16:00:18 <sgallagh> #startmeeting Server Working Group Weekly Meeting (2013-11-26)
16:00:18 <zodbot> Meeting started Tue Nov 26 16:00:18 2013 UTC.  The chair is sgallagh. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:18 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
16:00:28 <sgallagh> #chair sgallagh mizmo nirik davidstrauss Evolution Viking-Ice simo tuanta mitr
16:00:29 <zodbot> Current chairs: Evolution Viking-Ice davidstrauss mitr mizmo nirik sgallagh simo tuanta
16:00:33 <sgallagh> #topic roll call
16:00:39 <sgallagh> .hellomynameis sgallagh
16:00:40 <zodbot> sgallagh: sgallagh 'Stephen Gallagher' <sgallagh@redhat.com>
16:00:47 <tuanta> .fas tuanta
16:00:48 <zodbot> tuanta: tuanta 'Truong Anh Tuan' <tuanta@iwayvietnam.com>
16:00:56 <nirik> me is here, but kinda swamped and trying to catch up\
16:01:15 <mitr> Hello all
16:01:17 <davidstrauss> .hellomynameis davidstrauss
16:01:18 <zodbot> davidstrauss: davidstrauss 'David Strauss' <david@davidstrauss.net>
16:02:11 <simo> hello
16:04:41 <simo> sgallagh: are you driving today ?
16:04:46 * sgallagh waits a few more minutes for others to filter in
16:04:48 <simo> I do not see mizmo online
16:05:04 <sgallagh> simo: Yeah, and one of these agenda items is hers.
16:06:01 <sgallagh> #topic Working Group member selection
16:06:15 <sgallagh> So, Johann has chosen to step down.
16:06:53 <sgallagh> It was suggested on the list that we should approach the folks that previously volunteered.
16:07:20 <Evolution> did anyone reach out to see if those who volunteered are still willing?
16:07:22 <simo> sgallagh: you also porposed to open a call for volunteers ?
16:07:27 <sgallagh> I'd also like to suggest that we query the Fedora QA team for someone to represent their interests here (which was a role I was hoping Johann would fulfill)
16:07:44 * nirik nods sounds reasonable
16:08:14 <tuanta> should we call for more volunteers before?
16:08:14 <sgallagh> nirik: Just to confirm, that was agreeing to which part?
16:08:31 <mitr> For the record, our charter says "from the active Fedora Server community"; that's I suppose fairly irrelevant at this point
16:08:42 <sgallagh> Evolution: I haven't polled them yet, no. Had a busy day yesterday.
16:08:55 <nirik> all of it, in reverse order preference I guess... ie, ask qa first, if no one, call for/ping the people who stepped up before.
16:09:06 <mizmo> hi sorry im late
16:09:08 <mitr> I'd prefer a general call for volunteers as well; we've had some cases of people missing the original self-nomination period
16:09:12 <sgallagh> mitr: Yeah, at this point I think "responds to a request for inclusion" qualifies as "active"
16:09:13 <Evolution> +1 to nirik's suggestion
16:09:19 <mitr> sgallagh: yes
16:09:30 <Viking-Ice> I would assume based on previous discussion that you would simply pick the next candidate from the nominees to fill my place
16:10:11 <sgallagh> Viking-Ice: That was an initial pool, but in my mind we really want to have someone from QA working with us closely.
16:10:48 <tuanta> +1 sgallagh
16:11:17 <zerick> .fas zerick
16:11:17 <zodbot> zerick: zerick 'Erick Ocrospoma' <zipper1790@gmail.com>
16:11:32 <sgallagh> Proposal: Ask the QA community to recommend someone from their ranks. If no one steps forward, put out a general call, particularly to those who previously volunteered.
16:11:37 <Viking-Ice> sgallagh, As has been stated from the beginning reaching out to other communities should not be a problem  for the WG's
16:11:46 <nirik> sgallagh: +1
16:11:57 <tuanta> sgallagh" +1
16:12:01 <mitr> sgallagh: +1
16:12:34 <Evolution> sgallagh: +1
16:12:45 <sgallagh> +1 to my own proposal (for the record)
16:13:10 <sgallagh> I count a majority now. Any dissenting opinions?
16:13:50 <sgallagh> #agreed Ask the QA community to recommend someone from their ranks. If no one steps forward, put out a general call, particularly to those who previously volunteered.
