[00:55] dcope, hello [00:55] tuner support atm is with anything that will work in the kernel [00:56] might have to fetch firmware here and there but alot of cards work [00:56] dcope, you do not need a cable box at all [00:56] you just plug your cable line into you Ubuntu tv [00:56] depending on where you live and the provider [01:17] bobweaver: interesting [01:18] i didnt' realize that you could plug the coax directly from the line-in to the tuner [01:22] hmm [01:23] it's a bit more compliated than that probably [02:08] tgm4883: that's what i thought... i thought providers had some sort of authentication that the cable box performed. [02:16] dcope, that is a part of it [02:17] if you are in the US, it gets very dependent on who your provider is, what tuner you have, and what part of the country you live in [02:17] and what channels you want to record [02:17] oh wow [02:18] well, ubuntu tv still looks pretty cool even if i can't get an epg & dvr [02:19] The best is if you use a cable card tuner and your provider doesn't lock down the channels [02:19] although an HDPVR can be said to be equally as good [02:29] will the tv package be released to install upon an existing ubuntu instance? [02:30] dcope, yes [02:30] that is my understanding anyway [02:30] awesome [14:46] are there ubuntu tv images ready for download? [15:38] user82: not images, but you can install it [15:38] dcope, okay thanks..i have a set top box with arm(i think) but then it will not be easy [15:38] user82: what settop to you have? [15:39] a "teufel mediastation". noone knows that thing but the update image was similar to other ones with custom firwares existent [15:39] linux kernel etc included... [15:46] cool [15:46] i have a roku now, but i am wanting to build a media center pc with ubuntu tv now [15:49] maybe that is clever..pc always keeps you flexible [15:50] yep [16:46] Saviq, is any of the remote control stuff you worked on regarding the issues with X? [19:24] bobweaver, so I've got the guide data scope partially done, but I'm not sure I like the current implementation of it [19:24] it takes way too long to query I think [19:24] also, i haven't made it a lens yet [19:27] jhodapp, mhall119, do you know if the scopes run all the time in the background? [19:27] tgm4883: currently yes [19:27] currently? [19:27] you noticed that eh? [19:28] they're working on making the Dash smarter about starting and stopping scope processes [19:28] mhall119, well, I think I need to have the guide data get pulled down periodically, even when the dash is closed [19:29] tgm4883: then you'll probably want that to run outside of the scope process, using cron or something [19:29] otherwise it's too slow when it needs to retreive updates (which i've set to every 12 hours) [19:29] hmm ok [19:30] mhall119, so my second is regarding vala [19:30] I'm told vala is faster than python? [19:30] I believe that's true, yes [19:30] Vala gets "compiled" into C, then compiled into a native binary [19:30] currently, even with the guide data local, it takes 4-5 seconds to return a search result [19:30] i'm thinking vala might make that faster [19:31] maybe, especially for large amounts of data [19:31] tgm4883, issues with X as in sequence IDs > 255? [19:31] Saviq, yea [19:31] mhall119, yea that is what I'm dealing with [19:31] 14 days of guide data for all channels you have [19:31] tgm4883, there's workarounds, one of which is a patched evdev driver that lets you remap keys from the > 255 space to the <= 255 one [19:32] tgm4883, and the other is hidmapper, a small tool that talks to the USB HID interface directly [19:32] and lets you map HID calls to X11 keys (again, <= 255) [19:32] Saviq, yea, we've (mythbuntu) got a ticket into the v4l guys hoping they will carry a patch to fix some stuff until X gets it all worked out [19:33] tgm4883, so it's http://forum.xbmc.org/showthread.php?tid=88560 or http://www.mythtv.org/wiki/Remapping_remote_control_key_codes_greater_than_255 [19:33] Saviq, yea that is pretty much what we were working with as well [19:33] trying to drop lirc [19:33] tgm4883, looking at https://bugs.freedesktop.org/show_bug.cgi?id=11227 it's not promising [19:34] we might get wayland before then... [19:35] :/ [19:38] mhall119, I don't suppose vala can talk to python directly [19:38] no [19:38] bummer [19:38] I'll need to figure out a way to do previews then without recoding a bunch of stuff [19:38] there's always dbus :) [19:39] I'm not super familiar with doing stuff over dbus [19:39] but it might be worth looking into I suppose [19:39] tgm4883, re python vs. vala, I wouldn't bet vala will get you anywhere close to "fast enough" [19:39] tgm4883, proper data storage will [19:40] Saviq, proper data storage? [19:40] it's all json atm [19:40] tgm4883, exactly [19:41] Saviq, there is a better way to store that info in python? [19:41] tgm4883, I'd say there's many [19:41] tgm4883, re "if scopes run all the time" - yes, but data is only requested when the relevant scope is shown [19:42] tgm4883, i.e. only when the scope gets a viewType = Visible set [19:42] only then will the scopes be queried for data [19:42] that isn't to say that the scope might prepare that data in memory before then [19:42] Saviq, currently, it pulls json guide data into memory on a search. The first search takes 30+ seconds, but a second search takes 4 [19:43] so the guide data is in memory at that point [19:43] and from memory, it takes 4 seconds to search, which IMO is too long [19:43] tgm4883, yeah so here the biggest issue is that we transfer the whole set of data [19:43] tgm4883, instead of just the relevant part that's being shown [19:43] i.e. pagination [19:43] yeah, partitioning the data when you download it would help [19:43] Saviq, as in, not 14 days of data? [19:44] mhall119, not even that [19:44] or throw it into an sqlite db or something? [19:44] tgm4883, yes, currently there's no way to paginate in the lenses/scope infrastructure [19:44] mhall119, there's basically multiple places that need fixing [19:45] tgm4883: also remember that anything you add to the search results model is being serialized and passed over dbus [19:45] I don't think partitioning the data would work. For just viewing the guide yes, but searching would only see data in the loaded partition [19:45] doing that to 14 days worth of channel data is going to be slow in any language [19:46] i'm kinda leaning toward syncing to a local sqlite table, then on search just querying that [19:47] then I can do 1 day of data for the default "" search, although I'm wondering how much data was being passed over dbus for the searches [19:48] because while it was 14 days of data, I'm not returning everything to the lens, just the relevant stuff [19:48] which is maybe 20ish results [19:48] so it sounds more like it takes way too long to search though 14 days of data [19:49] tgm4883, yeah, that's 'cause you don't have indexing, even if you have it all in memory [19:49] it just goes through all of that data to compare [19:49] Saviq, sqlite should fix that, or is there a better way [19:50] it'll be faster, but I'd look at some proper free-text search engines like lucene [19:52] tgm4883, there's also http://voices.canonical.com/jussi.pakkanen/2012/12/03/introducing-libcolumbus-a-fast-online-approximate-matching-library/ [19:53] which should help with the searching in memore [19:53] memory [19:58] Saviq, is that something I would need to use explicitly, or is that making it's way into unity? [19:59] tgm4883, it can't really make its way into unity, it needs to be as close to the data as possible [19:59] ok [19:59] tgm4883, so in your case you'd initialize libcolumbus when initially reading the data from json [19:59] or from whatever storage you choose [20:00] then at least searching will be faster - you'll just need to have an ID by which to identify a data record [20:14] tgm4883: libcolumbus needs to be pre-populated with a list of "correct" values that it uses to match a submitted "incorrect" value against [20:14] so you'd need to initialize it with a list of program names, then you can give it a search term with typos and it'll get you it's best match [20:17] tgm4883, I've got an example program written in C++ that uses libcolumbus if you're interested [20:49] Wonder if anyone will integrate Google play (music) into ubuntu tv? [21:40] dcope: does it offer an API? [21:44] mhall119: i'm not sure [21:44] dcope, what are you 2 talking about ? [21:45] dcope: if it does, someone can write a scope for it, and it'll be available in Ubuntu (desktop, TV, and phone) [21:45] google music [21:45] ahh [21:45] IIRC, it has a non-public API [21:46] non-public? [21:46] meaning you have to authenticate using OAuth or something? [21:46] yeah Unoffical [21:46] https://github.com/simon-weber/Unofficial-Google-Music-API [21:46] oh, unofficial, that's different [21:46] requiring OAuth would be fine, we could make an Ubuntu Online Accounts provider [21:49] i'm so ready to get ubuntu tv installed on something to contribute and test [21:50] dcope, I have a 3rd party ppa if you like [21:50] for 12.04 [21:50] bobweaver, is that on the wiki [21:50] Not this one [21:50] bobweaver: sure [21:50] ~u2t/bleedingedge [21:50] Must be 12.04 [21:51] i'll get it in a vm now and then put it on a dedicated machine once i get one [21:51] * tgm4883 sees the irony in that [21:51] tgm4883: ? [21:51] https://launchpad.net/~u2t/+archive/bleedingedge [21:51] bleedingedge [21:51] Must be 12.04 [21:51] oh [21:52] what it is a move up from 11.10 [21:52] true [21:53] tgm4883, even has settings and loves you orig scope :) [21:53] bobweaver, sweet [21:53] tgm4883, AORN http://www.youtube.com/watch?v=qb8s7pSBu30&feature=youtu.be [21:53] is it running up to date scope? [21:54] I just made that video 3 minutes ago or something like that [21:54] bobweaver, I've been working on the guide data scope, but I've got some issues to work out [21:54] too much data [21:54] II will be finishing up the settings and making more and more myth stuff [21:54] tgm4883, dont worry about the guide if it is giving you that much trouble [21:55] bobweaver, nah, i've got it worked out I think with the help of mhall 119 jhod app and Sa viq [21:55] I can write a html5 app that connects to myth backend and read from that and launch also [21:56] I haven't done any lens work on it though, and IDK the status you got on the different layout [21:56] I have 6 different layouts now [21:56] only 3 are stable at this point [21:57] I kinda got caught up doing opengl stuff last night and particals and shaders so that set me back a little [21:58] bobweaver, ok, i'm going to work on speeding this thing up with the fixes we discussed today [21:58] as it takes too long for search results right now [22:00] bobweaver: I like the way you did the lightdm options [22:01] Yeah that is what I got all caught up in lol [22:01] I am posting a video of some experamental stuff [22:02] Turn down Volume [22:02] http://www.youtube.com/watch?v=Acjb5fkiCB4 [22:04] btw that reads the weather outside and returns icons that are preset for weather [22:04] I am going to remove it all [22:04] well the weather part and what not and just have a weather app [22:05] But yeah if it was sunny outside it would have shown little suns flaying around and not snow [22:06] because you can only call api so may times a day it uses cache for ever 6 hours [22:07] mhall119, do you think that it is a good idea to scale the icons in lightdm settings menu ? maybe even pulse them ? [22:19] bobweaver: I have no idea, maybe ask one of the design guys on G+? [22:27] thanks I am mhall119 :)