15:31:33 #startmeeting Helium M5 developer meeting 15:31:33 Meeting started Wed Aug 27 15:31:33 2014 UTC. The chair is phrobb. Information about MeetBot at http://ci.openstack.org/meetbot.html. 15:31:33 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:31:33 The meeting name has been set to 'helium_m5_developer_meeting' 15:31:43 regXboi: You just missed Karaf Happy Hour :) 15:32:00 #topic Project Contacts please #info in 15:32:06 #info edwarnicke controller 15:32:18 #info tbachman for Group Based Policy 15:32:23 #info Hideyuki for VTN 15:32:26 #info rafat ODL-SDNi App 15:32:28 #info oflibMichal 15:32:34 #info paulq for SFC 15:32:44 #info oflibMichal for openflowjava 15:33:14 #info Dana for bgpcep 15:34:06 #info regXboi for nobody... so I can hold the pen if need be (I can't believe I just volunteered for that) 15:34:13 #info Madhu 15:34:14 #info George Zhao for release 15:34:32 #info goldavberg for lispflowmapping 15:34:40 Madhu: are you representing both dlux and ovsdb? 15:34:57 #chair regXboi edwarnicke gzhao 15:34:57 Current chairs: edwarnicke gzhao phrobb regXboi 15:35:04 phrobb: good question... meeting conflict and am still figuring out :) 15:35:11 #info ttkacik for Yangtools 15:35:12 at least until someone shows up in dlux 15:35:17 but definitely ovsdb 15:35:19 regXboi: you are welcome to hold the pen :-) 15:35:32 * regXboi figures :) 15:35:48 #info CASP3R for Integration 15:36:02 * phrobb checking who all we have and who is missing… 15:36:46 #info abhijitkumbhare for OpenFlow plugin 15:37:54 did I miss aaa? 15:38:09 regXboi: correct, I don't see anyone from AAA 15:38:18 well I think liemmn just arrived 15:38:19 sorry... ran late 15:38:25 #info liemmn for AAA 15:38:32 #info catohornet also for integration 15:38:52 how about defense4all? 15:39:01 #info mlemay for reservation 15:39:13 l2 switch? 15:39:28 FYI -- having IRC client issues 15:39:31 did I miss anything? 15:39:40 tbachman_: no 15:39:44 regXboi: correct, I've not seen or heard from defense4all either 15:39:45 thx 15:39:53 tbachman: no, just finishing roll call 15:40:00 opflex? 15:40:02 phrobb: gzhao thx! 15:40:11 I guess I can cover opflex 15:40:17 PCMM? 15:40:21 #info tbachman opflex 15:40:24 tbachman: please #info that in 15:40:26 SNB-i? 15:40:48 opencontrail? 15:40:59 snmp4sdn? 15:41:10 table type patterns? 15:41:17 regXboi: We had someone turn up from SNBi last hour for Karaf Happy hour 15:41:19 toolkit? 15:41:22 arg 15:41:39 just looking at the #infos and the project list 15:41:43 and noting the differences 15:42:03 We are 11 after the start. Let's get started. If those here can ping those absent, please do 15:42:05 I think I'm at the end of the list now 15:42:21 #topic Status 15:42:22 #info for southbound plugin to OpenContrail platform 15:42:43 #info appear to be missing defense4all, l2switch, pcmm, snb-i snmp4sdn, tabletypepatterns and toolkit 15:42:47 I am not most up-to-date with table type patterns status - but I can relay back info to Colin/Curt 15:42:54 regXboi: I'm chasing l2switch folks 15:43:10 thx all 15:43:17 given that Colin is on a paternity leave 15:43:45 #info here from dlux project 15:44:39 #info Let's start with the Code Freeze requirement of M5. This should mean that no new features should be allowed by the committers starting on Monday. Only errors/bugs that are also identified in bugzilla are allowed from Monday forward. Is this an agreeable definition to the group and are there any projects that do not feel they can meet that requirement? 15:45:24 phrobb: Could we clarify that new *tests* are cool? 15:45:34 edwarnicke: and what about packaging? 15:45:42 * tbachman_ putting his karaf hat on 15:45:46 tbachman_: Good question 15:45:56 #info rovarga for tcpmd5 15:45:58 tbachman_: I would *really* like to get everyone in for karaf by M5... but I'm also a realist 15:46:02 edwarnicke: Yes edwarnicke, new tests are allowed and encouraged 15:46:06 phrobb: gzhao thoughts? 15:46:31 phrobb: Thanks :) Some folks had been asking :) 15:46:42 phrobb: Could we add that to our 'Code Freeze' statement so its clear to everyone? 