**** BEGIN LOGGING AT Mon Jun 16 02:59:59 2014 Jun 16 05:10:23 meeting is in 30? or my calendar gone crazy? Jun 16 05:10:38 in 17 Jun 16 05:10:57 backup meeting to get more time zones Jun 16 05:11:24 oh gonna be late for it Jun 16 05:11:33 need to go to work Jun 16 05:31:57 good morning Jun 16 05:33:55 av500: hey Jun 16 05:34:42 Morning all Jun 16 05:35:54 Hello all. Jun 16 05:36:10 hello praveendath92 Jun 16 05:36:30 Hi av500. Did you see the recent status? Jun 16 05:36:46 your mail? Jun 16 05:36:59 I didn't update the blog or mail about it. Just commits. Jun 16 05:37:07 Nope. Jun 16 05:37:10 ic Jun 16 05:37:30 is it possilble to flash a new image to the emmc using a pendrive ? in case somethings wrong with the sd card slot ? Jun 16 05:37:31 can you please write a short status update either by mail or blog Jun 16 05:37:51 anujdeshpande: not so easy Jun 16 05:38:03 it does not boot from USB like that Jun 16 05:38:22 you can use vvu's tools to boot from the micro USB and flash Jun 16 05:38:27 from a Linux PC Jun 16 05:38:28 Will do and ping you when I'm done. Jun 16 05:38:32 ok Jun 16 05:38:40 in fact, please make it not too short Jun 16 05:38:54 the status should be clear Jun 16 05:39:24 Will do so. I will actually write it on the weekly report thread. Jun 16 05:39:32 ok Jun 16 05:39:49 I'll be going to the hospital now, can't make it to the meeting :/ Jun 16 05:39:53 :( Jun 16 05:42:31 av500: will check. 4 of the 8 pins (i.e. one side) of the sd card slot have come off actually. thanks Jun 16 05:43:34 anujdeshpande: ouch Jun 16 05:43:59 you can get it RAM Jun 16 05:44:07 you can get it RMA'd but that will take time Jun 16 05:44:11 is this blocking you? Jun 16 05:44:20 if yes, please write an email to jason ASAP Jun 16 05:44:30 Kevin2314: you are with the beaglepilot, no? Jun 16 05:46:01 av500: dicey that i’ll get an rma. i tried last time for my xm and that didn’t pan out. i’ll try soldering myself, but that’s fine art with these pins. Jun 16 05:46:04 anujdeshpande: how that happen? Jun 16 05:46:09 also i do have another board, but it Jun 16 05:46:13 Yeah, but I haven't been able to do anything, life stuff so I'm trying to get caught up now Jun 16 05:46:18 it’s being used up somewhere. Jun 16 05:46:50 av500: not sure. but maybe it got thrown around a lot in my bag, plus the altoids tin can be harsh i guess Jun 16 05:47:42 av500: If I got a backtrace in a sysfs store that's calling devm_kzalloc that means the method slept, correct? How can I avoid that? Jun 16 05:48:31 *A warning backtrace, no message as such Jun 16 05:49:43 back here! Jun 16 05:49:52 mranostay: not sure. but maybe it got thrown around a lot in my bag, plus the altoids tin can be harsh i guess. Jun 16 05:50:59 anujdeshpande if u need help with the flasher just ping me Jun 16 05:51:04 it is a bit cryptic to use :) Jun 16 05:51:54 vvu: yepp. i already have a doubt :-D does this work off a macbook ? or will i need linux ? Jun 16 05:52:06 nop, no macbook Jun 16 05:52:08 just linux Jun 16 05:52:17 but "patches are welcomed!" :) Jun 16 05:52:31 it just needs libusb right ? Jun 16 05:52:36 it can be ported to osx but u need a headless kext Jun 16 05:52:40 yep, just libusb Jun 16 05:52:51 u have in the readme the deps Jun 16 05:53:31 cool. will try out. Jun 