[14:00] morning willcooke [14:01] afternoon ;) [14:01] how's it going? [14:03] fine, I figured I'd let you know I was the one proposing the Control from portable devices feature, though I wasn't available on irc during the discussion yesterday [14:03] So if you guys need any explanation about it, I'm available today :) [14:59] Escuse me...Can anyone tell me if the meeting will be here? [15:16] Thanks willcooke =) [15:17] :) [15:18] So, in about 1,5 hours will do that priority meeting =) [15:18] (At least, that's the plan) [15:23] I see some more people then yesterday? [15:34] Who's already around for the meeting? [15:35] MrChrisDruif: me :) [15:35] well I was here before sorta [15:36] Yeah, I saw on the log [15:36] You proposed the.. [15:37] ..control from portable devices [15:38] Apparently, but I don't think it really needs explaining, because it's pretty explanatory ;-) [15:38] yeah, basically control via mobile devices, but it's also key to allowing applications to run well across all devices [15:38] How do you mean? [15:38] so video editing project runs on several tablets and the TV simultaneously [15:38] Ahh, like that =) [15:41] Yep, Novacut really wants to make that possible [15:42] we suspect it will also be very good for other apps [15:43] But that is just one project, we shouldn't focus into one that much I think [15:43] MrChrisDruif: I'm not saying that Novacut should be tied in too much here [15:44] =) [15:44] Just demonstrating that building the right kind of flexibility into the system will let applications like that make UTV a better experience [15:48] I don't think any other platform can offer productivity on that sort of multidevice level [15:48] Well, if you noted the etherpad, we have plugin-system in the high segment =) [15:49] MrChrisDruif: Yes, though I'm wondering really what is meant by "plugin" [15:50] dmj726; it could be "anything" right now. It's not defined at all in this stage [15:51] I would guess that the right approach is to allow "plugins" to certain systems, perhaps via an api, plus a way to run normal applications in fullscreen mode. [15:52] Plugins can have all sorts of functions in UTV [15:52] Oh, how about: We publish a "remote control input" standard over HTTP and then people could use it in new and exciting ways [15:52] Video recording, gaming, I-don't-really-care-what-other-functions-you-can-think-of ;-) [15:52] In fact, let's not limit it to remote control [15:53] kenvandine in the house! Hey dude! [15:53] And he is special? [15:53] willcooke: if we have some standard methods for input over http, that sounds good [15:53] yo willcooke! [15:54] Actually very similar to how Novacut realtime collaboration works [15:54] local databases on both devices sync in realtime [15:54] dmj726, we could start with "Remote Control" functionality - that way people can write apps on any device [15:54] to input to the TV you would just have the tablet make changes in the database [15:54] what i really want to see is state syncing for media between devices [15:55] thats on the list [15:55] willcooke: yes, "remote control" using this system is a good way to start and provide general input [15:55] so watching something on my TV (online), pause it... whip out my tablet and pick up the playback :) [15:55] willcooke; remote control (at least for remotes) is in essential ;-) [15:55] kenvandine, yes- that will be awesome. [15:55] dmj726, "just have the tablet make changes in the database" sounds painful. API would make more sense? [15:55] willcooke: Then if applications want more, they can use the same system for their own use [15:55] and it should be content agnostic... if we can work that out [15:55] same for music, etc [15:56] willcooke: this would be a sort of REST API [15:56] dmj726, that sounds good [15:56] dmj726, are you the novacut guy? [15:56] So the client GUI, tells the local database what changes to make [15:56] Then the changes replicate automatically to the other databases [15:57] The software on the TV sees the change and behaves accordingly [15:57] callumsaunders1: I'm one of them. [16:00] I might be a brake around here, but I think it's better to first get things prioritized? [16:01] It sounds complicated but it actually helps simplify development in many ways. [16:02] Sounds like we need U1DB for that [16:03] willcooke: Yes, though U1DB currently doesn't do realtime sync [16:03] although I suspect they would be likely to add it if Ubuntu TV needed it. [16:04] U1DB is also document oriented as well, which is very good for this sort of thing [16:07] popey, are you chan op? [16:08] i am [16:09] o/ willcooke [16:09] MrChrisDruif: I think there's probably place to get an overall feel for the platform even without a priority list :) [16:10] where we want to go and do and all that [16:11] aquarius, tell us about how we can sync U1DB quickly between multiple devices [16:11] we want to sync bookmarks etc between devices [16:12] does "quickly" mean "each sync does not take very long", or "we can have a syncing thing up and running a short time from today"? [16:13] quickly as in fairly real time, yes? [16:13] ah, right. I should be clear that u1db is early in development right now. [16:13] Syncs are commanded by the application that's using the database [16:13] so if you want to fire off a sync every time a bookmark is written, you can do so [16:14] aquarius: So can one setup U1DB to automatically sync databases for each change? [16:15] well... doing a sync is just another u1db method. dmj726, are you a coder? [16:15] dmj726: yep, with Novacut, mostly on the gstreamer side though [16:16] dmj726, aha! then I've met you, I believe :P [16:16] First day of UDS? [16:16] ok, so you write a document to a u1db with something like db.put_doc(docid, document) [16:17] so there's nothing stopping your "save a doc" method doing "db.put_doc(docid, document); db.sync(server_location)" if you want [16:18] * willcooke loves it when people come together [16:18] aquarius: Are there any performance issues with the way U1DB syncs that would be an issue with doing many syncs? [16:18] Does everyone who writes code know everyone else who writes code?! [16:18] difficult to answer that question right now because the server implementation is still under discussion, but let me think about it for two mins :) [16:19] willcooke, yep. We all know one another. It is witchcraft [16:19] will U1DB provide anything like _rev in couchdb? [16:19] yep, there are revisions [16:20] willcooke: well, more