16:13:53 <Viking-Ice> and given that fesco is unwilling to budge with the separated repo which in turn hindrance us in QA to do much needed cleanup in the distribution in the process I will to what is in my power to keep resources from the QA community at minimum to other then the baseWG and the installer
16:14:02 <sgallagh> #action sgallagh to contact QA
16:14:29 <Viking-Ice> I will ensure that people will be knowing on which ground they will be dedicating their free time to
16:14:57 <nirik> you're welcome to spend your time however you like. Others can do the same.
16:15:01 <sgallagh> Viking-Ice: I don't think it's necessary to actively hinder anyone who wants to work with us.
16:15:24 <Viking-Ice> as well as the fact that I'm within my full rights presenting an alternative proposal to the community then what's taking place here
16:15:38 <sgallagh> Viking-Ice: No one is stopping you.
16:15:39 <nirik> sure, but thats off topic here, isn't it?
16:15:59 <sgallagh> Shall we move on to the Personas discussion?
16:16:00 <Viking-Ice> sgallagh, from the looks of it FESCo ricked rolled me in that relation+
16:16:10 <sgallagh> #topic Personas
16:16:14 <Viking-Ice> and it's time they start taking responsability for their action
16:16:22 <nirik> sgallagh: please
16:16:31 <sgallagh> Viking-Ice: I'll thank you to take this discussion elsewhere. This is not the venue for it.
16:17:11 <sgallagh> mizmo: Personas are your department. Kick us off?
16:17:32 <mizmo> sure
16:17:50 <mizmo> so i have really just a start at some rough ones here: https://fedoraproject.org/wiki/Server/Personas
16:18:02 <mizmo> (also I don't know why the full draft document status message isn't displaying, if anybody has ideas... )
16:18:08 <mizmo> anyway i based them off of nirik's suggestions from last meeting
16:18:32 <mizmo> also last week after the meeting sgallagh and i interviewed kdetony about his job as a sysadmin, i have notes from that i still need to post
16:18:40 <Evolution> I wasn't able to attend the last meeting due to a conflict. could I get a cliff's notes version for what personas are to accomplish?
16:18:46 <mizmo> i also need to post an FAQ of questions/answers about personas, i didn't get a chance to write that up
16:18:48 <nirik> yeah, I meant to try and come up with some more, but have been swamped.
16:19:11 <davidstrauss> I believe I'm queued up for an interview in the near future. I think we deferred it until I return to SF.
16:19:23 <nirik> so, the idea is we think of 'typical' groups and their needs and that way we can decide what we want to ship and can think about how it affects the various groups...
16:19:33 <mizmo> Evolution, sure - the basic idea is that we'd have a representative cross section of our target user base represented by the personas, so when we make decisions in the PRD we can go through each persona and make sure we're not hindering any of our target users with a given decision
16:19:43 <mizmo> later on it really helps marketing and design figure out the best way to present things to the users
16:20:00 <mizmo> davidstrauss, yep that's right!
16:20:21 <nirik> mizmo: we also would order them some? ie, these are are target, these others we don't want to harm, but not primary, these other ones we aren't targeting at all?
16:20:29 <sgallagh> Evolution: https://fedoraserver-wgblog.rhcloud.com/?p=32
16:21:01 <mizmo> nirik, yep exactly. We would probably have a couple of primary personas, a few secondary (maybe 3-4), and we can define anti-personas too
16:21:09 <Evolution> yeah. I read that, I just wasn't clear on the technical aspect for how this would help.
16:21:26 <mizmo> we don't need to define every possible anti-persona, but if there's one type of user that causes communication issues / conflation of issues, it might be worth defining an anti-persona to help document that
16:21:29 <sgallagh> Evolution: Right, it's mainly for decision-making. "Is a change going to help or harm the following groups?"
16:21:39 <Evolution> understood.
16:21:47 <mizmo> Evolution, if users were hardware, it's like a hardware testing matrix, maybe that makes more sense ;)
16:22:23 <simo> mizmo: wetware testing matrix ?
16:22:26 <tuanta> :)
16:22:27 <mizmo> hehe
16:22:47 <mizmo> right so, so far we have:
16:23:16 <mizmo> - Senior sysadmin / nonprofit org / she's classified as 'MacGuyver' because they don't have a lot of resources but try to get a lot done
16:23:58 <Evolution> I like the general concept, so long as we don't get caught up in arbitrarily pidgeon-holing things because of this.