16 05:53:51 puurfect Jun 16 05:54:14 :) Jun 16 05:56:30 praveendath92: you should adapt the probe function to be called again when the device is in ADK mode, to have only one driver for both modes Jun 16 05:56:55 and check in the probe the vid/pid and store relevant info about the bulk interfaces Jun 16 05:57:09 vvu: yes. Just need to check the pid. Jun 16 05:57:39 If pid isn't of accessory, do a setup and get it to that. Jun 16 05:57:40 yep, also in ADK you have just 2 endpoints: ep0 and ep1 with 2 bulk interfaces IN, OUT Jun 16 05:58:05 Hmm. Jun 16 05:58:18 also implement the structure I have to expose it in /dev Jun 16 05:58:24 it will be useful for testing purposes Jun 16 05:58:29 Actually, like I told you last night, I was working with the userspace code. Jun 16 05:59:20 This code works sporadically :/ Jun 16 05:59:26 * vvu goes to have exquisite coffee and breakfast Jun 16 05:59:42 praveendath92: talk in a jiffy Jun 16 06:00:04 hi Jun 16 06:00:08 hi Jun 16 06:00:12 is there a meeting right now? Jun 16 06:00:15 yes Jun 16 06:00:25 Take your time vvu Jun 16 06:00:26 it was the backup meeting that we scheduled last wednesday Jun 16 06:00:31 ahh ok Jun 16 06:00:37 to help some people in wierd timezones :) Jun 16 06:00:43 it's midnight right now in home Jun 16 06:00:48 vvu: https://github.com/praveendath92/bard-linux/blob/master/samples/simplectrl.c#L191 That's the line I'm talking about. Jun 16 06:01:09 so I wasn't sure Jason will be, it will be 2:00AM Jun 16 06:01:21 Gives me a NULL handle even when I can clearly see a device with new vid:pid. Jun 16 06:01:38 The smiley was a mistake. Jun 16 06:01:43 vid : pid Jun 16 06:02:16 DiegoTc: yeah, but im here Jun 16 06:02:29 ahh ok av500 Jun 16 06:02:30 in case there is anything organisation related Jun 16 06:02:43 as long as students and mentors engage its all fine Jun 16 06:02:54 av500, it will be here or via g+? Jun 16 06:02:57 hangout? Jun 16 06:03:03 I have my hangout open Jun 16 06:03:10 here Jun 16 06:03:16 irc meetings are on irc :) Jun 16 06:04:51 hello, This is Junaid over here. Jun 16 06:08:39 heh seems so Jun 16 06:27:29 praveendath92: i think you need to deattach the kernel driver Jun 16 06:28:11 http://libusb.sourceforge.net/api-1.0/group__dev.html#ga21bd343325f558987ca57e4e281a6d47 Jun 16 06:29:50 check what it returns Jun 16 06:39:33 and if you have LIBUSB_ERROR_BUSY as return code means that a kernel driver is attached to ur dev Jun 16 06:39:33 So, you think the reason I get a NULL handle could be this? Jun 16 06:39:33 yep, try it Jun 16 06:39:33 I get a NULL handle with vid and accessory_pid Jun 16 06:39:33 check the error code Jun 16 06:39:33 and a non-null handle with vid, pid Jun 16 06:39:34 libusb_open_device_with_vid_pid unfortunately doesn't throw an error. Jun 16 06:39:34 NULL or a handle Jun 16 06:39:34 try and deattach 1st Jun 16 06:39:34 Got SUCCESS for detach and still problem persists Jun 16 06:39:34 then recheck ur parameters, maybe you are doing smth bad Jun 16 06:39:34 The libusb is somehow still remembering the device list values before that line. Jun 16 06:39:35 rescan for devices Jun 16 06:39:35 https://github.com/ungureanuvladvictor/BBBlfs/blob/master/src/main.c#L217 Jun 16 06:39:35 so at that line the MLO has been sent to the board