18:01:45 #startmeeting tsc 18:01:45 Meeting started Thu Feb 26 18:01:45 2015 UTC. The chair is colindixon. Information about MeetBot at http://ci.openstack.org/meetbot.html. 18:01:45 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 18:01:45 The meeting name has been set to 'tsc' 18:01:53 #topic roll call and agenda bashing 18:02:08 #chair tbachman dfarrell07 regXboi phrobb gzhao 18:02:08 Current chairs: colindixon dfarrell07 gzhao phrobb regXboi tbachman 18:02:13 #info colindixon 18:02:21 * regXboi trying to reach webex 18:02:26 #info dfarrell07 for cdub and Red Hat (on TSC mailing list) 18:02:29 * regXboi is glutton for punishment 18:02:30 #info edwarnicke for fun 18:02:44 #info sdean778 asked for us to add a topic on how to hanlde vendor-specific code, it’s on the agenda 18:03:05 you might designate someone to be the alternate physical recorder if you can't get the recording to start 18:03:07 #link https://meetings.opendaylight.org/opendaylight-meeting/2015/tsc/opendaylight-meeting-tsc.2015-02-19-18.00.html meetings from last week 18:03:10 #undo 18:03:10 Removing item from minutes: 18:03:14 #link https://meetings.opendaylight.org/opendaylight-meeting/2015/tsc/opendaylight-meeting-tsc.2015-02-19-18.00.html minutes from last week’s meeting 18:03:32 * regXboi has made the webex 18:03:40 colindixon: we had an issue with that in SFC 18:03:42 didn’t work 18:03:43 * regXboi says mark the damn calendar! 18:03:48 #info regXboi 18:03:50 we couldn’t use the host key 18:03:58 same thing you saw lasst week 18:03:59 on TSC 18:04:00 yeah 18:04:08 it’s not the SFC meeting that’s the issue 18:04:15 the thing about webex for me is the webex "test" meeting works on linux but I can't join a real meeting with it :( 18:04:25 we can try it 18:04:34 I was wondering if it has to do with too many simultaneous meetings 18:04:37 with host role 18:04:44 #action colindixon to continue to follow AD-SAL deprecation between VTN, OVSDB, and controller (are ther others?) 18:04:44 * dfarrell07 braces to be punted from WebEx 18:04:48 lol 18:04:50 as an ongoing thing 18:05:09 #info kwatsen 18:05:13 #action colindixon and TSC to elaborate on proper procedure for removal of committers from a project 18:05:16 phrobb: saves the day 18:05:21 #action colindixon, zxiiro and others to coordinate figuring out how to set up opt-in/opt-out automatic version bump patches and tags 18:05:26 #action colindixon to start a discussion around possible use cases as focus areas for lithium on the mailing list 18:05:39 phrobb: the issue during the SFC meeting was that it said something about exceeding number of meetings 18:05:45 when we tried to use the host key 18:05:52 there were 3 meetings at the same time 18:05:53 AAA 18:05:55 DLUX 18:05:56 #info Chris Price (late again) 18:05:57 and SFC 18:06:08 (at least there are on the calendar) 18:06:13 #info Youcef Laribi 18:07:06 #info gzhao still hasn’t received response from Maple and Defense4All 18:07:16 #info phrobb is still trying to work out IPR issues with Maple project 18:07:33 #action phrobb to keep working with Maple (he is “cautiously pessimistic”) 18:07:38 color commentary ;) 18:08:02 #info LLDP with Auto-Attach was approved via the mailing list last week 18:08:07 #undo 18:08:07 Removing item from minutes: 18:08:19 #action phrobb to reach out to Rob Sherwood about closing down net-virt-platform 18:08:27 #info LLDP with Auto-Attach project was approved by the TSC into incubation via the mailing list last week 18:08:32 #action phrobb to reach out to plugin2oc, SDNi, and defense4all about JJB migration 18:09:24 #action phrobb to hunt down projects to see if they would continue to support Helium (at least) until Lithium is released (we are at 13 of 20 remaining projects agreeing to support Helium until Lithium releases) 18:09:49 #info colindixon asks if we got feedback on new tentative dates for new SR’s for Helium 18:10:02 #info jmedved 18:10:03 #link https://lists.opendaylight.org/pipermail/release/2015-February/001461.html < proposed Helium SR3 & SR4 18:10:06 huzzah! 