15:46:54 So should this be modified to only errors/bugs/additional tests/packaging/docs ? 15:47:06 abhijitkumbhare: Definitely also docs :) 15:47:08 Yes,… I'm make a note. 15:47:14 phrobb: For reservation we'll be late on Code Freeze... I'm trying to merge as fast as I can and we're getting to Yellow status.. 15:47:18 meant bug fixes (not bugs :) ) 15:47:22 phrobb: Just to clarify... things like log cleanups would be bugs, right? 15:47:33 * edwarnicke has a list of crappy logs he wants to fix 15:47:46 mlemay: does that mean code freeze is still up to projects? 15:47:47 abhijitkumbhare: Definitely. No committing new bugs after code freeze! :) 15:48:36 edwarnicke: I'd think log cleanups would be allowed if there is a defect filed in bugzilla 15:48:52 regXboi: I'm down with the bug in bugzilla... but man I want to scrub some logs :) 15:48:53 edwarnicke: yes log clean up is a bug… the packaging one I'm still grappling with... 15:49:00 phrobb: I don't believe OpFlex is going to meet this requirement, btw 15:49:07 phrobb: Could you speak your thoughts and maybe folks will have ideas? 15:49:17 tbachman: Do we have opflex guys here? 15:49:28 edwarnicke: I think I'm representing 15:49:32 alagalah is out this week 15:49:46 tbachman: I am told he's at the happiest place on earth, having breakfast 15:49:49 lol 15:49:50 he is 15:50:04 with a pair of black ears on top 15:50:17 edwarnicke: I'm headed to my next class, I see amit_mandke is here now 15:50:32 One of the major goals of code freeze is for integration testing and documentation to be allowed to go forward full steam. If packaging isn't finished, the integration team has a tough time testing… and the docs team has a tough time documenting packaging and install.. 15:50:36 evanz: That statements is *so* ODL :) 15:50:53 tbachman: I can code freeze a portion but I'd rather we have full functionality and was taking up by other dependency requirements that just have been going in (karaf etc..) now I can fully test end to end.. so we plan to stabilize that as soon as possible.. We'd like to flag ourselves as an experimental feature... 15:50:55 Thanks Evanz! 15:51:04 thanks for asking the basic question guys on code freeze :) 15:51:29 mlemay: not trying to be antagonistic... just trying to make sure we all come to an agreement :) 15:51:30 Madhu: definitly best when we all start from the same place :-) 15:51:42 phrobb: So... I can see that for the 'high level' features being packaged 15:51:55 phrobb: the ones listed in integration 15:52:21 phrobb: shall we have 2 different discussions ? 15:52:25 1. karaf related 15:52:25 phrobb: I'm assuming this is also a point where we decide what projects can or can't make the release? 15:52:29 2. non-karat related ? 15:52:37 edwarnicke: right, but you can't really build the high level features until the low levels are defined and constructed correct? 15:52:39 karat vs. stick 15:53:11 tbachman: M5 is approaching ... no times for jokes ;) 15:53:13 tbachman: yes, this is the precursor to it… 15:53:19 sorry ;) 15:53:34 * edwarnicke puts on a very very serious face :) 15:53:54 * regXboi doesn't buy any of this for a second 15:53:58 lol 15:54:10 phrobb: Does everyone know what they need to do for packaging 15:54:25 phrobb: We had at least one person at Karaf Happy Hour still unclear there 15:54:27 there are projects have reported met M4 milestone yet 15:54:28 Projects that recognize they can't meet the requirements can DOR (Dropped on Request) whenever they choose. The gate for the project is the TSC Release Review that will be scheduled for each project real soon noon post M5 15:54:49 phrobb: thx! 15:55:28 * edwarnicke looks distressed that regXboi doubts him 15:55:29 edwarnicke: is there anything that is unclear in the karaf step-by-step that you wrote regarding what the projects need to do? 15:55:52 phrobb: I don't think so... we continue to add things like 'Common problems and their solutions' as we find them... but not sure everyone is working from that page yet 15:56:26 edwarnicke: do all the projects understand that they need to be karaf'd? 15:56:27 phrobb: And we've had folks work through it successfully 15:56:31 (i.e. this is a requirement) 15:56:36 tbachman: That's part of what I'm asking 15:56:42 tbachman: I think they should at this point 15:56:47 tbachman: But want to check in and be sure 15:56:50 One question i have is yes i see that Karaf is slowly getting to a working solution, integration has kick off a job that deploy a karaf controller and test the base-of13 stuff 15:56:51 did we send emails or other form of contact? 