Web   ·   Wiki   ·   Activities   ·   Blog   ·   Lists   ·   Chat   ·   Meeting   ·   Bugs   ·   Git   ·   Translate   ·   Archive   ·   People   ·   Donate

#sugar-meeting, 2010-03-19

Index | Today     Channels | Search | Join

All times shown according to UTC.

Time Nick Message
18:39 mchua #chair sdziallas
18:39 I'm not sure the #chair command works here, but okay.
18:39 sdziallas: Let's finish the picking of Activities so we can build the iso and explain to folks what's going on.
18:40 sdziallas mchua: OH NOES I CAN HAZ NO CHAIR ;)
18:40 mchua: yeah, right.
18:41 alright, so. we've browse on our list and were discussing which other activities we wanted to add.
18:42 mchua kicks gobby to connect
18:42 Aha, there we go.
18:42 sdziallas: Yep. What's next after Browse?
18:43 sdziallas mchua: well... we were Physics / TurtleArt / Record / Write, last time.
18:44 (the latter ones are half-ish broken, though Record... *could* work)
18:45 #link https://fedoraproject.org/wiki/Sugar_Activities
18:45 (for reference)
18:45 mchua looking at the list
18:47 TurtleArt is exceptionall well-maintained.
18:47 er, exceptoinally.
18:47 AGH
18:47 sdziallas nods.
18:47 mchua very.
18:47 sdziallas LOL!
18:47 :)
18:47 yeah, so let's add that one.
18:47 mchua Physics also seems pretty on top of things.
18:48 I wonder if it's worth asking Activity maintainers to submit their things for consideration for SoaS, so we know they actually want to be on there?
18:48 sdziallas well, it bundles a good chunk of stuff, but since it's already packaged and certainly popular, we should go for it.
18:48 mchua (Maybe after this initial pick, though.)
18:48 nods.
18:48 sdziallas I suppose we won't hear from many folks then.
18:48 nods.
18:48 mchua So TurtleArt is in, Physics is in?
18:48 sdziallas yes.
18:49 mchua Okay. Now, Record... I'm a bit torn. It's a great Activity, but we can't... expect it'll work well on a lot of computers with its current state.
18:49 satellit__ chat?
18:49 mchua Plus a lot of computers don't even have webcams.
18:49 sdziallas yes, but...
18:49 mchua: so it works here for me with my webcam and I know that Fedora has continuously these "better webcam support" features.
18:49 mchua: but there *are* things that may randomly break Record.
18:50 (and it's not really maintained, either)
18:50 mchua How actively is it being maintained?
18:50 Ah.
18:50 sdziallas So alsroot has been fixing stuff.
18:50 mchua pulls up gitorious in the background, as well as aslo.
18:50 sdziallas But he says that the codebase is so complicated that a rewrite would make more sense.
18:51 mchua The last patch was in June 2009.
18:51 sdziallas Heh.
18:51 mchua Before then, it was April 2009.
18:51 sdziallas That's comparatively young... :p
18:51 semi-joking.
18:51 mchua It's... yeah, I can believe alsroot's statements.
18:52 So here's my question for both Record and Write, and for other things that may be borderline like this:
18:52 do we lose anything by adding these to the "you can install from ASLO!" encouragement list?
18:52 alsroot is hacking Record right now, to fix http://bugs.sugarlabs.org/ticket/1847 and http://bugs.sugarlabs.org/ticket/1848
18:53 mchua alsroot, what do you think? You've got a better handle on the state of Record code than we do.
18:53 I think that putting it as a "get from ASLO" does a few things.
18:53 (1) it removes the "ZOMG PACKAGE" extra step, allowing us to have less work and to be more agile about updating the code (maybe) because it only has to get updated in one place.
18:54 (2) it sets a different level of expectation than something that "ships by default" - try this thing out, it may or may not work with your webcam.
18:54 if it does, that's extra bonus cool. If it doesn't, you didn't necessarily expect it to, you were just trying that out.
18:55 sdziallas (side-noting that most of the activities on these lists are packaged, some aren't, but there are more we haven't even looked at. anyhow, generally +1.)
18:55 alsroot is wondering when soas switched from getting-from-aslo-as-is to repackage-all-activities
18:55 mchua: the thing is, we need ASLO in anycase
18:56 sdziallas alsroot: right now. but it's "package-a-small-number-of-activities" - or at least we're turning it that way :)
18:57 mchua alsroot: Well, it went from "let's stick .xo files in" for the first 2 versions to "we're a Fedora Spin for build infrastructure maintenance sanity" for this upcoming release.
18:58 alsroot: with the Fedora Spin stuff comes the "everything must be packaged" criteria - which is a little extra work, but has the benefit of making sure licenses, etc. are cleaned up, libraries split out, and such.
18:59 alsroot: But the number of Activities we shipped by default are way too many to turn into good packages, and a good portion of those Activities don't work anyway (lack of maintenance, etc) so we're paring down to a small number that *do* work, that will be packaged (or are already packaged) and maintained, and tested and (informally) supported
18:59 alsroot: and making really, really good "hey, there's this thing called ASLO... why not try it out?" resources for a bunch more.
18:59 I think if part of the out-of-the-box experience /is/ exploring ASLO, that's not a bad thing at all.
18:59 sdziallas: Feel free to jump in and correct me. ;)
19:00 sdziallas mchua: I'm totally +1 to that.
19:00 (works very well with our "explore" principle, too :)
19:01 mchua yup.
19:01 alsroot mchua: well, I guess answer is pretty common, just post activity bugs to bugs.sl.o, can't see better way
19:01 mchua alsroot: Yeah, but if there's no active maintainer (Browse is an example of this) the bugs don't get fixed. It's... a problem we need to fix.
19:02 alsroot: Basically, as the de facto maintainer of Record, what would you think about Record not shipping by default in SoaS, but instead being one of the "here's ASLO, try downloading this to see if it works" Activities?
19:03 or to rephrase - "any objections?" ;)
19:04 alsroot mchua: for me, ASLO is the main deployment scheme, I'm tracking what is current verison on ASLO etc. -- at the end I don't see (for me as current Record maintainer) any differences if Record will be shiped w/ soas or not
19:04 mchua alsroot: Ok, then I think that makes the decision easy. :)
19:05 sdziallas: I propose Record and Write be on our "install from ASLO!" list.
19:05 sdziallas mchua: granted, approved, noted! could you #action this?
19:05 mchua sdziallas: In fact, I propose that we just take a look at the remainder of the package list and simply list the ones that are up-to-date in well-maintained fedora packages and pick the shortlist from there.
19:05 #action Record and Write are on the "install from ASLO" documentation-ninjahood list.
19:05 sdziallas (for now, heh.)
19:05 alsroot anyway, due to nature of ASLO, public ASLO activities should work on all declared SPs
19:06 sdziallas (leaving out write is pretty hefty, but since it's not really working... heh:)
19:06 mchua sdziallas: Write is *so* broken right now.
19:06 sdziallas: I might be able to pump QA resources into checking it out and trying to push for dev resources to fix it up in time for GA, but... that's a slim shot.
19:06 sdziallas mchua: right. I found Mel's bugs and... it's so sad, yeah.
19:06 mchua: one step... ;)
19:07 mchua sdziallas: $5
19:07 alsroot: *nod*
19:07 sdziallas mchua: thx! :)
19:07 mchua sdziallas: okay, so on the remaining list on https://fedoraproject.org/wiki/Sugar_Activities are there any more Activities we know are well-maintained both codewise and as Sugar packages?
19:07 er, RPMs?
19:08 er, packages in Fedora?
19:08 sometimes does not actually speak English.
19:08 sdziallas: I'll note that we can also do a quick first pass here, and then explain stuff to the various lists with criteria for inclusion and how the few Activities that made it on there fit those, and then let people remind us of other well-maintained ones over the weekend.
19:08 Tuesday is Beta freeze, which is soon, but still a few days out.
19:09 So there is a little time for a few people to respond.
19:09 sdziallas possibly etoys? (pippy & read are broken, too) terminal? visualmatch?
19:09 mchua And I'd like a working minimal .ks asap.
19:09 sdziallas grins sadly.
19:09 mchua sdziallas: Do you know they work well and are maintained well, or are these things we should check into?
19:09 sdziallas looking at this list under that assumption shows just more drastically how broken all this is.
19:09 mchua sdziallas: Well, knowing the bugs is the first step to fixing 'em.
19:10 sdziallas mchua: etoys is taken care of by some RH guy + OLPC folks, so that should be alright.
19:10 mchua: no, I mean "being actively maintained"
19:10 mchua: I wouldn't classify many activities as that.
19:10 mchua: terminal is just... terminal.
19:10 mchua so there's "the code is actively being developed by an upstream with a responsive maintainer"
19:10 sdziallas mchua: and visualmatch is maintained by walter, saying all ;)
19:10 mchua and "the software is packaged in Fedora by a responsive maintainer"
19:11 sdziallas mchua: heh.
19:11 mchua sdziallas: you can do the Fedora package maintenance for a few, but not too many.
19:11 sdziallas: or... well, maybe you can, but you shouldn't.
19:11 sdziallas mchua: well, we need activities to be packaged in a day or two.
19:11 mchua sdziallas: so we can take a *few* Activities with really good upstreams that don't package, and etoys seems like a logical fit for that.
19:11 sdziallas: if you don't mind the packaging workload, I mean.
19:12 sdziallas mchua: I don't see many maintainers doing that responsively.
19:12 mchua Just like I can tackle Browse's general brokenness with a higher QA/support workload on my part in order to allow us to include it - but I can't do that for more than, perhaps, one other Activity. So the rest of the Activities need to have working code, for me.
19:12 sdziallas mchua: I think we can just jump on the etoys train, that should work (i.e. not much work for me, actually, since people package that alrady)
19:12 nods.
19:12 ...and goes adding etoys.
19:12 mchua sdziallas: Oh, okay. so Etoys definitely has an active dev/user community, the code's up to date, it's easy to reach them... and Someone Else is packaging it responsibly.
19:12 Ideal case situation.
19:12 Yay!
19:13 Pippy and Read are out, if they're broken.
19:13 VisualMatch...
19:13 looks at git
19:14 I know the CFS deployment has had successful test runs with it - Lynne May has played with it a lot and hasn't been able to break it yet.
19:14 Last updated 2/23, so fairly recent.
19:15 sdziallas Heh, cool. so it's in :)
19:15 mchua adds one more check - git, aslo, and trac
19:15 looks at bugs for visualmatch
19:16 huh, we're developing a way for quickly assessing activity health.
19:16 sdziallas nods. SOP time, heh? :p
19:16 mchua Yeah, eventually. :)
19:16 I'm going to be lazy and just leave the logs up and the next time this has to be done will probably write a SOP and get someone else to do it.
19:16 I think.
19:17 So, visualmatch not doing so well on the trac front - only one bug ever reported (by you, actually) and never closed, 3 months ago.
19:17 http://bugs.sugarlabs.org/ticket/1637
19:17 That having been said, it's not a bad bug.
19:17 And I think that's about on par for bug reporting for Activities, sadly.
19:17 Most of them don't have an active Trac workflow, is the impression that I get.
19:18 I'm happy with visualmatch being in.
19:18 sdziallas oups, I should have taken care of this. (though not many folks will use that way I describe there anyway.)
19:18 okay.
19:18 mchua sdziallas: let's talk about terminal, and then I want to look at our list of 5 (or 6) and do a quick sanity check.
19:19 sdziallas mchua: I'd like to run quickly through the list of non-included activities then to make sure we don't miss anything, yeah.
19:19 but terminal first.
19:19 mchua Ok - pulling up aslo, gitorious, trac...
19:20 http://activities.sugarlabs.or[…]/sugar/addon/4043
19:20 last update 1/16
19:20 http://git.sugarlabs.org/projects/terminal
19:20 sdziallas ah, cool.
19:20 mchua wade and sayamindu - sparse fixes, but looks like good ones.
19:21 http://bugs.sugarlabs.org/quer[…]ity&col=milestone
19:21 sdziallas oups!
19:22 mchua Not bad, actually!
19:22 I'd want to give it a quick smoke test run to make sure it's ok and that we can't casually break it.
19:22 sdziallas: unless you've used it recently and feel good about it.
19:22 sdziallas mchua: I haven't had issues lately.
19:23 (but didn't look too deeply)
19:23 mchua sdziallas: Would you feel comfortable about it? or do you want me to take a hammer to it this weekend and we can make the final call tomorrow or sunday after that?
