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

#dextrose, 2011-01-08

 « Previous day | Index | Today | Next day »     Channels | Search | Join

All times shown according to UTC.

Time Nick Message
01:57 m_anish is now known as m_anish_afk
08:07 dfarning_afk is now known as dfarning
09:06 dfarning is now known as dfarning_afk
09:20 dfarning_afk is now known as dfarning
11:47 alsroot_away is now known as alsroot
12:05 m_anish_afk is now known as m_anish
13:35 silbe <silbe!~silbe@twin.sascha.silbe.org> has joined #dextrose
16:04 satellit_afk <satellit_afk!~satellit@2002:d064:95e7:0:219:d1ff:fe73:14e6> has joined #dextrose
16:49 tch <tch!~tch@r190-132-66-45.dialup.mobile.ancel.net.uy> has joined #dextrose
16:56 dfarning is now known as dfarning_afk
16:57 dfarning_afk is now known as dfarning
17:15 dfarning is now known as dfarning_afk
17:31 dfarning_afk is now known as dfarning
20:30 m_anish alsroot, tch ping!
20:30 was preparing lunch + going thru sugar-meeting logs
20:33 alsroot m_anish: lets move to #sugar, bernie is worried that it is too silent these days :)
20:33 m_anish alsroot, ah just saw tht :)!
20:43 alsroot forgot about absent logs for #sugar
20:44 m_anish "kill the name picker that pops up when quitting activities (EASY)" already has a patch in the m-l if not build
20:44 "Figure out why so many control panels require restarting Sugar, and fix them not to (MEDIUM)" should probably be dx2
20:45 alsroot do we have some kind of consensus for that?
20:45 ..based on feedback from the field
20:45 m_anish kill the "Mute" function on the volume icon in the frame (EASY) -- probably has a patch :)
20:45 alsroot m_anish: maybe just one-by-one?
20:46 m_anish alsroot, we need a channel with bernie on it :-) , all these nitpicks are from bernie's time in py
20:46 tch pong
20:47 yup, the name picker was something bernie learnt from his testing at caacupe, no body where using it and it was causing confusion.. they thought it was annoying
20:48 m_anish alsroot, silbe tch i'm okay with discussing today or tomorrow
20:48 tch me too, the more we are the better the discussion, but lets not procrastinate it much
20:49 alsroot I think we need to handle every day meetings, not 1h but just 15-20min to sync our work
20:49 m_anish tch, alsroot lets go thru the list once, just to sync up...
20:50 tch, should have an idea abt some of the tasks :)
20:50 tch sure,
20:51 m_anish silbe, "Notification system (tch)" is done and included in build
20:51 bernie <bernie!~bernie@gateway/shell/sugarlabs.org/x-vqlbmpuszxyzxkkw> has joined #dextrose
20:51 silbe tch: most of us know (already knew) the name picker is a bad idea and needs to go. The real problem is replacing it with something better.
20:51 bernie hello. am i too late?
20:51 tch bernie: nope just began ;) haha
20:52 m_anish okay everyone's here for the moment :)
20:52 silbe m_anish: interesting. I really need a Dextrose fs tarball so I can try it out...
20:52 m_anish we're discussing the TODO w.r.t dx2
20:53 alsroot m_anish: lets start from the 1st entry then..
20:53 bernie silbe: I'm running a build
20:53 tch alsroot: +1
20:53 m_anish alsroot, okay!
20:54 starting from "Display lease info in 'About my computer' and notifiy if lease is about to expire"
20:54 silbe bernie: great! please post it on dextrose@ and I'll try it out tomorrow.
20:54 bernie m_anish: seems easy
20:54 alsroot +1
20:54 m_anish is being worked upon, is a requirement for dx2
20:54 bernie silbe: k
20:54 tch +99 very useful
20:55 m_anish "OLPC Microformat compatible activity updater" is done, lazy /me needs to upstream it
20:55 upstream=dx/mainline
20:55 bernie m_anish: yes please
20:55 m_anish: is anyone working on the aslo side of it?
20:55 alsroot +1, but what about removing/or-not regular ASLO one?
20:55 m_anish "Yum based updater for sugar" is in dx-py builds being tested
20:55 alsroot ..what the current situation, (I'm personally not sure do we need to remove it)
20:55 bernie alsroot: maybe you had already written the server-side code? or someone else?
20:56 m_anish alsroot, ah yes! we need to figure that out...
20:56 bernie, alsroot i have a list of activities with me they want included in the build and another list for the XS
20:56 bernie alsroot: if we add microformat support to aslo, then we don't need the old format in sugar
20:56 alsroot bernie: server side code works on activities-testing
20:57 bernie alsroot: cool
20:57 m_anish bernie, will i need to host it on the pyedu servers because internet would be slow
20:57 alsroot bernie: but we need to improve current microformat patch to handle all installed activities then
20:57 tch m_anish: actually thats what they want..
