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

#sugar-meeting meeting, 2012-03-06 15:03:06

Minutes | Index | Today     Channels | Search | Join

All times shown according to UTC.

Time Nick Message
15:03 meeting Meeting started Tue Mar  6 15:03:06 2012 UTC. The chair is erikos. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:03 Useful Commands: #action #agreed #help #info #idea #link #topic #endmeeting
15:03 erikos hi, who is around for the dev meeting?
15:03 silbe waves
15:03 benzea hello
15:03 gonzalo hello
15:03 manuq hello!
15:03 ajay hello
15:03 JT4sugar Hello
15:03 garycmartin waves
15:04 satellit_ listening
15:04 erikos great!
15:05 #topic status of the GTK3 port
15:05 - there is a bug in the screenshot taking patch, we are currently trying to fix
15:05 manuq yes
15:05 gonzalo good attendance today :)
15:06 cjl waves late
15:07 erikos http://lists.sugarlabs.org/arc[…]ruary/035637.html
15:07 this is the patch ^
15:08 benzea oh, long thread
15:08 hasn't read it
15:08 erikos when you switch the activity (go to another one, or go to the Journal)
15:08 instead of directly closing
15:08 silbe #link https://patchwork.sugarlabs.org/patch/1178/
15:08 erikos not the activity window is used
15:09 for the screenshot, instead the current window is used
15:09 if someone knows why, feel free to comment
15:09 now, or after the meeting
15:10 benzea hm, weird, though doesn't gtk+ now have a snapshot function itself?
15:10 erikos benzea: not anymore in gtk3
15:10 manuq benzea: is not in the GTK+ introspection
15:10 erikos benzea: and using cairo was the recommendation from company
15:11 benzea hm, well, you can't get a screenshot of an invisible window
15:12 so we need to render the content into a pixmap, instead of just grabbing the pixel data maybe
15:12 (what company suggested works great if you have a compositor ...)
15:12 gonzalo benzea, gnome screenshot is using Gdk.pixbuf_get_from_window to do the screenshot
15:12 http://git.gnome.org/browse/gn[…]shot-utils.c#n396
15:13 but the window isvisible
15:13 benzea gonzalo: yeah, but it wants a screenshot of a window that is visible
15:13 manuq benzea: I did a workaround using GtkOffscreenWindow
15:13 it works
15:13 benzea manuq: maybe one could simply call the draw() routine using a different target?
15:13 manuq but I'm afraid it can be slower and the widget relocation may do a mess
15:14 benzea I mean there are pdf screenshot tools ...
15:14 manuq benzea: I based my code in a pdf screenshot tool
15:14 gonzalo benzea, i think they use the offscreenwindow
15:15 benzea manuq: ie. simply calling gtk.Widget.draw() using a cr that is an offscreen pixmap (or even an image surface)
15:15 manuq: ah, ok
15:16 manuq benzea: ok I will test
15:16 benzea: yes I have not tested that
15:16 benzea manuq: of course, draw() may not work, if an activity draws directly to the window; but in that case the activity needs to override the preview routine
15:16 erikos benzea: you were saying, that we can not get the screenshot of a window not visible, but where does it get the information from of the current window
15:17 benzea erikos: hm?
15:17 manuq for the record, I was checking this Perl code: https://raw.github.com/potyl/W[…]ter/screenshot.pl
15:18 erikos benzea: I mean, the gdk.window we get with "self.canvas.get_window()" is the currently visible window?
15:19 benzea manuq: yeah, it just does a widget->draw() call
15:19 manuq benzea: exactly
15:19 benzea: my code too, pasting...
15:19 benzea erikos: well, it is the correct window
15:20 gonzalo has quit IRC
15:20 manuq benzea: http://fpaste.org/7Koo/
15:20 benzea erikos: but there is no pixel data stored for the obscured window
15:21 erikos benzea: but shouldn't it be black then?
15:21 benzea erikos: I guess it is just optimized to look into the same buffer
15:22 erikos benzea: hmm, ok
15:22 garycmartin Isn't the canvas thumbnail saved before the window goes away (e.g. as you switch/stop)? Probably missing something obvious...
15:22 benzea manuq: the thing is, that there is no need to even create the offscreen window and reparent the widgets
15:22 manuq benzea: that I haven't tested, I will
15:23 after the meeting
15:23 benzea yeah, lets move an
15:23 erikos garycmartin: if you go to another activity, the window gets shown first
15:23 garycmartin: should be async in the background the saving itself
15:24 garycmartin: otherwise you would need to block
15:24 benzea: thanks for those infos!
15:25 manuq yes, thanks benzea
15:25 benzea manuq: I'll write a small example
15:25 erikos benzea: I am currently doing the drawing for the borders of the palette
15:25 manuq benzea: cool
15:25 erikos benzea: will post to the ml hopefully soon
15:25 if you have any comments will be appreciated ;p
15:25 benzea erikos: I hope that it doesn't make much trouble :-)
15:26 erikos benzea: well, ....
15:26 benzea: the new code has it's downsides with the split in widget and window, palette
15:26 benzea: but I think I figured it out now