18:10:07 #topic Updates 18:10:28 #info phrobb says that everything moving along for April Hackfest — encourages folks to register 18:10:48 #info CFP has closed, received 90 submissions; notifying speakers week after next for those that were accepted 18:11:07 #info phrobb says that the CFP for the event in India has opened 18:11:14 * regXboi hey does that mean we can offer cookies? 18:11:18 :-) 18:11:35 regXboi: we have a plan — cookie cutters shaped like coins :) 18:11:41 lol 18:11:53 tbachman: how about cookies shaped like super cows? 18:11:53 #info Lithium and Stable/Helium 18:11:55 #undo 18:11:55 Removing item from minutes: 18:12:01 cookies are the coin of my realm!!! 18:12:03 #topic lithium and stable/helium 18:12:14 #info colindixon asks how we’re doing with picking SR3/SR4 dates 18:12:20 #link https://lists.opendaylight.org/pipermail/release/2015-February/001461.html < proposed Helium SR3 & SR4 18:12:32 gzhao: thx for that link :) 18:13:00 what was yesterday's hackfest regXboi ? 18:13:01 tbachman: sure 18:13:12 #info Proposed dates are March 26th and July 17th for SR3 and SR4, respectively 18:13:13 abhijitkumbhare: will explain on the call 18:13:18 abhijitkumbhare: was wondering the same 18:13:27 regXboi: switches hats 18:13:42 #info regXboi says there was a hackfest to try to get the openstack CI test jobs for ODL to work 18:14:03 #info regXboi says this involves patching stable helium; came across issues with UserManager 18:14:20 regXboi: switche hats again 18:14:35 regXboi has many hats :) 18:14:40 abhijitkumbhare: :) 18:14:49 abhijitkumbhare: I'm waiting for his cookie hat ;) 18:14:50 * ChrisPriceAB +1... 18:15:06 * Prem_ Missed counting the # of hats 18:15:12 * ChrisPriceAB both to the cookie hat and the date 18:15:27 is somebody offering a cookie hat? 18:15:33 RedHat 18:15:38 18:15:39 M7 lel 18:15:40 that would actually make eating one's hat relatively interesting :) 18:16:01 regXboi: http://www.hercampus.com/sites/default/files/2012/11/12/Pilgrim%20hat%20cookies_2.jpg 18:16:02 vote vote vote!!! 18:16:07 colindixon: 7 by my count 18:16:41 Rajeev is here 18:16:45 colindixon: ed 18:16:54 should we info again? 18:17:02 #info LuisGomez 18:17:04 #info 18:17:13 RajeevK_: it's #info ;) 18:17:16 #info RajeevK 18:17:18 #info RajeevK 18:17:33 #undo ? 18:17:41 #undo 18:17:41 Removing item from minutes: 18:17:43 #undo 18:17:43 Removing item from minutes: 18:17:45 #undo 18:17:45 Removing item from minutes: 18:17:50 * dfarrell07 doesn't have a strong opinion about the exact dates, seem fine to me 18:17:59 * ChrisPriceAB echo 18:18:02 gzhao: keeping it clean :) 18:18:03 thx 18:18:10 July 17th 18:18:10 #info RajeevK_ 18:18:33 ah, yes 16th 18:18:34 gzhao: Thx! 18:18:46 RajeevK_: np 18:19:15 #startvote Shall Helium SR3 release be on March 26th and tentatively SR4 be July 16th? -1, 0, +1 18:19:15 Begin voting on: Shall Helium SR3 release be on March 26th and tentatively SR4 be July 16th? Valid vote options are -1, 0, +1. 18:19:15 Vote using '#vote OPTION'. Only your last vote counts. 