15:57:10 edwarnicke: is there anything you would like to say here about pkging and karaf work to add clarity? 15:57:11 Now what if a project doesn't make it into karaf how do we package that? 15:57:19 CASP3R: good question 15:57:27 need to drop off Alex(alefan) would represent L2switch in this meeting 15:57:29 * tbachman worries we haven't made that 100% clear 15:57:57 I believe all projects understand the need to support Karaf and have acknowledged their committment to it. 15:58:03 phrobb: thx 15:58:31 gzhao: is that correct?… are there any projects requesting to not support karaf (that are java based)? 15:58:32 * tbachman apologizes for re-covering already-covered ground 15:59:37 phrobb: there are four projects is N/A to karaf, such as opflex, documentation, toolkit 15:59:44 CASP3R: do remember vtn and opflex have non java stuff 15:59:48 So it seems as though Reservation and OpFlex do not feel they will be at code freeze prior to Monday's M5. Are there any others? 15:59:57 VTN is a 3PP app so that's cool 16:00:04 i can second that for toolkit. it is a N/A 16:00:27 last email i got from gzhao was no from SNB and SFC 16:00:38 phrobb: Let me say my piece on karaf stuff here: 16:00:52 phrobb: SNBi SFC initially requested not to support Karaf 16:01:03 CASP3R: I am not sure SNBi last response was not ti support 16:01:20 #link https://wiki.opendaylight.org/view/Karaf:Step_by_Step_Guide Please read this and follow it through to get your project features into integration 16:01:37 #info If you need help, Karaf Happy Hour is 7:30am PST every morning this week on #opendaylight-meeting 16:01:48 I know that folks had reached out to SNBi and SFC to see if they could help them with their karaf work 16:01:56 gzhao: SFC folks have been seen working on features and asking good questions ... 16:02:12 phrobb: I think VTN will be at code freeze prior to Monday'S M5 barely. If it will not, I'll tell you. 16:02:15 I can check with Reinaldo and co. about Karak and report back 16:02:17 gzhao: And SNBi folks were at Karaf Happy Hour yesterday :) 16:02:22 just 1 thing... I think if folks work on it now... we can get Karaf working. 16:02:23 karak --> karaf 16:02:31 but we might see bugs post M5 16:02:33 * edwarnicke things we are deep in our cups at this point 16:02:45 Madhu: You mean packaging bugs? Most def 16:02:47 and we should be ready to accept bug fixes to make it solid before the release 16:02:49 edwarnicke: Do you know what are the Karaf status for yangtools, controller, openflowjava, openflowplugin, ovsdb 16:02:55 edwarnicke: no. it could also be runtime issues 16:02:56 gzhao: Yes 16:03:01 gzhao: ovsdb is ready 16:03:07 yet to push to integration 16:03:13 edwarnicke: that is great, IMHO, those project should go Karaf 16:03:19 CASP3R: what kind of issues will you see in integration testing as we move to RC0 if the karaf packaging is not completed at M5? 16:03:21 gzhao: yangtools is pulled in by controller. No high level user visible features listed because its supporting stuff 16:03:25 https://jenkins.opendaylight.org/integration/job/integration-master-csit-karaf-of13/4/robot/report/report.html 16:03:28 edwarnicke: phrobb it is just too last minute for getting all the tests done for karaf 16:03:32 don't u guys agree to that ? 16:03:41 gzhao: controller, the MD-SAL side stuff is done except for the clustering feature, which I'm working on 16:03:47 especially given the fact where we are at this stage 16:03:51 thats the current work, but there few patch waiting to be merged both in testing and the karaf release, we have a good update for TSC tomorrow 16:04:17 Madhu: thanks 16:04:25 gzhao: I believe the AD-SAL side stuff is in, and the adsal-compatibility patch is awaiting review in integration (with fixed commit message, thanks Madhu for pointing out the previous deficiencies there) 16:04:37 sure edwarnicke . 