**** BEGIN LOGGING AT Mon Mar 17 02:59:58 2014 Mar 17 05:27:30 morning Mar 17 05:54:41 mranostay: ping Mar 17 09:59:22 Need some help with kernel headers. Failing to compile kernel module. Anyone who can help ? Mar 17 09:59:29 Ping me. Mar 17 10:02:44 ? Mar 17 10:02:51 state your problem :) Mar 17 10:18:32 Hey. Mar 17 10:19:10 I was able to run a kernel module on my ubuntu and now i wanted to test it on beagle bone Mar 17 10:19:22 It is also running on ubuntu 12.04 Mar 17 10:20:01 To compile the module, i installed headers but it doesn't look like the installation went well Mar 17 10:20:08 Not able to compile. Mar 17 10:27:49 https://groups.google.com/forum/#!topic/beagleboard/l5RLNUGYrAc Mar 17 10:28:03 I tried this to install headers. Mar 17 10:28:21 I can send you the installation and error log if required. Mar 17 10:33:30 installed headers on bone? Mar 17 10:33:32 or on your PC? Mar 17 10:34:45 It worked on my PC. Shifting everything to Bone Mar 17 10:34:54 Issue running on bone. Mar 17 10:36:16 so you are running ubuntu on the bone? Mar 17 10:37:19 Yep. Mar 17 10:37:42 Do you have any specific recommendation for OS on bone ? Mar 17 10:39:48 if you go with the rcn stuff, I would suggest you use his debian build Mar 17 10:39:55 or, stock angstrom Mar 17 10:40:21 the beagleboard team cannot support random ubuntu builds Mar 17 10:41:13 https://github.com/beagleboard/kernel/tree/3.13 Mar 17 10:41:27 need to check which is the right kernel version to use Mar 17 10:43:11 I will go with the debian Mar 17 10:43:52 Just a sec. I will send you the log. If it something obvious may be, it saves some time. Mar 17 10:45:24 http://beagleboard.org/latest-images/ Mar 17 10:45:44 please use this one Mar 17 10:45:49 (the non-flasher) Mar 17 10:46:46 Sure. I will try this. Mar 17 10:46:52 Thanks :) Mar 17 11:00:05 Does these builds come with kernel header out of the box ? Mar 17 11:00:34 Can I compile modules directly m Mar 17 11:00:40 ?* Mar 17 12:16:20 praveendath92_: u can install the headers afterwards Mar 17 12:16:44 sudo apt-get install *insert magic* Mar 17 12:24:43 vvu|Log: ping Mar 17 12:25:27 praveendath92_: pong Mar 17 12:26:05 I didn't get that. 'Can't install headers afterwards' Mar 17 12:26:26 the headers are available as a installable package Mar 17 12:27:06 Okiee. Apt-get or different image file ? Mar 17 12:28:58 apt-get Mar 17 12:30:20 Actually, I installed as an apt-get package on my PC Mar 17 12:30:43 Can you tell me the name of the package for bone ? Mar 17 12:30:52 Ubuntu or Debian ? Mar 17 12:32:17 sudo apt-get install linux-headers-$(uname -r) Mar 17 12:32:30 I tried that Mar 17 12:32:50 It says package not found on regex search Mar 17 12:33:01 Something like that Mar 17 12:33:24 Let me try again. I will be able to tell you the exact error Mar 17 12:36:41 Couldn't find any package by regex 'linux-headers-3.8.13-bone30' Mar 17 12:37:02 so u need the headers for 3.8.13 Mar 17 12:37:19 This worked on my PC. Infact I wrote and tested some modules. Mar 17 12:37:22 Yeah. Mar 17 12:37:58 I did some search and found this Mar 17 12:38:01 https://groups.google.com/forum/#!topic/beagleboard/l5RLNUGYrAc Mar 17 12:38:25 av500 asked me try the latest debian build Mar 17 12:38:32 Doing that. Mar 17 12:38:52 so u have there the http://rcn-ee.net/deb/precise-armhf/v3.8.13-bone30/ latest headers Mar 17 12:38:54 as deb packages Mar 17 12:39:03 download them and dpkg -i package.deb Mar 17 12:39:20 On debian right ? Mar 17 12:39:40 if they match ur kernel yes Mar 17 12:39:48 Currently my BB has ubuntu. I suppose this has to done in Debian Mar 17 12:39:57 I think i will switch anyway Mar 17 12:40:17 switch and 1st try and see if u can install them via apt-get Mar 17 12:42:04 Will do. I will let you know the result. Mar 17 12:42:09 :) Mar 17 12:53:58 hi jkridner Mar 17 12:54:13 howdy Mar 17 12:54:45 find working on proposal and asking for your help Mar 17 12:54:56 could you give me an asnwer for this one Mar 17 12:54:56 our project is successfully completed, what will its impact be on the BeagleBoard.org community Mar 17 12:55:32 if the our project is successfully completed, what will its impact be on the BeagleBoard.org community is successfully completed, what will its impact be on the BeagleBoard.org community? Mar 17 12:55:41 What do you think it'll be? Mar 17 12:55:55 I have my answer, will copy paste from the proposal Mar 17 12:56:15 Creating this platform will have a great impact for the end users. BeagleBoard will provide a tool to the users for creating tutorials. This tutorials will benefit new users and enthusiasts to begin working with the beagle boards. Currently this is one weakness of the organization, there is no way to create tutorials or a full documentation that involves users. Mar 17 12:59:14 that's my answer :) Mar 17 13:08:47 I would do a bit more focusing on the positive.... Mar 17 13:08:58 and help define the audience a bit more... Mar 17 13:09:15 I think you are touching on two key benefits to the community... Mar 17 13:09:47 making easier to bring more people into the community and making it easier for them to contribute. Mar 17 13:10:17 btw, I had quite a few discussions with web developers last week at Fluent... Mar 17 13:10:34 great Mar 17 13:10:36 I was trying to recruit a mentor with significant web developer knowledge and didn't have so much luck. Mar 17 13:10:44 have details about that? Mar 17 13:10:50 the discussions? Mar 17 13:11:00 yeap Mar 17 13:11:03 or is private Mar 17 13:11:28 Adam Flahrety (http://makezine.com/author/adfm/) had some specific suggestions.... Mar 17 13:11:52 he thought doing the little tutorials as "cards" would have significant value. Mar 17 13:12:34 but using the same logic we have been talkiking Mar 17 13:12:35 I guess Mar 17 13:12:36 ? Mar 17 13:12:44 so, you would browse through the tutorials as cards you could lay out on the screen in different ways. Mar 17 13:12:45 yes... Mar 17 13:12:54 otherwise very similar to what we've been discussing... Mar 17 13:13:20 that sounds cool Mar 17 13:13:22 just the "wizards" would have each page be shaped as a BeagleBone. Mar 17 13:13:54 and, related to that was the idea that we store each of these cards in a "gist" (gist.github.com) Mar 17 13:14:04 yeap Mar 17 13:14:21 this would