of a general uttering incantations into a glowing talisman thing really [16:20] XD [16:22] so it sounds like U1DB could be used to provide input to the TV and feedback to tablets [16:22] sweet [16:22] yeah. Think of it as like couchdb but not requiring a separate server to be run, from one perspective [16:24] I mean, if you're writing a lot of docs, and you sync after every write, you'll be hammering the server [16:24] but maybe that's reasonable [16:24] the other alternative is u1 file sync, which is instantaneous [16:25] I would guess we should only write the docs that are changed [16:27] sync will only sync changed docs [16:27] I don't know what you're talking about using u1db *for* here, though :) [16:27] aquarius: mainly input from users for controlling the TV I believe [16:27] erm... how does input need a database? [16:28] plus applications interfacing between devices [16:28] willcooke: did you need something oppy? [16:28] I press the "left" button on a remote; a thing moves left. where's there a DB in this? :) [16:28] willcooke: I'll add you if you like? [16:28] applications interfacing between devices, sure :) [16:28] aquarius: let me link you to a mailing list [16:29] popey, nah I was trying to invite aq, but we're all sorted [16:29] gotcha [16:29] https://lists.launchpad.net/ubuntu-tv/msg00005.html [16:30] aquarius: ^^ [16:30] riiiight. [16:30] if you want actual real-time stuff like that, the way to do it is probably not to write locally and then sync every time. You can write direct to the server [16:30] so it's about controlling the TV from tablets and phones and such [16:32] so write direct to the server for input, and write locally with sync for collaborative stuff [16:33] that's sorta what I'm thinking. If you want to control a remote device from here in realtime, then you want a permanently-open communications channel, which is not what u1db is. [16:33] avahi ☺ [16:33] and trying to trick up u1db into being a permanently-open comms channel will just cause you misery :) This is why jabber exists :) [16:33] heh, jabber server in your telly [16:33] good idea [16:34] you could message it to say record xyz [16:34] all this assumes they're on the same network, of course [16:34] so you would send a message to the DB on the server via jabber? or am I being silly [16:34] ahh, yes [16:34] jabber can send all kinds of messages, so it's good at realtime stuff [16:34] you could chat with the telly [16:34] "Was eastenders any good on tuesday?" [16:35] "BARB Ratings for Eastenders on tuesday put it ahead of I'm A Celebrity" [16:35] check out Volity for an example of using jabber as a realtime comms channel -- they have a multiplayer games server which uses jabber as the comms protocol. It's not just for IM chat [16:36] you could do video chat inside the house, webcam on every telly [16:36] "Hey bob, your dinner is ready" [16:36] so jabber would be better than using a REST API? [16:36] the question is: why is this stuff going into a database on both the device and the tv? [16:37] if I'm controlling the tv to show people a youtube video, say... why do I need to record forever on both the tv and my phone that I did that? [16:37] it's just an RPC, right? [16:37] So i'm probably not going to be available for the meeting today [16:37] aquarius: that's true [16:37] I have to give two back to back interviews today starting in 20 minutes [16:37] Though it may sometimes be important to model the state of the system [16:38] I'm fine with the list so far, and I'm responding to the email from willcooke right now [16:38] So if the local device can for instance know that the TV is in "Play" mode, it can offer more intelligent interface on the client [16:38] dmj726, yeah, there's value in recording this stuff in a DB somewhere, but I can't see why you wuoldn't just have RPC from phone to TV and then have the TV stash away logs of commands (in a DB, or whatever) [16:39] sounds like the best option [16:39] dmj726, oooh, trying to keep the local device permanently in sync with what mode the TV is in is a real-time collaboration problem. At that point you're building World of Warcraft, and it ain't easy. If you're on the LAN, it's a lot easier to request http://tv/states/ and have it return {"state": "play"} :-) [16:39] tgm4883; too bad you can't come. Good luck with the interviews =) [16:42] Well, not nearly so intensive as WoW I wouldn't think, though not the easiest [16:42] aquarius: but then wouldn't the device have to keep pinging http://tv/states? [16:42] well, I'd make that a comet endpoint, but yeah [16:43] so you wouldn't ping it; you'd hang on it [16:43] so whenever something needs it, you wait for it to respond? [16:44] well, if you're hanging on it, then you get told whenever it changes [16:44] ah, okay, that hang [16:50] aquarius: Hmm, so you would requests http://tv/states/ parse the data, present it. Then when the user does something, send a jabber message to the TV, which then changes the contents of http://tv/states [16:51] I think that is how you should think of it [16:51] the question is whether you need lots of two-way realtime communication between device and tv [16:51] if you do, then use jabber for it [16:52] if it's mainly one-way (device sends a command, tv obeys) and occasionally the other way (tv changes something, device needs to know about it) then do it over http with a couple of hang-on-this endpoints on the tv [16:52] you shouldn't need both [16:53] Two-way. I don't know. It could for faster letting the device know that a change has happened [16:53] And the TV ALSO writes it to http://tv/states/ [16:54] the tv doesn't need to write it there, though, if it's sending jabber messages to everyone [16:54] no-one will ever read the endpoint [16:55] It could request the current state with jabber I think? [16:56] Authentication/pairing is also important to consider here [16:59] ubuntu-tv, MrChrisDruif ? [16:59] Yup [16:59] wat doet men precies? [17:00] I was announced with UDS-P (the previous one) [17:00] erkan^; This is an international channel, please use English =) [17:01] sorry, I will speak English language only :P [17:01] Alright everyone, shall we start the "meeting"? [17:01] sure [17:02] o/ [17:02] Alright, for who has missed it, here is the link to the feature list we want to get implemented: http://pad.ubuntu-uk.org/UbuntuTV [17:02] MrChrisDruif: want a meeting bot? [17:02] AlanBell, sure [17:02] AlanBell; could be usefull [17:02] -l [17:03] done :) [17:03] MrChrisDruif, #startmeeting [17:03] #startmeeting [17:03] Meeting started Wed Nov 23 17:03:18 2011 UTC. The chair is MrChrisDruif. Information about MeetBot at http://wiki.ubuntu.com/AlanBell/mootbot. [17:03] Available commands: #accept #accepted #action #agree #agreed #chair #commands #endmeeting #endvote #halp #help #idea #info #link #lurk #meetingname #meetingtopic #nick #progress #rejected #replay #restrictlogs #save #startmeeting #subtopic #topic #unchair #undo #unlurk #vote #voters #votesrequired [17:03] Welcome everyone, now we get "official" [17:03] Raise your hands for attendance [17:03] o/ [17:04] Indeed [17:04] o/ [17:04] o/ [17:04] o/ [17:04] o/ [17:04] tgm4883 is unfortunately missing because of two interview back-to-back [17:04] oh, is the meeting actually now? [17:04] Yup [17:05] you're not just testing the bot? [17:05] popey, lol yep [17:05] oh, sorry [17:05] ☺ [17:05] #link http://pad.ubuntu-uk.org/UbuntuTV [17:05] (Quiet link acceptance?) [17:05] And another one [17:05] #link http://irclogs.ubuntu.com/2011/11/22/%23ubuntu-tv.html [17:06] The first link is to the document we are about the further discuss. To know what has been discussed already, please look in the second link [17:06] Okay. Carryon [17:06] #topic Prioritizing the list [17:07] MrChrisDruif: yeah, #link doesn't do a massive amount right now, there is a bug about that [17:07] <_<" [17:07] It's still useful to mention the links =) [17:08] yes [17:08] We were going down the list from top to bottom and we got to Collaboration with Ubuntu devices [17:09] Does everybody understand the different sections? (They are pretty explanatory if you ask me) [17:09] hi! sorry for being late! [17:10] Welcome Wellark [17:10] Personally I think collaboration is high importance [17:10] It could be a pretty key differentiator for Ubuntu devices [17:10] +1 [17:11] I'm not sure to put it in the high section. [17:11] Mark seems fairly keen on the idea too [17:11] yes, +1 [17:11] i agree with with dmj726 [17:12] high [17:12] as I've understood the "Unity Experience" is also largely about collaboration [17:12] MrChrisDruif, Any reasons for that thought? [17:13] Well, I think it might be a very good feature to have, I do think we shouldn't put it in the high because I don't know how many devs we will have and we can't have it all in high [17:13] Well, I think it's important to design with it in mind than to tack it on after all other things are implemented [17:14] well, I think the rest can be other than "high" :) [17:14] but collaboration between devices as high [17:14] Wellark; How do you mean? The things that are now in high you mean? [17:14] I mean the uncategorized ones [17:15] Okay, noted [17:15] there are only two left after this one [17:15] the high priority list seems to be pretty good package [17:15] So collaboration in High? [17:16] high [17:16] +1 [17:16] Btw I had a note added after the meeting about the previous item [17:17] But you should consider what MrChrisDruif is talking about. Starting it with less number of developers will also lead to some wrong tracks and delay in work [17:18] We should try to figure out how many developers the project really has. [17:18] That is certainly something to keep in mind, but it's prospected to not have anything very functional until 13.04 or something [17:18] It's probably more important to get it right than get it done quick and have to backtrack [17:19] +1 [17:19] Indeed [17:19] I don't feel the HIGH section is too cumbersome [17:19] dmj726, True. But if you ask me, then I will say "What about finding the right developers be the high priority job? =]" [17:20] is there a hardware reference specification? [17:20] anyway the items in high section have to be prioritized [17:20] So we should decide what the core functionality needs to be for a compelling offering, and then work toward that. [17:20] is it ARM or Atom for example? [17:20] all of it. except home automation [17:20] dmj726; that was the main reasoning in the different sections [17:20] what screen resolution? full HD? [17:21] that list should be ubuntu tv 1.0 [17:21] 0.00001 [17:21] Haha, popey [17:21] popey +1000 ;D [17:22] also some things are easier to bolt on after the fact and others are much harder to change once you've gone far [17:22] What is absolutely required, is essential [17:22] That's why the plugin system is in high [17:22] It's not necessary, but it's a very useful way to add functionality later on [17:23] So essential is "what do we need for a working TV" [17:23] Yup [17:23] High is the "cool" system [17:23] We can even plan how it should be and can keep it for the next version (If needed) [17:23] Medium is the "awesome" system [17:23] high is "what do we need for a device that people will want beyond an ordinary tv" [17:23] yeah [17:23] And low is the "ICan'tDescribeThisInWords" system [17:24] for reference I suspect that control from portable devices will be very easy once we have Collaborate with other Ubuntu devices, just a little work on top of that framework [17:25] Yeah, that's why we initially put it in medium [17:25] and Push media to/from other Ubuntu devices / Media syncing capabilities would also fit within the framework [17:26] so implement collaboration framework and the other two are specific applications thereof [17:26] I tend to agree on that (I'm not a dev) [17:27] yeah, that sounds good [17:27] so we need the collaboration framework because that lets other devs make the TV very cool [17:28] this needs to be in tablet & phone & pc aswell [17:28] yes, it should be everywhere [17:28] So either all those should be in high OR in medium, and if we'd have to call it now I'd say high (seeing the functionality of collaboration with tablet/phones) [17:29] okay [17:29] Everyone agrees on that? [17:29] sure [17:29] sure [17:29] Alright, I'll move it up (and rephrase it after the meeting) [17:30] okies [17:30] I've put a note in the last one from the high (now at least) about collaboration with Ubuntu Phone & Tablet [17:31] I think we should also let them in on the "action" and pitch in where it's logical [17:31] yeah [17:31] (Luckily I've noticed that a lot from #ubuntu-phone & #ubuntu-tablet are already here) [17:32] But we still should give a shout-out to them on their mailing-lists [17:32] Also, input from other devices greatly enhances our 10' UX capabilities [17:32] (think typing on tablet vs typical TV keyboards for search) [17:33] its just gravy [17:33] True that (or even a laptop =) ) [17:33] Anyhow, shall we continue down the line? [17:33] maybe gravy, but definitely a pain point on traditional ui [17:33] yep [17:33] k [17:33] moving on [17:34] Alright, next one is: Sharing media with friends [17:34] it's a good way to get word out [17:34] With social media, think of youtube like/dislike, facebook, twitter, diaspora, what-have-you? [17:34] I think this sounds like a great plugin if done right [17:35] i would say low [17:35] Indeed, but now it's about prioritizing =) [17:36] low +1 [17:36] I would say medium as low is really "awful" things like home automation, which might even be more for server than tv [17:36] yeah, higher than home automation, but not high [17:36] But home automation should have a friendly shell for which a TV might be a good medium [17:36] So medium is good I think [17:37] Home automation would be a good app for a tv [17:37] +1 on that callumsaunders1 [17:37] Alright, next is Purchasing media [17:38] this would mostly be web interfaces for stores? [17:39] Yes, I would think so. Use case: search for a movie > if it is available for purchase it gets highlighted in the search results [17:39] I think of VOD (in high quality of course) [17:40] youtube do movies now, is that part of their api? [17:40] * MrChrisDruif doesn't know, nor care at this moment [17:40] lol ok [17:40] Is it in the right location in medium? [17:40] i would say medium [17:41] +1 on that [17:42] First a good basis/rich system, then we can think about financial stuff like buying content [17:42] So medium it is? [17:43] Silence says it all I think, moving on =) [17:43] Next has already been done I think: Home automation [17:43] Anyone disagreeing with the Low priority? [17:43] not even sure it should be there at all [17:44] That's the advantage of low priority, we can always decide to drop it all together in a later stage, but for now I'd say keep it in =) [17:45] If Ubuntu want a complete system, then home automation should get a place in there. For the front-end I think it's a good idea to have it on TV [17:46] Any other comments? [17:46] im happy [17:46] Good to hear =) [17:46] Alright, moving on [17:47] There has been an addition since yesterday: TV Guide [17:47] I think this one falls under Watching Media in Essential [17:48] Otherwise it could be easily implemented with the plugin system; devs what do you think? [17:50] dmj726? popey? IAmNotThatGuy? Others? [17:50] * AlanBell thinks mythbuntu does all of this already [17:50] it needs a unity theme [17:50] with lancher and lens type gui [17:50] then everything is either already done, or a mythtv plugin [17:51] already has the 10' /3m user interface [17:51] If we need some changes, then we can work on that plugin [17:51] TV guide is there, weather, home automation stuff like that [17:52] AlanBell; I think it was tgm4883 that already commented that we should try to use current software solutions [17:52] so a chunky launcher bar on the left selecting live tv, recorded stuff, playback of DVDs (which would already pop up in the launcher) [17:52] lens for current live channels [17:52] So if Mythbuntu already fits the bill, that would save a lot of effort [17:53] however, what are we targetting? [17:53] Hardware-wise you mean? [17:53] what is the hardware platform? does the OEM team have a customer hardware specification in mind? [17:53] hardware, broadband connectivity, screen resolution, screen size etc [17:54] is this for a settop box or something integrated in TVs? [17:54] Do we even have a OEM team? =/ [17:55] yes, canonical has an OEM team [17:55] Alright, then we haven't spoken to them about that [17:56] Back for 2 minutes before the next interview [17:56] were there concept drawings done for unity on tv when unity was being designed? [17:56] tgm4883; I wanted to close up, but if you have anything to say, do so =) [17:57] looks good, the TV guide data is necessary for any DVR type functionality [17:58] AlanBell, Mythbuntu does fit the bill for a lot of this, but I think we need to change the frontend [17:58] AlanBell; any suggestions for what we should do next? Speak to which teams, to get an idea of what kind of concept canonical already has envisioned for Ubuntu TV? [17:58] I don't think the MythTV frontend is capable/flashy enough for what we want [17:58] I suggested XBMC [17:58] fine, whichever [17:59] and time for interview #2 [17:59] we could contribute improvements upstream for one or the other [17:59] Indeed [18:00] Hey everyone, that meeting still happening? [18:00] is it for ARM or x86? [18:01] Does anybody feel for inquiring for this information? [18:01] * popey returns [18:02] And besides ARM or x86; settop-box or integrated [18:02] both and both [18:02] :D [18:02] there, that was easy [18:02] Probably [18:02] I dont think we're far enough down the line to make decisions about hardware platforms [18:02] full HD? [18:02] yes [18:03] where full = 1080p [18:03] yup [18:03] that makes a huge difference in gui design [18:03] yeah, needs a massive interface [18:04] big fonts, big icons/pictures [18:04] cover art etc [18:04] and crisp, svg all the way [18:04] That's the idea of 10' interface, right? [18:05] 17:58:10 < MrChrisDruif> AlanBell; any suggestions for what we should do next? Speak to which teams, to get an idea of what kind of concept canonical already has envisioned for Ubuntu TV? [18:05] Anyhow, I'm wasted. Shall we call it a day? [18:05] Design Team and Product Strategy [18:06] Thanks popey =) [18:06] (are who you might want to speak to) [18:06] Are we going to return to the idea of home automation at all? [18:06] Or did I miss that? [18:06] willcooke and myself both work in product strategy [18:06] imnichol; we will keep it low for now =) [18:06] and the design team live in #ubuntu-design [18:06] Ok cool. I've got some concerns about the right place for it, but I'll register them later [18:06] I think for a first run the main focus should be "choosing video to play" and "playing video" ☺ [18:07] imnichol; read the logs afterwards [18:07] where that content comes from who knows, how we play it, who knows.. [18:07] MrChrisDruif: reading them now. [18:07] popey; AlanBell; others; shall I #endmeeting ? [18:08] fine by me [18:08] yea [18:08] yup [18:08] Alright [18:08] Thanks everyone for showing up and raising your voice =) [18:08] #endmeeting [18:08] Meeting ended Wed Nov 23 18:08:46 2011 UTC. [18:08] Minutes: http://ubottu.com/meetingology/logs/ubuntu-tv/2011/ubuntu-tv.2011-11-23-17.03.moin.txt [18:09] Anyone