18:19:21 #vote +1 18:19:23 #vote +1 18:19:26 #vote +1 18:19:26 #vote +1 18:19:27 #vote +1 18:19:27 #vote +1 18:19:27 #vote +1 18:19:30 #vote +1 18:19:32 #vote +1 18:19:36 #endvote 18:19:36 Voted on "Shall Helium SR3 release be on March 26th and tentatively SR4 be July 16th?" Results are 18:19:36 +1 (9): jmedved, regXboi, ChrisPriceAB, LuisGomez, edwarnicke, dfarrell07, colindixon, Youcef, RajeevK_ 18:20:01 * ChrisPriceAB wonders if he can cheekily undo other peoples votes.... 18:20:05 #agreed Helium SR3 release will be on March 26th and tentatively SR4 be July 16th 18:20:13 phrobb: thx! 18:20:23 #action gzhao to put SR3, SR4 tentative date to Helium release plan 18:20:27 tbachman: my pleasure 18:21:40 #info regXboi notes that as a committer in neutron, he can’t merge backports to stable/helium, since it’s in the controller 18:21:50 #info regXboi and edwarnicke note that getting the fixes for Neutron for SR3 done will involve porting patches from Neutron to Controller and that will have complexities 18:22:43 #info gzhao says that defense4All and Maple are the only 2 projects with M2 status 18:22:57 #info gzhao says the Neutron project is still working on M2, but they are spinout so not due until M3 18:23:14 #info gzhao says offset 1 projects have M3 status due next Thursday 18:23:36 #info gzhao encourages all offset 1 and 2 projects to attend next Monday’s TWS, on integration project’s testing requirements 18:24:22 #link https://docs.google.com/spreadsheets/d/1PYxjiSYEks44uJByVO1P44rnI5xTJRulpKyrSsDQF9g/edit#gid=1598222722 < -- single feature test tracking 18:25:12 #info gzhao says that the spreadsheet provided in the link; AAA needs to submit a patch to turn on the single-feature-test 18:27:48 #action ALL PROJECTS need to make sure your single feature test is turned on and running 18:27:56 edwarnicke gzhao for OpenFlow plugin michal_rehak has done https://git.opendaylight.org/gerrit/#/c/15728/1 - have not updated the spreadsheet 18:28:37 #link https://lists.opendaylight.org/pipermail/release/2015-February/001488.html the mail ddescribing the issue and how to fix it 18:30:10 * tbachman fell down… uses clapper to turn on the lights to get back up 18:30:22 abhijitkumbhare: that spreadsheet is for each project to fill 18:30:27 #info https://git.opendaylight.org/gerrit/#/c/15773/ 18:30:27 #info rovarga_ asks for odlparent to merge two patches outstanding: one which fixes eclipse and one which makes the feature test run faster 18:30:46 yes - gzhao (will fill it) 18:30:48 #action dfarrell07 to get cdub to review and merge odlparent patches 18:31:14 abhijitkumbhare: don't worry, already done 18:31:29 #topic System integration and Test 18:31:43 thx gzhao 18:31:45 #info LuisGomez asks what to do when there are projects that block integration and are unreachable 18:32:58 #link https://lists.opendaylight.org/pipermail/tsc/2015-February/002643.html Email from LuisGomez on this issue 18:33:48 I'm all good. 18:33:54 * ChrisPriceAB limited trust issues 18:34:21 #info colindixon says he feels that it’s within integration’s scope to remove them in order to move forward 18:34:30 #undo 18:34:30 Removing item from minutes: 18:34:57 #info colindixon says he feels that it’s within integration’s scope to remove a project from integration build if they don’t respond within 48 business hours in order to move forward 18:35:32 #info LuisGomez asks if it would be possible to label/mark the projects “we know” nobody maintains/supports anymore (OSCP, Affinity, etc…) 18:36:04 btw ... there are about 16 projects still using yangtools features-test, afaict 18:36:17 rovarga_: thx! 18:36:27 #link https://wiki.opendaylight.org/view/Project_list this list says nothing about how “maintained” a project is, LuisGomez would like to add information about certain projects being abandoned, e.g., affinity and OSCP 18:36:47 #info rovarga_ says