The IRC channel is #sakai on the freenode network, you can join using a web client at http://webchat.freenode.net/?channels=sakai and you can view the complete logs at http://sakai.iriscouch.com/irc/_design/viewer/index.html
12:58 <SakaiGithub> [3akai-ux] nicolaasmatthijs pushed 1 new commit to 1.4.2: http://git.io/PvYLFQ 12:58 <SakaiGithub> [3akai-ux/1.4.2] [SAKIII-6178] Adjust the header size hierarchy to reflect the header level - Chris Roby 13:07 <SakaiGithub> [3akai-ux] nicolaasmatthijs pushed 1 new commit to 1.4.2: http://git.io/mGKzRg 13:07 <SakaiGithub> [3akai-ux/1.4.2] SAKIII-6176 - /system/me is not cache-busted for IE9 - Nicolaas Matthijs 13:30 <SakaiGithub> [3akai-ux] nicolaasmatthijs pushed 1 new commit to 1.4.2: http://git.io/-5gYdA 13:30 <SakaiGithub> [3akai-ux/1.4.2] SAKIII-6177 - Content doesn't appear to update when authoring with IE8 - Nicolaas Matthijs 13:37 <SakaiGithub> [3akai-ux] nicolaasmatthijs pushed 1 new commit to 1.4.2: http://git.io/jDsm0Q 13:37 <SakaiGithub> [3akai-ux/1.4.2] SAKIII-6143 - IE8 collection viewer doesn't update when adding content to the collection - Nicolaas Matthijs 13:40 <mrvisser> ping nicolaas 13:52 <nicolaas> pong mrvisser 13:54 <nicolaas> we don't support IE7 actually 13:54 <nicolaas> we only support IE8 and IE9 13:54 <nicolaas> and IE10 when it's out 13:55 <mrvisser> nicolaas: ahh I see. So I imagine we wouldn't have a lot of users on IE7. Are things known to not work well? 13:55 <nicolaas> I would guess we wouldn't have any 13:55 <nicolaas> I'm fairly confident the UI doesn't work at all for them 13:55 <mrvisser> haha, good to know. ok, thanks nicolaas! 13:56 <nicolaas> none of the supporting institutions still use IE7 13:56 <nicolaas> so it was deemed to be a safe thing to do 14:51 <SakaiGithub> [3akai-ux] nicolaasmatthijs pushed 2 new commits to master: http://git.io/zWg0Yg 14:51 <SakaiGithub> [3akai-ux/master] Merge branch '1.4.2' - Nicolaas Matthijs 14:51 <SakaiGithub> [3akai-ux/master] Merge branch 'master' of github.com:sakaiproject/3akai-ux - Nicolaas Matthijs 17:22 <GitHub169> [nakamura] arwhyte tagged org.sakaiproject.nakamura-1.4.1 at 1.4.1-release-tag: http://git.io/0MQ3cg 17:22 <GitHub169> [nakamura/org.sakaiproject.nakamura-1.4.1] NOJIRA fix pom version inconsistencies - Anthony Whyte 17:22 <GitHub169> [nakamura/org.sakaiproject.nakamura-1.4.1] NOJIRA fix javax.jcr version - Anthony Whyte 17:22 <GitHub169> [nakamura/org.sakaiproject.nakamura-1.4.1] release_post_process: Moving config files from 1.4.0 to 1.4.1-SNAPSHOT - Anthony Whyte 17:22 <GitHub25> [nakamura] arwhyte tagged 1.4.1-release-tag at org.sakaiproject.nakamura-1.4.1: http://git.io/0MQ3cg 17:22 <GitHub25> [nakamura/1.4.1-release-tag] NOJIRA fix pom version inconsistencies - Anthony Whyte 17:22 <GitHub25> [nakamura/1.4.1-release-tag] NOJIRA fix javax.jcr version - Anthony Whyte 17:22 <GitHub25> [nakamura/1.4.1-release-tag] release_post_process: Moving config files from 1.4.0 to 1.4.1-SNAPSHOT - Anthony Whyte 17:29 <GitHub184> [nakamura] arwhyte tagged org.sakaiproject.nakamura-1.4.1-RC1 at 1.4.1-release-tag: http://git.io/0MQ3cg 17:29 <GitHub184> [nakamura/org.sakaiproject.nakamura-1.4.1-RC1] NOJIRA fix javax.jcr version - Anthony Whyte 17:29 <GitHub184> [nakamura/org.sakaiproject.nakamura-1.4.1-RC1] NOJIRA fix pom version inconsistencies - Anthony Whyte 17:29 <GitHub184> [nakamura/org.sakaiproject.nakamura-1.4.1-RC1] release_post_process: Moving config files from 1.4.0 to 1.4.1-SNAPSHOT - Anthony Whyte 17:29 <GitHub115> [nakamura] arwhyte tagged 1.4.1-release-tag at org.sakaiproject.nakamura-1.4.1-RC1: http://git.io/0MQ3cg 17:29 <GitHub115> [nakamura/1.4.1-release-tag] NOJIRA fix javax.jcr version - Anthony Whyte 17:29 <GitHub115> [nakamura/1.4.1-release-tag] NOJIRA fix pom version inconsistencies - Anthony Whyte 17:29 <GitHub115> [nakamura/1.4.1-release-tag] release_post_process: Moving config files from 1.4.0 to 1.4.1-SNAPSHOT - Anthony Whyte 18:10 <mrvisser> thecarlhall: are you planning by any chance to disable the QoSFilter for 1.5.0? 