make it much easier to fork. Mar 17 13:14:22 jkridner: is Adam in this IRC or I have to contact him via email? Mar 17 13:14:26 no. Mar 17 13:14:52 Adam is an editor at O'Reilly and not participating as a mentor. Mar 17 13:14:53 via email yes? Mar 17 13:15:00 ahh Mar 17 13:15:03 iok Mar 17 13:15:06 you can try to email and encourage more ideas out of him though... Mar 17 13:15:12 he seemed very open. Mar 17 13:15:25 http://en.gravatar.com/adfm Mar 17 13:16:19 hey jkridner Mar 17 13:16:30 howdy Abhishek_ Mar 17 13:16:58 so much for me to catch up on https://groups.google.com/forum/#!forum/beagleboard-gsoc after traveling last week. Mar 17 13:17:18 We had our festival of colors - Holi today :) Mar 17 13:17:35 nice! St. Patrick's Day here today. Mar 17 13:18:27 jkridner: that means that if this project is selected you will not be the main mentor? Mar 17 13:20:34 jkridner: Could I have some updates on the Arduino-like library using Starterware project? I had mailed on the list with what could be the possbile directions to the project a few days back. Didnt get any reply. Could you please share some views on it? Mar 17 13:21:12 For this project, I most likely would be the primary mentor, but I really need to get another mentor engaged on this one. That was part of my motivation to go to Fluent. I have a couple prospects, but am concerned. Mar 17 13:21:35 * jkridner is getting to replying on the mailing list today. Mar 17 13:22:49 Oh great. But can I have your views on the possibility of taking up the project? I am in the process of drafting my proposal, should be able to finish it in another day or so. Mar 17 13:24:31 jkridner: the answer you gave was for electrobuz or for me? Mar 17 13:24:58 it was for you Diego Tc Mar 17 13:25:46 right? Mar 17 13:26:06 both Mar 17 13:26:09 I am guessing, just want to confirm :) Mar 17 13:26:16 in terms of responding to the mailing list Mar 17 13:26:33 oh... re: primary mentorship, DiegoTc Mar 17 13:26:56 responding to the mailing list : yes that part was for me :) Mar 17 13:27:38 jkridner: since we last talked about license issues for Starterware, are there any updates on it as well? Mar 17 13:29:15 jkridner returning to my original question. I can put this. Jason: You are touching on two key benefits to the community with this project. Making easier to bring more people into the community and making it easier for them to contribute and help define the audience a bit more. Or I have to wait for a formal answer from your part? Mar 17 13:31:09 electrobuz: no updates on StarterWare licensing, but views seem to be that it'd be fine even if incorporated into GPL code with the clarity *both* license restrictions would apply and the code wouldn't be merged into other upstream projects like u-boot. Mar 17 13:31:37 DiegoTc: You can use that as my formal answer. Mar 17 13:31:52 I hope that you further elaborate in your proposal. Mar 17 13:32:25 I read your comment, that was my starting Mar 17 13:32:31 I updated yesterday Mar 17 13:32:39 night(real night) Mar 17 13:32:46 waiting for your feedback Mar 17 13:34:57 i see. So is this a green signal for the project? (I have a little distracted since the issue came up, need some clarity, if you could) Mar 17 13:35:46 jkridner: should I upload my proposal for this project? Mar 17 13:36:35 Have you sent a draft to the mailing list to try to get feedback there first? I think more potential mentors are monitoring the mailing list than the proposal site. Some might only look at the proposal site when it is time to make final decisions. Mar 17 13:37:55 ahh ok Mar 17 13:38:04 jkridner: No I haven't. Okay then I should probably send it on the mailing list then. As for my first question? Mar 17 13:38:35 just the link of the proposal in the gsco group will be enough? Mar 17 13:39:10 electrobuz: as I mentioned, I'm working on going through the traffic on the list today. Mar 17 13:39:23 ^^Same doubt Mar 17 13:39:27 jkridner: hey, my name is alexandru, I am interested in working on Cross platform USB boot project, I am now writing the proposal , and I would like to complete the part where I have to receive some feedback from the members of the community regarding the project Mar 17 13:39:46 jkridner: cool. will wait then :) Mar 17 13:40:21 ^ I will follow your post :) Mar 17 13:40:49 jkridner: I already posted my proposal on the GSoC portal and notified on the group about it. Should I post a Google Docs copy as well? Mar 17 13:44:26 I have same question as Abhishek_ Mar 17 13:46:20 I believe we could send the google-doc link to the list first. get our reviews and then upload it again on the portal? seems okay? Mar 17 13:47:30 ^ I did the same. I was advised to also upload on melange soon. Mar 17 13:49:24 Oh i see. Mar 17 13:56:15 Abhishek_: can't hurt Mar 17 13:57:45 abarbarosie: if you don't get responses here, you might try #beagle and on the broad beagleboard at googlegroups dot com mailing list and other groups where people hang out related to your topic. Mar 17 13:59:30 jkrinder: I just posted my proposal draft on the mailing list Mar 17 15:13:23 Is there someone who could review my application draft and give me feedback? Mar 17 15:13:33 in melange? Mar 17 15:14:50 No, it is on the mailing list, it is just a draft Mar 17 15:15:27 ah Mar 17 15:15:39 you can enter drafts in melange too, you can update the application all the time Mar 17 15:16:00 I see, I was unaware of that Mar 17 15:16:38 So it is better to have it there rather on the mailing list? Mar 17 15:16:55 I would say so Mar 17 15:41:25 jkridner, "Active profile is required to access this page." error on the mentor signup page Mar 17 15:49:25 panto: you need to click once onto login Mar 17 15:49:29 then back there Mar 17 15:49:33 its a melange bug Mar 17 15:49:58 because if there is one thing hard for google, its writing SW Mar 17 15:54:57 who codes that POS? Mar 17 15:58:09 https://code.google.com/p/soc/wiki/MelangeIntro Mar 17 15:59:03 panto: fork, code, commit :) Mar 17 16:00:25 I'd rather pull my eyes out with a fork :) Mar 17 16:00:54 :) Mar 17 16:01:05 or jump into a shark tank? Mar 17 16:09:09 Doing GSoC to develop GSoC sounds a good idea ;) Mar 17 16:09:27 *GSoC portal Mar 17 16:09:53 yes Mar 17 16:10:10 if they find students :) Mar 17 16:16:13 