16:24:01 <mizmo> - Server app developer / freelance / i didn't come up with a nickname for him yet
16:24:10 <sgallagh> Evolution: We all agree on that score.
16:24:14 <Evolution> okay.
16:24:27 <mizmo> - Junior sys admin / huge company / lots of resources available, no nickname for him yet either
16:24:52 <davidstrauss> What makes the third one "junior"?
16:24:58 <mizmo> maybe the developer should be more devops, but i dont know as much about how devops work to fill that out
16:25:13 <sgallagh> I haven't been able to get it on paper yet, but I also want us to have  Persona that describes us here.
16:25:14 <nirik> I guess tester / docs writer would also fall under the app developer?
16:25:18 <davidstrauss> It seems strange to couple abundant company resources with the "junior" guy.
16:25:21 <mizmo> davidstrauss, my thoughts on that were he might not have as much decision-making power, he's not the team lead or anything like that. so he may be stuck with decisions made over his head
16:25:26 <sgallagh> I.e. "People who want to enhance the server platform"
16:26:02 <mizmo> davidstrauss, well not necessarily. e.g., if he needs more hardware he probably won't have a hard time getting it, but he's not going to be able to change the bureacracy at his org
16:26:06 <sgallagh> I think we need to keep that persona in mind or we're going to end up limiting our own community
16:26:06 <davidstrauss> Well, I guess I'm the contrast to that persona, anyway. We have lots of resources, and I'm CTO.
16:26:19 <mizmo> sgallagh, so kind of a meta persona?
16:26:31 <mizmo> davidstrauss, lol you are the super user! but i think you would be an excellent additional persona
16:26:50 <mizmo> davidstrauss, i would probably nickname your persona 'decision maker' right?
16:26:51 <simo> mizmo: yes I think we need the whole range
16:26:55 <sgallagh> mizmo: I'm not sure what that means.
16:27:02 <sgallagh> (meta-persona)
16:27:10 <simo> if we appeal to the junior but not the CTO we are useless as the junior will never bve allowed to use us :)
16:27:13 <mizmo> sgallagh, if we have a persona for someone who wants to work on fedora server platform itself, i think that's a bit meta isn't it?
16:27:27 <sgallagh> We're talking about something close to "Joe", but focused more on the platform than the apps.
16:27:33 <davidstrauss> mizmo: I don't really care if it's me, but we need someone with resources + power.
16:27:51 <mizmo> sgallagh, can you give me an example of what a platform developer would work on?
16:27:51 <mitr> mizmo: yes, a decision maker is an important case (... honestly, for RHEL)
16:27:54 <sgallagh> davidstrauss: If we had that, we'd be in better shape than we are now :-P
16:28:01 <simo> mitr: for anything
16:28:03 <mitr> Re: developer: I do agree we need a developer persona (particularly for the requirements on languages and deployment); can we clarify that Server is not the OS one uses to actually develop the application (i.e. run Eclipse / ask on stackoverflow ...) - that should be Workstation [... which requires inter-WG agreement on the languages and deployment mechanism)
16:28:12 <mizmo> okay i'm gonna mark davidstrauss' suggestion as an agreed cuz a lot of us seem on board with that
16:28:13 <simo> mitr: RHEL realy doesn't make a difference, decision makers decide
16:28:19 <sgallagh> mizmo: Perhaps development of new roles? Such as "Django Platform".
16:28:20 <simo> whatever the 'product' is
16:28:27 <mizmo> #agreed new persona: decision maker - should have resources + power, perhaps a CTO
16:28:41 <nirik> mitr: yes, but we shouldn't ignore that case... they need to install server software to test, etc.
16:28:45 <mizmo> sgallagh, but this persona would be someone working on fedora server itself right?
16:28:54 <sgallagh> mizmo: Please use #info for anything we didn't vote on.
16:29:02 <mizmo> sgallagh, or would they be making the next great web development platform?
16:29:06 <mizmo> sgallagh, okay how do i undo the agreed?
