[06:21] the last linaro with hardware acceleration was 12.03 [06:22] do you know how long it may take to have it in new versions? [12:28] jhodapp, hi [12:28] jhodapp, I realized you also posted on gst-devel [12:28] jhodapp, yes, you want to use gstreamer 1.0 [12:28] not only is most of the development concentrated there, but you'll also have serious performance gains on embedded platforms [12:32] bilboed: excellent, thanks for that feedback [12:34] bilboed: I'd read that, and that's really great to know someone with STB-like experience can speak to that [12:35] I'll give you a simple example : hw-accelerated decoding/encoding/display works out of the box with pandaboards [12:35] instead of having to butcher/patch gstreamer to make it work correctly [12:36] and another pro-tip if you are at the early stages : stay away from (gst-)openmax as much as possible [12:37] best case scenarios : it'll give you headaches. worst case scenario : you won't be able to have 100% accelerated playback/display [12:43] bilboed: ok, good tip [12:43] bilboed: so what was added to the 1.0 series that makes you say those things work out of the box? [12:44] becase it just does :) install gstreamer 1.0, install ducati gst plugins, works [12:44] where gstreamer 1.0 is from upstream (and not a modified version with patches for that specific platform) [12:45] it would take some time to explain the full details of the changes [12:46] well I just mean a summary [12:48] this is a wip summary : http://gstreamer.freedesktop.org/wiki/ZeroPointEleven/FeaturesArticle [12:48] and a talk by yours truly : http://video.linux.com/videos/gstreamer-10-no-longer-compromise-flexibility-for-performance [12:49] cool, thanks [12:50] and a slightly older one by wim : http://gstconf.ubicast.tv/videos/keynote-gstreamer10/ [19:21] tgm4883, you around ? [19:21] wanted to talk to you about a scaper and what not for the tv [19:21] get rid of xbmc as 1) it takes too much time and does not scan right [19:22] wanted to see if you had made or have friends that might have made something like this with tvdb and moviedb [19:23] If not I am just going to write a bash script for it [19:33] bobweaver, yea [19:34] bobweaver, yea we can do that fairly easily, but the results we'll get are XML [19:34] so we need UTV to be able to read that data [19:34] likely from a DB [19:36] bobweaver, I haven't heard much on the UTV front recently, is there work being done to get metadata out of the crappy .nfo files and into something useful? [19:36] eg. UTV/Unity needs to not look for .nfo files [19:38] tgm4883: the goal is definitely to get rid of .nfo files and replace it with a proper metadata index system [19:38] jhodapp, does UTV/Unity support metadata from other sources? [19:38] eg. mysql [19:39] tgm4883: not today, no [19:39] well UTV [19:39] jhodapp, is that something bobweaver can work on? [19:40] since he's getting familar with the code [19:41] jhodapp, or rather, are the plans for that public? eg. If we know it's going to a mysql database, we can work on the stuff needed to get the data into that DB [19:41] It's a bit premature for uTV [19:41] for Unity there's nothing stopping that since a scope can use anything as a data source [19:41] so really, we need to be looking at the videos scope then? [19:42] well what are your goals? [19:42] I'm guessing bobweaver's goals were to completely remove the requirement for XBMC [19:43] since the need for that is completely metadata based [19:43] so he pinged me to see how mythtv did it, which we have python scripts to pull the metadata down using the available API's at those websites [19:44] but the data is all XML, so if we already would have to do work to get it into the .nfo file, we might as well just get it into whatever the planned destination is and work on the scope to pull it from the right place [19:44] tgm4883: ah, that part I can speak to...our current goal is to use Grilo for this [19:45] * tgm4883 sighs [19:45] so if someone wants to create some additional Grilo plugins for certain online metadata sources not already created, that work could prove useful [19:46] tgm4883: why are you sighing? [19:46] jhodapp, this is the first I am