panto: malange was probably a gsoc project :) Mar 17 16:17:39 it used to be much worse in 2010 Mar 17 16:21:37 And still continues to be not so intuitive. Mar 17 16:36:35 jkridner, mranostay, panto: I just posted a GDrive pdf link to my proposal. Request your comments. Mar 17 16:36:45 k Mar 17 16:36:59 Abhishek_, I'm busy right now, will look at it after dinner Mar 17 16:37:21 sure Mar 17 16:37:39 will look during lunch Mar 17 16:38:31 Abhishek_: can you enter it in Melange instead? Mar 17 16:39:24 It's already in Melange, too, although this (the GDrive) is the latest one. Mar 17 16:39:37 Some edits were made Mar 17 16:42:20 edit the melange Mar 17 16:42:29 I dont see why to keep a PDF in parallel Mar 17 16:42:43 its melange that counts in the end Mar 17 16:43:19 Yes, I posted just for convinience of some of the other mentors Mar 17 16:46:58 hey all, was just curious and about people registering for the BeaglePilot project Mar 17 16:47:38 anyone here with the same ? Mar 17 16:51:58 av500: Done. Mar 17 16:52:41 thx Mar 17 17:14:12 mranostay, panto, jkridner, mdp : I have uploaded my proposals on melange! RFC please :) Mar 17 17:14:22 thanks karki Mar 17 17:14:44 beyond_: note #beaglepilot group in addition to this one. Mar 17 17:15:05 lots of interest in #beaglepilot Mar 17 17:16:00 vmayoral: ping Mar 17 17:16:26 your proposal are you using the PWM channels or PRU? or just userspace bitbanging... Mar 17 17:17:32 hmmm Mar 17 17:17:36 lots of PRU this year Mar 17 17:17:57 hello ds2 Mar 17 17:18:17 hey Abhishek_ Mar 17 17:19:26 I have uploaded my proposal to melange, as well as there is a PDF of it on the group; request you for comments Mar 17 17:20:01 looking at it Mar 17 17:20:06 saw your mails too Mar 17 17:21:08 ds2: i'm rather excited about the PRU love here :) Mar 17 17:21:19 Abhishek_: since there are a LOT Of logic analyzer stuff out there... a bit more details on benefits to the BB community would be good Mar 17 17:21:47 mranostay: I'd be more excited come Sept... not sure how much of that will last Mar 17 17:23:44 ds2: In the "About the project" section? Mar 17 17:25:50 Abhishek_: any place... the thing is this - why this project? why not get a logic shrimp, OLS, or some other thing... some of htose can be had for sub $10USD Mar 17 17:26:04 it is nice that we can do it with the PRU Mar 17 17:26:23 Abhishek_: also... any thoughts on the triggering? Mar 17 17:37:01 ds2: I think the strength of a BBB-based logic analyser would be the fact that we can process the data and do analysis on it simultaneously. The other logic analysers at this price point just capture data and pass it on to the host for processing. Mar 17 17:38:41 This is not meant to be another OLS. Mar 17 17:39:28 Abhishek_: Putting some of what you just said in the app might be good Mar 17 17:41:07 Abhishek_: My personal thoughts are that external logic analyzers might be cheap but it may not be in the lab where the Beagle Learner is. Flashing an LED is nice in that hte LEDs are on the board.... doing SPI/I2C is well an exercise in faith for the beginner; the LA should reduce the need for faith Mar 17 17:42:38 Yes, the idea will be to integrate this as default in the system image bundled with the BeagleBone Black, something that will run like Bone101 on board Mar 17 17:42:55 I do mention this point in my application, in the end Mar 17 17:43:48 (Impact on the community section) Mar 17 17:43:58 IMO, making that stronger might be good Mar 17 17:44:14 otherwise, I think it just gets lost in the sea of cheap LA's Mar 17 17:44:55 part of the app is to convince otherpeople why they should get your project going over others Mar 17 17:45:17 Yes, that is why I dedicate 3-4 weeks in my timeline for design of the web-based front end, which I consider as one of the USPs of this project Mar 17 17:45:26 it is like asking for investors to fund a new paper clip...got to convince them Mar 17 17:48:39 I have to run to Girls Who Code.... I'll be back on tonight trying to catch up with all of this. Mar 17 17:49:40 Hmm, for how much more time will you be there? Mar 17 17:50:02 several hours Mar 17 17:50:08 3-4 hours Mar 17 17:51:25 ds2: Okay, i'm on it, will get back to you once I make necessary amendments Mar 17 17:54:51 ds2: And for the triggers, I suppose you are talking about triggering on specific communication patterns like SPI and I2C beyond edge triggering, isn't it? Mar 17 17:55:32 well... Mar 17 17:55:34 not quite Mar 17 17:56:00 I am taking about triggering on patterns, yes. but not necessarily on protocol level (that is hard to do) Mar 17 17:56:02 the differnece is this: Mar 17 17:56:17 pattern trigger is like saying: 101101 Mar 17 17:56:42 SPI trigger would be looking for a series of bits over time to match a number Mar 17 17:56:44 much harder to do Mar 17 17:57:13 it would be cool to do SPI/I2C protocol level triggers but I don't think that is practical in the timeframe Mar 17 17:57:22 Yes, even I think so Mar 17 17:57:39 I used that feature on a Tek LA and it is handy Mar 17 17:57:46 Ohhhh Mar 17 17:58:03 I just figured out a possible way of doing trigger w/o loose of sample rate Mar 17 17:58:10 (vs the post you referenced the otehr day) Mar 17 17:58:21 wonder if it works in parallel input mode Mar 17 17:59:20 essentially sleep on the pattern Mar 17 17:59:27 when it comes, sample til buffer is full Mar 17 18:00:10 Would it be good to cover all these points in detail (Beagle Learner vs other LAs) in the "Is there anything else we can ask you?" question at the last, or should these points be inserted throughout the app? Mar 17 18:00:44 pick a way... I _personally_ don't have a prefernce on that...prehaps another mentor might Mar 17 18:01:45 As of now, you are the one who has given the most feedback on the idea, I am yet to hear what others have to say in detail. Mar 17 18:02:14 'k Mar 17 18:02:23 might be on later...got mtgs Mar 17 18:02:49 okay, will update my application and notify you. Mar 17 18:15:18 Ping @ds2 Mar 17 18:21:19 hi alexanderhiam Mar 17 18:21:46 hey Matt Mar 17 18:57:34 hi alexanderhiam Mar 17 18:58:53 hello Mar 17 19:02:37 I just looked at your comment, thanks for the leads. Will look into it. Mar 17 19:05:41 alexanderhiam : Hello :) I answered your question. Mar 17 19:06:12 Abhishek_: yeah, it might simplify things a bit Mar 17 19:06:53 karki: yes I see, just making sure about the documentation ;) Mar 17 19:07:46 just thinking whether I should have included optimization of PWM with C extensions in my proposal (instead of something else ;) ) Mar 17 19:08:42 alexanderhiam: If you remember me, I have had discussions with you on #userspace-arduino. I am pursuing the project Arduino-like library using starterware. can I talk to you regarding that? Mar 17 19:09:08 I need some assistance :) Mar 17 19:10:09 karki: there may be someone else looking to do C extension stuff (and you have a pretty big list already!) Mar 17 19:10:19 elctrobuz: sure, what up? Mar 17 19:13:17 *electrobuz: sure, *what's up? (spelling hard today!) Mar 17 19:13:53 alexanderhiam : oh! is there. (I know I have a lot on my plate, so thats why I said instead of something else ;) ). But I'm more than happy with what I have now :D Mar 17 19:14:06 alexanderhiam: I have had some discussions with jkridner as well. Here: https://groups.google.com/forum/#!topic/beagleboard-gsoc/OdRCMUyQYGA Mar 17 19:14:22 there seems to some licensing issues in using starterware Mar 17 19:15:00 what I want to ask is having a project where we create support for BB on energia would seem interesting and pursuable here? Mar 17 19:17:20 what's the licensing issue? would you have to implement it from the ground up if you couldn't use starterware? Mar 17 19:17:59 ds2 : ping Mar 17 19:18:31 ds2 : I replied to your question :) Mar 17 19:18:50 i am not really sure. jkridner mentioned : "but views seem to be that it'd be fine even if incorporated into GPL code with the clarity *both* license restrictions would apply and the code wouldn't be merged into other upstream projects like u-boot." Mar 17 19:19:16 he has'nt mentioned about what to do now. this is why i am in a bit of a fix. Mar 17 19:24:50 hmm, it looks to me like starterware has essentially an MIT or BSD license, and there's nothing in GPL that would prevent its license from being honored (i.e. keeping the copyright notice) Mar 17 19:26:20 TI doesn't seem to have any strong opinions about it, seeing as no one responded when this was asked a year ago: http://e2e.ti.com/support/embedded/starterware/f/790/t/252760.aspx :p Mar 17 19:26:48 This seems to be that standard 3-clause BSD Mar 17 19:28:17 so we could go ahead with it? :P Mar 17 19:29:42 I see no legal reasons not to (but I'm no lawyer!) Mar 17 19:30:49 ds2 : ping :) Mar 17 19:31:19 hmm.. i shall also go into talks with jkridner once he returns then. Mar 17 19:32:34 hey Guys this is my first time to participate in GSOC, and i want to enter "Generic Device Tree Creator" Project Mar 17 19:33:43 alexanderhiam: However I mention that python is internally used by libsigrokdecode for decoding protocols on the captured data, there seems to be no python bindings for libsigrok as such. The library is written in C. Hence the choice of the webserver backend is still open. Mar 17 19:34:04 <_av500_> ExtremeGTX: hi Mar 17 19:34:13 hi av500 Mar 17 19:34:32 i'm interested in "Generic Device Tree Creator" project Mar 17 19:34:37 <_av500_> ExtremeGTX: have an idea what DT is? Mar 17 19:34:56 Device Tree Mar 17 19:34:59 <_av500_> wel yes Mar 17 19:35:03 The one which can blend with C - written code for harnessing data from the PRU and post-processing easily will be the one I will go with. Mar 17 19:35:04 when i read about it Mar 17 19:35:04 <_av500_> apart from the name :) Mar 17 19:35:12 i think i got the idea Mar 17 19:35:33 i want to change pinmuxing and identified Modules on the fly Mar 17 19:35:37 <_av500_> http://elinux.org/BeagleBone_and_the_3.8_Kernel Mar 17 19:35:48 <_av500_> also the device tree docs in the kernel Mar 17 19:35:52 <_av500_> just for overview Mar 17 19:36:24 Abhishek_: the libsigrok wiki page links to some "experimental" Python bindings, I assumed that's what you wee looking at Mar 17 19:36:52 <_av500_> ExtremeGTX: you might also join the mailing list and introduce yourself there Mar 17 19:36:59 <_av500_> Jason Kridner is not here atm Mar 17 19:37:52 <_av500_> ExtremeGTX: have you looked more deeply into DT? Mar 17 19:38:21 this is my first time, but i read on the wiki about its config Mar 17 19:38:30 and how to use it with kernel Mar 17 19:39:26 i think what is needed is a tool with predefined DT Configs for BeagleBone Capes/Shields and an option for Manual Configuration Mar 17 19:39:27 <_av500_> DT is relatively new to many people even in the Linux world Mar 17 19:39:48 <_av500_> the whole topic of dynamin DT creation is also being discussed a lot Mar 17 19:39:53 <_av500_> dynamic Mar 17 19:40:18 <_av500_> so I would suggest your really spend some time on it Mar 17 19:40:35 that's good Mar 17 19:40:39 _av500_: would ACPI help here? Mar 17 19:40:40 <_av500_> also, introduce yourself on the mailing list Mar 17 19:40:50 <_av500_> mranostay: where is my @ Mar 17 19:41:15 ok thanks ... but did you see that i understand the topic correctly ?? Mar 17 19:41:26 mranostay : ping Mar 17 19:42:21 <_av500_> ExtremeGTX: roughly :) Mar 17 19:48:21 av500 can you please specify the level of this project ? Mar 17 19:49:13 <_av500_> define the levels :) Mar 17 19:49:20 <_av500_> gsoc projects are not for beginners Mar 17 19:50:51 ok i already know about Linux and its kernel and how things works, already ported u-boot to my DevKit8500D with Davinci SoC Mar 17 19:51:10 <_av500_> http://en.wikipedia.org/wiki/Diff#Unified_format Mar 17 19:51:13 <_av500_> oops Mar 17 19:52:35 karki, my melange setup is not complete Mar 17 19:52:39 can't see the proposal Mar 17 19:52:44 link here or send it via email Mar 17 19:54:51 alexanderhiam: yes I think that had been missed, however seems that the python binding is still missing some features like triggers which would be required by my project, but I think that can be worked upon as we go through. Mar 17 19:55:39 yeah, could be a good starting point at least Mar 17 19:55:47 vvu|Log: I forgot to tell. I got it working. With debian image. Mar 17 19:55:54 <_av500_> good Mar 17 19:56:26 praveendath92_: nice Mar 17 19:56:27 But not with apt-get. Still the same error. Manually installing of headers did the thing. Mar 17 19:56:31 :) Mar 17 19:56:47 yeah works Mar 17 19:56:51 Now whatever