16:04:41 as u guys can see... 16:04:52 gzhao: openflowplugin is in integration, but some of its servicability features, like test-provider and drop-test do not yet have features 16:04:53 both ADSAL and MDSAL controller base code came in last week 16:05:11 gzhao: openflowplugin nicera-extensions are not in as features at Madhu's request as he would like to do the features in ovsdb 16:05:14 it just doesn't give enough time for folks to write intgration tests with these 16:05:21 gzhao: openflowjava is pulled in by openflowplugin 16:05:29 edwarnicke: that is correct. it is being done there 16:05:49 so we have to clearly state the intent of code freeze for karaf. 16:06:00 get all the features in before M5 16:06:00 gzhao: I think that's what I now currently for the things you asked about 16:06:06 and expect bugs post M5 and fix them 16:06:24 " gzhao: openflowplugin nicera-extensions are not in as features at Madhu's request as he would like to do the features in ovsdb" <- can I ask why? 16:06:26 edwarnicke: looks major "core" projects are on track, next level are GBP, l2swtich, bpgcep, aaa 16:06:36 Madhu: we talked about running minimal and maximal testing of karaf bundles in integration testing using the existing integration tests… what additional integration tests are needed? 16:06:50 regXboi: I can only speak for myself on that. Madhu felt strongly, I didn't feel strongly, so I deferred. 16:06:58 regXboi: If other folks have strong feelings, I can reconsider 16:07:00 regXboi: thats because openfowplugin cannot depend on ovsdb (due to auto release dependencies) 16:07:00 Madhu: to be written? 16:07:01 gzhao: am working on GBP 16:07:09 edwarnicke: i didn't feel strong about that either :) 16:07:27 edwarnicke: because u pointed to the fact that it cannot be done on openflowplugin due to the auto release dependency stuff 16:07:39 edwarnicke: if it can be sorted out... i don't care where it resides 16:07:50 Madhu: Note, that doesnt' make it impossible to define the base nicera-extensions stuff in OFplugin, but I concur it would make it messy. 16:08:09 edwarnicke: regXboi am fully open to anything we decide guys 16:08:09 ok... got it 16:08:15 just that we need a clean way to do it. 16:08:26 the statement I saw looked a little backwards 16:08:28 and i volunteered to do it in ovsdb because that is the cleanest at this point 16:08:44 I'm ok with it being in ovsdb at this point while we figure a cleaner solution 16:08:44 regXboi: definitely not a project creep ;) 16:08:51 Just to be clear, I strongly feel, and I think Madhu concurs, that the nicera features should be named in a way that does not reference their current project of residence (either ovsdb or ofplugin) because its an open discussion where they should live and we don't want to bias future decisions 16:08:53 I think we need to address phrobb's earlier question about integration tests 16:09:00 Madhu: I wasn't thinking project creep - I was just confused :) 16:09:07 lol 16:09:25 regXboi: i make it a point to satisfy your valid scope creep pointers ;) 16:09:25 Madhu: Could you confirm or dispute that I represented your feelings accurately ? 16:09:40 edwarnicke: absolutely 16:09:54 We are also still waiting for netopeer netconf tool installation on LF integration server for testing netconf. 16:09:59 Madhu: Cool... I try to be careful representing other folks feelings and opinions, even when I'm pretty sure I got it right 16:10:05 Madhu edwarnicke regXboi - I think the Nicira extension can be thought of as a consumer of OF plugin 16:10:08 edwarnicke: just to be clear... we discussed this offline but just for everyone's sake : odl-openflow-nxm-extensions :) 16:10:11 catohornet: Whats the blocker there? 