=== vish is now known as evilvish [16:33] Pendulum: If you mean for unity, nothing definitive at this point, other than alpha 2. [16:34] TheMuso: are things still on track to make it in to alpha 2? (are you sprinting in Dallas, btw?) [17:37] Pendulum: Yes things are still on track. There are a few more people on board, as this is too much for one person alone, so things are going to move at a quicker rate in the coming weeks. [17:40] yay! [18:03] hola all ? [18:03] anyone here ? [18:04] codo: Hi. [18:04] hi codo [18:05] hello all. [18:05] hai TheMuso [18:06] hai codo [18:06] and leoquant [18:06] :P [18:06] so like If i want to get started on contributing to ubuntu accessibility where should I start ? [18:06] codo: Do you already have an idea of what you want to help with? [18:07] TheMuso: hmm may be some tools that help hearing impaired ? [18:07] codo: Right, do you know whether those tools exist already? [18:07] TheMuso: nope :P I'm a n00b in this domain :) [18:08] TheMuso: also i'm a hearing impaired guy from India. [18:08] TheMuso: it has been an ambition for me to get involved in some useful productive work to help some set of people because of my disability :) [18:08] codo: Ok, well I think the best bet at this point is to determine whether the tools exist for Linux already. If they do, then please let us know either via IRC, or the ubuntu accessibility mailing list. [18:08] We can ten look at ikntegrating those tools into the main distribution. [18:08] TheMuso: ah sure :) [18:08] codo: what sorts of tools would you find helpful? [18:09] (I'm genuinely interested as most of my friends with hearing impairments say that they don't generally have access issues with computers) [18:09] i already dropped a while ago the idea for sound vibration alphabet.every sound have his own vibration what can be felt [18:09] Pendulum: ah the most visible handicap is that we can't have transcripts for podcasts. [18:10] hajour: What hardware would do that? [18:10] Pendulum: when there is whole lot of videos on youtube that are interesting/useful we can't access them. [18:10] Pendulum: so say you want to hear to a google tech talk which is not even captioned how would you access it ? [18:10] if you put sound out loud it gives a kind vibration what deaf people feel [18:11] codo: Ah so you are thinking about speech to text recognition, and automatic text generation. [18:11] every hight have a differend vibration [18:11] hajour: Ah. [18:11] TheMuso: exactly. I doubt if people have achieved it now. [18:11] Just FYI, codo is a C programmer and gnome lover :) [18:12] it should be complete new so far i new anyway [18:12] Well speech to text is still very much a developing fied in open source, i.e there is not much that is possible at the moment. The commercial OSs still have way better software available in this regard. [18:12] I know nigelb who directed me to this list. [18:12] nigelb: Sweet. [18:12] :) [18:12] TheMuso: ah in that case are there any other ways i can contribute to ubuntu accessibility ? [18:12] TheMuso: like bugfixes or porting etc ? [18:12] TheMuso, you already seen the speechprogram to where we working on? [18:12] codo: Don't let what I said above put you off what you want to help with. [18:13] codo: the way to start may be to look at what speech-to-text programs need help. I think CMU Sphinx is in C [18:13] but I'm not sure about how to get involved in it or what they need help with [18:13] I suggest you get involved with one of the various speech to text projects that exist, and help them move forward with what you would like to see available. [18:13] Pendulum: ah ok. Will look at that [18:13] also i want to contribute to ubuntu /gnome accessibility [18:13] hajour: No. [18:13] even if it helps other disabled / blinds / spastics I'd be happy :) [18:13] so wher eto start :) ? [18:14] codo: CMU Sphinx http://cmusphinx.sourceforge.net/ [18:14] launchpad.net/~speechcontrolteam [18:14] sorry if i'm sounding like a total n00b here :) [18:14] ok thanks Pendulum hajour [18:14] codo: speech control is a new team that hajour's running to develop betting voice control for linux [18:14] I think it's less text-generation and more about controling the computer, but I could be wrong [18:15] Pendulum: ah ok so it helps motor impaired ? [18:15] to control computers ? [18:15] yep [18:15] yes we have people from simons listen festival and sonic [18:15] in the team [18:15] okay, i can help with that if i can learn the code base :) [18:15] codo: also, charlie-tca is in charge of bug triage for