20:57 m_anish alsroot, ?? the current patch handles existing+new activities
20:58 bernie alsroot: oh, I thought it was already that way
20:58 m_anish current=newly submitted :)
20:58 tch m_anish: because caacupe has a special optic fiber to Py edu offices
20:58 bernie alsroot: since we only have less than 300 activities, how about sending them all?
20:58 alsroot ok, then. /me was thinking it is only about collections
20:58 bernie alsroot: ah, I'm an idiot... it doesn't work like apt/yum...
20:58 alsroot bernie: ASLO can support "all" argument, it is not problem
20:59 bernie alsroot: ah ok.
20:59 m_anish: so I guess we need to make the url configurable via gconf?
21:00 m_anish bernie, okay, will make change before committing upstream
21:00 move to next item?
21:00 bernie +1
21:00 m_anish "Clear and complete error messages (even for crashes)"
21:01 alsroot thats the question...
21:01 tch yeah, we need to stop on this topic for a bit..
21:01 alsroot tch: in my mind it is pretty doable
21:02 simple scheme w/ uploading logs on actiivty fails and on sugar-session as well
21:02 bernie alsroot: I think all we need to do is catch exceptions in the shell (in the place where we currently log them) and bring up an alert
21:02 tch i think the 2 main features we need is to send backtraces from unhandled activities or even sugar itself.. AND all the information that could be related in case of a sugar crash..
21:04 mukul_afk: read this when you are back
21:04 alsroot tch: do we really need to handle crashes? maybe just start from python traces/logs
21:04 bernie: imho it will be to verbose and unexpected for regular users
21:05 tch alsroot: the problems that scares me the most are actually the one that crashes sugar.. because that is when users start to go into panic mode
21:05 bernie tch: in addition to reporting the bug, we also want to tell the user "oops, an error occurred", so they're not left there wondering why the file hasn't been copied (for example)
21:05 tch bernie: +1
21:06 alsroot tch: I guess, "Send an info" dialog after sugar crash won't hale much here:)
21:06 *help
21:06 m_anish alsroot, tch , bernie "oops, an error ocurred" doesn't give you a soln, but better than having nothing :)
21:07 silbe has quit IRC
21:07 alsroot m_anish: bernie: we can upload unhadled shell exceptions as well, not one-by-one, but at once (somehow)
21:07 bernie m_anish: we could tell them "open the Log activity for details"
21:08 alsroot thinks that alerts fro unhadled shell exceptions is too overkill for users
21:08 tch bernie:  hmm,
21:08 bernie alsroot: yes. were you thinking of resuscitating wadeb's patch?
21:08 alsroot bernie: similar, but more implicit (but configurable)
21:09 bernie tch: currently, things fail silently in case of any bug, which is very user-unfriendly...
21:09 tch: remember copying a file to a full usb stick?
21:09 alsroot ie, alerts and "Open Log activity" might be too much, but just a check box to send bug report...
21:09 bernie tch: the users think the file has been copied, but in fact there was an I/O error (which is nicely visible in the log)
21:10 tch i really dislike the alerts for this kind of situations,
21:10 bernie alsroot: if we already have the bug reporting feature, then yeah. but I thought that was too much work for dx2
21:10 tch and "open the log" relies on the wrong asumption that the user are able to understand the logs
21:10 bernie tch: well, who wouldn't? bugs shouldn't exist!
21:10 alsroot bernie: +1 for fails like "cannot copy", when I was talking about implicit reports I was thinking only about unhandled exceptions
21:10 bernie tch: but they *do* exist and so we have to tell the users somehow
21:11 alsroot bernie: I'm sure from tech pov, it is pretty doable for dx-2
21:11 tch that was one of the motivations for my messages notification system
21:11 bernie alsroot: well, all unhandled exceptions happen on some operation that the user wanted to do, and failed
21:11 m_anish bernie, alsroot how about just having implicit bug reporting, and having UI modifications go in later through the yum updater.
21:11 bernie alsroot: here's another case: user tried to connect to an AP with an accent in the essid. it fails silently, but there's a nice stack trace in the log.
21:12 m_anish feels UI needs design discussion
21:12 alsroot bernie: yup, we need to handle as much as possible, and for the rest, send reports
21:12 bernie m_anish: the error dialog was discussed months ago on sugar-devel, I'm pretty sure
21:12 alsroot: yes, when we don't handle it's a bug in sugar.
21:13 alsroot: but would you keep it quiet to the user? nothing on screen at all?
21:13 alsroot: it confuses even developers like me :-)
21:13 tch bernie: if we can use my messages notification to let the users something went wrong + send all relevant logs + a error dialog should be fine (?)
