19:00:06 #startmeeting refstack 19:00:07 Meeting started Mon Mar 21 19:00:06 2016 UTC and is due to finish in 60 minutes. The chair is catherineD. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:00:08 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 19:00:10 The meeting name has been set to 'refstack' 19:00:47 hello Catherine 19:01:11 dliu: hello welcome to IRC meeting 19:01:24 thank you :-) 19:02:06 o/ 19:03:45 o/ 19:04:48 alexandrelevine: rockyg: you there? 19:05:12 o/ 19:05:49 o/ 19:06:13 alright 19:06:15 #link meeting agenda and notes, please feel free to add items https://etherpad.openstack.org/p/refstack-meeting-16-03-21 19:06:51 #topic Make puppet pull releases from Pypi 19:07:09 #link https://review.openstack.org/#/c/281737/ 19:07:22 pvaneck: any update? 19:07:28 Congratulations to catherineD and alexandrelevine for their PTL wins 19:07:43 (had to get that nonsequitur in) 19:07:50 rockyg: Thx 19:07:51 will hopefully be merged today after having addressed some comments 19:07:53 rockyg: Thank you. 19:08:25 alexandrelevine: Congrats 19:08:39 pvaneck: that is great! 19:08:59 any question on this topic? 19:09:06 Congrats 19:09:11 rockyg: Where did you get this info about elections? 19:09:39 rockyg: I thought they'll talk tomorrow in the TC meeting. 19:10:04 They will. And if you can, you should attend. 19:10:42 #topic Vendor user management implementation 19:11:02 #link Add user management REST API https://review.openstack.org/#/c/285714/ 19:11:09 But, it's pretty much a slam dunk after all the email back and forth. And sorry for the hassles. I should have pinged you when I pinged catherineD to get your nomination in, but I wasn't sure it was you for EC2 19:11:09 rockyg: I will, but I'm wondering if I'm missing something now that you are congratulating already. I thought it's not quite decided yet. 19:11:26 rockyg: Thanks anyways :) 19:13:10 ready for the discussion on Add user management REST API https://review.openstack.org/#/c/285714/ ? 19:13:16 yes 19:13:47 alexandrelevine: for that patch .. I think it need a couple of unit tests ... 19:14:48 catherineD: Unfortunately at least until end of March, probably a little more we don't have any resources for that. 19:15:22 alexandrelevine: do you mind you someone else on the team help landing that? 19:16:11 catherineD: Which team? RefStack team? Of course 19:16:19 Hi, all! o/ 19:16:20 yea RefStack team 19:16:27 sslypushenko: Hi 19:16:48 alexandrelevine: OK 19:16:55 Hye sslypushenko ! 19:16:55 catherineD: I think it'd be fair to let this review in and add unit tests in the next one right after. Do you think it's possible? 19:16:59 sslypushenko:hi 19:18:31 alexandrelevine: I think it is best to add unit test ... 19:19:05 catherineD: +1 19:19:06 that is IMO 19:19:36 ok 19:19:48 moving on 19:19:49 I can add the unit tests onto the patch in a bit. won't take long 19:20:07 alexandrelevine: Andrey still be the owner of the patch 19:20:11 pvaneck: thx 19:20:12 we should keep unittest coverage at 100% 19:20:12 it's ok 19:20:33 #topic List RefStack Users 19:21:11 #link Catherine sent email to community http://lists.openstack.org/pipermail/defcore-committee/2016-March/001066.html 19:21:19 Got some feedbacks 19:21:30 also some comment on the Security meeting 19:21:50 #link Security team meeting http://eavesdrop.openstack.org/meetings/security/2016/security.2016-03-17-17.00.log.txt 19:22:55 In general, people are hesitating in listing the entire user list ... 19:23:34 funny comment. 19:23:53 that leading us to the next topic on how to add user reliably 19:24:24 Let's concentrating on this with Alex's recommendations 19:24:54 #topic Alex's recomemdation on adding user to vendor 19:25:06 I am in favor of option 4.2 19:25:47 in the future, when we have email notification iimplemented ... we could go with both option 4.2 and 4.3 19:26:13 but for now ... the most possible implementation at this stage is 4.2 IMO 19:26:26 what do you think? 19:27:30 * rockyg is digging through my mailboxes to find the email 19:28:03 I'm down for whatever keeps us moving 19:28:23 Easiest would be 4.1 19:28:48 And email is preserved so I don't see what's wrong with it. 19:28:49 4.1 gets us back to the discussion of list users ... 19:29:20 There is no problem with list of users. There is a problem with private data, which is email as I thought. I heard that name is not that private. 19:30:36 After all it's clearly written in the openstackid, isn't it? 19:31:02 we sent an email and get feedback for not list uers ... are we ignor the feedbacks? 19:31:17 I am just saying that for this release I want to avoid list users 19:31:57 so 4.2 give us an option that 1) we do not keep discuss about list user 2) it does not expose any privacy data 19:32:13 I just need us to move on ... 19:33:22 ok 19:33:45 why is it still an option for today then? 19:33:51 catherineD, reading those emails again, it seemed to be ok to list full name and openid, just not other info. 19:34:38 alexandrelevine: what is it still an option for today? 