accessibility bugs so he may have a good sense of what bugs you might be able to help with :) [18:15] but for that I guess I have to find a way to uggle my 5 day work with 2 days of contributions to open src [18:15] ok [18:15] too much info to handle :D [18:16] * codo should hang around here a bit more. [18:16] otherwise in a few weeks when alpha2 comes out for the next ubuntu release, we're going to need loads of people to test ;) [18:16] on the dev. speechcontrol is more information to on launchpad [18:16] yes! do hang around here [18:16] I also need to announce a meeting time. I think a week from wednesday [18:16] yaa sure. :) [18:17] hajour: would you be able to do a meeting on the 19th at 21h UTC? [18:17] if you want you may join to in speechcontrol codo [18:17] TheMuso: ^^ [18:17] charlie-tca: ^^ (meeting time) [18:17] 19 jan you mean [18:17] AlanBell: ^^ [18:17] yes [18:18] yes that good to me [18:18] Pendulum: Yeah I should be recovered enough by then to attend that meeting. [18:18] Pendulum: meeting time is alright with me :) [18:18] UndiFineD: I kinda figured you and hajour were probably a packaged pair :P [18:18] Pendulum: find by me I think [18:18] :) [18:18] I will be in Belgium but should have connectivity [18:18] Pendulum: should be okay here [18:19] \o/ [18:19] * Pendulum schedules [18:19] TheMuso, i would appreciate if you wanted to look at the speechcontrol project and say what you think from it [18:19] hajour: ya sure. [18:19] hehe yes Pendulum [18:19] hajour: I will have to see the site first [18:20] hajour: also im complete deaf without my hearing aids [18:20] hajour: so hope i can still contribute to some parts where there is requirement to hear sounds etc ? [18:20] hajour: No promises any time soon, but I'll see what I can do. [18:20] Pendulum: so meeting ? [18:20] Pendulum: i live in India. I work all 5 days monday to friday :( [18:20] we have more people with health issues in the team codo [18:21] Pendulum: not sure if it affects my work schedule the meeting that is [18:21] codo: I suspect it will be middle of the night for you :-/ [18:21] ok thank you TheMuso [18:21] Pendulum: ya that is fine. [18:21] nigelb: can you translate 21h UTC into local time? [18:21] Pendulum: if it is weekend it will be more helpful :) [18:21] (India throws me off) [18:21] Pendulum: :) [18:21] it's actually mid-week [18:22] charlie-tca: is not here ? [18:22] oops forgot to put me on back again === Pendulum changed the topic of #ubuntu-accessibility to: Ubuntu Accessibility Team https://wiki.ubuntu.com/Accessibility for more info | http://lists.ubuntu.com/mailman/listinfo/ubuntu-accessibility to join the mailing list | Channel IRC logs: http://irclogs.ubuntu.com/ - Updated every hour | Next Team Meeting, January 19, 2011 21:00 UTC. Meeting agenda: https://wiki.ubuntu.com/Accessibility/Team/MeetingAgenda [18:24] hajour: so, like where should I Start off ? [18:24] codo: yes, here, sort of [18:24] codo, if you want you may come to speechcontrol chatroom if you want. [18:24] to become more informed [18:24] charlie-tca: so someone here was mentioning that you are the person to contact for bugs related to ubuntu accessibility [18:25] hajour: ok server ? channel ? [18:25] freenode ##speechcontrol [18:25] codo: yes. Most of the bugs we have are related to at-api [18:25] what is at ? [18:25] Pendulum, i wonder when we become # channel [18:26] its for hole ubuntu accessibility this program and what is attached [18:26] charlie-tca: atapi ? wasnt that cdrom ? :P AT-SPI [18:26] hajour: what are you looking for as a channel name? [18:27] if you're just looking for #speechcontrol , I think that's a matter of registering with freenode? [18:27] codo: the core components of GNOME Accessibility. [18:27] charlie-tca: ah nice [18:27] charlie-tca: so how do I learn to contribute to such stuffs and help around ? [18:28] UndiFineD: yes, you are right. at-spi [18:28] hajour: if you want #ubuntu-speechcontrol, I'd ask about process in #ubuntu-irc (unless someone here knows AlanBell? charlie-tca?) [18:29] personally, I wasn't a fan of #ubuntu-accessibility-speechcontrol because it's just soo long [18:29] if i understand it good it was depending to on the accessibility team [18:29] codo: I triage the bugs, try to get them ready for the developer to fix them. [18:30] I don't do any programming, myself. [18:30] * AlanBell agrees with Pendulum #ubuntu-accessibility-speechcontrol is a bit long [18:30] ah, I got confused in there. well, my opinion is that you should go ahead with whatever name you