16:10:14 abhijitkumbhare: it is 16:10:30 Madhu: LOL... here we go again (we've explained it a few times, but its worth explaining again :) ) 16:10:32 (i know nxm has extensions ... but this seem way cleaner and readable) 16:11:01 guys back to the discussion on M5 code-freeze 16:11:10 lemme roll back to what phrobb has to say there 16:11:37 phrobb: karaf is a new runtime 16:11:48 phrobb: we have changed some start-levels 16:11:55 phrobb: we have changed some dependencies. 16:12:11 phrobb: and only after all the features are code-committed 16:12:24 we can do a meaningful testing on what happens when all the cooks are in the kitchen 16:12:45 Also one thing is what feature/project don't play nicely with others 16:12:51 I know Ed et all working on l2swtich, do bpgcep and aaa start Karaf process 16:12:53 CASP3R: +1 16:13:03 Madhu: start-levels? (I think I missed a beat... could you clarify?) 16:13:08 CASP3R: +1 16:13:09 gzhao: l2switch is in 16:13:24 edwarnicke: we have start-levels for the existing distribution (defined in config.ini) 16:13:28 with karaf we are removing it 16:13:31 (which is goodness) 16:13:35 we don't know how it is going to impact 16:13:37 CASP3R: Should we go over how that's being handled in the integration features.xml ? 16:13:50 Madhu: OK... cool, valid point :) 16:13:51 That would be nice edwarnicke 16:13:57 also... in the legacy distribution... we load ALL the bundles by default 16:14:03 Let me find a link... 16:14:05 edwarnicke: without netopeer installed on the tools VM, Netconf is not part of the integration testing. 16:14:09 but in karaf, we let it be loaded at runtime by the user 16:14:19 catohornet: Who needs to do what to get it installed? 16:14:23 though Osgi provides that awesomeness... the bundles must behave 16:14:26 Madhu: that could be brought back if needed... but right now it will install based on the dependency tree but still it all is the same start-level from a bundle perspective... 16:14:33 who knows what is hiding behind that closet ? 16:14:45 mlemay: that is my point 16:14:54 there are lot of unknowns at this point 16:14:56 madhu: do you remember which bundles needed specific start levels 16:15:02 which can be known only after all the bundles are pulled in 16:15:11 #link https://git.opendaylight.org/gerrit/gitweb?p=integration.git;a=blob;f=features/src/main/resources/features.xml;h=e20c8dc750dfc53213ffcca4b003e9d3bdf225c9;hb=HEAD <- integration features file, with instructions :) 16:15:28 so my request is for this team to consider that Post-M5 there must be a bug fix phase for Karaf 16:15:33 that are identified post M5 16:15:46 Madhu: Yes, the issue with start-levels is that, in a situation like ours they basically make the project specifying them non-compatible reliably with the whole world :( 16:15:47 Madhu: +1 defnitely 16:15:48 Madhu: Yes I understand we will find plenty of bugs once integration testing starts while using the karaf features. I was trying to see if *new* integration tests are needed and if so, what, and how many are there?… or is it that we will run our existing integration tests on the new karaf runtime (and find plenty of dependency/conflict issues etc. 16:15:54 Madhu: They can even break the container :( 16:16:01 There are the RC dates 16:16:02 maybe we should have some goals there? 16:16:03 Madhu: +1 as well… that makes good sense.... 16:16:05 for the integration team? 16:16:20 Madhu: I totally agree, we will see some packaging bugs 16:16:35 edwarnicke: not just packaging. we will see runtime bugs 16:16:43 because no one here has got any time to do integration tests 16:16:45 Is the integration team following this? 16:16:46 Sure 16:16:48 sorry 16:16:50 b/c they're the ones most affected 16:16:51 edwarnke: we have a request open with helpdesk@opendaylight.org. I think they've been pretty busy of late. 16:16:54 i cannot talk for others.. 16:16:57 i didn't do it myself. 16:17:00 tbachman yes we're following this 16:17:04 CASP3R: What have you seen so far in your testing? 16:17:14 https://jenkins.opendaylight.org/integration/job/integration-master-csit-karaf-of13/4/robot/report/report.html 16:17:16 * tbachman always feels sorry for those at the end of the whip 16:17:18 Madhu: I aslo assume that you are mandating that all karaf packaging be completed by M5.. as the tests/bug-fix cycle you describe cannot really occur until that is all done correct? 