21:14 m_anish in eben's words from m-l thread " I'm in favor of showing an error message for an uncaught exception (since
21:14 feedback is good!), though I caution against displaying the actual
21:14 exception string in this instance. It's unlikely to make sense to the
21:14 primary audience. Instead, I would generate some friendly message that
21:14 lets them know that the action they attempted didn't work as expected, and
21:14 alsroot tch: yup, it might be an option
21:14 m_anish then offering a link into the Log activity to learn more should they want
21:14 bernie tch: if we already have that... sure. but let's not block this on the notification system or the bug reporting system. it could be 3 lines of code.
21:14 m_anish to dig deeper."
21:15 bernie m_anish: I agree 100% with eben
21:15 tch bernie: the notification thing done and devel a error dialog should not take much
21:16 bernie tch: are they going to make it for dx2? it has to be out in less than 2 months.
21:16 tch i also do agree with eben's statement
21:16 alsroot so, do we have a consensus? "For unhandled exceptions, send a notification w/ "Send report" button"
21:16 m_anish tch, alsroot bernie everything seems to fit... flash the notification icon, see a simple error, link to log
21:16 +send report
21:16 +1 from me
21:16 bernie tch: plus, if we do the basic thing of just putting up an error message, upstream will accept it without complaining
21:17 tch we can always change the way we display the "friendly error message" ;)
21:18 bernie tch: sure, once it's in, it would be easy to extend / improve.
21:18 tch but the way we catch we unhandled exception + collect the logs + use a error dialog should not change
21:18 alsroot tch: for sending errors, I have two python snippets for cli and srv for bug reporting, ping me later..
21:18 bernie tch: err, you're right on this. probably abrt needs a completely different method. I have no idea how it does it
21:19 tch alsroot: i will :)
21:19 bernie oh, but we decided not to use abrt, right?
21:19 alsroot: cool
21:19 alsroot bernie: "but we decided not to use abrt" I'm +1 for that (at least for the first time)
21:20 tch +1 to go for simplicity for now..
21:20 bernie alsroot: curious to see your client code too
21:20 tch me too
21:21 m_anish alsroot, tch can you keep mukesh_dce in the loop on this, he'd be happy to work on it :)
21:21 alsroot it just HTTP POST code..
21:21 srv: http://pastebin.com/0yss13P4 cli: http://pastebin.com/LtSSJy6i
21:21 tch i pinged him at the beginning ;)
21:21 so he will read it when hes back
21:22 m_anish tch, i think u pinged mukul (not mukesh)
21:22 tch ops!!
21:23 m_anish next ... "Checks to prevent installation of incompatible activities"
21:23 alsroot tch: also about bugreporing, you can look to wadeb patches, http://bugs.sugarlabs.org/ticket/1439
21:23 m_anish tch, we can send him link to log
21:25 alsroot ok, I guess we are including this feature to dx-2, and going further?
21:25 tch alsroot: ok, i will give it a look too
21:25 m_anish alsroot, yeah, +1
21:25 tch alsroot: yep, we should try at least will be extremely useful to have this feedback
21:26 bernie I have a build. now testing
21:26 tch bernie: does it include last things like notif + updater?
21:26 + some bug fixes
21:26 bernie tch: I checked out code from git, but there were no patches
21:27 tch: I did not merge anything myself
21:27 alsroot m_anish: what the next entry in TODO..
21:27 m_anish bernie, hmm, do you want me to push the changes?
21:27 tch i see, maybe m_anish can send you what we have added lately locally
21:27 ?
21:27 bernie tch, m_anish: how about you resend me the patches you'd want me to include in the next build?
21:27 m_anish alsroot, Checks to prevent installation of incompatible activities
21:27 tch +1
21:27 m_anish bernie, i have them uploaded smwhr, un momente
21:28 bernie tch, m_anish: or, if you've already added them to your local repo, just do a push
21:28 tch momento* :)
21:28 m_anish bernie, there in my local repo yep.
21:28 tch jaja +5 for anish spanish skills
21:28 alsroot is "incompatible" about sugar version, we already have sugar ranges on ASLO
21:28 tch but does sugar block those activities from being installed?
21:29 or throw a warning?
21:29 m_anish bernie, http://people.sugarlabs.org/an[…]xo-1-builds.patch, http://people.sugarlabs.org/an[…]integration.patch , http://people.sugarlabs.org/an[…]repositorie.patch
21:29 alsroot tch: nope
21:29 tch should do it?
21:29 bernie m_anish: didn't you already have dextrose rpms built with these patches applied?
21:29 alsroot do we need this feature? I can implement a mimic for furture sweets' deps "requres" tag
21:29 bernie alsroot: +1, I'd postpone this to the next dextrose
21:30 m_anish bernie, you mean sugar rpm, yeah
21:30 bernie alsroot: checks on aslo aren't sufficient btw, because users are too clever about copying xo bundles on usb sticks and trying to load them on the wrong version of sugar. not having an error message is our fault.