19:24 sdziallas mchua: I'd be alright with adding it... temporarily.
19:24 mchua: so if it fails, we'll take it out. but for now, it could be good.
19:25 mchua Okay, i'll put it on my todo list to hit that hard this weekend.
19:25 #action mchua test the living daylights out of Terminal
19:25 and we'll see how it works.
19:25 sdziallas grins.
19:25 mchua sdziallas: Ok, so stepping back from the 6 activities we have so far... how many of those are going to be a nontrivial packaging workload on you? (i.e. how many are you going to have to maintain, and of those, are there any that are nontrivial?)
19:26 From my side, Browse may be a potential resource suck in terms of QA and support.
19:26 Terminal is an unknown quantity, I'm prepared for that to be my second (and last) QA/support heavy Activity, but don't expect it to be.
19:26 sdziallas mchua: etoys is a nobrainer, I don't need to do anything for that.
19:27 mchua The rest of the things - etoys, turtleart, visualmatch, physics - have other people taking very good care of them, and I can point folks in their direction, so I'm good there.
19:27 sdziallas browse can become pretty hefty if xulrunner breaks, otherwise it should be good.
19:27 mchua: ah, that's cool :)
19:28 mchua: the other five activities... can be kept up2date with a reasonable amount of work, I guess.
19:29 mchua So wait, you're packaging all of them?
19:29 sdziallas: (one sec, let me shovel a piece of pizza into my mouth)
19:30 sdziallas mchua: they are packaged, so I just need to keep them up2date.
19:30 :)
19:37 mchua back,  nmo
19:37 er, nom
19:37 sdziallas: Okay, so you've pretty much just got browse too.
19:38 sdziallas mchua: heh, yeah, browse is okay.
19:38 mchua sdziallas: So you feel like you have bandwidth to maintain a little more if needed?
19:38 sdziallas mchua: yes.
19:38 mchua doesn't want to push this too far - we should feel like we're shipping with far too *little* for us to do
19:38 okay.
19:38 sdziallas: you drive - we'll go through the rest of the list now.
19:38 sdziallas mchua: so I could still take another few activities.
19:38 mchua: since this is - compared to before - much better.
19:39 the James Simmons' ebook activities might make sense. Or Speak. Or...?
19:39 mchua Okay. Remember some of that bandwidth we need to use for things like "nagging inactive maintainers" and "finding contributors to help" and "documenting lots" and all.
19:39 sdziallas IRC, heh? let me look at our list again.
19:40 mchua sdziallas: Basically, what are the activities that it would pretty much break your heart not to include? Then we can pass those through a "does it work?" basic sanity check, and then look at the maintenance and packing state.
19:40 er, packaging.
19:40 sdziallas mchua: uh, before I forget, we should probably add another package (csound... and stuff, since otherwise people will complain when they use activities from a.sl.o and sound doesn't work)
19:40 mchua: infoslicer?
19:41 log?
19:41 dunno how popular chat is.
19:41 maze? *grin*
19:42 mchua Ohhhh, Maze.
19:42 sdziallas (playgo, but that's a personal one, so I'll grab that myself form a.sl.o)
19:42 mchua laughs
19:42 sdziallas (tamtam, because I wasted hours on packaging that sucker)
19:42 so the ebook stuff might... not make sense anyway, since I don't know how much read works.
19:43 mchua I've been meaning to get my friend Mark to teach me to play more Go.
19:43 sdziallas I've had issues with it when I tried, but I'm not sure whether it's entirely broken.
19:43 mchua But yeah, ASLO.
19:43 sdziallas We could use the PlayGo activity at some point... :)
19:43 mchua has never really gotten into using Read, so is probably biased here.
19:43 sdziallas heh. nope, it's alright.
19:43 (just sayin' that we don't need to whine about an ebook activity if we can't read them out of the box)
19:43 mchua nods.
19:44 That would also mean a lack of "ZOMG PANIC MUST FIND EBOOKS" rushing for this release.
19:44 Which would be lovely to avoid.
19:44 sdziallas LOL, yeah ;)
19:44 looks at list again.
19:44 tries to put them all in one line
19:44 satellit_ presence service?
19:45 sdziallas mchua: speak. IRC. infoslicer. log. chat (?). maze. tamtam.
19:45 mchua: for example log might make sense, since otherwise it'd be just hard to grab the logs.
19:46 mchua sdziallas: I like tamtam - though I'm not sure how active the upstream is. I like Maze - I'd like to do a quick sanity check on upstream health, but it seemed solid when I tried it (informally). Same with Speak, perhaps with a bit more testing. And the Help activity may be worth doing some legwork to update and get in there. Log is also a good one to have around.
19:46 will be shorter
19:46 sdziallas mchua: help is broken.
19:46 mchua sdziallas: speak, log, tamtam, maze, maybe-help-if-it-can-be-hamme​red-into-working-but-not-now.
19:46 sdziallas mchua: because it only works with 0.84, heh. :/
19:47 mchua: okay. no chance for infoslicer and irc? the latter upstream is... nice, I hear ;)
19:47 mchua Hm. it's an Activity worth maintaining, though, and one of the first I'd probably want to call for help with.
19:47 But that's later.
19:47 sdziallas yeah... $5 and stuff.
19:47 mchua sdziallas: Oh, IRC is poorly maintained by a total slacker who won't be updating it for this release due to lack of time.
19:47 sdziallas mchua: but I hear it works, no? ;)
19:47 mchua So if you really want to package it and ship it, all right, but don't expect a lot of dev help available from her. :)
19:48 sdziallas: It does, on a basic level. I'd want to add that to my smoke test list for the weekend, actually.
19:48 sdziallas (it's already packaged)
19:48 okay, so I'll just add that. if you smoke test stuff, that's... good :)
19:48 walterbender sdziallas: infoslicer? I had not seen a ticket for that one...
19:48 mchua sdziallas: Well, also - I'm getting a bit nervous about the size of this list.
19:48 sdziallas: some of these can be "please install via ASLO" things as well.
19:48 sdziallas walterbender: no no, no tickets... just whether we want to include it ;)
19:49 mchua I'd argue that if anyting goes in, it should be (1) maze, for quick demo happiness, (2) log, for debugging, and (3) IRC, for quickly getting help online.
19:49 Maybe TamTam as a 4th.
19:49 Because it demos very, very well.
19:49 walterbender sdziallas: I need to look at the backlog :)
19:49 mchua If we can smoketest and verify these all seem sane.
19:49 sdziallas: That brings us up to 10, and I would cap there, and put the rest as "hey ASLO is awesome."
19:49 sdziallas mchua: I like infoslicer, too, but... let me see.
19:50 mchua too, but... ASLO.
19:50 sdziallas mchua: 11, even.
19:50 mchua sdziallas: Everything we add brings more maintenance load.
19:50 11?
19:50 sdziallas walterbender: we could brief you ;)
19:50 mchua counts
19:50 sdziallas mchua: (gobby)
19:50 mchua: oh, no speak? yeah, we could kick that.
19:50 walterbender sdziallas: if you need some more "maintained" activities, there is alway: TurtleArt, VisualMatch, CardSort, Ruler, Sliderule...
19:51 sdziallas mchua: tamtam also gives you four activities in one (which I need to fix first)
19:51 mchua sdziallas: oh yeah. 11. I can count.
19:51 walterbender mchua: that is already 25% of the way to 20 activities :)
19:51 sdziallas walterbender: the first two of them are packaged, the other one are... waiting for me :)
19:51 mchua sdziallas: I'm nervous about this list, it's getting big
19:51 walterbender sdziallas: and anything Gray Martin in involved in is well maintained...
19:52 sdziallas: and unamindu and jim simmons are pretty diligent with theirs too
19:52 mchua sdziallas: I know the stuff in here is good, but for most of them I'm not *quite* sure, just 90% confident.
19:52 walterbender sdziallas: and alsroot is a machine :)
19:52 sdziallas mchua: mhm ;)
19:52 mchua And we can have strong pushes towards ASLO.
19:52 Actually, we should have a strong list of recommended ASLO programs
19:52 sdziallas walterbender: hehe, yeah - thanks for the help!
19:52 mchua: yes! you're right.
19:52 mchua that we know with good confidence are kickass and will work
19:52 to give people the "oh wait ASLO has good stuff, there is more quality to discover!" feeling.
19:53 sdziallas mchua: so, from the gobby list, do you want to kick more?
19:53 walterbender mchua: if you know with confidence that they are kick ass and will work, why not include them in the first place?
19:53 mchua walterbender: Because it's not confidence, it's "reasonably high certainty but I'm not entirely sure because we haven't thoroughly tested them."
19:53 And I can handle that for 2 or 3 Activities that ship.
19:53 I can't do it for 12.
19:54 walterbender mchua: you need to share the load...
19:54 mchua: of course you cannot do it yourself!!!
19:54 sdziallas walterbender: well, the same goes for packaging... and I haven't been able to share that load, yet.
19:55 walterbender: (meaning that I find myself doing a lot - probably too much - of everything right now)
19:55 walterbender sdziallas: that is a horse of a different color
19:56 sdziallas walterbender: the relationship lies close there, because if an activity is dead upstream, I can't do much as a packager.
19:56 walterbender sdziallas: I bet there are some of Jeff Elkner's students who could help you... we just need to ask them...
19:56 mchua walterbender: when have you known me to try to do something by myself? ;)
19:57 walterbender sdziallas: when you need a stick, just ask me :)
19:57 mchua I'm a lazy bum. :) I also try to underpromise and overdeliver.
19:57 sdziallas walterbender: hehe :p well, the issue with activity packaging is that it needs to be done *immediately* - like a day or two after another release is out.
19:57 mchua: gah. I heard that today already once :)
19:58 mchua sdziallas: So, gobby has my list of 6.
19:58 sdziallas: I think EToys is a *great* place for us to give a strong, strong call-to-action to not just ASLO, but to the upstream EToys community.
19:58 walterbender sdziallas: so give me a list of the activities that are causing you packaging issues and I will use my stick...
19:58 sdziallas mchua: so you mean we should have people pull it form ASLO?
19:58 walterbender: LOL!
19:58 mchua Which we can do more effectively in a "you - go get this awesome thing, and while you're at it check out these places with more resources!"
19:59 sdziallas: Yeah, I would propose that, *because*...
19:59 Etoys is complex and sophisticated enough that including it without instructions can confuse folks.
19:59 sdziallas walterbender: pick randomly one from here: https://fedoraproject.org/wiki/Sugar_Activities (I'm semi-joking)
19:59 mchua But if we set up a strong driving force to the EToys upstream, they get more participants - not just users - and are able to help give the people who take that extra step have a *really* good first etoys experience.
19:59 walterbender mchua: is Etoys not packaged because of questions about the license? It would be a shame not to include it preloaded...
20:00 mchua I want to make the list so slim that it's obvious that *not* being included on the SoaS image isn't a slight.
20:00 And that people *should* get more stuff from ASLO. It should be impossible to actually teach a class with the stuff preloaded on this stick.
20:00 walterbender mchua: I think you need to consider the user experience, not just slighting the developers
20:00 sdziallas mchua: LOL! (better not let anybody quote you to Sean on that)
20:00 mchua The point is that you need to poke around and customize, explore - because we are *not* in a position where we can deliver and support a preloaded experience.
20:01 walterbender mchua: I don't understand your logic
20:01 mchua: hopefully the stick will reach people with no internet access...
20:01 mchua walterbender: Hopefully, yes. Not for this release, though.
20:01 walterbender mchua: it is not production yet, but you are raising the user experience bar...
20:01 mchua walterbender: Um... I should probably back up. How much context do you have for what we're trying to do right now?
20:02 walterbender mchua: I read the meeting notes from this morning...
20:02 sdziallas walterbender: did you fell off your couch? :)
20:02 walterbender I didn't fall off my couch, but I have to say, I am in Sean's camp
20:02 sdziallas (fall, even. apparently I can't speak English tonight, either)
20:03 walterbender I guess I am not tuned in sufficiently to the time tables...
20:03 sdziallas: you are feeling F13 deadline pressures?
20:04 sdziallas walterbender: well, it's not simply the F13 deadline.
20:04 walterbender sdziallas: what else impacts your ability to set your own schedule?
20:05 sdziallas walterbender: I'm afraid that a lot of our core modules are broken and that I'll end up having to maintain 30 activity packages, to release-engineer the kickstart file and to do bug hunting and documentation work.