16:29:16 <simo> sgallagh: django platform is something I would defer to 'Software colleciton enablement' :)
16:29:19 <mitr> #undo
16:29:19 <zodbot> Removing item from minutes: <MeetBot.items.Agreed object at 0xb985390>
16:29:23 <mizmo> thanks mitr
16:29:28 <mizmo> #info new persona: decision maker - should have resources + power, perhaps a CTO
16:29:49 <sgallagh> simo: Well, "Django Deployment Platform Role"
16:30:10 <mizmo> mitr: agree with your suggestion on clarifying server as not the OS the developer uses to develop on, should be workstation for that
16:30:23 <simo> sgallagh: same comment from me, it's an inherently unstable platform due to very high turn over
16:30:34 <simo> sgallagh: we totally need to think *how* to address it
16:30:43 <sgallagh> simo: So maybe a bad example, but the idea remains
16:30:50 <simo> but I think it should be seen as a general problem with high turnover frameworks of any kind
16:31:13 <mitr> simo: yes.  A general problem for base/env+stacks (to have a general solution
16:31:18 <sgallagh> simo: Django is actually a lot nicer in that regard than others (they have an 18-month warning period on incompatible changes), but that's irrelevant at the moment
16:31:21 <nirik> we probibly don't want personas that are super tied to roles do we? or then we end up with too many?
16:31:23 <mitr> base/env+stacks (not sure which one)
16:31:43 <mizmo> sgallagh, i guess i could see your suggestion being one of three very different things: (1) A Fedora contributor who wants to make a new role or whatnot for Fedora server (2) A developer who wants to make the new hot web platform that's better than django and ruby on rails (3) A devops who wants to deploy a django app on server   (??)
16:31:50 <nirik> ie, database admin, webmaster, dns admin, etc... which could be too narrow I think.
16:31:50 <sgallagh> nirik: I was talking about a role-creation persona
16:31:51 <mizmo> sgallagh, which one is closest to what you meant?
16:31:55 <sgallagh> Not a persona for a role
16:32:35 <sgallagh> mizmo: Probably 1)
16:32:38 <mitr> sgallagh: Please let's not.   I don't want to make it _too_ easy to modify/restrict the product definition because "it will be easier to develop"
16:32:55 <sgallagh> But I'm hearing arguments that I agree with that this should perhaps be a Workstation responsibility.
16:32:58 <mitr> Developer of Server and user of Server should be strictly separate ideas.
16:33:05 <mizmo> sgallagh, okay i think we're on the same page then. it's not one persona per role. it's a persona for one of us :) someone who comes up with an idea for a new role for the Fedora Server
16:33:06 <sgallagh> mitr: I agree
16:33:16 <mizmo> so it is kind of meta
16:33:33 <mizmo> but important too. if it's too hard to make a new role, for example, then the roles won't be made :)
16:33:53 <mizmo> i would make that a secondary persona, maybe call it 'Fedora Server Role Creator' <= fair?
16:33:54 <sgallagh> I guess it's kind of a goal of the Project rather than the Product.
16:34:04 <mizmo> yeh that's true
16:34:57 <sgallagh> mizmo: Let's capture it as a secondary persona for now and then hopefully categorize it properly later.
16:35:09 <mizmo> seems fair to me, anybody have issues with that?
16:35:12 <simo> this disccussion is a little too meta for me
16:35:19 <mizmo> lol i knew someone was gonna say that
16:35:28 <simo> can someone take the task to summarize the points and why they are relevant offline ?
16:35:42 <sgallagh> If we keep this up, we're going to need to start passing out glow-sticks at the door...
16:35:44 <mizmo> #info meta persona for 'Fedora Server Role Creator' - secondary, not for product but for project
16:35:53 <mizmo> simo, i can post it in the meeting summary for today when i blog it
16:36:01 <sgallagh> mizmo: Thanks!
16:36:03 <davidstrauss> I don't think it's too meta.
16:36:07 <mizmo> okay so we have two additional personas now
16:36:16 <mizmo> should we have a devops persona too in addition to the developer?
16:36:17 <davidstrauss> It fits closely with my container goals.
16:36:33 <sgallagh> mizmo: I thought that was what your developer persona was, there
16:36:41 <mizmo> oh i also wanna note mitr's suggestion because i dont think i info'ed it,
16:36:44 <davidstrauss> "How does one create a service that plugs nicely into Fedora?"
16:37:01 <mizmo> #info developer persona: be careful to make it clear that server is the deployment platform, not the developer workstation / coding environmnet
16:37:02 <mitr> mizmo: What would be unique about devops?  Ability to deploy and to monitor should already be there
16:37:34 <mizmo> sgallagh, mitr: we could make him devops then, i think i just need more info about how the guy would work, i don't know enough about it
16:37:49 <mizmo> eg does a devops person get paged at all hours when there's an outage?