19:35:07 catherined: 4.1 19:35:16 rockyg: I am not sure that is the case ... the fastest way to get us out of this is to not "list user" 19:35:43 oops. My bad. Only members of org. can list users of same org. 19:36:04 rockyg: Don't think so 19:36:34 I think we can allow to any user list name+openID for other any user 19:37:01 alexandrelevine: here is one of the answer 19:37:16 In my opinion, listing users should work as follows: - Any user can list the users of the organizations (s)he belongs to. 19:37:24 that is from http://lists.openstack.org/pipermail/defcore-committee/2016-March/001067.html 19:37:44 that means that only user in a vendor can list user in that vendor 19:38:49 again team, please read the email threads ... to get us moving let's implement something else ... 19:38:55 catherineD: But how we will add user to vendor? 19:38:56 sslypushenko, I agree with you, just not in the email. 19:39:11 sslypushenko: we will use option 4.2 19:39:25 https://etherpad.openstack.org/p/refstack-meeting-16-03-21 19:39:30 oh! thx 19:40:06 a user will request to be added to a vendor ... since the user asks to be added ... he/she will provide all the necessary info 19:40:19 It is complicated 19:40:38 Do we really want this complication& 19:40:40 ? 19:40:59 sslypushenko: it is ... but eventually we will implement something like that any way when we get our notification framwork implemented 19:41:16 what I invision is for now 19:41:24 got it 19:41:24 user willl log in 19:41:27 Should be very similar to the vendor approval process with vendor admin instead of foundation admin 19:41:46 and in his profile we will provide a link for request to be add to a vendor 19:42:08 for private vendor .. he.she will need to get a vendor id from the vendor 19:42:09 Vendor users shouldn't be that many people 19:42:13 but... I'm not sure that RefStack is ready to become so complicated tool 19:42:23 for public vendor they can choose from a list 19:42:50 for what I just discribe ... I think it is doable 19:43:04 catherineD: Everything is doable) 19:43:16 to add a link on the user profile to request for being added to a vendor 19:43:18 It depends only on time and money) 19:43:35 sslypushenko: I understand 19:43:36 Another option would be to have a "create user" utility for the vendor admin 19:44:28 But in other hand... 19:44:32 Then there would be nothing in Refstack until the vendor admin has enough info from his company to create the account 19:44:50 I think the only option is 4.2 (which is complicated but doable) which does not ignore user feedback and does not expose private data 19:45:09 Now any openstack.org member can find openstackid for all openstack.org members 19:45:29 what the point to hide it in RefStack& 19:45:36 ? 19:46:12 sslypushenko: that is what we ask of the community ... the feedback is do not expose 19:46:13 sslypushenko, smaller group of people are in the Refstack list. But, really, if admin creates the user, all the issues go away. 19:47:20 we aill have to follow up with them onm reason ... but now that we have asked we need to respect the opinion ... meanwhile we can proceed with alternative 19:48:11 catherineD: let it be so 19:48:30 yea I am torn here :-) 19:48:34 Thx 19:48:45 now that we ask community 19:48:56 moving on .. 19:48:58 Your point of view is also sounds reasonable 19:49:31 thank alexandrelevine: to include option 4.2 19:49:38 catherineD: np 19:49:56 moving on 19:50:08 #topic Vendor registation 19:50:29 #topic spec: https://review.openstack.org/#/c/274837/ 19:51:04 the spec has merged ... just need implememtation 19:51:40 I really hope for us to have vendor resgistration before the summit ... 19:52:14 alexandrelevine: Andrey going to be tired up until end of March? 19:53:52 catherineD: At least 19:54:02 alexandrelevine: ok 19:54:06 catherineD: But he might have the code already 19:54:08 we have 7 mins left 19:54:25 let move to topic 7 19:54:36 #topic Pending review 19:54:36 catherineD, I need your input for the product roadmap! 19:55:04 And that's for item 8 19:55:09 7.1.1, 7.1.2 and 7.1.3 are all ready to merge 19:55:54 rockyg: did you put RefStack i18n discussion(optional discussion choice). ? 19:56:09 catherineD: I put 19:56:10 rockyg: that is 8.1 19:56:22 just for open discussion 19:57:13 but seems we don't have time today 19:57:15 rockyg: I have ahard time on product roadmap .. since we did not meet this cycle target 19:58:02 in fact we are far from the target ... I have a hard time to plan beyond Newton 19:58:26 doesn't matter targets. matters that lots got done3 and the plans for next cycle 19:58:53 rockyg: next cycle is OK ... but not beyond that 19:59:09 Newton plan is good enough. and pointer to spec that pretty much lays out all the work that needs to be done for at least two cycles. 19:59:27 Have to run. Bye everybody 19:59:39 rockyg: I will send you our plan .. sorry it take some time becaue I want to push some for this cycle ... we still have a month 19:59:45 thank you all 19:59:45 I feel like i18n should wait until we are in a more stable state with vendor registration being complete 19:59:59 pvaneck: I think so 20:00:05 #endmeeting