21:31 alsroot bernie: if it is real problem in the field, mimic impl is not a probelm for dx-2..
21:32 bernie: ok, I'll add simple "requires" tag, it will look like: "requires = sugar > 0.84"
21:32 bernie alsroot: yes, it was a big problem. users were blaming our new version if some old activity did not work on it. from their point of view, it looked like the new version of sugar doesn't work as well as the old one
21:32 tch and what sugar do with that? block or warning while installing?
21:32 bernie alsroot: cool that's great
21:33 alsroot tch: yup, it might an alert
21:33 bernie tch: just show an error like: "The bundle FooBar cannot be installed on this version of Sugar"
21:33 alsroot bernie: and for launching as well, I guess
21:33 bernie tch: if you want to be even nicer, you could add: "Please check $activity_website for an updated version"
21:34 tch yeah we need to cover both cases since they can just copy files
21:34 i like to try to be nice... hahaa
21:34 bernie alsroot: yup
21:35 alsroot ok, /me adds himslef as an owner for this feature
21:35 tch :)
21:36 m_anish alsroot, so this goes into dx2, right!
21:36 alsroot m_anish: yup
21:36 m_anish alsroot, great!
21:36 "Make activities work in Gnome without Sugar (controversial)" and "Make Gnome applications work seamlessly within Sugar"
21:37 both aren't dx2 goals IMHO
21:37 alsroot sounds too undoable for dx-2
21:37 tch we can put all gnome interaction aside for dx2
21:37 +1
21:37 m_anish "Webkit integration (lucian)"
21:37 doesn't know its status
21:37 tch either i...
21:37 bernie m_anish: are the rpms with those patches of yours already in the dextrose repository?
21:38 m_anish bernie,http://download.sugarlabs.org/[…]/rpms-py/i386/os/
21:38 dextrose-updater dextrose-updater-gui(tch) and latest sugar
21:38 tch we can discard that gui,
21:39 bernie m_anish: but they are only in the -py repo, right? do we want these in the intl builds as well?
21:39 tch not that specific rpms but yes the features it includes
21:40 m_anish bernie, only in py repo, we could move these to intl repo if no one has a prob with it :)
21:40 alsroot so, what about "Webkit integration", afaik it is not yet released (Browse), thus, sounds too early for dx-2
21:41 m_anish alsroot, yeah, i'll try an ping him abt his plans though.
21:41 tch lets move on then
21:41 :)
21:41 m_anish "Faster activity startup / memory savings (quozl)"
21:41 alsroot doesn't have any ideas about that
21:42 m_anish quozl had done 'something' on this front, but i don't think it ever took patch form on m-l
21:42 "Integration with social networks"
21:42 not for dx2 for me :)
21:42 tch controversy alert!
21:43 haha
21:43 alsroot +1
21:43 m_anish "Bidirectional Journal gateway for Gnome"
21:43 tch yeah, not for dx2
21:43 alsroot "Feature-complete packaging" will be released (I hope), in time w/ 0.92
21:44 so, next entry
21:44 m_anish OS: "Panic key" to restore default settings for GNOME and Sugar
21:45 alsroot that this feature about?
21:45 *what
21:45 m_anish has no idea :)
21:45 bernie, tch ^^
21:45 tch i think it started when kids accessed to gnome and modified manually sugar config files
21:46 so, we wanted to have some "reset to default" buttons..
21:46 but ..
21:46 since we are no using gnome in caacupe.. we could leave it for the future (?)
21:46 bernie m_anish, tch, SMParrish: are you guys ok if I continue the numbering from where we left with dextrose1? i.e. os430dx2
21:46 m_anish tch, yeah, i was told no gnome for the build i have to give
21:47 not for dx2 then from me
21:47 bernie someone in an olpcnews comment expressed that changing the naming scheme was confusing to them
21:47 m_anish: yes, I'm only doing nognome builds for now
21:47 m_anish bernie, np
21:47 bernie m_anish: I have not much intestest for gnome, but if a deployment wants it we can do it
21:48 tch yeah, so next :)
21:48 m_anish <rant>bernie, gnome for me runs insanely slow on xo1</rant>
21:48 Sugar: "Panic key" to force regenerating datastore index (hold both shifts on boot? hidden ctrl-r hotkey in the journal?) is done...
21:48 "Rainbow?"
21:49 and "Multiple item selection in the Journal"
21:50 tch, bernie re: ^, was it feedback from py?
21:50 alsroot maybe instead of panic key, having CP entry to restart sugar w/ clear ds index
21:50 bernie m_anish: at times, rainbow seems to be getting closer... but I'm not sure *when* it will be ready
21:50 tch alsroot: don't we already have that?