up for putting it on the weekend and mailing list? [18:09] I forgot to discuss weekly meeting, but I'll put that on the mailing list myself [18:09] ML link please [18:10] launchpad link of the team will be a better one ;] [18:10] IAmNotThatGuy; ubuntu-tv@lists.launchpad.net ? [18:10] https://launchpad.net/~ubuntu-tv [18:10] See topic@! [18:11] thanks man! /me was busy with the other machine [18:13] Anyone wanting to put it on the wiki? [18:15] Not all at once please... [18:15] =( [18:16] I can make a post of this meeting and share it to my friends but I am not good with meeting logging :[ [18:17] It's not really though IAmNotThatGuy [18:18] Where it our wiki actually? [18:18] MrChrisDruif, Yeah! I know. But I am preparing for Lubuntu meeting :[ [18:18] is* :P [18:18] Good going, I might also attend, just to hang around [18:19] And be obnoxious? ;-) [18:19] I think there is no wiki yet [18:19] for TV [18:22] =( [18:23] Maybe we could just send the link meetingology send us? Saves us the trouble I would think =) [18:24] I don't want to setup an entire wiki area right now, just to put up meeting notes [18:28] lol [18:37] I'll send the link to the mailing list and MAYBE put it up later on the wiki [18:48] Alright, 2 mails out [18:52] Thanks IAmNotThatGuy for your email response =) [18:59] MrChrisDruif, I'm here :P [19:00] M0hi; ;-) [19:00] MrChrisDruif, running another machine for other works which are tbd =] [19:02] MrChrisDruif, we need some branding for the teams [19:02] team* [19:03] Well teams, also tablet and phone ;-) [19:03] But that are concerns for there ;-) [19:04] Hmmm [19:08] I'veposted it in my site. try telling your friends to do the same [19:08] * MrChrisDruif is off TTYL [19:59] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=469397 [19:59] xbmc isn't in debian yet, and that is the ITP bug report [20:04] AlanBell, I've looked through that before and it appears that Andres Mejia is working to fix any of the issues (and most/all of them were fixed as of the latest entry) [20:05] I'm also unsure how well XBMC would work on the ARM platform, as I've not tested it myself and I've heard conflicting reports [20:06] the latest entry from Andres Mejia was a year ago [20:06] hmm, so it is. I apparently cannot read :) [20:06] Might be worth reaching out to him and find the current status/road block [20:08] so why is xbmc considered better than mythtv? [20:09] AlanBell, regarding your earlier comments about Mythbuntu (since I now have more time), I think as a DVR backend that it would work great, but for the scope of this project since TV recording isn't a requirement MythTV might not fit the bill. Also, the MythTV frontend looks a bit dated and is missing some of the functionality that XBMC has (likely because XBMC has more developers) [20:09] Full disclosure, I'm a Mythbuntu developer [20:09] ok, I ask because I don't know much about this area at all :) [20:10] what is the scope of this project? [20:10] I believe it is to be a more open apple tv type device [20:11] I think recording needs to be an option, but not required [20:11] ok [20:11] AlanBell, are you on the mailing list? There has been some good discussion there regarding scope [20:18] ok, read the archives now [20:26] Hmm...I would think recording should be fairly central if there is available storage [20:27] maybe, maybe not [20:29] Why would recording not be a desirable feature these days? [20:30] if your storage is in the cloud and you have loadsa bandwidth [20:31] but wouldn't that just mean recording to the cloud? [20:32] dunno, maybe the cloud records it for you and streams it to you as required [20:32] maybe the cloud is a box in your house with storage [20:32] well, the TV would record it to whatever media is available [20:32] anyhow, I think listing specific features is a bit of an odd place to start [20:33] AlanBell: I'd agree with that statement [20:33] just saying it seems like an obvious use case: UbuntuTV acts as a PVR [20:33] yeah, I have nothing against it :) [20:35] Can pretty much guarantee that an Ubuntu TV that could do scheduled recording would be attractive for my parents house [20:35] so the vision for this project is Unity on a telly [20:36] Yeah, I can see Unity working really well if adapted for arrow key interface [20:36] there has been very little discussion of unity so far [20:37] I think the starting point for this project is to figure out what the best platform is for implementing Unity on a telly [20:37] AlanBell: I'm thinking that people are still trying to see the big picture of what this thing is trying to be [20:38] Basically, you're talking about the how, whereas discussion of features is more of the why [20:38] both are important [20:39] and knowing the use cases makes how we use Unity more clear [20:42] We were using the etherpad yesterday to try to figure out what we're trying to do [20:42] I think the data's still up there [20:43] Yup it is [20:44] imnichol: Yeah, I think that's a useful document to gauge what we need to make a unity-centric system do [20:44] Although now that i look at it, it still suffers from "how do we do this" syndrome [20:44] When we really need more "what are we doing" [20:45] yeah, but we are not going to build a new media centre [20:45] we can integrate and distribute one [20:45] I don't want Yet-Another-Media-Center [20:45] Exactly [20:45] What *do* we want? [20:46] http://www.markshuttleworth.com/archives/820 that is the starting point really [20:46] What I want is something that is a TV and can replace a media center, but provides a great platform for social/collaborative experiences. [20:47] brb [20:49] The PC, the phone, and the tablet are the personal screens, while the TV is the shared screen that connects them. [20:50] The TV should be simple for passive use like watching TV, and seamless for active use (think how people use video game consoles) [20:50] http://musl1m.deviantart.com/art/My-Ubuntu-Phone-267986471 [20:51] that is a reasonable stab at a concept for unity on a phone [20:51] yeah [20:51] I saw a nice one on the ayatana list too [20:51] I think there should be some sketches of unity on a telly [20:51] AlanBell: Have you used or seen Maemo or Meego? [20:52] for phones they got multitasking spot on [20:52] not used them, seen screenshots [20:53] and their menus worked well too [20:55] Obviously not quite right for a TV [20:55] do we have anyone in the channel to do some sketches [20:56] anything I do would just get the gist of it, not