I do will be on BB. Mar 17 19:57:23 av500, I will update my proposal tomorrow. Mar 17 19:57:32 <_av500_> ok Mar 17 19:57:34 <_av500_> np Mar 17 19:58:34 It's already late. So, nite guys :) Mar 17 19:58:39 <_av500_> nite Mar 17 19:58:44 panto : can I have your email id :) Mar 17 19:58:54 Or morning ;) Mar 17 19:58:58 <_av500_> post on the mailing list Mar 17 19:59:12 av500, I almost forgot again.. Mar 17 20:00:14 For my proposal I need some comments about how this project would be useful for BB community. Mar 17 20:00:32 Could you please write a comment for me ? Mar 17 20:01:28 what project? Mar 17 20:01:29 <_av500_> sure Mar 17 20:01:36 <_av500_> android remote display Mar 17 20:01:45 Yep Mar 17 20:01:51 <_av500_> "it saves people from buying an HDMI cable" Mar 17 20:01:54 <_av500_> :) Mar 17 20:01:58 "Can we customers expect Ubuntu-Phones with Declarative Firmware and Upstreamed Kernel?" Mar 17 20:01:59 Yay! ds2 is also here Mar 17 20:02:03 sorry, ignore :) Mar 17 20:02:17 Haha. $6 saved ;) Mar 17 20:02:30 saves people from buyng an hdmi cable and helps them buy a nice android tablet i think should be the right version Mar 17 20:02:31 ? Mar 17 20:03:16 ds2: I was hoping that you could also write a comment on this. Mar 17 20:03:16 praveendath92_: can you in a few sentances tell me why your project instead of USBnet + VNC? Mar 17 20:04:46 VNC is heavier. While this interface shall be lighter. Mar 17 20:05:04 It is not going to be running in the userspace Mar 17 20:05:19 what's wrong with running in userspace? Mar 17 20:05:40 <_av500_> also, no X involved be default, can be /dev/fb only Mar 17 20:05:55 it sounds like from your argument, we should implement a VNC server in the kernel just like they implemented a webserver in the older kernels Mar 17 20:06:06 you can have /dev/fb based vncservers Mar 17 20:06:16 <_av500_> mayne Mar 17 20:06:20 <_av500_> next year :) Mar 17 20:06:37 Kernel modules have higher precedence interms of resources Mar 17 20:06:43 <_av500_> ds2: erm Mar 17 20:06:49 <_av500_> USBnet does not exist on android Mar 17 20:06:55 <_av500_> so, no can do Mar 17 20:07:17 ds2 : hi! I replied back to your comment on melange :) I want to discuss further possibilities ! Mar 17 20:07:18 <_av500_> idea is to run on stock android, no changes on phone side Mar 17 20:07:18 _av500_: eh? I have seen them in the gingerbread and earlier era Mar 17 20:07:27 <_av500_> ds2: on some devices Mar 17 20:07:35 <_av500_> in tethering mode Mar 17 20:07:41 <_av500_> usb tether Mar 17 20:07:41 karki: which project? (sorry, lots of students, haven't memorized their irc handles and projects yet) Mar 17 20:08:02 <_av500_> here you just plug the cable and your app pops up Mar 17 20:08:02 isn't usb tether standard? or is that a CM only thing? Mar 17 20:08:07 <_av500_> since its the ADK intent Mar 17 20:08:13 thank you @_av500_ Mar 17 20:08:14 *nod* Mar 17 20:08:19 <_av500_> ds2: it is, but it needs the user to enable tether Mar 17 20:08:30 <_av500_> does that work on wifi only tablets? Mar 17 20:08:56 oh so an additional requirement is it must work out of the box on android devices? rooting, sqlite commands, or even settings changes? Mar 17 20:09:09 <_av500_> ds2: that is the nice thing about ADK Mar 17 20:09:13 <_av500_> it should need zero mods Mar 17 20:09:42 Yeah. I got it tested on my N10 with just an apk insta Mar 17 20:10:02 I see a problem right there with that statement Mar 17 20:10:10 <_av500_> ? Mar 17 20:10:13 you cannot install an APK on most android devices w/o a setting change Mar 17 20:10:16 usb tether is not by default anywhere, just on nexus devices is enabled by default Mar 17 20:10:20 <_av500_> ds2: ? Mar 17 20:10:22 Install* Mar 17 20:10:24 <_av500_> from google market Mar 17 20:10:25 ou need it to be signed Mar 17 20:10:26 <_av500_> works for me Mar 17 20:10:31 <_av500_> no settings change :) Mar 17 20:10:44 okay, fair enough if it is in the market Mar 17 20:10:47 ds2 : Program PRU using high-level scripting languages :) (The botspeak interpreter on PRU) Mar 17 20:10:52 <_av500_> eventually, it can be Mar 17 20:11:18 karki: got it... the question I had was more of for the other signals you are generating...say a sine wave...how do you plan to do that? Mar 17 20:11:23 puting into market is easy, just 10$ or smth small to google for the account and it's done Mar 17 20:11:45 $25 Mar 17 20:11:54 One time fee that too ! Mar 17 20:11:54 <_av500_> praveendath92_: go to bed, I will handle ds2 :) Mar 17 20:11:57 yes but w/o that, users have to enable install from unknown sources Mar 17 20:12:09 <_av500_> ds2: that is still better than a kernel mod :) Mar 17 20:12:23 i can vouch for hating kernel mods Mar 17 20:12:26 No. It's fine Mar 17 20:12:30 * vvu|Log is a live example Mar 17 20:12:38 fair enough. Mar 17 20:13:04 <-- mods kernels daily Mar 17 20:13:08 nothing to hate Mar 17 20:13:10 ;) Mar 17 20:13:15 The app can be put up on playstore once done Mar 17 20:13:16 ds2: last year when we were trying to debug usb communication ROM boot with nexus device was because kernel on nexus did not have RNDIS driver Mar 17 20:13:36 took me like 4-5 days to figure that out :( Mar 17 20:13:56 vvu|Log: part of the complication with android devices is the lack of a real console Mar 17 20:14:08 also debugging for this project is going to be really bad Mar 17 20:14:20 praveendath92_: research a bit how to enable adb once adk is running Mar 17 20:14:53 I was looking at usb snooper for Android this evening Mar 17 20:14:56 what kills adb when adk starts and how to bring it up again cuz without that u will have super much pain making it work nicely Mar 17 20:15:04 this discussion has gone off on a tangent...the reason I pointed that out was the list of requirements... I personally think toggling exposed settings in the settings app is reasonable Mar 17 20:15:05 Not much luck though :/ Mar 17 20:15:22 vvu|Log: did you find the adb manager app? Mar 17 20:15:33 it might help Mar 17 20:15:36 no, did not look into it Mar 17 20:15:42 otherwise, test this on a device which has the console exposed Mar 17 20:15:47 Checking... Mar 17 20:16:02 i.e. certain old Archos devices have a simple mod to bring out a console Mar 17 20:16:02 console exposed ? Mar 17 20:16:09 yes, the kernel console Mar 17 20:16:16 