21:50 alsroot: i mean in dextrose..
21:50 bernie: ^
21:50 bernie m_anish: I spoke with m_stone about it last week. he seems to be waiting that the sugar developers do something to revive it, but I'm not sure what this something should be
21:51 m_anish tch, alsroot we have that IIRC
21:51 tch checks..
21:51 bernie alsroot: we have a menu item on the journal icon, because we couldn't find a better place for it.
21:51 alsroot ok, we can exclude this entry then..
21:51 bernie alsroot: we felt that adding one more control panel for the journal would have been overkill
21:52 tch ack, its in the journal icon at the top frame bar...
21:52 in the journal icon palette.
21:52 bernie m_anish: the things we're not doing for dx2, can you please push them to a TODO/Dextrose3 page?
21:52 m_anish bernie, hmm, so re: py, this is -1,
21:52 bernie, sure, will do
21:53 alsroot multiple selection might be too late for dx2.. (keeping in mind testing and getting feedback)
21:53 m_anish "Updater similar to puppet (plain ceibal)" -- we are using yum updater so not for dx2
21:53 bernie m_anish: I think we can cancel this feature and use rpm post-install scriptlets for  the same purpose
21:53 m_anish "3G connection sharing in Sugar" is already done tch ?
21:53 tch yup, done and working
21:53 m_anish bernie, yup
21:53 tch if you don't trust me call yo my mom, hes using it rofl..
21:53 to*
21:53 m_anish "Automatically register to Schoolserver"
21:53 tch shes*
21:53 hahaha
21:54 thats an interesting feature..
21:54 bernie tch: at OLPC-SF, dsd seemed to think that sharing over ad-hoc networks wasn't possible for some reason
21:54 tch it is
21:54 i can tell because i did it and i have been using it :)
21:54 bernie tch: he also thought that ad-hoc could not scale to more than just a 3-4 laptops, which I'm pretty sure is wrong
21:55 tch well, that another issue
21:55 anywyas 3G modems does not provide huge bandwidth in 3rd world..
21:55 so in practical terms does not matter
21:56 bernie m_anish: when you roll new py builds, could you please upload them here?
21:56 http://download.sugarlabs.org/[…]esting/dxo2/*/py/
21:56 m_anish bernie, will try from beertarium... not possible from office :)
21:56 bernie m_anish: well, you don't have to do it for every new build, because it would take you too long... but it would be good to archive significant builds
21:57 m_anish bernie, yep, for eg, starting from the build i hand them over on this tue
21:57 bernie m_anish: when the b/w at the office sucks, maybe ask me to check and I could find out who's hogging the net. sure there are people who like torrenting
21:58 m_anish bernie, okay
21:58 tch just blame on carlos by default
21:58 haha
21:58 bernie haha
21:58 m_anish haha
21:58 bernie deme
21:58 tch statistics never lie!
21:59 m_anish lol :)
21:59 tch oh.. something funny.. can you block his mmorpg tpc/udp port? lol
21:59 tcp*
21:59 bernie tch: lol
21:59 tch just kidding
22:00 m_anish tch, "Automatically register to Schoolserver" is -1? what abt "Make Schoolserver visible in neighbor view"
22:00 tch ;);)
22:00 bernie m_anish: too much work, I think, for dx2
22:00 m_anish: but it would be AMAZINGLY useful
22:00 tch i like it.. but i have no idea how much effort could take
22:00 bernie m_anish: if we disregard the security problems, we could do it from a cron job
22:01 tch yeah, we just need someone to config his spy XS and wait for kids to connect to it..
22:01 m_anish bernie, we can always roll patches in via the updater! lets leave for now?
22:01 bernie m_anish: dsd might have ideas on how to do the registration securely and non-interactively, but I fear the complexity
22:01 m_anish: ok
22:01 tch lets move then,
22:01 m_anish one table down, two to go
22:02 "kill the name picker that pops up when quitting activities (EASY)"
22:02 tch i don't see any benefit of seeing the xs there.. anyone else?
22:02 m_anish tch, u had a patch for this?
22:02 tch yup
22:02 bernie m_anish: the reason why this is important is that many users don't register manually, and thus they loose backups and collaboration
22:02 tch i just added a gconf value.. so you can disable it.. and a few lines of code to use it..
22:02 bernie m_anish: oh yeah let's kill that crap!
22:02 m_anish: it won't be missed
22:02 m_anish bernie, so this should be high pri feature for dx-3 or 'dx2-updates' ;)
22:03 bernie tch: is the patch in dx already?
22:03 tch if anish added my patch, yes.