20:05 walterbender: so I'm alright with the deadlines. because even if we had unlimited time frames, this stuff simply wouldn't get fixed all of a sudden.
20:06 walterbender: I'll get you a bug with a scary example.
20:06 walterbender sdziallas: I guess I am confused... a chicken and egg problem? F13 and Sucrose 0.88 need to settle BEFORE you can make the next SoaS release.
20:06 and there may be packages in F13 that are not sympathetic to some Activities
20:07 sdziallas walterbender: http://bugs.sugarlabs.org/ticket/1767
20:07 walterbender: write is currently broken. and nobody would have noticed, because write has no owner in trac.
20:07 walterbender: (the next SoaS release will happen together with F13)
20:08 walterbender: but I'm happy to step back for a minute and listen :) - might make sense to hear what drives you to the other side's camp and how you see this. :)
20:09 mchua walterbender: yeah, I'm also curious how you see the situation, because we apparently have a different viewpoint here, and maybe there is stuff I'm missing.
20:11 walterbender sdziallas, mchua: I think that we have a bit of a panic behavior going on right now and that we need to slow down and communicate the situation with the community in order to find the help you need.
20:12 sdziallas: Write is broken on the current SoaS builds? But not in general... because I use it all the time with the latest Sugar...
20:12 so it is probably just a matter of getting the correct pieces into place.
20:12 sdziallas walterbender: broken, as in: "the screen is obfuscated, you can't really see what you type"
20:12 (the latest snapshots, yup)
20:12 mhm.
20:13 mchua walterbender: We were actually trying to get a quick draft of the kickstart and the resulting iso out this (EST) morning so we could run it by the various lists and the rest of the community as the start of a conversation.
20:13 walterbender sdziallas: I'll grab a snapshot and take a look...
20:13 mchua walterbender: Initially, I was thinking that the whole "revise .ks, make .iso, email the lists and say "hey, what do you think?" would take... maybe 2 hours, 3 at most."
20:13 walterbender it could be a mismatch with either abiword or metacity...
20:13 but I doubt it is a problem in Sugar itself.
20:13 mchua walterbender: and we were having that convo (logged) in public channels so that folks could see what we were doing.
20:13 walterbender Gary is the maintainer of the Sugar side of the problem.
20:14 mchua walterbender: but we kept getting (very good) questions on why, and stopping to explain what was up, and so it stretched out into... longer than 3 hours.
20:14 walterbender mchua: just because it is public doesn't mean anyone is tuned in :)
20:15 mchua walterbender: Right. Hence the intention to quickly make the kickstart, then the iso, then hit the lists and then listen.
20:15 walterbender mchua: I don't think it is widely known that you guys are in a bind.
20:15 mchua walterbender: Looking back on this morning, we should have said "just wait 2 hours and let us make the kickstart and explain what we're trying out, then we'll talk."
20:15 walterbender: I agree - it isn't, it should be, that is a problem, and we need to be better in the future about communicating those deadlines.
20:16 walterbender mchua: OK. I'll get out of your hair (and try to look into the Write situation tonight)
20:16 mchua walterbender: Doesn't change the beta freeze date being Tuesday though. :)
20:16 walterbender needs to make dinner...
20:16 mchua walterbender: Well - before you go, this is something I /am/ concerned about... do you know if we're going to be actively screwing anyone over if this SoaS version ships with very few Activities
20:17 (even if there's a big push for a lot of supporting documentation that's kid-and-teacher-tested)?
20:17 walterbender mchua: not screwing anyone over, but limiting the amount of testing and feedback you'll get
20:17 mchua Do you know of any deployments who actively expect this next version to have a lot of activities, and can't for some reason use the earlier versions, or aslo?
20:17 walterbender: Okay. That's what I thought, and what I hoped, and that's all right.
20:17 walterbender mchua: I don't think this is a deployment issue.
20:18 mchua We'll get testing and feedback on a smaller number of Activities, most likely, but we can drive the testing/feedback process far better for those few.
20:18 walterbender mchua: deployments for the most part aren't using SoaS...
20:18 mchua walterbender: (I don't want to keep you from cooking dinner - I'll be around for a while if you want to come back later.)
20:19 walterbender: Yep, deployments aren't using SoaS. We would like them to eventually. The purpose of SoaS is to make a product for deployment, and each release should get us closer to that.
20:19 Ideally, we'd have a large number of high-quality releases with actively engaged upstreams and resources that "pull" users into contact with those upstreams.
20:19 er, high-quality activities, I mean.
20:20 But we don't have that yet.
20:20 And my premise is that:
20:20 walterbender but SoaS is a great way to introduce the idea to more people and to expect that after all the pain of making a stick and figuring out how to boot and then finding the wifi drivers because they aren't in F13 and etc to then have to go to ASLO to get activities enough to get the idea is asking too much. we already loose too many people at step 1: downloading SoaS...
20:20 mchua: at least half the machines I run SoaS on have no network...
20:20 mchua instaed of having the intermediate step of "a large number of mixed-quality activities with mostly not-so-actively-engaged upstreams and nearly no resources that "pull" resources into contact with those upstreams..."
20:21 the intermediate step of "a small number of high-quality activities with actively engaged (blah blah blah)" is more likely to get us to the ideal point.
20:21 reads up
20:21 walterbender mchua: well, again, I don't know what is driving the timetables... I'd personally opt for slowing down and getting the activities qualified.
20:21 but if that is not possible...
20:21 mchua walterbender: Hm - your case sounds similar to Sean's, a self-contained demo.
20:22 walterbender: The timetables are driven by the Fedora release schedule, which we have to adhere to as a Spin.
20:22 walterbender: (which brings us a lot of infrastructure advantages - the load of maintaining a build system is off our backs, etc)
20:22 walterbender mchua: well, that may be the problem... we perhaps are one release too soon for being a spin...
20:23 or a few hands short...
20:23 it is too much pressure on sdz
20:23 it is a fine goal, but maybe that is where we bend
20:23 mchua walterbender: How about this - it sounds like we really do want two separate things, a demo stick (where 'quantity' is the important thing, and mixed quality is more ok because it'll be a more "try it out" thing, or in the hands of an experienced demoer)
20:24 walterbender: and a "working towards a solid deployment" stick (where 'quality and upstream engagement' is the important thing, and quantity of activities is small because we only have enough resources to help a few cross that high bar on a short timescale)
20:24 walterbender: the second, imo, is the spin
20:24 walterbender: the first could quickly be done later - at any time, really - as a quick remix
20:24 sdziallas: ^^ ?
20:24 sdziallas mchua:I like that idea, actually. it might make sense to promote  the Fedora version as "rock-solid deployment but can haz  help?" thing and to build a different one using that as a  base... just thinking, not sure.
20:25 mchua: so yes, I'd say that could be an option - and would even prevent us from "trying to innovate and stablize at the same time" :)
20:25 mchua walterbender: That way, you and Sean have the demo sticks you need. And we can market that as... well, it'll basically be SoaS with a bunch of stuff added to it, so it's like "SoaS after you've customized it a lot more"
20:26 walterbender: that way also, the additional stuff on the demo stick wouldn't need to be Fedora packages - we could throw in random .xos, change things on a whim, not be tied to the Fedora release schedule for it, etc.
20:26 walterbender mchua: well, if we will do the remix, then it really doesn't matter... make the "spin" be as austere as you want... but don't advertize that as what end users should be exploring...
20:26 sdziallas mchua: exactly.
20:26 walterbender mchua: for that matter, I think we should have two versions of the helper cd available,,, one with selinux=0
20:26 mchua chuckles
20:27 walterbender mchua:  I am serious.
20:27 mchua walterbender: Oh! Why?
20:27 is missing context here
20:34 Well, either way we should go finish the kickstart file and get this convo out to the lists, albeit some hours belatedly.
20:54 (for those watching in the channel, sdziallas and I are just attacking gobby right now and trying to finish *a* kickstart file so we can actually open up the convo)
21:03 satellit__ sdziallas:The Fedora Unity Project is proud to announce the release of new ISO Re-Spins of Fedora 12. from Bernie in soas lists
21:19 bernie satellit__: yeah, I cc'd him
21:19 mchua bernie: you can make remixes too, you know ;)
21:21 bernie mchua:  I'm already cooking too many builds for paraguay  :-)
21:22 mchua The kickstart file is done and building and we'll see what iso we get.
21:24 drafting explanation in http://etherpad.com/soas-experiment
21:24 sdziallas: ^
21:25 sdziallas mchua: awesome! I'll join you in a sec.
21:25 putting the kickstart in git right now
05:05 satellit__ sdziallas: IRC will not start on soas-3-20100319. workaround is to delete /usr/share/sugar/activities/IRC/activity directory, then DL from ASLO..then it works
09:07 sdziallas: I sent e-mail to soas and you  about IRC on new .iso did you see it?
09:59 mchua satellit__: Thanks for the heads-up on the ticket - we're a bit swamped right now, but I'm going to be looking at the IRC Activity this weekend.
09:59 As part of the "wow, it's been a long time since I was a proper maintainer for that thing" remediation.
10:00 satellit__ mchua: a quick fix is not to include it in build (ks) for now till you have time to fix it
10:01 it can be downloaded from ASLO and it works
10:01 I filed ticket
10:01 mchua satellit__: Good to know - and thanks for the ticket!
10:01 satellit__: I'm staggering under the load of $dayjob stuff that I'm behind on right now, but when I clear that, I'll be taking a look at the activity.
10:02 satellit__: So... just hold tight for a few days.
10:02 satellit__ thanks for all of your great work : )
10:02 * lurking to help test  : )
10:03 mchua satellit__: I may take you up on that. ;)
10:03 satellit__ ok
10:36 christianmarcsch hi everyone
14:20 soasUSB-9c2d test of edited irc auto channels on soas-3 USB
15:03 satellit__ sdzaillas: I have a question. I have built a soas-3-20100319.img of a 2 GB USB that has IRC correction and updates plus an updated TA and analyze. 7 applications. I can upload a compressed file to Tgillard. Everything works should I upload it for testing?
15:08 soasUSB-9c2d this is the USB
15:09 on an my EeePC900 netbook
10:06 erikos_ tomeu: do we start the meeting?
10:06 is on fire!
10:08 tomeu hi!
10:08 hurries
10:08 #startmeeting
10:08 heh, guess it cannot say who started it?
10:08 guess mchua or sdziallas
10:09 sdziallas wth?
10:09 #endmeeting
10:09 this isn't me, it seems.
10:09 mchua: hullo! ;)
10:10 mchua whoops.
10:10 #endmeeting
09:51 sdziallas #chair mchua sdziallas
09:51 #topic kickstart history storytime
09:51 just for reference, we're trying to work out how to keep SoaS sustainable.
09:51 so yeah, early days. strawberry.
09:52 we shipped the major fructose activities as .rpm files (you couldn't remove those without root permissions because they lived in /usr/share/sugar/activities) and a chunk of other activities as .xo files.
09:52 mchua Would that have made it not eligible for spinhood?
09:52 the .xos, I mean.
09:52 sdziallas for blueberry, we moved to an .xo file-only approach.
09:53 mchua: ayup, exactly.
09:53 mchua: neither strawberry nor blueberry would have qualified as a spin.
09:53 mchua How did the .xo approach work out?
09:53 Why are we a spin now?
09:53 sdziallas since mirabelle is going to be a Fedora Spin, we're locked down to use .rpms.
09:54 mchua: it didn't work too badly. basically, we hadn't to worry about activities. from a release engineering point of view, it was half-ish crap, though.
09:54 mchua: maybe this is also a lack of... well, here's the thing:
09:54 before we simply pulled the latest activity from a.sl.o (for blueberry).
09:55 that worked or didn't. but if something didn't work, it was the activity maintainers fault (at least with quite some certainty).
09:56 now since we moved to packaging them (basically since Fedora needs a solution anyways and since bundling libraries is Not Such A Good Idea), we're lacking... ressources.
09:56 (oh, and the blueberry release was still worse than the strawberry one, btw)
09:56 mchua Worse in what way?
09:57 sdziallas people got excited about strawberry. there was nothing to get excited about in blueberry. it sucked. the whole release process was screwed (time-wise). I'm still telling the story of me rebuilding the final images frantically in the middle of the night before I took the plane to Toronto.
09:58 I'm having a hard time defining what it was, but yeah...
09:58 mchua Ok, yeah, I remember you doing that. :)
09:59 And being a spin... is a good idea because... packaging is good?