18:11 <thecarlhall> mrvisser, hadn't thought about it. Should it go away? 18:11 <thecarlhall> mrvisser, I can never remember if it's on or not. Seems like it was supposed to be turned off at some point. 18:12 <mrvisser> thecarlhall: I believe it's on. I've seen logging coming out of it when performance tests get out of hand and it exceeds the maximum suspended requests limit. 18:12 <mrvisser> thecarlhall: if it's not disabled, I'll have to implement more stuff for the JAX-RS stuff — specifically, around handling thread-managed contexts. 18:13 <thecarlhall> mrvisser, oh yeah. Is QoSFilter the one that uses jetty continuations? 18:13 <mrvisser> thecarlhall: yup 18:13 <mrvisser> it will store away request threads and prioritize them when things get bad 18:13 <thecarlhall> ugh 18:16 <mrvisser> thecarlhall: so, if we have no need for it, I can go ahead with disabling it to avoid the additional work it will incur us to maintain it. 18:16 <thecarlhall> mrvisser, it's been in our code long enough that I think it may warrant an email to the list. 18:17 <thecarlhall> mrvisser, I haven't heard how anyone is using it so we may be surprised to hear about it 18:17 <thecarlhall> lancespeelmon, does rSmart tune or turn off the QoSFilter? 18:17 <lancespeelmon> no I don't think so 18:18 <lancespeelmon> let me double check 18:18 <lancespeelmon> yes we have it enabled in both 1.1 and 1.3 thecarlhall 18:19 <thecarlhall> lancespeelmon, have you done testing to tune it or do you use the defaults? 18:19 <lancespeelmon> one min 18:20 <lancespeelmon> so if you look at the settings in console: 10, 50, 60000, 10, 2, 2, 60000 18:22 <lancespeelmon> ./environments/acad-prod/modules/localconfig/manifests/init.pp: $qos_limit = '10' 18:22 <lancespeelmon> ./environments/acad-staging/modules/localconfig/manifests/init.pp: $qos_limit = '10' 18:25 <mrvisser> yea those are the defaults. 18:31 <thecarlhall> mrvisser, zacht or anyone else with a mac, would you mind trading me a jira for kern-3141? Safari on linux is undependable at best 18:33 <zacht> thecarlhall I am close to solving KERN-3126, so I think I'd better hold onto it. 18:34 <zacht> thecarlhall But you can take KERN-3128 18:34 <zacht> thecarlhall It just needs to be cherry-picked over to 1.4. 18:34 <zacht> 1.4.2 18:34 <thecarlhall> zacht, I can totally cherry pick that. 18:35 <zacht> cool. 18:39 <zacht> kentfitz Do we have an environment where KERN-3126 is happening right now? 18:39 <kentfitz> zacht: It's happening on qa20. Authenticated users 18:40 <zacht> kentfitz Ok. I'm down to six or so lines of code where this could be happening, but I need to see this bug in action. 18:40 <kentfitz> zacht: makes sense. 18:58 <zacht> kentfitz Is it alright if I bounce qa20 to test this fix? 18:58 <GitHub4> [nakamura] thecarlhall pushed 1 new commit to 1.4.2: http://git.io/cF3N4A 18:58 <GitHub4> [nakamura/1.4.2] KERN-3128 JSONObject["message"] not found. org.apache.sling.commons.json.JSONException - Lance Speelmon 18:58 <kentfitz> zacht: Yeah, I'm trying to figure out a search oddity, but I can take a pause 18:59 <zacht> kentfitz Ok, it'll be back momentarily. 18:59 <kentfitz> zacht: no worries. 19:00 <zacht> kentfitz which startsakai.sh did you use? 19:00 <zacht> kentfitz I think I found it. 19:00 <kentfitz> zacht:/home/sakai/142 19:00 <kentfitz> from that directory 19:01 <zacht> kentfitz Coming back up. 19:02 <kentfitz> zacht: Cool 19:22 <mrvisser> hey zacht: do you have any resources or information in your brain that can help me understand HTTP filter ordering in Nakamura? 