so adb isn't fully needed Mar 17 20:16:34 heh Mar 17 20:16:44 maybe this weekend i will do some hacking around Mar 17 20:16:55 even some of the nexus devices can do it if you have the right cables Mar 17 20:17:05 I'm willing to do any modifications on N10 Mar 17 20:17:21 first need to do crash course to a student who is making a proposal for the multi-boot project Mar 17 20:17:33 donno off hand if the N10 can do it Mar 17 20:17:54 the N4 I think can do it over the headphone jack Mar 17 20:18:03 i would stick to software checking 1st then #hardwareHack Mar 17 20:18:58 ds2 : In case of a sine wave I will have to send along with the logic the sample output values of the rising quarter of the wave form. The logic reads the samples at equal interval and generates the waves. Both code and data together should take around half a KB Mar 17 20:19:37 karki: so you will need a DAC? Mar 17 20:20:06 ds2 : yes Mar 17 20:20:37 karki: okay... you might want to clarify that you are doing DDS instead of PWM Mar 17 20:21:00 karki: do you know how big the code is likely to be on the PRU? Mar 17 20:21:24 sorry I meant PWM Mar 17 20:21:43 <_av500_> ds2: praveendath92_ : yes, getting a device with a serial console should be doable Mar 17 20:22:06 karki: can you please re-answer my question about how you plan to generate a sine wave? (you seem to have given me 2 contradictory answers) Mar 17 20:22:12 <_av500_> and there is alwas toast debugging :) Mar 17 20:22:25 worse case - use the android terminal and a rooted device Mar 17 20:22:34 <_av500_> yes Mar 17 20:22:38 throw in a cheap BT keyboard and it is usable for debugging Mar 17 20:22:44 Worst case is rooted device !? Mar 17 20:22:49 <_av500_> anything you can run logcat on Mar 17 20:22:54 *happy* Mar 17 20:22:59 there are even apps that will do logcat for you :D Mar 17 20:23:08 <_av500_> needs root to Mar 17 20:23:12 <_av500_> since 4.something Mar 17 20:23:15 for JB Mar 17 20:23:20 for ICS no Mar 17 20:23:22 <_av500_> ah Mar 17 20:23:33 <_av500_> I forgot what what version does what Mar 17 20:23:36 <_av500_> too many of them Mar 17 20:23:40 Rooted all my devices. I'm trying with terminal emulater and su right now. Mar 17 20:23:48 think it is the same one that you need to dance on the screen to enable ADB Mar 17 20:26:11 Just checked. I could read kernel logs with su permissions Mar 17 20:29:01 praveendath92_: the problem is not there Mar 17 20:29:32 there are 2 options, get adb working while adk is online or get serial console via cable, as we do it for BBB Mar 17 20:30:12 or do how av500 suggested, print to the android screen all logs from ur app/toasts Mar 17 20:30:15 adb is for logcat messages ? Mar 17 20:30:36 adb is how you get into the device remotely Mar 17 20:31:13 because you will use the usb connection to communicate with your host so you cannot use adb via usb, u need to use it via wi-fi *ds2 correct me if i'm wrong* Mar 17 20:31:48 Generally, yes. Mar 17 20:31:50 ^same understanding Mar 17 20:32:00 unless you have some special device with 2 usb ports Mar 17 20:32:37 Unlikely. But here, we need adb to control or view logs ? Mar 17 20:32:46 Or both ? Mar 17 20:36:27 both or either Mar 17 20:36:42 vvu|Log: Just found out that there are ways to use adb over WiFi Mar 17 20:36:53 yes, there are ways but once adk gets active Mar 17 20:36:59 your adb daemon gets killed Mar 17 20:37:02 so those ways are useless Mar 17 20:37:05 @ds2: will that help ? Mar 17 20:37:21 Got it. I missed this. Mar 17 20:37:45 Sorry but how did you use to debug ? Mar 17 20:38:00 adb over wi-fi Mar 17 20:39:10 Okiee. Does reading systemlog help ? Mar 17 20:39:59 vvu|Log: you tested that it gets killed ? I'm curious to run a test otherwise. Mar 17 20:40:21 yes, it gets killed Mar 17 20:40:43 i did a version of ur project but with screenshot and userspace usb communication :) Mar 17 20:40:46 Bad adb. Mar 17 20:41:19 Yeah. You mentioned this. Mar 17 20:41:32 :) Mar 17 20:41:33 i could not make it work nice because did not have ok debugging interface Mar 17 20:41:42 so you should really take good look at this Mar 17 20:42:01 You also used ADK for that ? Mar 17 20:42:01 cuz hacking kernel and speaking with android without debug won't be nice at all Mar 17 20:42:05 yes Mar 17 20:42:09 adk communication Mar 17 20:42:43 mainly for adk the tricky part is to send data to the phone over isynchronous usb transfer to tell the phone "dude switch to adk mode" Mar 17 20:42:59 and in adk mode it goes with some google default identification values Mar 17 20:43:18 after u have the phone in adk mode you do basic bulk transfer over the endpoint and you are done Mar 17 20:43:29 Thanks for the heads up Mar 17 20:43:35 i think i gave you some userspace implementation of ADK Mar 17 20:43:51 look over those and try to simulate them in the kernel with usb api from there Mar 17 20:44:57 I was working on these lines on my PC but I decided to switch over to bone today Mar 17 20:45:32 praveendath92_: sorry not isynchronous transfer, Control transfer Mar 17 20:45:38 Few hiccups in the beginning but now its all good Mar 17 20:45:47 don't even know if isynch transfer exists for usb, i'm a bit sleepy Mar 17 20:45:50 ds2 : sorry I was away (it is quite late in the night here), the thing is I already will have a pwm command; so all that will be happening is reading the list of samples and writing is using the pwm call that already exists. But I haven't really done too much research for the sine wave part of it as such, I will give a further thought and get back to you :) Good night! Mar 17 20:46:11 Same here. 2 am already. Mar 17 20:46:20 karki: okay... no prob... good night Mar 17 20:46:27 I am here on and off at odd hours anyways Mar 17 20:46:35 praveendath92_: https://gist.github.com/ungureanuvladvictor/9607986 Mar 17 20:46:37 :) Mar 17 20:46:46 praveendath92_: line 108 Mar 17 20:47:00 that is the function that tells android to switch to adk Mar 17 20:47:29 you need to do smth like that in the kernel once android is connected to the host Mar 17 20:47:55 Wow ! That's a great resource ! Thanks :) Mar 17 20:48:34 also the VID is going to be the same for all adk states Mar 17 20:48:39 but PID will change Mar 17 20:48:45 dunno why multiple ones Mar 17 20:49:23 so the main thingies u need