09:59 sdziallas And now I've moved myself into a situation where I've even more to do (expecting that more people jump on board, but that didn't happen) - which means that Mirabelle has a good chance of sucking, too.
09:59 Well, I *need* to do it.
09:59 It's the only way of satisfying both Fedora's and Sugar's needs at the same time.
10:00 And yes, it's a good idea because we've nightly builds (which aren't being thoroughly tested) and the release simply *happens* at one point.
10:00 Maybe I'm just afraid of failing. But blueberry was already fail. Time to learn how to do it better.
10:01 Maybe I should have listened to Peter and folks. Gah.
10:01 What if we moved SoaS out of Fedora again and had simply a Sugar Development Spin?
10:01 One that contained only very few activities but something that people could test stuff with?
10:02 Sugar on a Stick wouldn't be locked down to a release date and we could still ponder with .xo files as we wanted.
10:02 mchua What does being a spin get us?
10:03 What I vaguely recall is that we get to use Fedora's build infrastructure (which is nice, we don't have to maintain that)
10:03 sdziallas (Mel is in question mode, I like that :)
10:03 mchua which means that nightly images magically appear based on the uploaded kickstart file.
10:03 sdziallas right. so nightly builds (which we could setup ourselves)
10:03 yeah. and we wouldn't need to push the *build* button for the final release.
10:03 but I think it'd be actually good to be able to do that.
10:03 mchua Aye, but that's extra infrastructure maintenance and being a sysadmin is not nor should it be part of our core capacity.
10:03 It'd be good?
10:03 How?
10:04 sdziallas We wouldn't release crap just no matter what.
10:04 And people could still test the nightly builds.
10:04 (if SoaS wasn't a spin, but "Sugar Something" was)
10:04 but that would mean two solutions.
10:04 I was originally pretty happy to have achieved the situation I'm currently in.
10:05 Because I felt I'd be doing a good chunk of work more for some time, but would be able to hand that off then.
10:05 (which is mostly: *packaging*)
10:06 The current situation is quite reasonable from a simple rel-eng POV.
10:06 But the more I think, the more I look back at the Strawberry times.
10:06 And then I realize that this approach simply didn't scale.
10:06 Which is why I made this step.
10:06 And then I notice that this thing here doesn't scale either.
10:06 Because nobody cares.
10:07 But to have people care, to get them involved, they need to be excited.
10:07 Mirabelle is not going to have anything exciting.
10:07 And then I wonder where we're heading.
10:07 mchua thinks sustainability is exciting.
10:08 or rather, "watching Sebastian burn out" would be... exciting in a VERY BAD WAY
10:08 so, ah, lack of that excitement is... good. And thus exciting. Being able to grow a community to help out with this is exciting.
10:08 sdziallas Heh. Living on the edge.
10:08 Yes it is.
10:08 But it also needs to be exciting for the contributors.
10:08 I don't worry much about me being exciting.
10:08 s/exciting/excited
10:08 mchua Oh, we can give them shiny pictures of happy cute kids. ;)
10:09 sdziallas grins. heh ;)
10:09 mchua One of the reasons I'm so keen on getting deployments contributing.
10:09 sdziallas yeah, right.
10:09 mchua Anyway. For this release, SoaS is a spin.
10:10 sdziallas ayup. and presumably thereafter, too, unless somebody pays a me a million bucks to screw college. ;)
10:10 mchua Because of the necessity of reducing infrastructure maintenance and workload.
10:10 sdziallas nods.
10:11 mchua Okeydokey.
10:11 So we're stuck with a spin for this release.
10:11 Which means we have to package everything that goes in SoaS.
10:11 Which in some ways is extra work and all, true.
10:11 But in other ways is good, because it means all the code will need to pass Fedora guidelines and be relatively clean and modular, and properly licensed.
10:12 speculates a bit, is that the right reasoning?
10:12 sdziallas this is the reasoning behind it.
10:12 if you look at some stuff, it's quite horrible, though, but it works.
10:12 mchua Okay. So the issue with the kickstart file now is that a lot of the packages that are listed in it just Don't Work
10:12 and the maintainers aren't updating them
10:12 and there's no way you can do it all by yourself.
10:12 (By Tuesday.)
10:13 (which is, for those following along with this log, Beta Freeze for Fedora, which we need to follow since we are a Spin.)
10:13 sdziallas (I figured that final freeze is the cut-off date, so we've some time, but having stuff working by the end of the week would be good)
10:13 well, the thing is that I should be doing most of the packaging related issues...
10:13 mchua #link https://fedoraproject.org/wiki[…]s#Creating_a_Spin
10:13 sdziallas ...and then NAG everybody about the issues I know nothing about.
10:14 however, reducing some kind of load might be good, too.
10:14 mchua Ok. So stuff needs to be trimmed so that you can handle the packaging stuff and the nagging-load.
10:14 (I'll help with the nagging.)
10:14 sdziallas Okay. Thanks :)
10:14 , reducing some kind of load might be good, too.
10:14 mchua sdziallas: is this the point in time where we go ruthlessly slashing the kickstart file? ;)
10:15 sdziallas uh, bad c&p. sorry.
10:15 mchua: mhm, possibly.
10:15 mchua opens up ze gobby
10:15 sdziallas (I'm currently trying to figure out in the background why we won't have a sugar bootscreen for this release, heh.)
10:15 does, too.
10:16 mchua: you want to go for sunjammer? I can run a sobby session there.
10:16 mchua Ok, I have the doc  up, it's called fedora-livecd-soas...
10:16 ...oh, that works too
10:16 sdziallas mchua: or which server should I connect to?
10:17 mchua sdziallas: Sure, just tell me what to connect to when it's up
10:18 sdziallas mchua: sunjammer.sugarlabs.org
10:19 mchua sdziallas: Ok, I have a doc up.
10:19 I'm going to work under the assumption that we both have the .ks file locally available to us, so I'm going to...
10:20 (gimme a second to go SLASH SLASH SLASH)
10:20 sdziallas (good to add yourself there, yes :)
10:20 what is important to know is that @sugar-desktop still contains activities.
10:20 namely, all fructose activities.
10:21 mchua sighs
10:21 sdziallas so we'll either need to do -sugar-write or so or take the base sugar packages and name them directly (which should work)
10:21 LOL!
10:21 mchua How about that, then?
10:21 sdziallas pulls up comps
10:21 lols
10:22 alright, I'm pushing the package list in gobby now for you to see.
10:22 mchua oh holy god.
10:22 sdziallas we can kick all the optionals
10:22 they are not getting pulled anyways.
10:23 mchua how about lines 18 - 28?
10:23 can they go without breaking anything that's left?
10:23 sdziallas well, they are the fructose activities.
10:23 mchua so? ;)
10:23 sdziallas so we should probably just add "sugar"
10:24 and remove these.
10:24 mchua ta-dah
10:24 do we need gdm
10:24 ?
10:24 sdziallas BURN BABY BURN
10:24 or some kind of replacement, yes.
10:25 we *might* want to be able to login, right? ;)
10:25 mchua What for? It boots directly into Sugar, I don't see a GNOME login screen.
10:25 is ignorant.
10:25 we can't do that in a virtual terminal?
10:25 sdziallas yes, but you need something to log the liveuser account automatically in.
10:25 mchua Ohh.
10:25 Gotcha.
10:25 Okay. Is gdm working, is it a pain to maintain in any way?
10:25 sdziallas just starting X doesn't work, I hear (because of some policy / security foo things)
10:25 I packaged this: http://www.enricozini.org/sw/nodm/
10:26 looks like exactly what we want, but I haven't been able to get it to work.
10:26 mchua It's used by... just about everyone else in Fedora, so I would expect it to be well-maintained without our intervention, but I could be wrong.
10:26 looks
10:26 what's the benefit of that over gdm?
10:26 sdziallas it has NO dependencies.
10:26 or at least it doesn't pull in $random_bloaty_gnome_crap.
10:26 mchua whereas gdm has... right. Gotcha.
10:26 sdziallas grins.
10:26 but since it's non-working, we'll need to go with gdm for now.
10:26 mchua Okay, um... well, yeah, I'll file nodm as an enhancement we want for v4.
10:27 sdziallas (and probably put a comment to look at nodm at some point)
10:27 ah, awesome!
10:27 mchua you gobby, I'll trac
10:27 sdziallas deal!
10:29 (I'll do a ticket cleaning session in trac at some point this weekend, too)
10:29 mchua sdziallas: what do I do for the distribution/version field, and the milestone field? how do I make sure this is a soas bug?
10:30 sdziallas mchua: SoaS is the *component*
10:30 (this is important - when I clean up the wiki, I'll make sure to put this up highly)
10:30 mchua: distro is fedora - at least here
10:30 and milestone is currently unspecified
10:31 mchua looks
10:31 sdziallas: and I should be ccing sdz, right?
10:31 or do you automagically get it?
10:32 sdziallas you don't need to, because he and mchua automagically get tickets assigned to the component "soas"
10:32 :)
10:33 mchua: btw, here's what I'm thinking for v3 - something like "Do Your Own SoaS! - we included the base, now it's your turn. EXPLORE! - grab whatever you want from our portal, activities.sugarlabs.org"
10:33 mchua Bam.
10:33 yep.
10:33 sdziallas grins. BAM. :)
10:34 satellit_ browse and terminal will be included?
10:34 mchua sdziallas: btw, do we know why nodm doesn't work?
10:34 sdziallas: actually, nm, I'll put the ticket in and you can put more details later if there's time, it's not for this release
10:34 satellit_: Maybe, we haven't gotten to that decision point yet. ;)
10:35 sdziallas I think it might not be up2date for Fedora or there's some other strangeness going on, maybe some missing config file. will need to investigate at some point.
10:35 mchua Okay, I'll just put in the bug... one sec...
10:35 sdziallas: anything else that's needed to start SoaS and run Sugar (with no Activities yet), period?
10:36 sdziallas mchua: well...
10:36 satellit_ they are needed to go to aslo and edit
10:36 sdziallas mchua: gvfs is needed to show external hard drives.
10:36 mchua: gnome-power-manager is needed for... guess what? :) power management
10:37 and you don't have audio right now, but since we don't have any activities, that's alright.
10:37 mchua #link http://bugs.sugarlabs.org/ticket/1849
10:37 for the record
10:38 sdziallas mchua: you rock!
10:38 mchua satellit_: Yep, and those are both good arguments for adding them back in later, but we haven't gotten to that part yet.
10:38 satellit_: Right now, we're trying to start with the bare bones of "what will make Sugar work at all?"
10:38 satellit_ ok  anaconda (for liveinst)?
10:39 mchua sdziallas: I'm noticing we're using lots of GNOME stuff here, is that where we want to end up?
10:39 has nothing against GNOME, just likes using slimmer things if available
10:39 (...says the person who runs KDE)
10:39 sdziallas mchua: here? you mean when talking gvfs and power-manager?
10:39 mchua: (right now) :)
10:39 mchua sdziallas: Yeah, g(anything)
10:40 sdziallas mchua: well, that's probably because we're close to gnome... so I guess we'll need to live with that.
10:40 mchua: I recall tomeu advocating for these two packages.
10:40 mchua satellit_: hold on a bit - write up a list of packages you'd like considered in the background, and when we're done with this first pass through "what do we need to make Sugar work at all?" then we can take a look at yours, to see if we're missing anything.
10:40 tomeu yeah, rewriting those wheels doesn't look like a good idea to me
10:40 mchua satellit_: (we're doing a non-Activities pass first, and then an Activities pass.)
10:41 tomeu: Agreed.
10:41 sdziallas: Okay, they're in, where do they go?
10:41 sdziallas tomeu: I'm trying to have less on my plate, heh :) rewriting gvfs doesn't sound like... a good idea, agreed.
10:41 mchua: I'd put them into system, maybe with an explanation why we want them?
10:41 mchua: (I like that you added that for gdm, too)
10:41 mchua and by doing this I get to learn about kickstart files yay!
10:42 sdziallas hehe :)
10:42 mchua sdziallas: what else is needed to run Sugar?
10:42 Is that it?
10:42 sdziallas I think that's it.
10:43 mchua looks at old ks just in case
10:43 satellit_: Okeydokey, what's your list?
10:43 sdziallas Because the rest gets pulled in through dependencies.
10:43 mchua satellit_: (non-Activities first)
10:43 sdziallas (as, installing sugar pulls $good_chunk_of_stuff)
10:43 satellit_ mucha: liveusb-creator, gedit, wget, yum, gvfs
10:43 mchua sdziallas: Woo boy. Okay.