there are about 16 projects still using yangtools features-test, by his count (applied towards previous topic) 18:36:57 rovarga_: if they are not on https://docs.google.com/spreadsheets/d/1PYxjiSYEks44uJByVO1P44rnI5xTJRulpKyrSsDQF9g/edit#gid=1598222722 could you add them? 18:37:38 after 1.5 years 18:37:40 * dfarrell07 directly pinged cdub about patches, also added him as a reviewer on Gerrit /cc rovarga_ 18:37:42 I would like some description added. Even "Last Modified" could be useful 18:38:04 #info edwarnicke suggests that we maybe reach out to aks projects to terminate themselves if projects seem like they shoudl abe archived 18:38:07 * tbachman missed that 18:38:07 hehe 18:38:25 yes 18:38:31 it's the 1.5 for an outside termination 18:40:57 * tbachman fell down again 18:40:58 back 18:40:59 sorry 18:41:02 #action colindixon to send a mail to the list and (possibly) put it on the agenda for next week about ways to trackwhich projects are active 18:41:47 #topic Vendor Specific Code 18:42:10 #link https://lists.opendaylight.org/pipermail/tsc/2015-February/002618.html the thread that sparked this topic 18:42:10 #link https://lists.opendaylight.org/pipermail/tsc/2015-February/002613.html email from Steve Dean on vendor-specific code 18:42:16 #undo 18:42:16 Removing item from minutes: 18:42:18 :-) 18:42:19 oops 18:42:26 lol 18:42:33 * tbachman is sleeping today 18:42:42 #info sdean778 asks what the TSC’s opinion is on this 18:42:55 #info regXboi asks what the code license is 18:43:31 #info colindixon says this is to submit code under the eclipse license 18:43:41 Does it not just go through review and get added under eclipse based on that review? 18:43:59 * ChrisPriceAB +1 18:44:50 #info regXboi says it should be split into a legal and a technical part, on the legal side the code must be able to be EPL code if we’re going to host it in ODL repos 18:45:14 I think on the technical side, we may want to enforce an API/IMPL separation 18:45:17 #info edwarnicke says from a technical perspective, he cares more that things can be side loaded without changes to the core repo than anything else 18:45:28 #undo 18:45:28 Removing item from minutes: 18:45:45 e.g. the vendor-specific code is pure plugin and does not wag the dog 18:45:46 #info edwarnicke says from a technical perspective, he cares more that things can be side loaded without changes to the repo than anything else 18:46:49 #info is this "vendor specific code" being provided as free and open source code for ODL? That can be developed beyond "vendor specific"? 18:46:54 #info sdean778 asks if we want the distribution to support vendor devices out-of-the-box 18:47:16 #info phrobb asks if there’s a construct in our repo and test environment to enable this — we need the vendor devices to do this 18:47:51 #info ashaikh wonders if there’s a way to indicate that this isn’t licensed the same way 18:47:56 ashaikh: ^^ is that what you said? 18:48:08 ashaikh: hi btw :) 18:48:25 #info colindixon says there’s the question of licensing, but also how to do we make sure it’s tested 18:48:34 tbahchman:(hey!) yes, using a third-party directory convention 18:48:43 Would a ttp definition in the ttp project not also be "vendor specific" code? 18:48:45 #info abhijitkumbhare says it doesn’t have to be a separate licesnse 18:48:47 that should be :-) 18:48:51 rovarga_: yes, the DIDM design already has separate api/impl. a vendor plugin could be part of the project or kept separate already. 18:48:55 ChrisPriceAB: it’s a good question, quite posibly 18:49:06 #info regXboi says licensing is where you get tripped up 18:51:38 Do we have Open Source device drivers for proprietary devices in Linux? 