16:38:18 * nirik thinks devops is a buzzword much like cloud. ;)
16:38:26 <sgallagh> mizmo: Probably a good idea for the two of us to sit down with rbergeron and hash that out
16:38:38 <mizmo> nirik, agreed :) the buzz doesn't help!
16:38:43 <mitr> I'd slightly prefer not restricting ourselves to devops (multiple platforms, and speed to deploy are equally important when deploying a 10-year old J2EE application), but I don't feel strongly about it
16:39:25 <mizmo> well i used to work on a big java web app, the development team was really separate from the ops team, not the same people
16:39:26 <sgallagh> The primary difference with DevOps is the concept of "Mean time to recovery" is more important than "Uptime"
16:39:30 <mitr> devops kind of captures nicely "the Ruby problem" (sorry, I'm unfairly picking on a single language for simplicity)
16:39:43 <mizmo> if there was an outage it was ops responsibility, not the devs
16:39:58 <simo> mizmo: devops combines the 2 figures in one
16:40:03 <mizmo> mitr, what's the ruby problem?
16:40:06 <sgallagh> Right, no more wall.
16:40:19 <simo> sgallagh: I say it is just a dev'startup'ops thing
16:40:21 <mizmo> simo, does devops necessarily mean a smaller org? cuz i dont know how you'd have them combined into one role for a huge app like the one i'm thinking of
16:40:23 <sgallagh> mizmo: The ruby problem is mainly "the bundling problem" rehashed
16:40:33 <mizmo> sgallagh, oh okay
16:40:37 <sgallagh> Except with a community that has no stake in backwards-compatibility (in the general case)
16:40:38 <mitr> what sgallagh said, + the culture that "that's how things are done"
16:40:54 <simo> mizmo: well devops people would say no, but in reality I see real devops only in startups :)
16:41:16 <mizmo> so with ruby you need devops, because the platform isn't backwards compatible, so you need a dev to port things forward if something goes wrong bc of a bug in the platform or whatnot
16:41:43 <sgallagh> mizmo: Precisely
16:41:47 <simo> mizmo: it is not just ruby
16:41:59 <sgallagh> Right, ruby is just a representative example
16:42:08 <sgallagh> This is true of a lot of Java shops too
16:42:11 <simo> it is also: oh cloud provider X we use is going to change API in 10 days
16:42:24 <sgallagh> And Google APIs...
16:42:25 <simo> need to develop a different connector for the production servers asap
16:42:37 <mizmo> ah i see
16:42:45 <simo> language really doesn;t matter
16:42:45 <sgallagh> So it's really about resiliency not just of the service but also the deployment itself
16:43:03 <simo> it's about people using *new* platforms that keep changing and shifting all the time
16:43:15 <simo> and they use so much of *everything* it is all in costant motion
16:43:21 <sgallagh> Well, it's also about a perception change.
16:43:22 <mizmo> so basically if we have both a dev and an ops persona we'll probably not cause any issues for a devops person since they do both types of things, but maybe we make sure that one of the personas is using a platform less dev-oppy and one that is
16:43:34 <sgallagh> That absolute stability is not as important as rapid recovery.
16:43:49 <mizmo> eg so we keep the developer persona on ruby on rails so he's working with a platform that represents less stability in terms of backwards compat, etc.
16:43:52 <davidstrauss> It's part of the perception change I'm referring to in the migrations vs. support duration
16:44:01 <simo> mizmo: it is a lot about automation
16:44:06 <mizmo> and then we have one of the admin personas primarily working with a platform that's more stable maybe?
16:44:10 <simo> mizmo: becasue devops will redeploy a lot
16:44:21 <simo> a lot more than traditional ops used to
16:44:28 <davidstrauss> How easy is it to keep the app on a supported foundation? *versus* How long can the foundation we've deployed to remain supported?
16:44:28 <nirik> "thing x is hard and we only do it once a year" "so, lets do it 10x/day now and fix all the problems so that can keep going"
16:44:50 <sgallagh> Right, the devops folks pretty much live in the "livestock" side of things, except possibly for their hypervisors.
16:45:11 <davidstrauss> Hypervisors and container hosts.