15:27 benzea: have to clean it up
15:27 benzea erikos: yeah, question is which widget draws the border
15:28 erikos benzea: right, I do draw the palette border part in PaletteWindowWidget
15:28 benzea yup, sounds sane
15:29 erikos benzea: but I have to get the max-width information from the Palette
15:29 otherwise when the palette does build up slowly you get something like"
15:29 [back]
15:30 [link to website]
15:30 [another link to website]
15:30 instead of
15:30 [back         ]
15:30 [link to website]
15:30 hopes his approx drawing is readable ;p
15:31 manuq erikos: it is :P
15:31 benzea yeah, the width is not supposed to change when the secondary palette opens
15:31 erikos benzea: right
15:32 benzea: the problem is that the palette code does a show-calculate-width-hide
15:32 before the title is drawn
15:32 and that info I have to upstream to the PaletteWindowWidget
15:33 benzea not sure how the whole thing works exactly
15:33 true, it needs to be passed up somehow
15:33 erikos benzea: in total, it is s mess ;p
15:33 benzea erikos: it has always been :-)
15:33 manuq: http://fpaste.org/qUv5/
15:34 erikos benzea: yeah, the upside is, that upstream is interested in a new gtkmenu
15:34 benzea manuq: pretty much only your code, with the reparenting removed
15:34 erikos: that is great to hear
15:34 erikos benzea: if we work with them, we can raise our needs and design the thing right for us from the start
15:34 manuq test
15:35 erikos benzea: yup
15:35 benzea erikos: yeah, the whole thing with the correct width will still require some hacks I expect, but just having one window widget would be *so* nice
15:36 erikos stops dreaming
15:37 cool, anyone else with updates from the gtk3 world?
15:37 manuq not me, I hope I can finish the PDF inline and get into the theme again
15:37 erikos manuq: yes, good point
15:38 #topic current status of the development cycle
15:38 manuq so, turning my engines on now!
15:39 erikos I released a few tarballs last week
15:39 hold off the toolkit-gtk3 one because I wanted to land the palette/preview stuff
15:39 I hope that will be done soon now
15:40 when packaging I got pinged from a Fedora dev about cjson
15:40 see http://bugs.sugarlabs.org/ticket/3142
15:41 ahh, nice manuq did some testing!
15:41 http://bugs.sugarlabs.org/ticket/3142#comment:4
15:41 manuq erikos: so yes, I did some
15:42 erikos reads
15:45 manuq: in 2.6 simplejson looks quite a bit faster
15:45 manuq erikos: cool
15:45 erikos: ah simplejson you mean
15:45 erikos manuq: from your test, yes
15:45 manuq yes, a friend of mine helped testing
15:45 silbe erikos, manuq: I've seen your work on that bug (thanks!). As long as nobody shows that JSON parsing performance matters in some use case (something I doubt, given where and how we use it), we should prefer the module shipped with Python 2.6+ and only fall back on something else for older systems if we still support Python 2.5 (I wouldn't be surprised if some 2.6ism already crept in).
15:46 manuq silbe: +1
15:46 erikos silbe: yeah, maybe it is ok for sugar 0.96 upwords to only work on 2.6+
15:47 even Debian should ship 2.6 by now...
15:47 couldn't resist
15:47 loves Debian
15:47 silbe erikos: it's even in Debian sta[b]le by now ;)
15:47 erikos silbe: :)
15:48 so, I would say: 2.6+ as a Sugar dep
15:49 manuq I am sorry, I got to leave for half an hour :(
15:49 erikos and we just use the built-in json
15:49 manuq: no worries, thanks for your attendance!
15:49 manuq erikos: thanks!  I will check benzea snippet when I'm back
15:50 erikos manuq: nice!
15:51 silbe: agreed on that?
15:51 others?
15:51 waits for the Python old-skoolers
15:52 silbe erikos: +1. It's easy enough to fix up if it turns out some distro has all other dependencies (esp. recent Telepathy), but not Python 2.6.
15:53 garycmartin 2.6+ dep sounds reasonable (fwiw OSX Lion, where I tinker most, is at 2.7.1)
15:54 erikos garycmartin: great, thanks for that info
15:55 silbe: do you want to port the datastore?
15:56 anything else from someone?
15:56 silbe erikos: to Python 2.6 json? Sure, can do that. I should even have a patch floating around already, not sure why I didn't send it to sugar-devel yet...
15:56 erikos silbe: yes 2.6 json - great
15:57 JT4sugar Few quick questions about my proposal if ok
15:57 erikos JT4sugar: sure, please go ahead
15:57 JT4sugar Know you guys are in 0.96 and this would be 0.98
15:58 Have had few email suggestions for these to be possible GSOC projects need to know from you guys if that would be recommended
15:58 erikos should maybe add a point to the agenda for 'future development'
15:58 JT4sugar Not sure how hard these would be to technically implement
15:59 So any feedabck very helpful
16:00 erikos JT4sugar: do you have a link to the email thread?
16:00 about your proposal
16:00 JT4sugar one sec
16:01 http://sugarlabs.1511649.n2.na[…]ar-tt7327661.html
16:03 erikos great, see that now
16:03 JT4sugar I'm trying to best understand sine Im not a developer how best to approach