19:22 <mrvisser> I've seen "filter.priority" and "service.ranking" properties on threads that seem to attempt this… do you know about these? 19:23 <zacht> mrvisser thecarlhall Would know. 19:23 <mrvisser> delegation++ ! :) 19:23 <thecarlhall> mrvisser, service.ranking is an osgi thing but is not enforced by the framework 19:24 <thecarlhall> I'm trying to find what introduces filter.priority 19:25 <simong> I think that's Sling 19:25 <simong> thecarlhall / mrvisser: http://sling.apache.org/site/filters.html 19:25 <mrvisser> thecarlhall: The QoSFilter has set, if that helps at all.. apparently the higher number, the sooner it executes in the chain. 19:26 <mrvisser> simong: good find! 19:27 <thecarlhall> simong, I don't see where filter.priority is 19:28 <simong> I don't know, I've never used that 19:28 <simong> For the Google stuff I just used filter.scope and filter.order 19:28 <thecarlhall> I don't see "filter.priority" in the sling code base either. I can't recall where that comes from though 19:28 <thecarlhall> I'll keep digging. 19:29 <simong> Might just be a mistake? 19:29 <mrvisser> thecarlhall, just found this http://markmail.org/message/quxhm7d5s6u66crr#query:+page:1+mid:otdvpfa6voq67auo+state:results 19:29 <mrvisser> simong, thecarlhall: "The order property has already been deprecated in favor of the OSGi standard "service.ranking" property (as of FELIX-1213) 19:32 <thecarlhall> mrvisser, looks like filter.priority is used when registering the filter to the http service. I think it's just something we made up 19:33 <mrvisser> thecarlhall: ahhhh we manually register with the http service.. how the heck did I not see that. 19:34 <thecarlhall> mrvisser, there are a couple of filters like that. it's a subtlety that is maddening 19:34 <thecarlhall> mrvisser, while we're on it, also note that a servlet filter is different than a sling filter 19:34 <thecarlhall> request -> servlet filters -> servlets (SlingMainServlet) -> sling filters 19:36 <mrvisser> interesting. 19:36 <thecarlhall> servlet filters are part of the regular servlet stack whereas sling filters happen in the sling context 19:37 <thecarlhall> another interesting point I found out the other day while reading a bunch of code is that sling doesn't extend HttpServlet. It extends GenericServlet 19:41 <mrvisser> thecarlhall: I think I need to make the QoSFilter filter.priority property private (i.e., not modifiable from the system admin console). Need to make a guarantee about execution order of that guy. 19:42 <mrvisser> WDYT? 19:42 <zacht> kentfitz Sorry, can I do one more restart? 19:42 <kentfitz> zacht: sure thing. Whatever it takes to track it down 19:42 <zacht> kentfitz This bundle does not seem to want to refresh on its own. 19:42 <zacht> kentfitz At least we get a stacktrace in the log now. 19:42 <kentfitz> zacht: that seems like it should help 19:43 <zacht> kentfitz It helps. 19:52 <zacht> kentfitz ah, much better! https://qa20-us.sakaiproject.org:8088/categories 19:53 <kentfitz> zacht: shiny! 19:59 <thecarlhall> mrvisser, sounds good to me. If you want it truly unmodifiable, remove it as an osgi property 20:00 <thecarlhall> mrvisser, making it private just removes it from the metatype service but doesn't disallow it from being configured via the config admin 20:00 <mrvisser> thecarlhall: ahh, good to know! I'll do that. Thanks! _______________________________________________ oae-dev mailing list oae-dev@collab.sakaiproject.org http://collab.sakaiproject.org/mailman/listinfo/oae-dev