18:51:47 #topic Best Practices 18:52:11 colindixon: Can we consolidate the code style pages? I created this https://wiki.opendaylight.org/view/ODL_Coding_Guidelines 18:52:12 #link https://wiki.opendaylight.org/view/Developer_Best_Practices Developer Best Practices wiki page 18:52:27 #link https://wiki.opendaylight.org/view/Logging_Best_Practices Logging Best Practices wiki page 18:52:43 #link https://wiki.opendaylight.org/view/General_Style_Guidelines General Style Guidelines wiki page 18:53:05 #link https://wiki.opendaylight.org/view/CrossProject:HouseKeeping_Best_Practices_Group:Main Housekeeping Best Practices wiki page (main) 18:53:38 #link https://lists.opendaylight.org/pipermail/controller-dev/2014-July/005766.html email describing commiter checklist 18:53:52 Is edwarnicke's audio borking or is it just me 18:54:00 alagalah: it’s just you 18:54:06 vunderbar 18:54:06 alagalah: not just you 18:54:15 tykeal: huh 18:54:21 breaking 18:54:38 colindixon: tykeal Its entirely possible its me... the time delay from my position out in orbit causes issues :) 18:54:43 * alagalah space cadet 18:54:52 colindixon: only audio problems I've been having have been with edwarnicke for some reason 18:55:25 * rovarga_ hears JNI and puts on his TCP/MD5 hat 18:56:01 * alagalah there's more hats in here than a rodeo 18:56:45 lol 18:56:55 as the person doing a ton of deployment stuff for Integration team, major +1 to that colindixon 18:58:01 I would be comfortable 18:58:05 * ChrisPriceAB either document or formally vote... 18:58:31 non-portable dependencies should always be optional 18:58:35 * tbachman is totally unhelpful today 18:58:45 (for a particular release) 18:59:58 Shall the TSC acknowledge there are best practices such as not using JNI and maintaining just the “download and run” model is strongly preferred 19:00:30 the TSC strongly suggests that projects avoid things that will break cross-platform compatibility, e.g., the use of JNI, and avoid breaking the simple “download, unzip and run”, i. i.e., allow for external dependencies, but don't require them 19:00:52 the TSC strongly suggests that projects avoid things that will break cross-platform compatibility, e.g., the use of JNI, and avoid breaking the simple “download, unzip and run” deployment model, i.e., allow for external dependencies, but don't require them 19:01:01 * ChrisPriceAB ok with me 19:01:03 yes 19:01:15 * ChrisPriceAB all good to agree 19:01:24 * ChrisPriceAB but that's cause I would vote +1 19:01:45 the TSC strongly suggests that projects avoid things that will break cross-platform compatibility, e.g., the use of JNI, and avoid breaking the simple “download, unzip and run” deployment model 19:02:02 I'm down with the above from colindixon 19:02:46 #actoin colindixon to start a discussion for the TSC to agree on a variant the statement “the TSC strongly suggests that projects avoid things that will break cross-platform compatibility, e.g., the use of JNI, and avoid breaking the simple “download, unzip and run” deployment model” on the mailing list 19:02:53 #action colindixon to start a discussion for the TSC to agree on a variant the statement “the TSC strongly suggests that projects avoid things that will break cross-platform compatibility, e.g., the use of JNI, and avoid breaking the simple “download, unzip and run” deployment model” on the mailing list 19:03:03 Maybe append: Where this isn't possible, please document the situation clearly 19:03:15 #action colindixon start a thread to have some kind fo subgroup deal with other best practices 19:03:36 #topic Cookies 19:03:40 thanks 19:03:40 #endmeeting