to master are control transfer and bulk transfer for USB Mar 17 20:49:40 PID changes for ADK mode and normal connection ? Mar 17 20:50:11 yes, both VID/PID will change once android goes from normal mode to adk Mar 17 20:50:32 they will change to VID of Google and some random PID *check list at line 17* Mar 17 20:50:50 VID at line 16 is google VID Mar 17 20:51:36 Random as in one of those in the list right. Mar 17 20:51:48 try and search why to those Mar 17 20:51:53 they for sure have a meaning Mar 17 20:52:39 Hopefully ADK datasheet has some answers. Mar 17 20:52:50 do u have link to that datasheet ? Mar 17 20:52:59 This would be next task. Mar 17 20:53:26 I have the source code and I think I have that too. Mar 17 20:53:46 Not on my PC right now. Mar 17 20:54:30 the documentation is so crap for ADK Mar 17 20:55:53 Surprisingly no datasheet with the source code Mar 17 20:56:01 You have it ? Mar 17 20:56:22 nop Mar 17 20:56:51 I will share it if I get lucky. Mar 17 20:56:53 ds2: some magic info here except the google website ? Mar 17 20:57:31 donno off hand Mar 17 21:00:39 anyway praveendath92_ i think you have enough info for now Mar 17 21:00:45 to get some reaserch done Mar 17 21:01:21 Yeah. Thanks vvu. I will try with these. Mar 17 21:01:48 :) Mar 17 21:03:05 anytime, if any questions ping here Mar 17 21:04:04 Definitely :) Mar 17 21:04:25 I should sleep now. Class at 9 am Mar 17 21:04:39 no worries quiz at 9 am also :) Mar 17 21:04:42 Thanks again. Nite :) Mar 17 21:04:45 nite Mar 17 21:04:48 my classes at 730 :P Mar 17 21:04:58 Good luck :) Mar 17 21:05:02 10x Mar 17 21:06:01 You should get some sleep too @Abhishek Mar 17 21:06:05 :) Mar 17 21:06:24 Good night all Mar 17 21:06:42 *night Mar 17 21:07:21 mranostay, mdp, panto and jkridner: I will be looking forward to your comments when I check in tomorrow :) Mar 17 21:13:47 ds2: i'm helping a student from my university to write a proposal for the multi-boot project Mar 17 21:14:05 what things do u think the project should include more ? Mar 17 21:14:24 i did a crash course with him through my project and src code these days :) Mar 17 21:46:16 wb jkridner Mar 17 22:31:23 Hello, are there any mentors on? Mar 17 22:31:32 <_av500_> yes Mar 17 22:31:34 <_av500_> some Mar 17 22:32:00 Hello, I posted my proposal on melange, is it possible to get some feedback on it? Mar 17 22:32:34 It is the one for Cross platform USB boot Mar 17 22:33:49 <_av500_> hmm, strange Mar 17 22:33:53 <_av500_> I thought I left a comment Mar 17 22:33:59 <_av500_> or was that on the mailing list? Mar 17 22:35:00 I think it was on the mailing list Mar 17 22:35:47 <_av500_> also, there are other comments on the mailing list Mar 17 22:36:23 Yes, I tried no answer those as well Mar 17 22:36:32 <_av500_> ah ok, DFU Mar 17 22:36:38 <_av500_> did you update the proposal? Mar 17 22:37:48 Sure Mar 17 22:37:51 <_av500_> also, why DFU? Mar 17 22:38:32 <_av500_> if you download a flashing kernel, you can use any USB protocoll Mar 17 22:39:05 <_av500_> was last years flashing done via uboot or a kernel+ramdisk image? Mar 17 22:39:05 _av500_: usb all things eh? Mar 17 22:39:33 last year was using kernel+ramdisk Mar 17 22:39:57 well DFU will alove writing not only the kernel but as well to the other parts of the filesystem Mar 17 22:41:10 <_av500_> that is not a specific feature of DFU Mar 17 22:41:41 <_av500_> once you have a kernel and ramdisk running, you can do a lot of things Mar 17 22:41:47 <_av500_> usb-ethernet, usb mass storage Mar 17 22:41:48 <_av500_> etc.. Mar 17 22:42:44 <_av500_> does the Linux version of this already exists? Mar 17 22:43:05 <_av500_> I mean, is this just a win+mac port, or is there work needed on a linux tool too? Mar 17 22:43:14 <_av500_> (Using last years work) Mar 17 22:43:27 Yes, but from what I understood it uses some hacks which should be reimplemented Mar 17 22:44:33 <_av500_> so, step 1 would be to finalize a Linux tool Mar 17 22:44:49 <_av500_> can be command line, but should implement the needed features Mar 17 22:44:54 Yes, and then proceed with it on other platforms Mar 17 22:44:59 <_av500_> transfer a boot loader and kernel Mar 17 22:45:04 <_av500_> flash the sd/emmc Mar 17 22:45:08 <_av500_> flash a partition Mar 17 22:45:13 <_av500_> replace a file only Mar 17 22:45:23 <_av500_> then port that to windows Mar 17 22:45:28 <_av500_> cmdline first Mar 17 22:45:30 <_av500_> GUI later Mar 17 22:45:42 <_av500_> if gui, I would argue for Qt Mar 17 22:45:49 <_av500_> since that can be reused on lin+mac too Mar 17 22:46:25 <_av500_> and really the effort should be on functionality, not making the GUI shine Mar 17 22:46:30 Yes, I was considering using QT as well, at least for linux Mar 17 22:47:08 _av500_: i was thinking we can lose the hassle of transfering kernel+ramdisk Mar 17 22:47:18 cuz last year doing so if u remember we tried gadget_fs Mar 17 22:47:23 I see, I mean in the project shedule I left as less time as possible for the GUI so that I could focus on the actual implementation Mar 17 22:47:31 and it failed because of the usb code in the kernel Mar 17 22:47:49 you don't want to write the GUI twice, you want something cross-platform like Qt Mar 17 22:48:02 <_av500_> yeah, thats why I proposed QT Mar 17 22:48:12 <_av500_> there is enough hassle to handle windows oddities Mar 17 22:48:27 <_av500_> non need to get into c#/MFC foo Mar 17 22:48:40 <_av500_> vvu|Log: I forgot all about it Mar 17 22:48:45 <_av500_> vvu|Log: remind me :) Mar 17 22:49:02 <_av500_> what did you end up using to tx kernel+image? Mar 17 22:49:02 True, but .NET offers a wrapped for WinUSB which might make the workd easier and since the GUI is not the main feature it can be easily rewritten I believe Mar 17 22:49:30 i did not know about DFU at that point Mar 17 22:49:47 and thought of having a small kernel+ramdisk there as a good option to flash stuff around Mar 17 22:50:08 but when i eneded up to transfer the image we tried to make gadget_fs work and expose the full device to the host Mar 17 22:50:14 and it failed badly, some awkward errors Mar 17 22:50:24 <_av500_> ok, so what did you use? rndis? Mar 17 22:50:35 until kernel was on the board yes rndis Mar 17 22:50:39 <_av500_> and after? Mar 17 22:50:43 after kernel was loaded i had g_serial