hearing of grilo's use here [19:46] and...? [19:46] am I not subscribed to the right areas for discussion? [19:47] I'm always here and subscribed to the UTV mailing list [19:47] tgm4883: it has only been discussed between Saviq and I for now [19:47] yea, that was what the sigh was for [19:47] so you are on the cutting edge [19:47] I don't have an issue with grilo (how can I, I don't know anything about it yet), but there doesn't seem to be much direction/leadership in this project [19:48] tgm4883: no need to sigh...you're really not missing much of anything yet [19:48] and that makes me sigh :/ [19:48] tgm4883: I understand your frustration...I share it [19:48] I'd ask for an update on the Unity 3D move, but I don't think I'd like the response [19:48] the project has some issues to work out before we pull all stops out for implementation...for example...I haven't really begun implementing myself yet [19:49] tgm4883: though I don't know many details on the move to Unity 3D, that part has the most traction [19:49] jhodapp, AFAIK, there still hasn't really been any information on any of the DVR type stuff [19:50] which also makes me sad [19:50] tgm4883: yes that part is particularly frustrating [19:50] jhodapp, so who do I have to buy a drink to get this info? [19:50] I haven't seen willcooke here in a while [19:51] he pops in, but he hasn't said much in here...you can ping him any time he's online [19:51] is he still the project lead? [19:51] he's UTC +1 [19:51] which is like 10PM right now [19:51] maybe I'll just shoot him an email [19:51] tgm4883: he is the business manager, Saviq is the technical lead [19:51] ah, so I need to buy Saviq a drink then :) [19:52] wassup? [19:52] lol [19:52] Apparently I need to buy you a drink [19:52] yes! [19:52] I was about to tell you that [19:53] jhodapp, (psst. why?) [19:53] Saviq, so I'm hoping you can give some direction. jhodapp answered a bunch of questions, but there are still some unknowns [19:53] Saviq, there doesn't seem to be much information coming out regarding this project [19:54] I'm not looking for timelines on things, but there are a bunch of people ready to help in their own areas, but we don't know what to do, because there isn't any specification we're working toward [19:54] for instance, regarding things like metadata lookup. We just learned that Grilo is going to be used for that [19:55] I haven't looked at that much, but it doesn't appear to have any metadata grabbers for TMDB or TTVDB websites, which we could certainly write if we knew we were using Grilo [19:56] We don't know the deal with DVR lens/scope. (we've discussed this before, but no conclusions) [19:56] tgm4883: personally I think that's a great idea [19:57] tgm4883, we will definitely try and communicate as much as possible, as soon as we know it [19:57] jhodapp, I'm fine with whatever is chosen to do metadata lookups, provided it's not super terrible. I'm even willing to work on plugins to add support. I just need to know what we're doing there ;) [19:57] tgm4883, which is now much closer than it was even a week ago [19:57] heyguys just got back reading ^^ [19:58] tgm4883> bobweaver, yea we can do that fairly easily, but the results we'll get are XML << UBTV uses QML which uses Javascript this would tie in nice [19:59] Saviq, A very large issue with the way information is being handled right now, is that it can be interpreted one of two ways. 1) Everything is done super secret and when done will be released to the public (see landscape, launchpad) 2) The project is dead [19:59] tgm4883: like I said before, Grilo is the likely choice but not 100% yet...you are on the cutting edge [19:59] Saviq, neither of those are good things [19:59] eg. UTV/Unity needs to not look for .nfo files << I have fixed that [19:59] bobweaver, you'll want to keep reading, as it seems things keep changing [19:59] will do [20:00] tgm4883, there's also a third option - c) we're still in the process of defining all that [20:00] yes, I know it's been long, and we feel the same [20:00] too long [20:00] * jhodapp agrees [20:00] the biggest setback was the decision to switch to Nux [20:00] Saviq, which is unfortunate, since we're in