22:03 ;)
22:03 bernie haha, now I can do builds in just 7m48s
22:03 m_anish tch, oh, right! will do
22:03 ""Figure out why so many control panels require restarting Sugar, and fix them not to (MEDIUM)""
22:04 tch that is a huge of work.. because we need to change the whole CP options logic..
22:04 m_anish not probable for dx2, btw, have a dev from india who might be interested in working on it
22:04 tch m_anish: +1
22:05 m_anish next ... "kill the "Mute" function on the volume icon in the frame (EASY)" i remember a patch for this on m-l but probably not dx/mainline
22:06 tch, did you work on this? ^
22:07 tch nope, it was a guy from india i think
22:07 alsroot anyway it should be trivial to implement, if it is ok from user experience pov
22:07 tch i think everyone barked when they saw the patch,
22:08 alsroot: +1
22:08 personally i don't care.. seems to be too trivial to not move on to the next topic haha
22:09 m_anish "Notification system (tch)" is in build and probably going to dx/mainline real soon...
22:09 "Automated bug reporting tool" already discussed ;)
22:10 tch lets cross fingers.. ;)
22:10 bernie m_anish: today I spoke with gary on #sugar. he didn't seem to like the notification system at all
22:10 m_anish 2 tables down 1 to go
22:10 bernie m_anish: if it doesn't go upstream, it can still go in dextrose and maybe olpc
22:10 m_anish bernie, any particular concerns?
22:10 bernie tch: is it ready?
22:10 tch bernie: i actually implemented based on all his ideas
22:10 bernie: it is alreayd working
22:10 bernie m_anish: he said he thinks it makes the UI more complex
22:11 tch bernie: it is not exactly what he designed , but if you check his design you would understand
22:11 bernie m_anish: while I do agree on keeping things *simple*, it doesn't mean that we should not provide feedback to the user for important events. that's not simpler, it's harder.
22:11 m_anish bernie, gary was working on a new HIG rev right?
22:12 tch bernie: just check what i have already implemented
22:12 bernie tch: where can his design be reviewed?
22:12 tch bernie: it was on the list
22:12 bernie: i really like his idea..
22:12 bernie m_anish: yes, I think so. he's a very good guy... just a little too much of an interface nazi :)
22:13 tch bernie: is just too complex to implement now.. and i just want to display text messages!!!
22:13 bernie: but my implementation is a realistic version of his idea
22:13 bernie m_anish: for an explanation of what I mean, google for "gnome interface nazis" :-)
22:13 m_anish bernie, :) , maybe when the updated HIG is out, we'd actually know how to design to get upstreamed easily instead of falling into endless discussions
22:13 bernie tch: well, if the dbus side is compatible, why not?
22:13 tch it is. ;)
22:14 bernie tch: (why not adopt your version now)
22:14 tch not clue..
22:14 just test my notif and let me know what you think
22:14 i tried to take into account ALL the comments from the ml
22:14 bernie tch: where's the patch?
22:14 tch on dx ml
22:15 m_anish bernie, should be in the sugar rpm on the rpms-py repo
22:15 tch are 2 patches, the initial one and some enhancement patch
22:15 bernie tch: I know you've done a good job interfacing with upstream lately. it's that upstream is very hard to work with
22:15 m_anish tch, ^ (the patch you sent me was the latest)
22:15 ?
22:15 tch yup
22:16 the current implementation is really similar to how the current notifications works..
22:16 bernie m_anish: ah, ok. but not in dextrose's git? do you have push access on dextrose? I don't mind you pushing your patches because I know you're one of those rare developers who test their shit before giving it to others
22:16 m_anish: otoh, I'd rather not bypass peer reviews
22:16 m_anish bernie, http://download.sugarlabs.org/[…]o.fc13.noarch.rpm has tch notification sys
22:16 tch check the ml for usage examples :)
22:16 bernie m_anish: yes, but I'm afraid the default dextrose builds fetch from the intl repo only.
22:17 tch or just use the update calls
22:17 m_anish bernie, i do have push access, i'll do as you say :)
22:18 bernie m_anish: I think stuff should go in the py repo only for 2 reasons: (1) to avoid slowing down paraguay when upstream dextrose is blocked for some reason and (2) for stuff that only paraguay wants (but at this time, hopefully we shouldn't have any of those)
22:18 m_anish: oh, and (3) as a warranty for the future that paraguay will always be able to make changes to their builds (but this will be another repo in the release dir, not testing).
22:19 m_anish bernie, if I have freedom to commit tested patches upstream, i have no problems in doing it
22:20 bernie m_anish: yes, I prefer it this way because nowadays I'm always busy at the fsf while you guys work, so I'd be a bottleneck
22:21 m_anish: normally, it's saner if there's a guy acting as a maintainer in the middle, but we can't have it now
22:21 tch m_anish: you can always ask me for a second opinion :)
22:21 m_anish tch, of course!