10:43 Ah, package management.
10:44 looks at time
10:44 sdziallas: shall we move to Activities and try to wrap before the SLOBs meeting?
10:44 satellit_ plus browse and terminal
10:44 sdziallas mchua: sounds good to me.
10:44 mchua satellit_: (it's mchua, btw - my screen won't flash for 'mucha' ;)
10:45 satellit_: ok - rationale for each? we can take them one by one.
10:45 satellit_ mchua: sorry : /
10:45 gvfs for disk access
10:45 mchua satellit_: np :)
10:46 satellit_ terminal for editing and access to yum
10:46 mchua satellit_: we have gvfs already, so that's set.
10:46 satellit_: good to have that confirmation.
10:46 satellit_ wget for uploading
10:46 gedit for easier editing than vi
10:46 mchua satellit_: actually, the terminal thing brings up a good point... do we need the terminal *activity* or can the current things in the kickstart file give us a VT?
10:46 sdziallas: ^?
10:47 satellit_ liveusb-creator which can be "sugarixed" as an applicaion in sugar
10:47 mchua satellit_: That would be nice, but it's not necessary to *run* Sugar
10:47 I'd think of that as an enhancement for later if we have bandwidth to maintain more.
10:47 Right now we're going for *super* minimalist.
10:47 sdziallas mchua: you could press alt + Fx and do stuff there, afaics.
10:47 satellit_ http://wiki.sugarlabs.org/go/C[…]e_liveusb-creator
10:48 mchua sdziallas: Ok, so we don't really need any other terminal stuff.
10:48 sdziallas: does the image already come with yum? Because if it does, that solves just about everything else
10:48 sdziallas: people can install their own stuff, and we're not on the hook for supporting it
10:48 sdziallas mchua: we're including a base kickstart file - and need to do that - so that you get a good chunk of packages directly.
10:49 mchua satellit_: A sugarized liveusb creator - and gedit - and so forth - would be nice to have, but that's exactly it; they would be "nice to have," not "nothing will work if we do not have them."
10:49 sdziallas (such as yum, which comes in one of these groups we pull)
10:49 @base-x
10:49 @base
10:49 @core
10:49 @hardware-support
10:49 @fonts
10:49 @input-methods
10:49 kernel
10:49 memtest86+
10:49 mchua satellit_: So I'm going to say we should pass on them for this round, because they aren't stricly necessary - and then perhaps provide instructions for how people can install it themselves, for this version.
10:49 satellit_ how do I access yum without terminal?
10:50 mchua satellit_: if maintainers step up to make sure those things work for future versions, great! Then those can be added to the v4 kickstart.
10:50 notes we should probably set criteria for inclusion in soas
10:50 #note We need to set inclusion criteria for SoaS for v4 and beyond.
10:50 sdziallas that sounds like a good plan!
10:50 mchua satellit_: Virtual terminals. if you're running Linux right now, press Ctrl+F5
10:51 satellit_ ok
10:51 mchua satellit_: (or Ctrl-F(anynumber))
10:51 satellit_: (Ctrl-F1 will usually get you back to GNOME or KDE or whatever wm you're using)
10:51 ...gee, I hope he read the "how to get back" directions before he... went off to virtual terminals.
10:51 satellit_ in soas...how
10:51 mchua oh, good. :)
10:52 satellit_: Same thing. Ctrl-F(key).
10:52 This is standard stuff on Linux systems.
10:52 satellit_ thanks...will test later on ubuntu netbook here
10:52 mchua satellit_: Whoops. It may be Ctrl+Alt+F(number).
10:53 ...yep, Ctrl+Alt+F(number). Sorry about that, I mix up my keyboard shortcuts sometimes.
10:53 satellit_ : )
10:53 mchua satellit_: it just takes a moment, you could easily check it out now.
10:53 sdziallas: Anyway, in the interests of time - can we say "yum is already included, people can download whatever else they need from there"?
10:53 sdziallas mchua: ayup!
10:54 mchua #agreed people can yum install further programs for this release if needed, we should provide instructions on how to do so
10:54 foresees fun with publican
10:54 forsees? foresees? whatever the right spelling is.
10:54 sdziallas YAY YAY YAY :)
10:54 mchua sdziallas: Okay, wow, 8 minutes left
10:54 sdziallas: ACTIVITIES GO
10:55 sdziallas alrighty!
10:55 do you just want to head through each or how do we do this best?
10:56 mchua sdziallas: Start with 0 and then add them back - what's the 1st Activity you'd add back, the 2nd, and then so on?
10:56 sdziallas: actually we could pause here and take up the question again when SLOBs is over.
10:56 sdziallas mchua: that would probably give us some time... yeah.
10:56 sounds good.
10:56 mchua #info where we're leaving off: Activity selection
10:56 Aaaand I will take the logs and put them on some wiki page somewhere...
10:56 #endmeeting
10:56 sdziallas awesome! mchua, thanks :)
10:57 mchua ...wait, meetbot?
10:57 sdziallas mhm.
10:57 #endmeeting
11:56 #chair mchua sdziallas
11:56 #topic kickstart part 2
11:56 here we go!
11:57 mchua When we last left our heroes, they were DRASTICALLY SLASHING the SoaS kickstart file to make the mainenance load more sane.
11:57 sdziallas so this is going to be the activity part!
11:57 nods.
11:57 mchua *introductory animation, cheerful trumpet music*
11:57 Anyway.
11:58 sdziallas :)
11:58 mchua sdziallas: do we need to put anything in the "multimedia" or "internet" sections or should we just delete those?
11:58 (also, hardware, sounds, and fonts)
11:58 sdziallas well, unless we need alpine, internet can go awawy.
11:59 mchua why would we need alpine?
11:59 sdziallas multimedia might be needed, depending on which activities we add.
11:59 to read emails :)
11:59 j/k
11:59 so multimedia is a little tougher, since some of the activities might actually need or not need stuff from there.
12:00 mchua Okay, so we'll leave the section in and if we add Activities that need things we can put them in.
12:00 and if the section's empty at the end of this exercise, it goes way.
12:00 sdziallas hardware was just there to support more hardware (which would be nice to have, I guess)
12:00 mchua er, away.
12:00 sdziallas nods, that sounds fair
12:00 mchua Mm, yeah, hardware support sounds good... how are those packages to update/maintain?
12:00 sdziallas (we could make sound and multimedia go into the same category, though)
12:01 mchua are those hardware packages being taken care of by other people for normal-Fedora anyway, and we can piggyback along?
12:01 sdziallas well, the b43 thing is maintained by... somebody (so non-sugar), i think Peter was looking into that.
12:01 so this is... exactly! the case there.
12:01 and the libertas firmware is for the XO, so we shouldn't need to worry there.
12:01 nukes sound (and will put that stuff later into multimedia)
12:02 mchua sdziallas: your call, I'd say. If you're confident they will take 0 maintenance, more hardware sounds good.
12:02 sdziallas so I add the two packages for additional hardware support?
12:02 yup.
12:02 nothing to worry about, I guess.
12:03 next one. fonts.
12:03 Walter filed a ticket on that some time ago, asking to include these.
12:04 http://bugs.sugarlabs.org/ticket/1119
12:04 should we simply add them and link this ticket for reference?
12:04 mchua looks
12:04 sdziallas: as long as $someone_else is maintaining the package, sure.
12:05 sdziallas mchua: yup, that's the case.
12:05 mchua sdziallas: yay packages that require no work from us!
12:05 oaky.
12:05 er, okay.
12:05 can't type this morning.
12:05 sdziallas mchua: mostly, the activities are what cause trouble, I guess
12:05 mchua: mhm :)
12:05 mchua: uh, actually we need to add the boot screen package, too (can wait, though)
12:07 mchua Oh, go for it
12:07 I wanted to save Activities for last
12:09 sdziallas: lemme know when you've got everything except activities in, I suppose.
12:09 sdziallas alright!
12:09 well, the rest is... depending on the activities, or negotiable.
12:12 (which is like "do we need a flash replacement?" or "a screen recorder?" "or a second installer?" or "a way of duplicating the image?")
12:12 mchua Activities, then.
12:12 Ah, okay.
12:12 sdziallas (if the answer is no to all of them, I think we're GO for activities)
12:12 mchua I'd say no to all of them.
12:12 and if people want that stuff, *they* can maintain it.
12:13 sdziallas heh.
12:13 mchua This is "operation: let's be lazy bums."
12:13 sdziallas well, it's not like I'm maintaining all of this, but it doesn't make sense to ship stuff that doesn't get used, either.
12:13 mchua er, I mean... "in order to ensure the high quality of the software that ships with Sugar on a Stick, we are making strategic decisions to..."
12:13 Right.
12:14 sdziallas can we turn that operation in a movie at some point? :D
12:14 j/k
12:14 alright. so. activities.
12:14 SeanDaly hi mchua, sdziallas
12:14 sdziallas waves to SeanDaly
12:15 mchua Hey, SeanDaly. We cancelled the SLOBs meeting.
12:15 (about an hour ago)
12:15 sdziallas: Yep, Activities.
12:16 SeanDaly ah, didn't know, I don't get invites anymore and didn't see a mail
12:16 mchua sdziallas: let's start with 0 and then add things in until we're not wincing *too* much anymore.
12:16 I'd like to try to stay at 5 or less, but... we'll see.
12:17 sdziallas mchua: whoa... heh, okay ;)
12:17 mchua: browse is a must, I hear.
12:17 mchua Yeah, if we can only ship with one, I would ship with Browse, from a "this makes Sugar functional" point of view.
12:17 It's also used to install other Activities, which... es muy importante.
12:17 sdziallas nods. it enables you to do a.sl.o... right.
12:17 goes adding that.
12:18 mchua sdziallas: so we know how much bandwidth we have to maintain other activities... how bad is Browse to package and maintain as a package?
12:18 Code-wise, the bugs scare me. It's not an easy activity to develop.
12:18 sdziallas mchua: well, browse is orphaned upstream.
12:18 mchua: so there are no updates anyways, heh.
12:19 mchua sdziallas: dumb question - why not use $another_browser? What's special about Browse?
12:19 (I think I know the answer, but humor me)
12:19 satellit__ firefox sugarized it is in aslo?
12:19 sdziallas mchua: well, if surf (as in: the webkit one) was better integrated with sugar, I'd go for that. but it's pretty dead, too.
12:20 mchua: to answer your question about maintaining it: browse is insofar special as the slightest change in xulrunner (in Fedora) gives us pain. but since xulrunner is pretty strongly maintained by the RH folks (except that they failed to care for F12), I think we... can't do much here.
12:21 mchua winces slightly to hear about that history, but... well... okay.
12:21 sdziallas mchua: basically: if anybody develops and maintains a new activity for browsing the net, I'd be more than happy to go for that instead.
12:21 mchua satellit__: Dunno, you could check in ASLO. ;)
12:21 satellit__: Firefox is huge and slow compared to Browse, though. So I'd be reluctant to make the swap. The UI also doesn't translate well to smaller screens, in my experience.
12:22 sdziallas: Okay, so... should we put Browse in this time, live with it, file a ticket to either find a maintainer or to look for something better for v4?
12:22 sdziallas mchua: yup, sounds good! against what will the ticket be filed, though? SoaS? Browse?
12:22 mchua: (the sad thing is: our most important activity is orphaned!)
12:22 satellit__ it is in aslo....need password to do dl
12:23 sdziallas (heh, I think I'm both of us might be blogging about this session, so... heh.)
12:23 mchua sdziallas: SoaS, because we need to find "a web-browsing solution" which may or may not be Browse. As part of that ticket we'll have to be poking the Browse folks, but we'll also explore other options.
12:23 sdziallas mchua: sounds reasonable!
12:24 mchua goes to file that ticket
12:24 sdziallas thanks :)
12:26 mchua So it's basically no maintenance work from the packaging side this round.
12:26 But may be a heavy load on testing and support.
12:26 That's okay with me.
12:26 sdziallas nods. exactly!
12:26 yeah... we should be good for v3 by now.
12:27 mchua Okay. If we *just* shipped Browse, would we be totally unsatisfied?
12:27 sdziallas Heh. Possibly.
12:28 SeanDaly what's the topic?
12:28 sdziallas SeanDaly: sustainability :)
12:29 mchua sdziallas: What's the 2nd Activity you'd add?
12:29 SeanDaly always a good one :-)
12:29 mchua SeanDaly: The idea is to ship a *much* smaller SoaS Activity list - bare minimum, providing instructions on how people can install more.