the second dev cycle since anouncing the project [20:00] tgm4883: how long is a cycle to you? [20:00] jhodapp, half a year [20:00] jhodapp, 6 months [20:00] jhodapp, remember it's Ubuntu [20:01] oh right, that's what you mean [20:01] tgm4883, truth is we're not thinking in Ubuntu cycles that much, at least not yet [20:02] the biggest setback was the decision to switch to Nux I dont care about that anymore [20:02] I tried to do that for a week [20:02] Saviq, thats fine, but everyone else is [20:02] the lib is not strong and paints things weired for me [20:02] well not as strong as qt [20:02] bobweaver, that should be better every day now [20:03] but untill then I want my tv [20:03] especially since the decision was made to stick with Nux, much more manpower goes into it, especially regarding stability and documentation [20:03] so [20:03] I have made new stuff for UBTV all QML [20:04] if you look at my branch you will see that there is abour 12 new "pages " [20:04] pluged in networkmanager [20:04] bobweaver, yeah, I saw some of that [20:04] bobweaver, you need to be aware, though, that you're alone with that [20:04] bobweaver, unity-2d is basically getting decomissioned [20:04] made a TV page that is all for youtube watchseries and project free tv which is all have lens [20:05] yeah but untill things are stable .... [20:05] already the lp:ubuntutv code is incompatible with 12.04, it's only going to get worse [20:05] that is a shit ton of work porting all that qml to C++ [20:05] Saviq, as much as I hate blogging, this project needs a weekly update blog [20:05] tgm4883, don't they all? [20:06] Saviq, this project more than others [20:06] plus it is harder and longer work to do the NUX route there is no good IDE ect [20:06] tgm4883: would you be willing to help me in that effort? [20:06] jhodapp, blogging? [20:06] bobweaver, don't get me wrong, I co-wrote lp:ubuntutv [20:06] tgm4883: yes [20:06] and the decision was a tough one for me as well [20:06] jhodapp, what kinda help do you need? [20:07] tgm4883: first of all, helping make sure that it happens weekly would be a great start [20:07] I also like the idea of nux too. But to port this much code is going to take 6 months [20:07] only that I didn't really have a say - we needed to choose a single technology, and the choice was Nux [20:07] at least [20:07] bobweaver, it would've taken that much to take the QML stuff production-ready as well [20:08] tgm4883: second, you can even help me write blog entries...I'd be happy to have a quick online meeting with you weekly to give a status report of the project [20:08] but at least there is framework that is soild [20:08] it was a proof of concept, sure it went surprisingly well, still it was a long way away from being ready [20:08] ehhh.. not a biggie. [20:08] so true ^^ [20:08] Saviq, that is ^^ [20:09] bobweaver, tgm4883, guys you can be sure we will be communicating as much as possible, as soon as possible [20:09] we really want to be as public with all that as possible [20:09] indeed [20:09] I am sure of that but .. there is alot of work to happen [20:10] bobweaver: you're telling me :) [20:10] jhodapp, let me ping a blogger buddy of mine [20:10] and it is good work but. there is no code .... [20:10] bobweaver: design comes before code [20:10] there is so much to paint and nothing to paint on :) [20:11] bobweaver, there will be, real soon, we have been nailing things down these past weeks [20:11] there are things that I am seeing with my tv that are going to get in the way v.soon [20:11] and are really trying to get our hands dirty now [20:11] like if other's want scopes that are not in the SimpleLauncher [20:12] which will be the launcher when C++ comes into play [20:12] there needs to be "Landing Pages " for scopes [20:12] like one for TV [20:12] One for Music [20:12] ect [20:12] that way there can be lens everywhere [20:13] but one will not have to choose for 50 lens on the launcher [20:13] like in music pandora scope local scope grooveshark scope [20:14] jhodapp, pm? [20:14] bobweaver, of course, the number of lens needs to be limited [20:14] bobweaver, much more so than on the desktop [20:14] Saviq, let me take a screen