drewwills> ah yes I had seen that in some logs... nice feature
[12:16:59 CDT(-0500)] <EricDalquist> and the log messages include a better "short" class+package string:
[12:16:59 CDT(-0500)] <EricDalquist> INFO [uPortal-TaskExecutor-841-aggregateRawEvents] o.j.p.e.a.PortalEventProcessingManagerImpl 2012-08-13 12:14:32,881 - Aggregated 2362 events created at 13.0497 events/second between 2012-08-13T12:09:35.781-05:00 and 2012-08-13T12:12:37.626-05:00 in 84699ms - 27.8870 e/s a 2.1370x speedup
[12:17:38 CDT(-0500)] <EricDalquist> really the only user facing change is that instead of log4j.properties we have logback.xml
[12:18:16 CDT(-0500)] <EricDalquist> oh and logback by default looks for logback-test.xml first and ignores logback.xml if it finds the test version
[12:18:25 CDT(-0500)] <EricDalquist> so when running unit tests it is easy to have a seperate test config
[12:18:31 CDT(-0500)] <EricDalquist> without classpath overrides