want, my only objection is the length of #ubuntu-accessibility-speechcontrol [18:30] charlie-tca: oic. [18:30] I would like to see it in the #ubuntu- namespace though, if it is an Ubuntu project [18:30] #ubuntu-speechcontrol would work [18:31] thats good for me ubuntu-speechcontrol [18:31] codo: you can look at bugs tagged a11y, that should be almost all the accessibility bugs. [18:31] AlanBell: do you know the procedure for getting an ubuntu namespace IRC channel approved? [18:32] AlanBell: Pendulum we have time to make that decision, next meeting would be fine [18:32] charlie-tca: ok [18:32] by the way, I hope you guys will come talk about speech control at the next meeting! [18:32] charlie-tca: thank you. So in case if I have some queries I can ping you etc ? [18:33] yes, feel free to [18:33] absolutely [18:34] codo: also, in terms of passive things you probably want to join the ubuntu accessibility mailing lists and the gnome one :) [18:34] the procedure for creating new channels is here https://wiki.ubuntu.com/IRC/CreatingChannels [18:35] and naming convention here https://wiki.ubuntu.com/IRC/ChannelNaming [18:36] Pendulum: ok. [18:36] hajour: UndiFineD if you decide you want #ubuntu-speechcontrol, I'm happy to talk to the IRC Council in support of that :) [18:37] ok thanks Pendulum :) [18:38] also, I've rejoined the ##speechcontrol channel [18:38] (I fell out of it in some weird computer and health stuff a few weeks ago) [18:39] hajour: UndiFineD if you decide you want #ubuntu-speechcontrol, I'm happy to talk to the IRC Council in support of that ツ within the set topic a linkage to accessibility? [18:39] leoquant: sounds good [18:40] great leoquant [18:40] yes, the topic should be adjusted ;) [18:41] btw, one thing I would suggest is that if you are in an ubuntu namespace ultimately, you should have an ubuntu logbot with public logs. I've found public logs to be really useful all over the place :) [18:43] Pendulum: while we understand the take from Ubuntu (CoC) there is much rivalry in this area, and we should hold of the big corp here [18:44] via rt@ubuntu.com brad marchall [18:44] ==>logbot [18:45] http://irclogs.ubuntu.com [18:45] its very important that this project is and stays open source [18:46] and that means package and license it first [18:46] leoquant, can you pm that things to me.completly lost here in that really much lines [18:46] UndiFineD: sorry, I'm not quite sure what you mean by that [18:47] I can explain at some other time, I have a meeting to attend [18:47] sorry and ok hajour [18:47] * AlanBell is puzzled too [18:47] UndiFineD: no worries. now that I'm mostly actually back around, I should be in better shape to answer :) [18:48] we are afraid big company s use what ubuntu have worked on.then it lose open source [18:49] I've never heard that concern when it comes to publicly logging channels before. In fact, in the Ubuntu community I'd guess that 95-100% of the technical channels are logged [18:49] when it is not longer open source they ask money for it to use it.then the main group of people who need it cant longer use it because they have no money for it [18:49] the only channels I don't know of being logged within the ubuntu community are ones where people are worried about personal things being discussed (such as Ubuntu Women's concerns about providing a safe space for people who are being harrassed) [18:50] indeed [18:50] packed means it stays open source if i bin good informed [18:51] *nods* [18:51] thats the first goal thought [18:51] hajour we could encrypt the logs, and even give this channel apassword... [18:52] and we have a very good team now [18:52] thats a good idea leoquant [18:52] a channel pass/ [18:52] I don't know if the IRC Council would be okay with that [18:53] if it are packed its no problem anymore thought? [18:53] I think it's kinda either you log or you don't log [18:53] in terms of logs that aren't personal logs on people's computers [18:53] advice is very welcome at this part [18:55] its that .... license part what could make us project fail [18:56] I don't know that the license bit is threatened by logging, though [18:56] i agree with Pendulum [18:56] and public logs will in fact strengthen ability to show "hey, that came from us" if it were to ever come up [18:56] wow too much backlog [18:56] to keep track with :) [18:56] im not sure where to start :d [18:57] I'm also not sure that speech control is something that a lot of commercial companies would be interested in since Windows and MacOS both have their own