12:30 SeanDaly: That way we have the bandwidth to really test and "support" the small number of core Activities we're putting out.
12:30 And people don't expect us to support the rest - and they also get to feel a little more participatory by going out and exploring other Activities to install.
12:30 (part of this would be making very, very good docs and guides for many different sorts of audiences on how to do exactly that.)
12:30 (welcoming them to the community as explorers, etc.)
12:30 SeanDaly mchua: that's a bit of a crazy idea, considering we tell everyone that SoaS has a selection of ready-to-use Activities
12:31 mchua SeanDaly: But the truth is that it doesn't, really - a lot of the Activities are broken/unmaintained right now.
12:31 SeanDaly No argument there, but that's not the point
12:31 mchua SeanDaly: SoaS will still have a selection of ready-to-use Activities, just a smaller one. And all the Activities on it will actually work, work well, and have responsive maintainers.
12:32 SeanDaly SoaS serves a vital goal in overcoming the two main barriers Sugar faces: the install barrier and the unfamiliarity barrier
12:32 mchua SeanDaly: No argument there either.
12:33 SeanDaly It serves an extremely useful purpose to recruit teachers, parents, and contributors
12:33 mchua SeanDaly: Yep, I agree.
12:33 SeanDaly leaving out Activities raises the install barrrier unfortunately
12:33 SoaS for a deployment would be a different story
12:34 But when I am in front of 100 people like I was 2 days ago at the education roundtable of Solutions Linux, I boot up Sugar and point to Activities and launch a few
12:35 mchua SeanDaly: Oh, we're not cutting *all* Activities. We'll certainly leave a couple that will make a compelling demo.
12:35 SeanDaly: And the instructions for installing more Activities will be so good, first graders can follow them.
12:35 SeanDaly: Actually, my plan is to get first graders to write them. ;)
12:36 SeanDaly Booting it up with 3 Activities and saying "fihure out by yourselves how to use this unfamiliar interface which may or may not work with your wireless card to get online and donwload and install more Activities" defeats the entire purpose of SoaS
12:36 fihure/figure
12:36 what possible advantage is there?
12:37 mchua SeanDaly: Well, let me try to describe what will happen if we *don't* cut the list.
12:37 SeanDaly: We have... what, sdziallas, 20 Activities on the list now? Something around there?
12:37 SeanDaly: 8 of them simply don't work, as of right nowl
12:37 er, right now.
12:37 sdziallas mchua: probably something around that, yup.
12:38 mchua SeanDaly: Nor do they have active maintainers. So they will, in all probability, continue to be broken.
12:38 SeanDaly: The remainder aren't necessarily in much better shape. Some are extremely well-maintained (say, TurtleArt), but those are few and far between.
12:38 And things we would consider "core activities" - Browse, Write - are among the broken, unmaintained ones.
12:39 SeanDaly so 240 Activities in ASLO, and we can't find 20 that work?
12:39 mchua SeanDaly: That work well and are actively maintained by responsive developers? Yes, we can't find 20 of those.
12:40 SeanDaly the maintenance matters less than whether they work
12:40 tomeu also, I don't think we have enough resources to go through those 240 activities and actually see if they work or not
12:40 mchua tomeu: +1
12:40 tomeu in the past, we have included activities just because they managed to startup
12:40 mchua SeanDaly: The maintenance matters, because we also want them to work tomorrow.
12:40 tomeu and later turned out that were useless because of some later error
12:40 SeanDaly I mean, 1.4 millions XO-1s in the wild, and most of those Activities don't work?
12:41 mchua SeanDaly: Yep.
12:41 tomeu well, those xo-1s don't care if an activity works in mirabelle
12:41 SeanDaly it's news to me that most Activities used by children today are not working, where does this info come from?
12:42 tomeu often, we cannot say if an activity "works" because there's no documentation about how the author intended it to work
12:42 sdziallas tomeu: right, test cases are something we desperately need, too.
12:42 mchua SeanDaly: I think you're assuming that we have (or had, in the past) a sufficient amount of development and QA manpower to do things properly. ;)
12:42 tomeu SeanDaly: anecdotal experience from reading the bugtracker and the mailing lists
12:43 mchua SeanDaly: As the person who used to be responsible for running the OLPC QA community team as half of my $dayjob, and did deployment support as the other half... I can tell you with great confidence that This Is Not The Case.
12:43 SeanDaly I'd be surprised if the major Sugar deployments hadn't mentioned to anybody that most Activities are not working
12:43 sdziallas SeanDaly: booting the latest snapshot. knowing that browse and record were broken *for months* without anybody caring about that. (to be continued)
12:43 tomeu SeanDaly: they aren't using mirabelle
12:44 mchua SeanDaly: They're using old, old images that we shipped more than a year ago.
12:44 tomeu the problem is not only if something "works", but also in which sugar version it works
12:44 mchua SeanDaly: And they do report bugs sometimes (the ones who even know they can report bugs - not all of them do).
12:44 SeanDaly tomeu: mchua just confirmed that most Activities in the 1.4m XO-1s are not working, I'm just surprised to hear that
12:44 cjb SeanDaly: the main argument is that activities *did* work, and continue to work on XO-1s, but do not work on current SoaS
12:44 (as I'm hearing it)
12:44 tomeu cjb: ++
12:45 SeanDaly I'm confused at this point
12:45 cjb: but put like that I'm less confused
12:45 tomeu SeanDaly: also, it's not only if an activity works on a sugar version or not, we need to take into account different versions of the same activity
12:45 mchua SeanDaly: the set of deployments we care about Activities working on SoaS for is "the set of deployments actually using SoaS"
12:45 SeanDaly: which is extremely small at this point, afaict
12:45 SeanDaly This seems to be a case of throwing out the baby with the bathwater
12:46 mchua SeanDaly: the N-million kids using XO-1's are not part of our userbase for SoaS.
12:46 SeanDaly: How so?
12:46 SeanDaly As I say there's a difference between SoaS for lowering barriers and SoaS for deployments
12:46 mchua sdziallas: Which one would you say we are working on now?
12:47 (I would say the latter.)
12:47 SeanDaly SoaS fulfills a vital role for demoing Sugar, allowing teachers and parents to get past the two biggest hurdles
12:47 mchua SeanDaly: Ok - SoaS as demo tool vs SoaS as deployment tool are different uses, I agree with that.
12:47 tomeu suggestion: make a call for testing activities for inclusion into mirabelle
12:47 sdziallas mchua: I believe that with the position statement I issued for Blueberry (and I hear most people agreed with), we moved more towards the latter point.
12:47 mchua SeanDaly: The intent of demoing, though, is that we hope people will eventually deploy SoaS, right?
12:48 SeanDaly SoaS without Activities may be a good base for a pilot deployment version, but will fail in its role for demoing Sugar
12:48 mchua SeanDaly: so what I would do, if I were to stand in front of 100 people, is to have my SoaS image with minimal Activity set, and go to ASLO and load it up myself.
12:48 tomeu SeanDaly: the root of the problem is that if we want something done, we need to find someone who will do it
12:48 mchua SeanDaly: Then, when I gave my demo, I would say "I installed these additional Activities from ASLO - here's how I did it. Isn't ASLO awesome?"
12:49 SeanDaly mchua: and every nongeek teacher there will think "too complicated"
12:49 tomeu SeanDaly: until now, we have been trying to do more and more without worrying enough about who will do it, and this lowers quality and stresses people like sdziallas
12:49 mchua SeanDaly: And right now, I think that's fine.
12:49 SeanDaly tomeu: not just sdziallas :D
12:49 tomeu well, in the soas case, most of the pressure has been on him
12:49 mchua SeanDaly: We don't have bandwidth for support.
12:49 satellit__ could we have a script to install sets of activities from aslo
12:49 mchua satellit__: That's one possible solution that may be worth exploring.
12:50 SeanDaly mchua: I showed ASLO at Solutions Linux too
12:50 mchua satellit__: It still makes the assumption that we start with a very small set of Activities and let other people figure out what else *they* want to add and worry about themselves. :)
12:50 satellit__ like fedora net install preset activity sets
12:50 SeanDaly but when I handed out SD cards, I was always asked if "content" was present or if had to be added, and teachers are invariably happier to find out that they don't need to do anything complicated
12:51 For example, I installed Fedora from DVD on 2 netbooks and getting Sugar to work on them has been difficult
12:52 and also disorienting, since there are no Activities in the base install; and no indication that ASLO exists
12:52 mchua satellit__: A script like that would be outside the scope of SoaS itself, but it might make a nice thing to make as a tool to act *on* the finished SoaS image (though I'm not sure how easy/possible it is, tbh; it may be a simple 'yum install activities' bash script, or it might be ubercomplicated.)
12:52 satellit__: in other words, "if you want to make a script that takes the SoaS image and does stuff to it to add more Activities, that's great, but that script itself isn't part of SoaS."
12:53 SeanDaly We've been dealing for some time with our limited resources issues
12:53 mchua SeanDaly: I don't disagree with your assessments - I just think that we do not have the capacity to support teachers who can't figure out a lot of this stuff, at this point.
12:53 As a project - as SoaS.
12:54 Individual deployment supported by individual people can do that.
12:54 satellit__ could the script be on a startup menu plus install to hard disk  (usb) like trisquel and fedora
12:54 SeanDaly If we have that many orphaned Activities, perjaps we need to raise priority of recruitment
12:54 mchua For instance, that's why I've got the CFS SoaS deployment - because I can guarantee that *I* (Mel) will help that classroom.
12:54 SeanDaly mchua: Teachers need to know Sugar exists
12:54 mchua SeanDaly: We do need to raise the priority of recruitment - but that's going to help us make a more full-featured SoaS later.
12:54 SeanDaly: Right now, we need to slim down many things because the capacity we have to make things well *now* is small.
12:55 SeanDaly: I'd rather do a small, high-quality release than a large crappy one. I want to have a reputation for being Extremely Good.
12:55 SeanDaly: Teachers need to know Sugar exists, yes. And when they discover Sugar, I'd like them to discover a great, working product.
12:55 SeanDaly Well, hearing that most Activities in ASLO don't work seems to me to be a critical situation
12:56 mchua SeanDaly: Right now, if they don't know about Sugar, they have no expectations about what it does - so we set expectations for the few things we know we can do well, and then as we gain the capacity to do more things well, we go "and now we have more things!" and celebrate that.
12:56 SeanDaly: It absolutely is a critical situation.
12:56 tomeu a critical situation is that people think that SLs does things by itself, without needing investment
12:56 SeanDaly mchua: a solid bulletproof Sugar would spread very quickly, by word of mouth. But it seems to me we can't have a solid bulletproof Sugar with new versions every six months, unless I'm missing something?
12:57 tomeu this has drained the energy of the people who managed what we have today, without giving us a way forward
12:57 mchua SeanDaly: With our current manpower for SoaS, we can only have a *very small* solid bulletproof SoaS every 6 months.
12:57 sdziallas tomeu: +1.
12:57 mchua tomeu: +1
12:57 tomeu we have shown that we are capable of doing stuff, now we may need to switch to show how other people can make it possible that we make more stuff
12:58 SeanDaly The recruitment problem could be partly solved with funding
12:58 mchua SeanDaly: That's actually the goal - it's why sdziallas made SoaS a Fedora Spin this time around, it's why we're slimming down the Activities list for SoaS right now... we *want* solid bulletproof Sugar.
12:58 SeanDaly: But we don't have funding now.
12:58 SeanDaly: The thing is, beta freeze is *Tuesday.*
12:58 tomeu SeanDaly: but not just money, the money should be coming from those who actually use and need sugar
12:58 mchua SeanDaly: So our problem right now is "okay, between now and Tuesday, what's going to work - and work well?"
12:58 tomeu otherwise we won't invest it wisely
12:58 SeanDaly mchua: without Activities, there's no point for SoaS
12:59 mchua SeanDaly: There'll still be Activities, just not 20 of them.
12:59 SeanDaly As I say, no point
12:59 mchua SeanDaly: We're going through the list here in public precisely so that people can help us prioritize which Activities we need to spend our time on shipping.
13:00 tomeu sees a big point in what mchua and sdziallas are doing
13:00 mchua SeanDaly: Criteria include things like "impact on education," "demoability," "effort needed to package and maintain the package," and "state of bugs/ease of development."
13:00 SeanDaly might be better to deal with this on-list, it's an important topic (unless I missed the mail, I have had connectivity issues)
13:00 tomeu and I'm sure deployers like simon and children like his class do as well
13:00 mchua SeanDaly: It's going to be hitting the list as soon as we're done with the first cut of the kickstart file.