shot of what I am talking about [20:15] Saviq, I agree, there needs to really be only like 6 lens [20:15] maybe less [20:15] tgm4883, I'd say less [20:15] tgm4883: pm? [20:15] tgm4883, videos, music, photos, tv, maaaaybe contacts [20:15] tgm4883: oh sorry, private msg [20:16] OT: yay, a PR1.3 update for my stillborn phone :D [20:16] see here is the thing thou gents to have scopes of this magnitude you need to have lens also [20:16] Music, Videos, Pictures, tv [20:16] not just one scope per lens [20:16] Saviq, heh, we have the same idea's [20:16] I'd probably forgo contacts as a lens in TV though [20:16] because the lens need to use css to grab stuff [20:17] bobweaver, of course [20:17] bobweaver, css? [20:17] tgm4883, I only mean for VoIP [20:17] the python code to grab from each site there stuff [20:17] brb [20:17] tgm4883, might be a separate app, though, but would be nicer to have all your contacts aggregated, not split for skype, google chat, jabber etc. [20:18] http://imagebin.org/219361 [20:18] take a look at that [20:18] each one is its own lens [20:19] that is what I mean by "Landing page " [20:19] Saviq, I can see a reason for contacts or applications lenses, but probably not both [20:19] Saviq, I'd lean towards applications rather than contacts though [20:19] the user should be abl to install whatever lens they like [20:19] bobweaver, sure, lenses should be apps, basically [20:19] yeah [20:20] tgm4883, I [20:20] 'd rather get rid of apps in mine [20:20] but whatever [20:20] so if I click on watchseries it loads the lens [20:20] I have a altered apps launcher also that I am wprking on [20:20] Saviq, I'd agree, but then you might have apps that don't have a reason to be associated with contacts [20:20] that wraps the apps in the dash [20:20] which is why I went the other way [20:21] I think that each "Launcher" should be to a landing page that has Launding page [20:21] Landing * [20:22] this way other lens to not get mucky like the video lens and the video lens and also the other video lens. aka the Ubuntu tv video lens the Ubuntu 12.04 video lens the Youtube Video Lens [20:23] launcher==>laundingpage ==> lens ==> dash [20:24] there is just to many cool lens not to do it that way [20:25] bobweaver, so by "landing page" you mean an intermediate stage [20:25] not sure [20:26] this is what I did [20:26] where you select the ~content type, then the lens [20:26] I hacked Home.qml [20:26] so that lens are only showing up [20:26] correct Saviq [20:27] bobweaver, I'd say that's too much complication, why not just have the scopes bind into the Video lens? [20:27] like select Videos "Launcher " and you get page that can take you too a number of lens [20:27] then you can select the source via filter s [20:27] -[ ] [20:27] because it will not work that way [20:27] what won't? [20:27] and why won't it? [20:27] some of the scopes Need there own lens [20:28] like the youtube one and watchseries one this would also solve are 12.04 video lens trouble [20:28] youtube is already included in the 12.04 video lens [20:28] rename the Ubuntu tv lens to ublens ect [20:28] that is a joke right ^^ [20:28] There needs to be a single Videos lens for Ubuntu TV/Desktop/etc [20:29] +1 ^ [20:29] bobweaver, youtube works in my 12.04 desktop videos lens [20:29] If you go that route you are going to miss out on code that a=is already out there [20:29] and re-inventing the wheel [20:29] bobweaver, it's already there for the video lens, too [20:30] the Ubuntu video lens has Nothing on my youtube lens [20:30] bobweaver, the whole idea behind Unity is aggregation, unification [20:30] we don't want to have separate lenses for the same content type, unless necessary [20:30] sure but that is impossible how can css be read to render to screen from one site and another in one lens [20:31] css? [20:31] bobweaver, not following [20:31] why are we rendering websites? [20:31] Saviq, oh, I get what he is saying [20:31] to get Icons and all sorts of info [20:31] Saviq, the questions was, how is it possible to render two different websites in a lens [20:31] ok so on the lens that I just made [20:32] bobweaver, that's what scopes are for, each is specific to