be pretty [20:58] It definitely needs to autohide both launcher and panel [20:58] and I'm not entirely sure the panel is a good idea in its current form [21:00] yeah, I don't think the panel quite fits [21:00] the clock might sometimes [21:01] and things should probably always be fullscreen [21:02] so I think we should figure out the visual design of it, then figure out what media centre can do that, then find out what features or applications that does [21:02] so if it has a weather app then yay, we get a weather icon in the launcher [21:03] if it has games/home control/etc. they go in the launcher, if not then we don't have them [21:04] if it does recording of live tv, then awesome, we get that [21:04] if it doesn't right now then we don't have that until someone (possibly canonical) contribute such code upstream [21:06] Well, I suspect we'll have to write a fair amount of front end code [21:07] So Unity and our interface scheme plus whatever media center [21:08] front end and integration code, yes [21:09] tgm4883: Do myth and xbmc allow themselves to play nice with a composited window manager? [21:09] and depending on the media centre there might be opportunities to fit in other ubuntuish stuff like a U1 file browser or streaming from U1 thing [21:09] I think with launcher we can handle multiple apps fairly well [21:10] dmj726, mythtv does not play well with compiz unless you turn on legacy support, I'm unsure of how xbmc handles it [21:10] but take a broad view of what "unity" is [21:10] if it looks like unity then it is unity :) [21:10] so if our media center plays well as being one window with several others [21:10] +1 [21:10] I don't think Unity desktop really fits into a media center [21:10] unless you are discussing theming a 10' UI to look like Unity, then maybe it works [21:11] then we can have someone switch to another fullscreen task [21:11] it doesnt have to look exactly like current unity [21:11] I still disagree with that though [21:11] it has to have a launcher on the left with buttons on it [21:11] tgm4883: does myth play well with any window manager [21:11] doesn't have to be NUX or Qt or anything like that [21:11] or does it need to be the root window [21:11] :) [21:12] dmj726, mythtv runs fine on any window manager, you can run it with compiz you will just see the gnome panels [21:12] or the top panel in unity (not sure about the launch bar) [21:12] Mythbuntu uses XFCE [21:13] tgm4883: ah, that's fine then [21:13] did you see the ask mark session earlier? [21:13] so we just slide any unity UI elements out of the way when not in use [21:13] dmj726, AlanBell so are you guys discussing having a 10' UI that launches other applications for video/music/etc? [21:13] tgm4883: yes [21:14] that is a bad idea IMO [21:14] one of (probably the main one) being the media center ui [21:14] how do I control banshee and totem with a IR remote control? [21:15] Not to mention it looks crap when switching between applications [21:15] tgm4883: We don't launch apps made for desktop [21:15] dmj726, so what apps do we launch then? [21:15] all of this would be 10' centric [21:15] launch things that are designed to work with a media centre framework [21:15] so apps that we develop specifically for this media center? [21:15] tgm4883: yes [21:16] us and third parties [21:16] Does anyone have a link to a definition of "10' UI"? [21:16] dmj726, AlanBell so plugin support then :) [21:16] which is already there, yes [21:16] imnichol: A UI that can be used from across the room [21:16] I've seen it thrown around a lot and I'm not exactly sure what it means [21:16] imnichol, http://en.wikipedia.org/wiki/10-foot_user_interface [21:16] imnichol: 10' means about 3 meters [21:16] Well then I don't see how there are any problems with Unity as a 10'UI [21:16] Except the top panel [21:17] Which we've discussed removing [21:17] Americans and their silly measurements [21:17] There is a lot of wasted space [21:17] imnichol: you also can't count on mouse and keyboard always with a 10' UI [21:17] a big chunky unity launcher [21:17] i dont think anyone would run banshee on their telly [21:17] dmj726, imnichol in fact, count on not having either [21:17] FWIW, the Unity launcher is hidden unless you call it up [21:17] you might only have the equivalent of a dpad and enter key [21:18] Usually that's done using the "windows" key or moving the mouse [21:18] But I don't see why you couldn't just use the "menu" key on the remote [21:18] yes, menu to bring in the launcher from the side [21:18] remote control cursors to navigate it [21:19] perhaps a button to press the BFB at the top and get a lens to pop out [21:19] What is the point in hiding it though, other than that is what the desktop does? [21:19] so you can watch telly :) [21:19] If we engineer this for an input device that is too simple, then we'll lose all of our ability to do the cool stuff [21:20] tgm4883: because you don't need to see it unless you want to change what you are doing, it's wasted space if you're just watching TV [21:20] imnichol: That's why I've been proposing additional input methods [21:20] imnichol: We should allow all basic functions using a standard remote [21:20] I agree [21:20] However, we should also allow applications to be controlled in a more advanced way [21:20] imnichol, TV should be full screen, and have an OSD that is only brought up when requested via remote [21:21] exiting TV should bring you back to the media center [21:21] So I'm confused about the hate for unity in this situation, since the entire point of unity is that it's out of the way until you need it [21:21] tgm4883: OSD? [21:21] imnichol: I don't think there's hate here [21:21] On screen display [21:21] Ok [21:22] just discussion that a few adjustments need to be made to go from the desktop to a really good TV UI [21:22] what is the use case for having the ability to pop up the unity launcher while I'm watching a video? [21:22] "hate" was an analogy [21:22] :) [21:22] an analogy for hate! :) [21:22] tgm4883: So you can go do something else [21:22] FWIW, I like Unity for a desktop [21:23] Unity is mostly fine. Global menu bugs me for 1920x1200 displays [21:24] tgm4883: Most of my questions were directed at you're comment that "I don't think Unity desktop really fits into a media center". I'm just curious about what you think :) [21:24] dmj726, so the two things we are discussing are A) You are watching a video and want to listen to music instead, so you open up