13:01 SeanDaly: Freeze is Tuesday; between now and then, there are 4 daily builds.
13:01 SeanDaly I don't know what a kickstart file is, I'm sorry
13:01 mchua SeanDaly: We just need to get a first small working version up, and then tell people the process for other things being added to it.
13:01 SeanDaly: Whoops, sorry. Geek-brain slip. :)
13:01 SeanDaly mchua; that's fine for deployments, but fails as a Sugar demo
13:02 SoaS is the central pillar of our marketing because it overcomes the two greatest barriers Sugar faces: installation and unfamiliarity
13:02 mchua SeanDaly: You can think of the kickstart file as the list of Activities that will ship in SoaS. It's in a special format so that when you run a script on a kickstart file, it generates an .iso.
13:02 sdziallas: (I think. Correct me if I'm wrong.)
13:02 sdziallas mchua: that's a good description, I think.
13:03 mchua SeanDaly: If we have no solid product, though, we have nothing to market.
13:03 SeanDaly: We need to market what we do have - what we can produce - which is this small thing that works well.
13:03 SeanDaly mchua: no, that's not correct
13:03 mchua willing to be corrected - I'm an engineer, not a marketer, though I've been trying to learn the latter lately. :)
13:04 SeanDaly although a bulletproof reliable product would make marketing extremely simple - think Google, which never had to advertise for its first 10 years in business,
13:04 we can't say "stop spreading the word about Sugar, it's not ready"
13:05 mchua SeanDaly: oh, no, I wouldn't want to say that.
13:05 SeanDaly: I'd want to say "spread the word about Sugar - here's what we've got right now, and you can help us build the rest and make it even better."
13:05 SeanDaly Everyone who evaluates Sugar - journalists, ed tech buyers, the EU contacts for example - everyone knows Sugar is not ready
13:06 What they want is to see the signs of how it is developing and improving
13:06 But, my concern is not with them
13:06 it is with the 7 million or so primary school teachers in the world
13:06 We need to inform them that there is an alternative
13:07 mchua There's an alternative, and it's not ready for most of them yet.
13:07 They can help make it more ready for them.
13:07 SeanDaly OLPC has been thoroughly slagged in the press these past couple of years, for several reasons,
13:07 tomeu is confused about the subject of the discussion
13:08 SeanDaly but one of the main ones is: journalists couldn't get their hands on one
13:08 mchua Hey, rkabir!
13:08 tomeu SeanDaly: I think everybody in this channel wants sugar and soas be wonderful products and cater all the educative needs of teachers and students
13:08 SeanDaly tomeu: SoaS is fulfilling two roles, and if Activities are absent from it, it will not work in one of those roles
13:08 tomeu SeanDaly: but: 1. we cannot do more than what we can do, 2. we need to think about expanding our capacity
13:09 mchua SeanDaly: I think we're drifting from the original purpose of this conversation, which was taking a hard look at the engineering reqs for SoaS for this release cycle (with the upcoming Tuesday beta freeze) and being realistic about what we can and can't do well.
13:09 tomeu SeanDaly: I suggested before making a call for people testing activities for inclusion in the next release
13:09 SeanDaly: or who do you expect to do that work?
13:09 SeanDaly tomeu: yes, which is why I said earlier we should raise priority of recruitment and funding if situation that bad
13:10 mchua SeanDaly: That's a great medium-term solution, and I'd love to do that. And in the short term - between now and Tuesday - we have to be realistic about our engineering capabilities.
13:10 On that note... sdziallas, what's the 2nd Activity you'd add back in? ;)
13:10 tomeu SeanDaly: so, as I said before, I think first step is stop giving that image of we being able to do much more than what we can really do
13:10 SeanDaly tomeu: we've made that call before (cf. press releases) without much success, we need to try another way
13:11 tomeu: I'm not aware of that image...
13:11 tomeu SeanDaly: I'm pretty certain there is
13:11 sdziallas mchua: turtleart? physics? not sure.
13:11 mchua sdziallas: wait, over Record and Write?
13:11 pulls up activity list
13:11 sdziallas mchua: write is broken.
13:11 mchua realizes that Record and Write are broken, but yeah :)
13:11 sdziallas mchua: record... well, works half-ish, could go in.
13:11 SeanDaly Look, if situation is this critical we can skip the launch for Mirabelle
13:12 mchua SeanDaly: You're welcome (actually, it would be *great*) to chime in on the Activity selection... which ones demo best?
13:12 SeanDaly: I think we can redirect a lot of the publicity push for Mirabelle for users/consumers
13:12 SeanDaly What we should do is number it v2.5 and position it as strengthening stability for our deployments
13:13 mchua SeanDaly: but we need to "release early, release often" to enable developers to help out - it's part of what we have to do in order to increase our capacity, and make recruiting possible.
13:13 SeanDaly mchua: as I say, if the situation is this critical we shouldn't do any publicity
13:13 mchua SeanDaly: to users, no.
13:13 SeanDaly mchua: yes, as I say we need to raise priority of recruiting/funding
13:13 mchua SeanDaly: To developers who might help out, we need to do a *lot* of recruiting.
13:13 SeanDaly: so maybe that's not called 'publicity,' but somehow we have to reach those folks. :)
13:14 satellit__ fix spin v3 then when it works do publicity
13:14 SeanDaly mchua: by "users" I guess you mean my 7 million teachers? :D
13:14 mchua sdziallas: ok, I'm looking at https://fedoraproject.org/wiki/Sugar_Activities
13:14 SeanDaly: Yeah... market less to teachers and more to techies who can help make it ready to market to teachers
13:14 sdziallas mchua: I'm there, too (and have the rest of the activities in the old kickstart file)
13:14 mchua SeanDaly: Or to techies who can then opt to support the teachers near them, kinda like what erikos and I are doing now
13:14 SeanDaly satellit__: betterr to renumber v2.5 maintenance release and save v3 number for media launch in the fall (or whenever ready)
13:15 satellit__ yes i guess I agree...sadly
13:16 SeanDaly mchua: well, a diffferent approach could be to recruit geeks to pilot in schools (previously we had been thinking "geeky teachers")
13:16 mchua SeanDaly: Hold on a moment, I want to make sure we're on the same page here...
13:16 here's what I have in my brain, correct me if I'm wrong.
13:16 satellit__ I still like the minimal idea....with scripts to install on startup menu
13:16 mchua SeanDaly: The facts of the situation are that we have a looming deadline (Tuesday beta freeze) before which to get engineering done, and few resources, so we need to (temporarily) scale back the scope of our engineering efforts.
13:17 SeanDaly satellit__: it's a good approach... for a school admin setting up a deployment
13:17 mchua SeanDaly: which is what sdziallas and I are trying to do right now, in the form of reducing the list of activities that ship with SoaS by default.
13:17 SeanDaly mchua: I have no objection (I mean, we can't spin gold from flax), but dong so make SoaS unusable in its demo role
13:18 mchua SeanDaly: So there will be an image created that is 100% solid working SoaS - but very minimal SoaS, with few Activities, because we will be spending our efforts on making sure those Activities are working really well.
13:18 SeanDaly: Right, getting to that.
13:18 satellit__ strawberry works for demo?
13:18 mchua SeanDaly: This image is a lot less full-featured than the previous one - it has fewer Activities, and as SeanDaly pointed out, doesn't demo as well because... well, there's less stuff to demo.
13:19 SeanDaly: And so it's less appealing to a large group of teachers - the kind who want and need the extra leg-up of having Activities ready-to-go on the stick they're trying out.
13:19 Which is fine, 'cause teachers have a lot of other things to do. :)
13:20 SeanDaly: And so the question now is - okay, we're making this slimmer image... what do we call it and how do we market it when it's released (in May, because we're tied to the Fedora 13 release date, as a spin)?
13:20 SeanDaly mchua: we don't market it
13:20 mchua SeanDaly: I agree.
13:20 SeanDaly: I think we use that release as a recruitment release, rather than a marketing release.
13:21 SeanDaly mchua: we use it in a recruitment campaign, and the less said in our PR the better
13:21 mchua SeanDaly: +1.
13:21 SeanDaly: Okay, so I don't think we actually disagree on anything. :)
13:21 SeanDaly: (wait, isn't recruitment sort of a form of marketing, though?)
13:21 SeanDaly: (marketing the experience of participation, anyway.,)
13:22 SeanDaly Well, it's a disappointment not to keep press release rhythm going, but I have other PR in the pipeline
13:22 Marketing is what I call reaching 7 million teachers, and recruitment is what I call reaching 1000 developers
13:23 mchua SeanDaly: Yeah... well, between May and the start of the school year, there's some time to fix stuff. I'm not sure what'll happen there, but hopefully we'll have enough contribution from that recruitment push to give you something really good to market for v4.
13:23 SeanDaly naturally, there's a communications strategy in recruitment :-)
13:23 satellit__ I think strawberry is great for demo's most applications work. use v2.5 for recruitment for developers
13:23 SeanDaly Shouldn't be v4, that will confuse people. We should renumber v2.5 and save v3 for the fall when hopefully it will be ready
13:24 mchua SeanDaly: Yeah, I think we're going to need a lot of help with that when the time comes. :)
13:24 SeanDaly satellit__: Ii still use Strawberry for demoing
13:24 tomeu SeanDaly: I don't think recruitment is just reaching developers, there's much more to it. also, we don't only need to recruit, we need to manage them
13:24 satellit__ : )
13:24 tomeu iow: we need a community manager
13:24 mchua SeanDaly: Actually, can we talk about the version numbers for a bit?
13:24 SeanDaly tomeu: no disagreement from me, but mchua was talking marketing
13:25 mchua SeanDaly: As a (former) developer, calling this 2.5 is confusing and goes against all the FOSS dev conventions that I know.
13:25 SeanDaly most FLOSS projects think marketing is recruitment, and miss opportunities for breakout
13:25 tomeu SeanDaly: I just say this because if we reach those developers but aren't in place to welcome there, we'll miss the chance
13:25 mchua SeanDaly: Maybe what we need is a way to mark releases as "for teachers" - the same way, for example, Ubuntu marks some releases as LTS.
13:25 satellit__ if use 2.5 then have 1 year releases. better for school periods?
13:26 SeanDaly mchua: exactly the opposite: I argued for the renumbering to v1, v2, v3 precisely because the v0.84 numbering was incomprehensible to nongeeks teachers
13:26 mchua tomeu: Yep. So any outreach/contribution campaign has a dependency of making sure we can take in the influx of whatever that campaign will give us.
13:27 SeanDaly v3 is commonly considered a version that is stable and has the bugs out
13:27 it would be a big mistake for us to use v3 on a maintenance release
13:27 mchua SeanDaly: Hm, okay. I thought that's what the names (strawberry, blueberry, mirabelle) were for - so that there would be a series of user-facing names independent of the engineering-development numbers.
13:27 satellit__ v3 RC?
13:27 tomeu mchua: and perhaps there's even more to that, we need to change our community's culture to be one where it's more clear where more work is needed, how that work can be contributed, and what will happen if nobody steps up
13:28 mchua SeanDaly: from my pov, as an engineer, I need some way of marking when large changes in the product happen - for technical compatibility and maintenance reasons.
13:28 SeanDaly mchua: if you look at the PR, we have always associated the ice cream flavor with v1, v2
13:28 mchua SeanDaly: and in my mind, what we're doing now is a *huge* change, relatively speaking
13:28 SeanDaly mchua: removing nearly all Activities is a pretty major change
13:28 mchua SeanDaly: so I'm going "oh hey, this is something different altogether - this is a new major version, engineering-wise."
13:29 SeanDaly: Maybe the engineering numbering needs to be separate from the marketing numbering, but I'm not sure how we can separate those two counters.
13:29 tomeu: Yep, culture of contribution.
13:29 SeanDaly then best to start an Nth engineering numbering system
13:29 v1, v2, v3 have always been "marketing" version numbers
13:30 mchua SeanDaly: in the engineering numbering system, the-image-that-will-come-out-in-May will, I think, be the-third-version...
13:30 tomeu should restart reading jono's book
13:30 SeanDaly This was discussed in depth last May before Strawberry release
13:30 mchua SeanDaly: what can we use for counting engineering releases that won't collide with the marketing version numbers?
13:30 sdziallas: thoughts?