16:45:22 <sgallagh> Right
16:45:26 <simo> sgallagh: they concentrate on the changing pieces but abslutely rely on the stable infra as well
16:45:27 * mclasen spreads the rumor that sgallagh calls devops cattle
16:45:36 <sgallagh> mclasen: moo
16:45:37 <simo> DNS ? DHCP ? Hypervisor ? etc ...
16:45:59 <mizmo> so devops people are more automated, the traditional people are more omg don't break it?
16:46:00 <davidstrauss> I rather like the livestock metaphor, aside from how the vegans at my office will take it.
16:46:03 <sgallagh> simo: Sure, some of the critical infrastructure will likely remain on stable platforms.
16:46:12 <davidstrauss> Maybe farming versus a bonsai tree?
16:46:16 <sgallagh> mizmo: That's a pretty fair analogy
16:46:23 <simo> davidstrauss: I like more the bees and hive analogy
16:46:34 <simo> livestock can't be replaced as fast in real life
16:47:05 <mizmo> the extreme traditional example would be maybe nasa where the platform has to be stable enough to float in space for many years
16:47:12 <mizmo> no devopping a satellite or shuttle
16:47:15 <simo> davidstrauss: farming versus bonsai works too :)
16:47:30 <sgallagh> mizmo: Actually, NASA is doing a lot of DevOps type stuff now
16:47:41 <sgallagh> But the communications channels are certainly stable as hell
16:47:45 <davidstrauss> In any case, I think making devops a key part of our thoughts is healthy.
16:48:00 <sgallagh> davidstrauss: +1
16:48:01 <mizmo> yeh this discussion has really helped me understand the difference here, thank you very much for that
16:48:02 <simo> mizmo: well that kind of stuff also uses a lot of very custom development, so I am not sure we really care for that extreme
16:48:32 <mizmo> i think ill make the developer persona a bit more dev-oppy to represent that and potentially talk to robyn for some guidance on that
16:48:35 <mizmo> does that seem fair?
16:48:39 <davidstrauss> When it comes to server and cloud usage, I want Fedora to leapfrog the current popular choices philosophically.
16:48:48 <sgallagh> mizmo: Sounds good to me
16:49:03 <mizmo> #info modify developer persona to be a bit more devoppy
16:49:04 <nirik> I think we can add a lot of value in standardizing/best practices...
16:49:04 <sgallagh> davidstrauss: Sorry, that was a bit over my head.
16:49:11 <davidstrauss> Our competition is, in some ways, CoreOS more than Ubuntu.
16:49:40 <sgallagh> Let's also not get in the Cloud group's way too much either
16:49:52 <sgallagh> I think we probably want to be focusing on the infrastructure to support their product.
16:50:04 <sgallagh> Fedora Cloud is likely to be Fedora's answer to CoreOS.
16:50:31 <mizmo> did anybody have a chance to read through some of the stuff on the developer persona? i kind of really just made it up but i dont know if i was getting at the wrong thing
16:50:37 <mizmo> eg under frustrations "Spending cycles porting code to an endless array of platforms - it's time-consuming and uninteresting work. "
16:50:45 <mizmo> but i guess a devop wouldn't do that?
16:51:06 <mitr> I'd generally prefer if a Cloud application and Server application were bit-for-bit the same (but it hasn't been discussed yet)
16:51:23 <davidstrauss> mitr++
16:51:24 <mitr> mizmo: Java/Python/Ruby eliminate >50% of this problem (but not all)
16:51:26 <sgallagh> mizmo: No, devops would pick a minimalist platform and just VM/Containerize it
16:51:37 <mitr> (note: not necessarily Server Role)
16:51:57 <mizmo> do we care about devs/admins who would need to do porting work like that on a server? do such people exist?
16:52:01 <sgallagh> mitr: I agree as well
16:52:46 <sgallagh> mizmo: Traditional app developers will need to be served, yes.
16:52:48 <mizmo> i was thinking about an app like, i don't know, mediatomb, where there's arm builds for NAS devices and linux builds and maybe bsd builds
16:52:55 <sgallagh> The Oracle DBs and SAPs of the world, for example.
16:52:55 <simo> sgallagh: tbh I see less and less separation between server and cloud product
16:53:14 <simo> I wonder if they really will be that different if we keep pushing hypervisor stuff that way
16:53:17 <mizmo> sgallagh, okay so i think we need an additional traditional server app dev persona as well
16:53:23 <Evolution> there's a large amount of crossover in the two
16:53:32 <simo> I walways thought cloud os would care about guests more than bare metal
16:53:36 <mizmo> #info potentially another: traditional server app developer persona
16:54:08 <Evolution> simo: but the guests are often-times running the server bits we're interested in.