versions already [18:57] yes [18:57] and they are not free :) ? [18:57] I suppose ? [18:57] we are further then first Pendulum [18:58] hajour: *nods* [18:58] the GPL is what protects code and keeps it Free [18:58] right [18:58] we have now help from someone wo worked with that kind of things 7 years [18:58] logging chatter about code is nothing to do with licensing [18:58] AlanBell: +1 [18:59] the speech control landscape is actually rather complicated in license terms [18:59] thats 100% sure i hope [18:59] different things have dependencies with different licenses [18:59] you can have free software with a non-free voice file for example [18:59] so loong its not company s like mi so [19:00] wo have info from the project [19:00] to be in Ubuntu it generally needs to follow the Debian Free Software Guidelines (although there are exceptions) [19:02] hajour some "big"compagnies" gave their code away: IBM viavoice afaik their not evil perse [19:02] its the main important thing it is and stays open source [19:02] mm in netherland dislection bad sight program cost 1295 euro [19:03] i never forget that ever [19:03] thats more then a minimum month income [19:04] here [19:04] so important to do it good but also in ubuntu rules [19:04] hajour: I definitely think licensing open source is a good plan [19:05] help?this is bothering me al weeks how to do it good === API is now known as API_afk [19:06] AlanBell: do you have any good resources for open source licensing info? [19:06] hajour: I can try to help you look at licenses, but I have no idea which would be useful for speech control (although I kinda assume the GPL is likely too restrictive) [19:07] GPL might work for some things but not for others [19:07] but I'm happy to at some point go over with you what some of them say and translate from legal-speak into plain english [19:07] JanC: do you know of any good resources on open source licensing? [19:08] to understand licenses you have to start with the GPL http://www.gnu.org/licenses/gpl.html [19:08] all the others add and remove bits from that [19:08] : http://www.opensource.org/licenses/category? [19:09] the speech control licensing landscape is *complicated* because it is not a monolithic program that can be built [19:09] yeah [19:09] yep [19:09] it will be using other general purpose voice recognition and speech synthesis engines and libraries with different licensing situations [19:09] right [19:10] most of which I assume require things to stay open source anyway. [19:10] I don't think there is one resource that has all possible info about licensing, I got most of my knowledge from reading lots of discussions about it spread all over the internet... [19:10] the chances are that any code written would have to be something like LGPL, possibly with some exception clauses [19:10] hajour: it might be useful to start by reading the licenses of the programs you'll be building on top of [19:11] I can help advise on the likely licensing implications of certain combinations, but I fear it would end up as a long thread on the debian legal mailing list [19:13] which is where simon-listens is stuck, last time I checked [19:13] *nods* [19:13] I suspect the trick is to read through all the licenses of the base libraries and stuff [19:13] and then create a chart [19:13] and figure out which have what requirements, etc. [19:13] simon listens is resolved [19:13] it also depends on how thinks are used [19:13] excellent news [19:14] things [19:14] its only julius .. (forgot how to write it )now [19:14] bedahr have resolve simons listen [19:14] debian bug 596511 [19:14] Debian bug 596511 in wnpp "ITP: simon -- Open source speech recognition" [Wishlist,Open] http://bugs.debian.org/596511 [19:19] hajour: do you have a link to the resolution, the ITP seems to be open still? [19:19] JackyAlcine knows all of the details he have talked with bedahr.i was half here yesterday.