a backend, there's a scope for youtube, a scope for vimeo etc. [20:32] project free tv [20:32] Saviq, which is a misunderstanding of how that works [20:32] I know how to make lens guys and scopes [20:32] What's with all this activity tonight? =P [20:33] MrChrisDruif, people having grief with us not telling what we're doing ;) [20:33] bobweaver, I'm just not understanding why we need to render any website at all? [20:33] I can completely understand [20:33] It's like a secret group within a community, making their own decisions [20:33] the lens just shows whatever the scope is offering it [20:34] the scope takes the data it gets from the website and condenses it down to just the info that the lens needs [20:34] no need for css [20:34] if a.get('href').startswith("http://http://www.free-tv-video-online.me/internet/"): if counter % (2): [20:34] how to pull the links with out css ? [20:34] for a in tree.cssselect('tr td a'): [20:35] bobweaver, so first thing is: we will never support scopes that do site scraping [20:35] if only for compliance with terms of service [20:35] bobweaver, you're misunderstanding. The css and all that would be handled by the scope, not the lens [20:35] but 3rd party ? [20:35] tgm4883, no I am not [20:36] bobweaver, which is why you are doing it wrong [20:36] bobweaver, the scopes are responsible for translating "websites" into: title, description, image [20:36] bobweaver, the lens only sees those three [20:36] you can not have a bunch of scopes calling and scraping it will crash I have tried [20:36] bobweaver, of course you can [20:36] bobweaver, that's exactly what they're there for [20:36] bobweaver, that is what it does right now [20:36] I have tried over and over again and get nothing [20:36] bobweaver, that means you did something wrong [20:36] could be [20:37] bobweaver, a single lens calls into whatever number of scopes you have installed [20:37] I know that [20:37] I have wrote lens and scopes [20:37] I understand how they work or at least the idea [20:38] you can scrap all them sites in one scope then parse to vlc ? [20:38] no [20:38] sorry one lens [20:38] bobweaver, no, one scope per site [20:38] well, maybe [20:38] sure you could, but that's not the idea [20:38] one lens references multiple scopes [20:38] the idea is: one source, one scope [20:38] one lens, multiple scopes [20:38] the aggregation happens on the lens level [20:39] all scopes query in parallel (which should help speed up data flow) [20:39] so you are saying that if I take my three scopes youtube watchseries and prokect free tv out in /unity/lens/videos they will all work with native UBtv video scope ? [20:39] bobweaver, yes [20:39] no workie [20:39] bobweaver, they should, at least [20:40] ok So I just tried to put Ub TV scope into 12.04 video lens and no workie [20:40] bobweaver, can't say why, but that's how it should work [20:40] renamed dbus also [20:41] bobweaver, you can see there's a video-remote.scope there [20:41] correct [20:41] bobweaver, did you implement the "source" filter in your scopes? [20:42] this is where if you add libunity picks up on dbus path [20:42] correct ^^ [20:42] I did [20:42] bobweaver, can't really say what happened, but that's exactly the path you should take [20:42] if it didn't work, there's something wrong, but not the general approach [20:43] I fought with that for a month [20:43] to get both 12.04 video and also Ubuntu tv video to go togeather [20:44] no workie maybe me but the minute I make new lens Name and rename things both scopes work [20:44] bobweaver, so the ubuntutv video is different in the sense it implements both the lens _and_ the scope in the same process [20:44] mostly because it's a quick hack-up of a lens [20:44] ahh [20:45] bobweaver, so you would need to chuck the lens parts of it out, and make it a proper scope [20:45] I hade to rename everything in Ubuntu tv scope and now both work for 12.04 [20:45] correct one that ties to the video scope [20:45] just write into that scope [20:46] This was my thought 1 month ago [20:46] it is hard to do [20:46] maybe I am not skilled enough [20:47] http://imagebin.org/219365 [20:47] bam [20:47] that is how it works right now with