16:17:18 catohornet: OK... so its just blocking on infra then? 16:17:23 that the current report 16:17:29 phrobb: exactly 16:17:33 catohornet: I've got someone looking into that now. Hopefully we'll have it for you today 16:17:39 M5 is the gate for any new feature to be added to Karaf 16:17:42 phrobb: Could we add that to the TSC agenda for tomorrow? 16:17:49 But I would recommend most folks presume it is so 16:17:50 any bug fixing off there MUST be allowed post M5 16:18:01 edwarnicke: yes… I'll make sure it is a topic for TSC 16:18:02 Madhu: Bug fixing of all kinds is fair game post M5 :) 16:18:11 edwarnicke: yes 16:18:18 edwarnicke: i mean those identified post M5 :) 16:18:29 Madhu: Again, same with every other kind of bug 16:18:39 edwarnicke: this is too abstract 16:18:40 although given the rapidfire irc here edwarnicke could you please define "that" for the TSC agenda for tomorrow? 16:18:45 shall we set some ground rules guys 16:18:49 Madhu: OK... feels concrete to me 16:18:52 we don't want arbitrary hand waves 16:19:11 Madhu: We did a lot of clarifying already 16:19:36 edwarnicke: maybe it is clear for u 16:19:43 can u please explain what are the ground rules ? 16:19:46 so everyone is clear too 16:19:47 ? 16:19:58 phrobb: Could you read back the stuff you did earlier in the meeting? 16:19:59 #info Thomas Kee PacketCable PCMM 16:20:08 xsited: Welcome! :0 16:20:22 thx 16:20:23 what I read is a statement that Karaf features be finalized at M5 16:20:41 and then a question of whether bugs found post M5 are allowed to be fixed? 16:20:51 did I miss something or did I distill that correctly? 16:21:06 regXboi: Yeah... I think the big point is: get your features in for M5! 16:21:09 regXboi: that is my request / hope 16:21:09 Are bugs filed post M5 not allowed to be fixed (after M5 before the release)? Do we say anywhere? 16:21:24 so... my answer would be to the statement: +1 16:21:25 abhijitkumbhare: That's the confusion I'm trying to avoid 16:21:32 Post M5 we clearly are testing, finding bugs, and fixing them 16:21:57 to the question: I agree with what edwarnicke just said 16:22:04 and was going to say it myself 16:22:20 Which is part of the clarificaiton early in the meeting about new tests being acceptable post M5 16:22:22 ed: looking at the projects most of them will not have proper freatures in my monday unless some miracle happens 16:22:29 edwarnicke: u said this before : "[08:47am] edwarnicke: abhijitkumbhare: Definitely. No committing new bugs after code freeze! " 16:22:30 Yes - I would think bug fixes at any point should be a fair game - regardless when bug is found. I mean what if a blocker is found just before the release (due to some bad code pushed in of course) 16:22:32 ed: I just redid the list 16:22:32 what does that mean ? 16:22:50 abhijitkumbhare: edwarnicke Correct. A bug is a bug regardless when it is found. The gate through the RCs and final release if if a bug is critical enough to stop a release. 16:23:00 Madhu: I think he was joking there 16:23:02 it mean don't create any new bugs 16:23:15 tbachman: lol. i wish 16:23:38 Madhu: it was a joke in response to abhijitkumbhare correcting himself from talking about "new bugs" to "new bug fixes" :) 16:23:41 phrobb: +1 16:23:58 edwarnicke: now i need a translator to find out what is a joke 16:24:08 lol 16:24:11 and what is being said seriously :) 16:24:12 lol 16:24:13 yes - edwarnicke Madhu :) 16:24:23 Madhu: if it's from me, it's a joke, but it's not funny 16:24:25 Madhu: Apologies, I'll try adding in the future :) 16:24:38 Madhu: because your point is quite valid 16:24:55 am just perpetually confused i think :) 16:25:00 since most of u got that as a joke 16:25:08 am a sinister sitting here and thinking about that ;) 16:25:27 Madhu: sleep will do us all good 16:25:40 I will also add the :) 16:25:42 Madhu: Truly, apologies, I should be more cautious about not being misunderstood in this context 16:25:48 tbachman: wtf r u talking about ;) 16:25:51 lol 16:26:04 #joke in meetbot ? 