22:22 bernie, okay i'll start committing changes upstream
22:22 next on the list ... ""Power has only 2 options, they could go into a context menu on the battery icon in the frame""
22:22 tch not a priority imho
22:23 alsroot +1, would be not bad to leave it as is
22:23 bernie m_anish: yeah, dx3
22:24 m_anish okay
22:24 "Backup/restore -> journal (or a schoolserver icon in the mesh view)"
22:24 bernie alsroot: it would be nice, long term, to reduce the number of items in the control panel. now it scrolls and it's pretty unfriendly
22:25 alsroot: inexperienced users can't find the hidden icons, and the experienced ones don't want to go such a long way just to toggle power management
22:26 tch that would be a good reason to add XS icon in the mesh view, but i don't think it is priority either
22:26 m_anish tch, okay, moving on...
22:26 "touchpad -> device icon on the frame + global keyboard shortcut to switch between resistive and capacitive"
22:26 wasn
22:26 bernie dx2 really should be small
22:27 m_anish wasn't this done long ago?
22:27 tch i already sent a patch for the hotkeys part
22:27 only the hotkeys part left, which i already done
22:27 m_anish walter's patch for the frame icon?
22:27 tch, in dx/mainline?
22:28 tch walter's patch is already in
22:28 in dx
22:28 m_anish tch, yeah, and yours?
22:28 tch i just sent a patch to have the globalkeys to switch touchpad mode
22:28 m_anish tch, okay... next topic...
22:28 "accessibility -> fold into keyboard"
22:28 tch not sure what it means
22:29 any idea, someone
22:29 ?
22:30 m_anish tch, no idea
22:30 bernie tch: oh, we already have the notif system!
22:31 tch bernie: told ya ;)
22:31 m_anish bernie, any idea abt ^^
22:31 alsroot, ^^
22:31 alsroot m_anish: I guess we can leave "fold into keyboard" feature for dx3 :)
22:31 tch i am still not sure what it means
22:31 lol
22:32 bernie m_anish: we have to add one more todo for dx2:
22:32 m_anish just wants to be sure that this wasn't py feedback/requirement
22:32 bernie "change color scheme of boot/shutdown splash screens"
22:32 tch but does anyone actually understand what it means?
22:33 bernie m_anish: you choose the colors. maybe use india...
22:33 m_anish bernie, ok :-)
22:33 bernie m_anish: orange and green look good
22:33 m_anish: the uruguay flag colors are not that nice for a sugar logo (light blue and white)
22:34 tch bernie:  is just a matter of taste
22:34 haha
22:34 m_anish tch, bernie, :-)
22:34 "modem config -> context menu on the modem icon in the frame" is next. I'll check the prev item with roberto
22:35 tch i think it relies on the "not priority" find..
22:35 field*
22:35 bernie tch: but the bg is also white... you see, it's really just one color :)
22:35 tch: korose!
22:35 tch bernie: deme!
22:35 m_anish ^^ prob. not dx2!
22:36 bernie tch: I feel like I'm doing a new step in japanese... now I understand some adverbs and more verbs
22:36 m_anish bernie, tch, s/japanese/english/ :P:P
22:36 bernie tch: even some conjunctions. like, "however" or "that can't be helped"
22:36 shikashi...
22:37 tch bernie:  lol.. yeah i recognise some of those sometimes
22:37 m_anish (here we go again...)
22:37 bernie m_anish: yameru!
22:37 tch m_anish: this is non-sense!
22:37 m_anish translate.carlos.py isn't around, sorry
22:37 bernie m_anish: they say "yameru!" when they mean "stop that!" or "enough!"
22:38 tch: impossible!!
22:38 tch checking http://users.tmok.com/~tumble/amfaqs/101glos.html
22:38 alsroot basta
22:38 lets talk about TODO
22:38 bernie m_anish: ah yeah, carlos is the best otaku
22:38 tch m_anish: chigau!
22:38 m_anish bernie, just two more """Register Again creates duplicate xs icons on the volumes toolbar""" and """Journal sort by creation date - journal entries are being created with ctime=0 (41 years ago)"""
22:39 bernie tch: what's chigau?
22:39 m_anish: I thought this was done?
22:39 m_anish both seem like bugs, /me dunno if fixed?
22:39 tch i fixed the duplication thing months ago
22:39 bernie m_anish: for the second one, we need to refresh the journal patches, I guess. the version that went upstream was slightly better
22:40 alsroot "sort by creation" should work fine, will recheck ds and journal code
22:40 bernie yeah, I wonder why they are in the todo list
22:40 m_anish bernie, ah!, will do!
22:40 bernie maybe recheck the register again bug
22:40 I can't from here
22:40 tch i dont have an xs here
22:40 m_anish bernie, i think silbe has a list of such patches, which are different b/w dx and upstream
22:41 thinks we need a acronym for 'upstream' (so many chars)
22:42 tch utrm?