the videos scope, only there are a lot more videos scopes in use by default [20:47] tgm but local on both scopes do not like ech other [20:48] bobweaver, so make the scopes not local? [20:48] I have looked at that code over and over again but get no where [20:48] bobweaver, I'm no expert on it, but the mythtv scope is a separate scope from the videos lens [20:48] tgm4883, I am talkingabout taking the scope from Ubuntu tv and implanting into that I tried for a month [20:48] no expert on scopes, I'm an expert on the mythtv scope ;) [20:48] it is not as easy as it seems [20:49] bobweaver, you know http://developer.ubuntu.com/resources/technologies/lenses-and-scopes/ I presume? [20:49] yes [20:50] bobweaver, for the mythtv scope, I actually had to write a third utility to query the API [20:50] not 100% that was necessary, but it's doable [20:50] bobweaver, afraid there's no better source, but maybe you should go through that again focusing on the "remote scope" approach [20:50] IIRC, to do the mythtv scope, I basically just learned from the youtube scope [20:51] Yeah it might just be me. But at any rate I have both of them running just as 2 lens and not one [20:52] hmm, apparently lunch was 2 hours ago [20:52] lol [20:52] I blame Saviq for showing up and answering all these questions :P [20:52] I apologize, will now go away and not show up again during lunch time ;P [20:53] I do say the minute that there is Nux code I am all ready to paint but until then I am going to keep on making qml pages [20:53] it's actually 11pm here, so I've to go [20:53] no need for me to waste more time [20:53] bobweaver, if you need help with the scope, don't hesitate to ask, if not me, I'll try and find someone who can help [20:53] all I want is a good looking projector :) [20:54] thanks Saviq [20:55] so there is one other thing that I would like to say about the tv [20:55] the Video Preview [20:56] this should have "dialog" pop ups that show the movie actors fan-art in background and also the nfo files that javascript is reading need to be placed in a background image with a scollbar [20:56] it runs of the page :) [20:56] that is all I ahve to say thanks guys [21:01] bobweaver, so to answer your initial question, it looks like we need to create grabbers for grilo to get metadata from the right sites [21:03] yes that's right [21:08] eye-yia-yai [21:08] what is wrong with moviedb and tvdb I am looking at grilo now [21:09] bobweaver, grilo is a framework for pulling metadata (among other things) [21:10] it seems similiar to Ubuntu TV [21:10] you query for metadata for a piece of content, and it queries a bunch of places [21:11] tgm4883, you are right about a weekly blog [21:11] lol [21:11] bobweaver, it's going to be done [21:11] starting this friday [21:11] I just spent two days reading up on tvdb and moviedb [21:12] bleeding edge = not working for everyone [21:13] Meaning I do not see the 2 most imporant places that meta data has to be hade [21:13] moviedb and tvdb [21:14] bobweaver, currently those plugins for grilo don't exist [21:14] so this means altering gnomes code which means reinventing the wheel again [21:14] so we'd have to write them [21:14] bobweaver, grilo works with plugins [21:14] you mean re-write [21:14] no [21:14] I mean plugin [21:14] a plugin that is already there in other programs ? [21:15] yea [21:15] I mean why girlo I should say [21:15] because this..... [21:15] If you use grilo, then when you write ubuntu TV code and you need to get metadata, you code "ask grilo" [21:16] if you don't, then when you write ubuntu TV code and you need to get metadata, you code "ask ttvdb", "ask tmdb", "ask apple movie trailers", etc [21:17] er [21:17] 15 ago (make new grilo code so that 20 apis will be one ) 15 minutes later there is now 21 apis [21:18] no, from the Ubuntu TV perspective, there is 1 API [21:19] so that is the plan then ? [21:19] apparently so [21:19] make new plugins for grilo for moviedb adn tvdb [21:19] any other ? [21:20] IDK, I only care about metadata at this point and I don't know of any better sites for metadata [21:21] https://live.gnome.org/Grilo/PlannedFeatures [21:22] lookslike there are plans already [21:22]