16:26:05 tbachman: What is this sleep of which you speak? 16:26:06 anyways 16:26:13 Madhu: Totally #joke 16:26:16 phrobb: can u please summarize ? 16:26:21 Madhu: They could have their own section in the minutes :) 16:26:31 * tbachman notes how patient phrobb is with all of us 16:26:58 LOL... sometimes we are like hyperactive children :) 16:27:03 sorry too many meetings this morning… reading scrollback 16:27:06 sometimes? 16:27:07 * edwarnicke looks around for pixie sticks... 16:27:17 paulq: lol... 16:27:38 paulq: we can hardly recognize folks here if they change their hyper activism :) 16:27:57 let's go back to Karaf status, any project has started Karaf at this time 16:28:10 madhu sorry, what do you want me to summarize? 16:28:15 gzhao: or is it better to be "has not"? 16:28:17 gzhao: I think you are the designated adult 16:28:44 phrobb: just the obvious :) on what is a code freeze ? LOL 16:29:02 sorry phrobb ... we need it because of all the API freeze confusions that happened post M4 ;) 16:31:19 * regXboi amazed at sixty seconds of IRC silence 16:31:23 hahaha 16:31:35 Code Freeze means that no new features/functionality are to be allowed into the Helium code base. Only errors/bugs identified in the bugzilla system should be allowed. The exceptions to this include new tests, and documentation. Packaging, ie Karaf packaging must be completed by M5. Errors/bugs found post M5 are still bugs and they may be created and worked on post M5 16:31:37 gzhao: I am going to have a list in a couple of minutes 16:31:40 That sound about right? 16:32:01 phrobb: one more loophole to close 16:32:05 phrobb: thanks. 16:32:39 phrobb: In response to Madhu's comments, could we make it clear that packaging bugs can be fixed post M5 as well? 16:32:48 by "errors/bugs" you mean filed defects, *not* a feature add masquerading as a bug :) 16:33:00 The committers on each project are responsible for adhering to the code freeze. They must not allow new features in… no matter how much the contributor tries to spin the new feature as a bug fix. 16:33:05 I think Madhu raised the valid point that folks might be confused by 'packaging must be completed by M5' meaning 'you can't fix your packaging bugs' which I don't think was the intention. 16:33:09 phrobb: thx 16:33:43 Yes, bugs in packaging (which we expect plenty) found post M5 are still bugs and can be created and worked on Post M5 (Code Freeze) 16:33:43 i agree guys. 16:33:49 to me freeze is to help us 16:33:52 not threaten us :) 16:34:06 * mlemay is concerned by the current state of karaf features in majority of projects... all the core is in (level 2) priority but the level 3 is far from complete 16:34:10 so anything that works best for the developers must be considered 16:34:12 We should info this stuff in 16:34:27 tbachman: Congratulations, you are now the designated adult ;) 16:34:30 phrobb: +1 16:34:32 lol 16:34:37 Would a bug requesting additional debug/logs/specific troubleshooting tool be classified as a feature or a bug? Hypothetical question (don’t have anything in mind). 16:34:58 abhijitkumbhare: i would just use common sense in these cases 16:34:59 abhijitkumbhare: That's a really good question 16:35:07 if it is not going to impact anyone, we should 16:35:08 Madhu: I tend to agree with you 16:35:24 we are letting committers for each project to mature up :) and see things through 16:35:29 Madhu: But I would point out, folks should be cautious about overlogging as that *does* impact folks 16:35:32 including regXboi point 16:35:34 * edwarnicke stares guiltily at his shoes 16:35:41 * tbachman sees the good-intentioned print statement leading to NPEs in corner cases 16:35:53 tbachman: You bring up a good point 16:35:55 For Hydrogen 16:35:59 We did concerted log cleanup 16:36:10 Including rooting out System.out.println's outside of tests 16:36:12 tbachman: edwarnicke this is where common sense comes through :) 16:36:15 I think we may want to do that again 16:36:17 :) 16:36:22 code-reviews, committers, etc... can catch it 16:36:23