22:42 m_anish bernie, tch alsroot so these two are done?
22:42 alsroot I guess everybody agree if m_anish will be an "maintainer in the middle" as bernie said, to test push tested patches to dx git
22:43 m_anish tch, ugly :( maybe translate upstream to japanese and then try
22:43 tch in japanese would be upurusteruamo
22:43 way longer !
22:44 alsroot people will email/review patches on dextrose@ and m_anish will apply/test/build them for testing purposes
22:44 tch alsroot: +1, *2, ^2
22:45 alsroot m_anish: could you sort TODO-temp as well :), btw maybe having Dextrose/2 and Dextrose/2/Roadmap on wiki, like w/ sucrose
22:46 m_anish alsroot, ya, will do
22:46 tch yatta!!!
22:46 m_anish and Dextrose-3/TODO
22:47 tch btw.. anything else you guys would like me to ask/do in Ceibal offices?
22:47 m_anish tch, till when are you thr?
22:47 tch m_anish: for sure, the full next week
22:48 m_anish tch, just interested to know if they
22:48 're keeping a todo list
22:48 :)
22:48 alsroot m_anish: ok, will send my patches to dextrose@ for reviewing and keeping in mind that you will apply them to git repo and create rpms
22:49 tch m_anish: would be great to have the priority list to show them.. that might regain interest to work closely with us
22:49 m_anish i mean, feature requests, etc.
22:49 alsroot, +1
22:49 alsroot m_anish: btw wiki might look like http://wiki.sugarlabs.org/go/0.90
22:53 tch well, time to me go afk,
22:53 m_anish alsroot, dx/TODO-Dx3  dx/TODO-Dx2-py and dx/Roadmap-py would be the new/modded pages
22:53 tch great meeting :)
22:54 m_anish tch, bernie alsroot thanks a ton!!! for the meeting :)
22:54 alsroot back to obs patch, a bit of work before going to bad...
22:55 m_anish goes outside to enjoy weather :)
22:59 bernie alsroot: +1 to make m_anish our middleman!
23:05 m_anish: did I gave you my "uploadrpm" script?
23:08 m_anish bernie, yep
23:17 dfarning tch, for the big picture we want to _show_ other deployments that onsite of m_anish/martin + upstream team of silbe/alsroot is the most effective way to help them solve their problems so that they will hire an onsite team from us.
23:17 satellit_afk is now known as satellit_
23:17 dfarning onsite team.
23:18 part two will be showing that multiple onsite teams (at different deployments) working together is more effective than working alone.
23:20 but, I don't really expect them to 'get it' quit yet.  because it seems faster in the short run to use quick hacks to fix bugs.
23:21 this leason of upstreaming and working together is usually not learned _until_ a group or company has tried to update to a new version and sees how difficult it is.
23:22 10 years ago every company that work with linux on embedded devices -  used custom versions of the kernel.
23:23 now, all the major embedded company stick very close to upstream.
23:24 tch dfarning: +1
23:24 dfarning I expect the next deployment to join dextrose will the olpc.au.  Their CTO is very sharp and understands the collaborative development process.
23:25 tch, nice work today.  you guys got a lot of important stuff done.
23:27 tch the really valuable thing in my eyes is that you all were asking eachother 'do we have the resources to do that?' whenever someone suggested doing something cool.
23:33 tch :D
23:50 m_anish tch, bernie plan to watch a anime video tonight, have a link??
23:52 tch http://www.animefreak.tv/watch[…]ubbed-online-free
23:52 watch it subbed
23:53 bernie m_anish: yes, take this very weird one: http://people.sugarlabs.org/be[…]untain_sleeps.avi
23:54 m_anish bernie, tch i'll watch only one, which should I pick from the two?
23:54 tch if it takes too long to download watch it online, using the link i just sent
23:54 bernie m_anish: careful, this one isn't a light novel... it's for true lovers of the japanese culture.
23:54 tch both are the same
23:54 hah
23:54 bernie m_anish: ah, both are the same thing ;)
23:55 tch m_anish: everyone starts with one ;)
23:55 m_anish :)
23:55 bernie m_anish: I could also upload another one which is targeted at a wider public
23:55 m_anish: btw, are you sharing the apt with scs? you should ask him to show you a bunch of his japanese movies. careful!
23:55 tch m_anish: and then... sometimes... they become: technoRogy Niiinnjas!
23:56 bernie tch: suburashi chinchin mono!
23:56 ye! ninja, kimazu!!
23:56 tch lol..
23:57 but beware.. there is no way back...
23:57 m_anish bernie, wget ting

 « Previous day | Index | Today | Next day »     Channels | Search | Join

Powered by ilbot/Modified.