16:03 since
16:03 erikos and there have not been any replies yet, ok
16:04 ok, I will note that you want feedback in the meeting minutes
16:04 and if I get around it reply myeld
16:05 f
16:05 JT4sugar Know you guys are concentrating on 0.96 so thats understandable. Appreciate it.
16:05 erikos any other topics, or direct comments to John?
16:06 JT4sugar: yeah, sometimes it takes a while
16:06 garycmartin JT4sugar: FWIW 'Discovery mode' reminds me of the old Apple help bubble system.
16:07 silbe JT4sugar: what I'm lacking is clear descriptions of a) what you want to do and b) what you want from us (i.e. the Sugar core devs).
16:07 JT4sugar Thats ok. With GSOC timeline coming up wanted to guage the possibility realizing 0.98 is 6/7 months off
16:08 silbe, I am hoping to open up the writing inquiry process through a global chat feature-Is it doable technically with present resources?
16:10 silbe JT4sugar: Depends on a) what you actually mean ("open up the writing inquiry process" doesn't have a meaning for me) and b) what resources you have available.
16:11 JT4sugar Silbe, In discovery mode trying to offer teachers a way to get up to speed to make a better constructivist experience for learners. Can we have a palette mode that has descriptions and possible use cases so teachers dont have to lleave Sugar
16:11 silbe JT4sugar: if you'd like us to do the implementation work, then I'm afraid the answer is that it's not feasible with the present resources, as we're already booked out with the Gnome 3 port, bug fixes and a couple of other ongoing projects. We're having trouble keeping up even with that (see the growing number of open tickets).
16:12 garycmartin JT4sugar: Also worth mentioning that there has been occasional discussion over some time now to design.implement an awards mechanism to try and a) encourage users to explore, b) provide an additional way of evaluating how deep a users understanding has gone.
16:14 JT4sugar Silbe,Ok that is what I am trying to ascertain. My resources come from the pedagogical and learning side
16:14 silbe JT4sugar: there's been a similar project in the past, I believe it was called Tutorius. Maybe you should work with them?
16:15 JT4sugar silbe, Am aware of tutorious my understanding was college class and individuals have moved on.
16:16 Any feedback on how much technical work think it would take would be helpful. Then I will see if there is way to facilitate resources.
16:17 Appreciate your feedback and let you continue your great work on 0.96
16:17 silbe JT4sugar: I'm sorry I don't have a more positive answer for you, but given the scope of Sugar we have ridiculously few developers.
16:18 JT4sugar: we have that much too often, with GSoC projects as well. :-/
16:20 JT4sugar Silbe,I understand I am just attempting to push a few features that could really help learners and teachers get more out of Sugar Reflection-writing and peer questioning through Global chat. But understand the difficulties and time for GTK-3 work
16:20 silbe JT4sugar: but it may still be worth picking it up. If nothing else, it'll prevent you from doing the same mistakes over again.
16:20 erikos JT4sugar: I think you can still find people around that looks for a challenge
16:21 JT4sugar: devs that want to work on a project
16:21 JT4sugar erikos, I will be looking and trying
16:21 erikos JT4sugar: great
16:21 ok, let's close here
16:22 JT4sugar: and feel free to update your email thread with the info from today
16:22 maybe this is already the door opener for someone
16:22 starts countdown
16:22 JT4sugar Thanks
16:22 erikos 5
16:22 silbe JT4sugar: as for how much work it's going to be: It's a major project, so don't assume less than a year (real-time)...
16:23 ... even for several developers, especially since they need to get up to speed with Sugar first.
16:23 JT4sugar silbe, Ok thanks for heads up-that is information I am trying to get at
16:24 silbe JT4sugar: closing note: GSoC might be a good way to get the project started, but you'll need someone who will continue working after that for quite some time, or your project will suffer the same fate as others, including Tutorius.
16:25 erikos it is true that GSoC has to be handled with care
16:25 silbe JT4sugar: not trying to discourage you (I'd love to see this kind of feature in Sugar!), just speaking from experience
16:25 erikos it is quite short
16:25 in time
16:26 anyhow
16:26 4
16:26 silbe 3
16:26 JT4sugar Silbe, Not discouraged when trying to bring powerful tools to learners and teachers hands
16:26 silbe JT4sugar: great!
16:26 erikos 2
16:26 1
16:26 #endmeeting
16:26 meeting Meeting ended Tue Mar  6 16:26:59 2012 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot. (v 0.1.4)
16:26 Minutes: http://meeting.ole.org/sugar-m[…]-06T15:03:06.html
16:26 Log:     http://meeting.ole.org/sugar-m[…]12-03-06T15:03:06

Minutes | Index | Today     Channels | Search | Join

Powered by ilbot/Modified.