the launcher and launch the music app while you are in the video. VS B) You are watching a video and want to listen to music so you exit the video and launch the music app [21:24] does that about sum it up [21:25] basically, being able to launch another app while still in another application vs having to exit first? [21:25] tgm4883: ok I think I get what you're saying [21:26] I would say better to be able to keep both up [21:26] Yes [21:26] but only have playback of whatever happen in whatever has focus [21:26] I disagree with that [21:27] I'd like to be able to browse the internet temporarily while still listening to the big game in the background [21:27] imnichol: okay I'll amend the statement [21:27] (I have never watched the big game in my life) [21:27] I think Internet on the TV from a media center perspective is overrated [21:28] have default apps respect certain signals [21:28] so the music player doesn't keep playing while you start a video [21:28] tgm4883: I disagree [21:28] It's only overrated if your browser sucks and you only have a remote control to control it from [21:28] dmj726, do you have a keyboard and mouse connected to your media center? [21:28] dmj726: why not just pause the music then? [21:29] have a look at what media centres look like e.g. http://xbmc.org/skins/confluence/ and think about how things could be mapped onto a launcher/lens type look and feel [21:29] If I had a tablet to control the browser and wanted to show something to a group, it would be awesome [21:29] dmj726, use case, why use the TV to temp browse the web? [21:29] possibly including playback controls in the panel area [21:29] AlanBell, is that directed at me? [21:30] not directed at anyone in particular [21:30] so I wouldn't regard a web browser as something to use with a remote, but a TV certainly can be useful [21:30] dmj726, that is completely different and something we actually have on the wishlist [21:30] tgm4883: I know it's on the wishlist [21:30] dmj726, the ability to shoot content from a tablet to the TV [21:30] dmj726, but that doesn't require a web browser on the TV. Even if it did, not a full web browser [21:31] I wish we could video chat, I think we're missing each other's points here [21:31] video chat would be something I would want on the telly incidentally :) [21:31] tgm4883: Actually a web browser controlled via the phone but running on the TV would be a good thing [21:31] Hahah [21:31] dmj726: I agree [21:31] AlanBell, I agree, I think we could get something similar to Unity using something like XBMC, but I don't want to see this on my TV http://lh3.googleusercontent.com/_1QSDkzYY2vc/TWbk116J3FI/AAAAAAAADIo/BdBOuK9NnSQ/unity-screenshot.png [21:31] you don't want everything to be a bunch of vnc sessions over the wifi [21:32] +1 to video chat on the TV [21:32] tgm4883: no, the launcher should be about twice as chunky [21:32] it shouldn't be a desktop, it should just look unityish [21:32] AlanBell, meh, it should probably go all the way across and have banners for each item [21:33] Actually, the launcher does look about like your typical bluray OSD [21:33] also think about the dash and lenses [21:34] AlanBell, What about something simlar to this for each item http://wiki.xbmc.org/images/f/fb/Wide1.jpg (only looking at the bottom right section) Having a wide banner for "Videos" "Music" "Pictures" etc [21:34] That could be made to look unityish [21:35] yes! [21:36] I think that would be a brilliant design and could be made to resemble Unity [21:36] or videos/library/tv shows as launcher buttons, the programs in a lens (groups for heros, house, it crowd with episodes in the groups) and the stuff on the left as lens filters [21:36] I'm going to have to learn more about lenses, I just switched to Unity from Gnome-shell not too long ago [21:37] tgm4883: Spoilers: THey rock [21:37] so click the dash button, then more apps [21:37] you have goups for most frequently used, installed and apps available for download [21:38] hmm, I could see that being *Really* useful. If you went into recordings, and you had just wanted the first three episodes of some show it could list the next couple episodes [21:38] depending on viewing habit and such [21:39] If anyone uses MythTV, this would be similar to the watch list [21:39] AlanBell: so you're advocating for basically reskinning xbmc? [21:39] but yes, I think selecting one of the items and then showing a few views in the lense would be great [21:39] reskinning and integrating $mediacentre [21:40] imnichol, I've advocated that a few times [21:40] tgm4883: Yeah I know, I wanted to make sure I had Alan's point right [21:40] views being (next in recently watched, new content, etc) [21:40] imnichol: yeah, that is right, but I am not totally convinced mythtv is out of the running yet [21:41] I'm not sure we should tie everything into being part of the XBMC UI [21:41] dmj726, XBMC is an example of something that already does (or is in the process of developing) most things that we want [21:42] I actually prefer the idea of just tweaking Unity slightly and then having all of the various things we want to do (watch tv, video chat, etc) as applications on the unity menu or whever we want to put them [21:42] mythtv has the advantage of acutally being in the repos, so it got through all the legal and licensing issues and is supportable [21:42] AlanBell, I just think that the requirement that MythTV places on needing a recording device doesn't fit in line here [21:42] not saying XBMC shouldn't be a key part of what we're doing [21:42] same for Myth [21:43] AlanBell, but I DO think we need to support existing DVR backends for the recording/scheduling functionality [21:43] MythTV being one of those backends [21:44] mythtv has its issues no doubt, XBMC not being in ubuntu is a bit of a show stopper though [21:44] I prefer keeping Unity, making tweeks as necessary [21:44] can we get a quick idea about what media center's people currently use and what control devices they have? [21:44] integrate it well with whatever media center we use [21:44] dmj726: that is one possible implementation of the design :) [21:45] I use MythTV, although I've started to use XBMC as a frontend to that. I have an IR remote control and also an android app on my phone [21:49] I don't currently have a media center, though I do have an Asus Oplay, which is similar [21:50] Just basic cable [22:25] http://people.ubuntu.com/~alanbell/unitytelly/programlens.png [22:25] http://people.ubuntu.com/~alanbell/unitytelly/controlsinpanel.png [22:25] http://people.ubuntu.com/~alanbell/unitytelly/progresspanel.png [22:29] the .ep files that open in pencil are in that folder too [22:30] ??? [22:31] pencil.evolus.vn [22:32] Three quick mockups for Unity UI TV thing? [22:33] yeah [22:34] * MrChrisDruif looks at them more closely tomorrow, off to bed now [22:35] Sleep well AlanBell when the time cometh [22:41] AlanBell, looks great. I think the background needs to change based on currently selected item (or have the option to switch between the dynamic and static backgrounds). Also, should have the ability to remove main window entry options (eg. removing movies, or weather) [22:57] AlanBell: so the BFB brings up a dash that lets users browse via lenses for applications or shows? [22:58] (or whatever lens they pick) [22:58] I was thinking the background would be the thing playing [22:59] tellytubbies in that example [22:59] can you name them all? [22:59] makes sense to have background always be the playing thing (or whatever the user was just doing) [23:00] the launcher should be translucent too probably [23:00] Yeah, probably do the blurring thing it does on desktop [23:01] I think it would be great to see more designs like that for unityish user interfaces mapped to media centre concepts, on a 1920x1080 proportioned layout [23:01] which is 480x270 at quarter size [23:02] btw, what did you use to make those pics? [23:02] pencil.evolus.vn [23:02] it is a firefox plugin [23:02] or extension [23:05] ah, there's a standalone version too [23:05] I put the source files in http://people.ubuntu.com/~alanbell/unitytelly [23:06] cool [23:06] yeah, standalone version is a bit broken as it depends on xulrunner which isn't in oneiric [23:06] will try it out later [23:06] I think [23:06] hmm...I do have a maverick box [23:07] it works basically like a standalone app, you just launch it from firefox [23:07] you can even close firefox after you have launched it [23:08] I am going to try to get the design team to use pencil and make an Ubuntu widget set for it, so you have a thing for drawing an ubuntu window with buttons on the left etc [23:09] and an easy lens widget for example [23:09] yeah, easy launcher too [23:17] yes, mine looks a bit rubbish [23:17] it is important for these things to look a little sketchy [23:20] AlanBell: why's that? [23:20] so people don't feel shy about pointing out problems [23:21] ah, and don't get too attached to the ooh pretty [23:21] yeah [23:21] one thing I notice that doesn't feel right is the playback controls at the top of the screen [23:22] if you google paper prototyping you sill see stuff about it [23:22] for some reason it's always felt more natural at the bottom, might just be familiarity [23:22] yeah, I saw one of the xbmc things had controls at the top [23:22] I think the progress bar as the top panel kind of works [23:23] progress bar didn't phase me where it was [23:24] that would kind of be your position in the program if it was a recorded thing [23:24] I once prototyped a mobile video editor UI on a 3.5" piece of paper taped to a (box of a) deck of cards [23:25] ah, progress bar as you call it should be shown with the playback controls I think [23:26] If you're not messing with playback, you maybe don't need to be looking at your progress? [23:26] I was thinking of it as a draggable or clickable thing [23:26] We also may have to deal with TV safe region btw [23:26] and it would fade away when you are watching [23:27] launcher would slide off to the left, top panel (whatever is in it) would slide or fade upwards [23:27] Do have an option for permanent timecode, or whatever overlay I think, but make that a special toggle [23:27] clock top right could be permanent overlay [23:28] and any other indicators [23:47] What if the launcher automatically showed any apps/projects that somebody had sent a message to open [23:47] So somebody says "Open this project on TV" on their tablet, and it opens up in the launcher [23:48] and the launcher flashes that it's available for the moment [23:49] cool, tablet/phone/computer integration with the TV interface sounds nice [23:49] That seems ok, but why not just have it open up automatically? [23:49] could be done over avahi or something [23:49] sketch it up :) [23:49] FWIW, Apple's new file transfer thing in Lion uses avahi [23:50] yeah, apple invented most of the avahi stuff [23:50] imnichol: I'm debating the automatic vs flash in the launcher [23:50] Airdrop. That's the name, it was really bugging me [23:50] That actually might be a legitimate case for a preference. [23:51] Yeah [23:51] depending on if you have annoying people in the house that like to mess with the tv [23:52] (roommates) [23:52] three options: 1) switch to new task immediately 2) stay on current task but display launcher with flashing icon 3) place icon in launcher but do not show launcher automatically [23:53] Those all sound good. [23:53] Do you think there should be an option to just disable it entirely? [23:53] 1) is useful for several people working together 2) and 3) are useful for households with an annoying individual who thinks its funny to interrupt someone's viewing experience [23:54] imnichol: 3) is effectively disabled since you wouldn't notice it unless you specifically checked the launcher [23:55] #3 would allow an annoying person to spam the launcher with a bunch of tasks [23:55] if you really don't want stuff appearing in the launcher (which basically shows everything running on the system) you shouldn't give said person access to your system [23:55] Although it may not be necessary depending on how the phones/tablets/computers are authenticated to the tv [23:56] probably best to focus on the single device use-case first [23:56] we do want authentication. Random person can just use the remote, not access all your apps with their tablet [23:56] TV+remote control [23:57] AlanBell: Well I want to make sure this is compatible with the expanded use case, even if it's not implemented right away, otherwise there will be backtracking [23:58] ok [23:58] at least in general design [23:59] I would say for authenticated devices (however we authenticate them), allow #1, #2, or #3 to be an option. [23:59] for the phone/tablet linkup I would think that just a vnc connection (maybe a wayland thing . . .) to just clone the display to your phone/tablet would give you touch control for the big screen [23:59] picking whichever default seems sanest