(have a sick child here ) [19:19] sorry i know not yet all [19:19] i just bin here again just a hour have read loads of lines in chat [19:19] ok [19:20] and i have not yet read all pff [19:20] thats fine [19:20] but that part i had read already [19:20] hajour: the licensing issues are really not going to come from people wanting to make stuff proprietary, it is going to be interactions between different free software licenses [19:21] which is complicated and frustratingly slow, but not evil, just people wanting to do things right [19:22] o UndiFineD remember something [19:22] rx007 can speak japanese [19:22] we have rx007, who could communicate with the japanese makers of julius [19:22] grrr i was typing already [19:25] maybe the makers of Julius want to change their license a bit to make it less ambiguous? [19:26] as I understand that is a problem? [19:29] basically yes [19:29] slightly complicated by the fact that simon is a GPLv2 app with julius (BSD 4 clause modified) as a dependency with an exception clause to allow this [19:32] next to that, simon comes with its own version of an updated Julius, with patches. Bedahr tried to get them merged, but there is little to no response to the impressive improvemenets [19:33] looking at the copyright-holders (several universities), re-licensing might be somewhat complicated... [19:33] unless somebody has contacts there [19:34] speech research is basically done in three places [19:34] MIT (recognition) Edinburgh (synthesis) some japanese uni beginning with N [19:35] Nagoya [19:35] and there are some smaller projects like openMary from a German uni [19:35] thats the one [19:35] wow [19:35] I never knew this stuff is haappening :) [19:35] wow wow [19:35] i seriously hope something happens for deafs eventually [19:36] it isn't undergrad stuff, or even a regular regional university postgrad research area [19:36] I don't know if the Belgian university is still working on things (where MBROLA was developed)? [19:37] Faculté Polytechnique de Mons [19:37] Carnegie Mellon in the US was also working on it at one point (they're the people who did the Sphinx stuff) [19:37] I know some of those people went on to a French speech tech company [19:37] s/MIT/Carnegie Mellon/ [19:38] CMU is where it is at, not MIT [19:38] MBROLA is non-Free [19:38] "This synthesizer is provided for free, for non commercial, non military applications only." [19:39] and binary only [19:40] I know, but IIRC results from MBROLA can & have been used or something [19:40] voices created with tools that use MBROLA or whatever [19:40] yeah, that starts getting really complicated [19:41] and leads to debates on whether a voice file is data or complied object code [19:41] it's a bit comparable with fonts or images designed with non-free tools [19:42] yes [19:42] you can maybe edit them with free tools, but not in an optimal way [19:46] actually, the company based on MBROLA research is Belgian/French & is the company that made the voices for the Tux Droid: http://www.acapela-group.com/text-speech-contact.html (they still have an office in Mons) [19:53] AlanBell: looking at the HTK license, I guess the main problem is that you can't redistribute it? [19:55] yes [19:55] and sub-license which is the same thing [20:02] so is there a nice doc around for creating coc key ? [20:02] anywhere ? [20:03] : http://screencasts.ubuntu.com/taxonomy/term/89 [20:03] screencast team [20:04] duane hinnen [20:04] yaa it is slow [20:04] and i have slow internet :D [20:04] it hasn;t even loaded for me :) [20:04] I"m not sure if it's captioned, either [20:04] ok got it :) [20:04] Pendulum: yaa someone do it :) [20:05] i saw it, it works and is outstanding [20:05] Pendulum: it doesnt help for hearing impaired :) [20:05] leoquant: is it captioned ? [20:05] :O [20:07] i am on three meetings now sorry.... [20:08] http://screencasts.ubuntu.com/taxonomy/term/89 whats the meaning of this link?cant see or hear something on it [20:08] i can hear duane [20:08] i can see the pgp coc proces [20:08] strange i hear nothing [20:09] :/ [20:09] got all volume on max [20:09] the video and my eeepc [20:09] allso no image [20:10] strange hajour [20:11] tried it a minute earlier sound and visual [20:11] i really don't see or hear sanything [20:11] black image [20:11] no sound at all [20:12] mouse on the screen, left button? [20:12] youtube like [20:12] codo also problems? [20:12] already done leoquant [20:12] thats on max [20:12] ok [20:13] no sounds [20:13] and no image [20:13] sorry to hear [20:14] just reported it sorry [20:21] hajour: you can also download the video on that page [20:22] hm, bar moves across the bottom, but no video or sound here [20:22] same here charlie-tca [20:23] no problem here [20:23] but downloading & playing should work I guess [20:23] i don't need it JanC but for others who do needed it must work good [20:24] hajour: what browser ? [20:25] the one above in page has a moving bar the one below not .boths have no image and no sounds [20:25] Mine is ff3.6 [20:25] Chromium web Browser [20:25] hm, FF 3.6 works here [20:26] beta [20:26] no beta here [20:27] firefox version 3.6.13 [20:27] 3.6.13 too [20:27] it uses HTML5