Mar 17 22:50:47 <_av500_> ah Mar 17 22:50:53 and with g_serial super small transfer speed Mar 17 22:51:01 <_av500_> right Mar 17 22:51:01 cuz at big transfer speed it just failed at some point Mar 17 22:51:06 some babble errors Mar 17 22:51:11 <_av500_> yeah :( Mar 17 22:51:17 was i though of ditching kernel idea Mar 17 22:51:19 <_av500_> f*cking musb ;) Mar 17 22:51:29 maybe DFU from u-boot is nicer Mar 17 22:51:36 <_av500_> well, same usb controller Mar 17 22:51:42 <_av500_> shared driver code afaik Mar 17 22:52:12 <_av500_> vvu|Log: did not try g_ether? Mar 17 22:52:19 abarbarosie: but wouldn't you then be implementing both the gui and the usb part twice? Couldn't you just have a single Qt+libusb program? Mar 17 22:52:27 as far as i remember no :-? Mar 17 22:52:31 why did not i have that idea Mar 17 22:53:04 but i'm still afraid of the usb stuff at bigger speed Mar 17 22:53:11 <_av500_> and you mean g_storage Mar 17 22:53:17 <_av500_> not gadget_fs Mar 17 22:53:22 <_av500_> no? Mar 17 22:53:34 gadget_fs Mar 17 22:53:38 <_av500_> hmm Mar 17 22:53:52 you said smth that u have tried it before and it works Mar 17 22:53:58 alexanderhiam: haven't thought abut it that way Mar 17 22:53:59 don't have the logs tough Mar 17 22:54:11 <_av500_> but what protocol on top of it? Mar 17 22:54:24 <_av500_> yes, we used that in the past to implement MTP Mar 17 22:54:44 <_av500_> gadget_fs gives you usb client in user space Mar 17 22:54:51 <_av500_> then you need to run some protocol Mar 17 22:54:55 <_av500_> maybe you just send the data Mar 17 22:54:59 <_av500_> from the host Mar 17 22:55:04 <_av500_> no protocol Mar 17 22:55:30 <_av500_> anyway, it we fail, we can fallback to g_serial Mar 17 22:55:39 <_av500_> and wait for somebody to fix the kernel :) Mar 17 22:55:47 yeah...anyway i'm going leaving you with abarbarosie Mar 17 22:55:55 i gave him a crash course through my project/code Mar 17 22:56:03 nite Mar 17 22:56:06 alexanderhiam: indeed, since libusb is cross-platfrom complatible that is a better idea, fot some reason I was looking to do something different on windows Mar 17 22:58:03 <_av500_> abarbarosie: btw, we have CareBear\ here who is involved with libusb as I understand Mar 17 22:58:34 <_av500_> ok, I need to sleep :) Mar 17 22:59:19 gn :) Mar 17 23:03:10 abarbarosie: why QT on Linux? Mar 17 23:04:08 (vs just using mono) Mar 17 23:04:11 I have some experience with it, and as it was mentioned above as well I can easily port it to Windows Mar 17 23:04:25 but if you are doing .NET already.... Mar 17 23:04:36 @ds2: it was an example of something cross-platform Mar 17 23:04:53 this ain't twitter :) Mar 17 23:05:09 (instead of .NET + something else on Linux) Mar 17 23:05:17 alexanderhiam: yes but mono would let him run the .NET app and allow for cross-platform Mar 17 23:05:32 ds2: tehnically I am not doing anything yet :) Mar 17 23:05:33 ds2: I made amendments to my proposal on Melange. RFC Mar 17 23:05:34 but would it work with WinUSB? Mar 17 23:05:52 which was the original plan I believe Mar 17 23:05:54 donno Mar 17 23:06:08 Abhishek_: okay, I'll look at it later or tomorrow Mar 17 23:06:18 okay. Mar 17 23:06:37 maybe .NET + libusb + mono could work as well Mar 17 23:07:00 I am not fan of .NET Mar 17 23:07:12 cuz 1 app should simplify things a little Mar 17 23:08:07 I'm very much not a fan of .NET myself, which is why I was thinking something 'truly' cross-platform would be nice Mar 17 23:08:14 ah okay Mar 17 23:08:47 ds2: assuming mono would support WinUSB, why would I go with it, what would be the advantage? Mar 17 23:09:31 I am totally unfamiliar with it Mar 17 23:09:34 abarbarosie: not suggesting you should or should not... was more of wondering why have multiple apps...it just means in time, one will get left behind Mar 17 23:09:41 mono is a .NET run time for Linux and other platforms Mar 17 23:11:32 I see Mar 17 23:12:50 native support is always better though, so definitely look into what alternatives there are Mar 17 23:15:23 from what I understand it is important to find the golden middle between less coding and stable software for both platforms Mar 17 23:16:16 <_av500_> yes Mar 17 23:16:47 that would certainly be a good goal Mar 17 23:17:19 in this case QT + libusb sounds good for me Mar 17 23:20:47 btw, I saw that in the project specs the idea was to have a platform for mac as well, since I don't own one, maybe you can suggest me something, how can I hack this i.e. code for mac without a mac Mar 17 23:22:33 I think Qt and libusb both support os x, so you'd really just need someone with a mac and a bbb to test it Mar 17 23:22:47 alexanderhiam: here i can help Mar 17 23:22:54 i tried to port my code to osx last year Mar 17 23:22:59 but had some trouble with it Mar 17 23:23:15 libusb has deattach_kernel_driver function that works on linux but for mac is just a stub Mar 17 23:23:27 so a codeless kext *as i know* must be used here...this is kinda tricky Mar 17 23:23:42 and now i have read that kexts for mac need to be signed with an official apple dev key Mar 17 23:23:59 good old apple... Mar 17 23:24:34 had major problems here Mar 17 23:24:49 and i am not sure about the status of RNDIS kext...also they have issues there Mar 17 23:26:56 have you looked at libusbx? looks like it was merged into libusb this year Mar 17 23:27:11 might be better os x support now Mar 17 23:27:30 last year i have done stuff, now i do not know even if i follow their ML Mar 17 23:27:35 have not seen many things for OSX there Mar 17 23:27:56 alexanderhiam: yep max support both, just cheked Mar 17 23:29:11 yeah, looks like they claim full cross-platform support Mar 17 23:30:57 from claiming to having there is a lot of work Mar 17 23:31:31 quite true Mar 18 00:12:11 ds2: any clue what was fixed for usb in the latest kernel ? Mar 18 00:12:29 sorry, no...haven't really looked at that Mar 18 00:12:36 i have seen a post somewhere where it was stated "rock solid usb" for 3.13 or what is the latest version Mar 18 00:13:18 jkridner: any hints here ? ^^ Mar 18 00:23:27 hi jkridner Mar 18 00:23:33 do you have time? Mar 18 00:42:12 CareBear\: ping Mar 18 01:37:07 morning Mar 18 01:55:21 Abhishek_: so looking at your proposal i suspect you aim for too much **** ENDING LOGGING AT Tue Mar 18 02:59:58 2014