16:54:21 <simo> Evolution: right
16:54:25 <mizmo> so are these personas gonna care about the server or will they primarily care about the cloud product?
16:54:25 <sgallagh> simo: The "roles" are where I see the primary difference.
16:54:35 <simo> Evolution: so I wonder if it really make sense to have separate stuff in the end
16:54:47 <mizmo> maybe an app dev isn't a great primary persona since they're using cloud product as their platform
16:54:50 <simo> mizmo: both :)
16:54:56 <davidstrauss> We deploy Fedora to bare metal as a container host. Once companies like GitHub, Etsy, etc. grow to a certain point, they often substantially leave "cloud" for bare metal.
16:54:58 <sgallagh> Cloud I expect to be focused on delivering as little as possible so app writers don't have a lot of stuff potentially conflicting from the base system
16:55:05 <simo> they need infrastructure to support their cloud instances
16:55:29 <sgallagh> Whereas our position is to deliver sturdy infrastructure both for traditional roles and to support Cloud deployments
16:55:45 <davidstrauss> And once they leave cloud, they either deploy to hypervisors on their own hardware or something like the "guest images" to bare metal directly.
16:55:49 <simo> davidstrauss: true, many company have mixed env, with in house clouds + overflow on public clouds after some time
16:55:50 <mitr> (I have a hard stop in 5 minutes)
16:55:51 <mizmo> so we have folks who are 'graduating' from cloud (the github, etsy growth case)
16:56:04 <mizmo> and we have folks who are doing more traditional stuff (examples?)
16:56:05 <simo> once you are big enough, managing your own cloud is cheaper than buying wholesale from amazon
16:56:28 <mitr> ... but it still looks like a cloud, quacks like a cloud, only isn't paid for like a cloud
16:56:30 <nirik> simo: not to mention the 'I control my own data' argument, but thats drifting off topic. ;)
16:56:42 <simo> nirik: yeah that will come later honestly
16:56:47 <davidstrauss> simo: Some companies opt to run their own clouds. Others move to PXE deployments and similar.
16:56:50 <simo> and unevenly across geographies imo
16:56:53 <simo> but diff. arg.
16:56:57 <simo> lets close the meeting
16:56:57 <mizmo> before we end can someone give me a solid canonical example for traditional server deployment?
16:57:08 <davidstrauss> The line's getting blurred even more with OpenStack bare-metal deployment support.
16:57:09 <sgallagh> mizmo: Domain controller
16:57:12 <sgallagh> DNS server
16:57:16 <mizmo> okay thanks!
16:57:17 <mitr> Email infrastructure
16:57:22 <simo> mizmo: File server
16:57:32 <sgallagh> Storage cluster
16:57:32 <davidstrauss> Office DNS and web proxy
16:57:35 <mizmo> i'll modify the personas based on the discussion today, and mail the list when they're updated, if you folks could review i'd really appreciate it
16:57:44 <simo> mizmo: all those services that are set in stone and won't change in the short term
16:57:53 <mizmo> any other additional persona ideas please post to the list too
16:57:59 <sgallagh> Will do
16:58:33 <nirik> we probibly should try and finish personas in not too long...
16:58:47 <nirik> given our short timeframe to come up with deliverables/document
16:59:03 * sgallagh nods
16:59:15 <mizmo> nirik, i think we're pretty close
16:59:35 <mizmo> i didnt see a whole lot of contention here about the ones we have so fair, just some refinements / expansion
16:59:36 <nirik> yeah, hopefully
16:59:41 <sgallagh> Shall we aim to have it ready for a vote at next week's meeting?
16:59:52 <mizmo> i think that's going to be difficult for me given the holiday
16:59:57 <sgallagh> Ah, good point
17:00:05 <sgallagh> The 10th, then?
17:00:24 <mizmo> sounds good, i think i can probably have a good draft by next fri to give time for folks to read before meeting
17:01:01 <sgallagh> That would be fantastic.
17:01:30 <simo> +1
17:01:37 <nirik> cool.
17:01:38 <sgallagh> I'll offer to do some early draft reviews for you as well
17:01:51 <sgallagh> If you can deal with my pedantry :)
17:02:18 <mizmo> sure that'd be great
17:02:22 <mizmo> #endmeeting