13:30 SeanDaly mchua: not for teachers, so should be v2.5
13:30 mchua: you could use SoaS F13/v0.88
13:30 sdziallas (I've been following this convo)
13:31 mchua SeanDaly: well, if we were using the same numbering system for marketing and engineering, sure. But we're not, because engineers need a different numbering system than... oh, hey. SoaS F13/v0.88 would work for me.
13:31 sdziallas I'm leaning towards arguing that it's still the third release.
13:31 mchua sdziallas: Why's that?
13:31 SeanDaly mchua: I argued for v1 instead of "F11/v0.84", and you saw the press coverage?
13:32 sdziallas mchua: Because... it is. Whether it has a different audience from a PR point of view doesn't decrease the value of the release.
13:32 mchua SeanDaly: Refresh my memory?
13:32 satellit__ why cannot it be labeled V3 release candidate?
13:34 SeanDaly sdziallas: journalists who reviewed v2 Blueberry and in the fall hear about v4 Cloudberry will say: "Whoa! I missed a version!"
13:34 mchua satellit__: Because a release candidate is "the final release, we think, except it hasn't been fully tested yet" - implying that the feature set in the release candidate is the feature set that will be marketed to users.
13:34 SeanDaly But they didn't, because this version is not intended for demoing Sugar
13:34 sdziallas SeanDaly: well, not necessarily.
13:34 SeanDaly: here's the thing:
13:34 satellit__ ok  I thought is was for debugging and testing
13:34 mchua satellit__: In this case, we have a very tiny feature set, and we don't *want* to market that to users, we want to wait another cycle to have more features in a release we *do* want to market to users.
13:34 satellit__: at least that's my understanding - SeanDaly can probably correct me if I'm wrong. :)
13:34 sdziallas (gah. gotta find a quote, hang on.)
13:35 SeanDaly sdziallas: surely you're aware that most tech writers review a "v3" more in depth than first two versions?
13:35 sdziallas SeanDaly: "Do Your Own SoaS! - we included the base, now it's your  turn. EXPLORE! - grab whatever you want from our portal,  activities.sugarlabs.org"
13:35 SeanDaly: together with a "we're calling for more developers"
13:35 SeanDaly: plus a "we're collaborating with Fedora to make v4 more awesome - v3 is already rock-stable"
13:35 ...call should make a pretty decent marketing call from my point of view.
13:36 mchua I do think there's a way to calibrate expectations for this as a release such that reviews based on those expectations will be good.
13:36 SeanDaly sdziallas:as I say i agree we can work on recruiting developers, but a SoaS without Activities will fail as a demo for teachers
13:36 sdziallas SeanDaly: we are *not* removing all activities.
13:36 SeanDaly: and you were talking about journalists a second ago.
13:36 I'd like to tackle these issues step by step.
13:37 mchua listens
13:37 SeanDaly sdziallas: it's just not realistic to expect someone who has never seen the interface and doesn't know what a wifi driver is to connect to ASLO and install Activities
13:37 sdziallas SeanDaly: that is why there will be awesome documentations.
13:38 SeanDaly: such as included instructions in the journal and a redesigned wiki page.
13:38 SeanDaly sdziallas: it's a major change from previous launched versions
13:38 That's why we shouldn't do any media campaign, just outreach to developers
13:38 sdziallas SeanDaly: I'm not sure I understand that point.
13:38 SeanDaly but v1, v2, v3 numbering should be tied to media launches
13:39 sdziallas SeanDaly: Change is good. And as long as we can justify why we're doing this - and we certainly can - it'll be alright.
13:39 satellit__ we need an accompanying DVD .iso to DL with the instructions and demos
13:39 sdziallas SeanDaly: If we explain that v3 introduces a cooperation with Fedora, making everything more stable, I don't see why we shouldn't do that.
13:39 SeanDaly sdziallas: it will be fine, just not for media launch and teachers, that can wait until the fall
13:39 sdziallas SeanDaly: such as we did with Strawberry.
13:40 SeanDaly: I still don't understand why it shouldn't be fine for a media launch.
13:40 SeanDaly sdziallas: developers will care about that, but teachers will ask: "what's Fedora?"
13:40 sdziallas SeanDaly: if you select the target audience wisely, it won't be a problem.
13:40 SeanDaly: that's why you say: "to make it more stable."
13:40 SeanDaly: you give rationals for *what* we're doing.
13:41 SeanDaly I think the target audience should be developers
13:41 sdziallas SeanDaly: we're telling the people that we're making these steps to make SoaS *more* awesome.
13:41 mchua pipes up. I was just reading backscroll in #fedora-mktg and saw a potentially relevant conversation - short excerpt at http://fpaste.org/ZUWG/.
13:41 SeanDaly sdziallas: most of my 7 milllion teachers will ask: "what's Sugar?"
13:42 to be clear, it is a major challenge to build a brand from zero
13:42 mchua It's a huge challenge, and you're doing an amazing job at it. I'm not sure what we'd do without you, SeanDaly.
13:42 SeanDaly we are encountering success because we are smart and depsite having no marketing/promotion budget
13:42 despite
13:42 mchua And because we have been producing *good stuff.*
13:43 We fundamentally have a good product. It's rough. It's definitely nowhere near complete. But it's got plenty of potential.
13:43 SeanDaly mchua: yes, that's always the case - you can only market colored water for a limited amount of time
13:43 mchua It is also a major challenge to build a software learning environment - or a distro - from zero.
13:43 SeanDaly Our shortcut to raising awareness with tiny budget is: news coverage
13:44 mchua that's one of many possible strategies, but yeah, it seems to work well. :)
13:44 SeanDaly We've had success with tech news and some general outlets, but not with education press yet
13:44 mchua Anyway, the version numbering - I think this is something we can come back to, really. Right now, we have to figure out our engineering scope.
13:44 Otherwise we won't have anything to assign a name or number *to*.
13:44 SeanDaly mchua: I'm always open to strategy discussions :-)
13:45 mchua SeanDaly: as am I. ;)
13:45 sdziallas (uh, I'm, too. heh :)
13:45 mchua how about this - sdziallas and I and whoever else wants to hang around will blaze through the first round of the slim package list, make that iso
13:45 tell the list, attempt to summarize this conversation
13:45 we'll tell 3 lists in 3 different ways.
13:45 soas, to say "okay, here's what we're doing, engineeringwise, and why"
13:46 SeanDaly If we can't offer a SoaS with Activities, it's a pity, but no big deal, we just delay the media launch to the fall
13:46 mchua activities (or maybe ieap and devel or... wherever we think activity devs hang out) saying "here's why these things made soas this time and why the rest did not, and how you can get your stuff in next round by stepping up to do this work."
13:46 SeanDaly But journalists and analysts watching us should not get confused by us skipping a number
13:46 mchua and marketing, to say "here's what's going on in the other two lists," and then let you take it from there and figure out what marketing should do.
13:47 first topic can be version number. ;)
13:47 SeanDaly: how's that?
13:47 SeanDaly mchua: many approaches possible: "adopt an Activity"
13:47 mchua SeanDaly: Yep yep. But we gotta get the first version of "slim SoaS" done first. (...I'm sure we'll think of a, um, better codename soon.)
13:48 SeanDaly mchua: what is clear is that calling a maintenance release v3 makes my work much harder
13:48 media launches are a lot of work
13:48 mchua SeanDaly: Right. so we'll go ahead with the engineering, but we won't call it anything in particular until you have a chance to figure this out with the marketing team.
13:49 SeanDaly no big deal to call this one off if it's not ready, after all it's not like we have made media investments already
13:49 mchua or we'll call it "slim SoaS" or something less... silly sounding than that in the meantime. But not Mirabelle, not v3... for now.
13:49 right.
13:49 good to find this out early on.
13:49 SeanDaly: I'm glad you popped in today.
13:50 SeanDaly Mirabelle is cool, it's just "v3" I have a problem with for a non-media launch release
13:50 mchua sdziallas: (does that sound like a reasonable temporary compromise - we'll just use a codename for this until we can figure out the version/name thing with the marketing folks?)
13:50 SeanDaly mchua: I popped in for the SLOBs meeting
13:50 mchua SeanDaly: wait, so you'd be fine with us calling it mirabelle, but not... v3?
13:50 SeanDaly mchua: Yes, we had alrready agreed to save "Cloudberry" for fall version
13:51 mchua SeanDaly: basically, "what should we call this slim-Activity-set SoaS version we're making, until such time the Marketing list can discuss what to actually call it?"
13:51 SeanDaly: if that's mirabelle, then I think we're set.
13:51 SeanDaly call it v2.x Mirabelle where x is any number you want
13:52 mchua Heh, I usually think of Marketing wanting the names and engineering wanting the numbers :)
13:52 SeanDaly This is not a big marketing team issue which requires lots of discussion, since it is a media launch cancellation
13:52 sdziallas I believe rel-eng has a call in the versioning, too.
13:52 SeanDaly mchua: no, the v1, v2, v3 came from me, for marketing purposes
13:52 sdziallas I also don't get why this is a "media launch cancellation".
13:52 is quite a bit lost in this convo.
13:53 SeanDaly sdziallas: we had planned a media launch for Mirabelle, but it's clear it won't be ready
13:53 sdziallas SeanDaly: it will be ready.
13:53 mchua so we'll do a recruitment push instead, is my understanding
13:53 sdziallas and it will be released in May.
13:53 mchua sdziallas: ready for recruitment, not for "hey end users, new stuff for you, out of the box!"
13:54 SeanDaly sdziallas: with or without Activities?
13:54 without Activities, there's no point
13:54 sdziallas mchua: right. but for people reading stuff also with "we're doing this BECAUSE..."
13:54 SeanDaly However, it's now clear to me that recruitment and funding needs to be higher priority than before
13:54 sdziallas SeanDaly: certainly with activities. but neither broken nor crappy activities nobody needs.
13:54 mchua SeanDaly: So, clearly, not everyone is on the same page re: version numbering and naming and marketing stuff :) I'd like to have that convo on the marketing list
13:55 ...I'd also like to finish the engineering scoping of what Activities we will be shipping asap
13:55 so we know what we will be marketing and/or recruiting for.
13:55 If we don't know that, at some level, we're just talking hypotheticals.
13:56 Let us figure out the product first, because we're under some hard resource constraints in terms of engineering firepower.
13:56 And then we'll figure out the marketing on-list with you.
13:56 SeanDaly: sound fair?
13:56 SeanDaly Look at the press page... the next PR media campaign needs to be v3. That launch should happen in the fall if we can't ship with a set of Activities
13:57 sdziallas SeanDaly: we are NOT saying that we're shipping "without a set of activities"
13:58 SeanDaly mchua: the key takeaway is... an engineering crisis situation needs to have marketing in the loop asap, and not as an afterthought
13:58 sdziallas SeanDaly: we are saying that we're shipping with a *well-selected* set of activites which can be adjusted and enhanced easily by following clear instructions.
13:58 SeanDaly: what do you think does having three core modules (out of 12) for Sugar mean?
13:59 SeanDaly sdziallas: again, great for deployments, but useless for first-time tryers
13:59 sdziallas SeanDaly: so we're trying... to target everybody at the same time?
13:59 mchua SeanDaly: That's why I'm glad you're here, and why we're going to be sending this to marketing right after we finish picking the activity set.
14:00 SeanDaly: but right now, Sebastian and I have to scope the feature list for engineering so we can do the engineering and actually have a product.
14:00 SeanDaly sdziallas: No, we are accepting that next SoaS can't fulfiill demo role, no big deal we just roll back the media campaign to the fall. But the v3 number needs to be associated with the next campaign.
14:00 mchua So I'm going to request we table this convo for a couple hours while sdziallas and I work that out, and then we'll take this to the marketing list.
14:01 satellit__ sdziallas:  off topic :we also need to fix the jabber.sugarlabs.org It has been down quite a time now. Needed for testing and demos
14:01 mchua Okay, guys, I really need to get some food.
14:01 sdziallas same here.
14:01 SeanDaly me too need to go
14:02 sdziallas satellit__: daveb is the guy for that. if nothing helps, emailing him should help.
14:02 mchua sdziallas: Meet you back here when we're done eating and we'll go through the rest?
14:02 sdziallas: (I might be an hour or two)
14:02 sdziallas mchua: sounds good!
14:02 mchua: I'll be around.
14:02 mchua sdziallas: #endmeeting?
14:02 yeah, sam here
14:02 er, same here
14:02 sdziallas #endmeeting

Index | Today     Channels | Search | Join

Powered by ilbot/Modified.