and u-boot starts its usb stuff Jun 16 06:39:35 and i just get_device_list again Jun 16 06:39:35 Line #220 praveendath92 Jun 16 06:39:35 Is it wrong? Jun 16 06:40:55 I didn't follow that. L220. vvu Jun 16 06:41:05 just rescan for devices Jun 16 06:41:18 in my link i was refering to #220 Jun 16 06:41:51 L191 should work with VID, ACCESSORY_PID but it's not. However it works with VID, PID Jun 16 06:42:07 Checking. Jun 16 06:42:10 i need to go now, work calls :) Jun 16 06:42:13 have fun! Jun 16 06:42:27 I referred lines in my code. Oops. Jun 16 06:43:23 Go on. Jun 16 06:43:40 Once I go past this. It will be fun :) Jun 16 06:44:05 I actually went past before. With same code surprisingly! :/ Jun 16 07:33:24 Good morning everyone! Jun 16 07:33:53 Arriving a bit late for the additional meeting but here I am in case somebody wish to comment something Jun 16 07:39:58 vmayoral|pc: hello Jun 16 07:40:10 as long as you are not blocked in your project all is fine Jun 16 07:41:09 av500: greetings. All good on my side. I will start with the ROS integration within ArduPilot. I shared a preliminary analysis with the drones community and they provided nice feedback Jun 16 07:41:10 https://groups.google.com/forum/#!topic/drones-discuss/o7Z5UWlpMlQ Jun 16 07:41:56 great Jun 16 08:15:48 Hi panti Jun 16 08:15:56 *panto Jun 16 08:16:51 av500: for the framebuffer stuff should we just keep the displaylink stuff and just include our custom usb comm ? Jun 16 08:18:05 because in 1-2 days i think praveen is going to finish adk stuff and then just communication stuff all the way Jun 16 08:18:40 panto: The BeagleLogic Driver has been cleaned up and detached from pru_rproc driver. Would you like to see right now? Jun 16 08:18:46 vvu: I Jun 16 08:18:54 vvu: I'm fine to reuse stuff that exists Jun 16 08:19:18 for a first shot, copy it and replace the protocol Jun 16 08:19:27 long term, we can see if that can be merged somehow Jun 16 08:19:44 next years project: "get last years upstream" :) Jun 16 08:19:48 now communication is just one usb_bulk_transfer call away :) Jun 16 08:20:31 good Jun 16 08:22:47 i switched all my work env on a beagle just to have some linux over here Jun 16 08:23:16 :) Jun 16 08:23:19 subversive Jun 16 08:23:20 Abhishek_, I will later on the day Jun 16 08:23:39 #underTheRadar Jun 16 08:24:03 at least you are using TI gear Jun 16 08:24:23 yeah when they saw the board they were "ok it has a TI chip it is good" Jun 16 08:28:23 :) Jun 16 08:29:33 vvu: Connected via HDMI or ssh? Jun 16 08:32:36 ssh Jun 16 08:32:50 old plain ssh Jun 16 10:06:35 panto: The very first DMA mapping on the buffers operation takes long, subsequent mappings happen quickly; what could be the reason? Jun 16 10:09:31 define long Jun 16 10:09:53 and is that for input or output? Jun 16 10:09:53 <-----------------------------------------------> Jun 16 10:09:55 20ms for a 16MB buffer Jun 16 10:10:10 you might have to fault that memory in Jun 16 10:10:23 how was that memory allocated? Jun 16 10:10:27 kzalloc Jun 16 10:10:33 I mean devm_kzalloc Jun 16 10:10:39 hmm Jun 16 10:10:48 try this for kicks: Jun 16 10:11:02 memset the memory with 0xff after the allocation Jun 16 10:11:25 and, since you're going to dma in, don't zero it Jun 16 10:12:40 BTW, I added a sysfs attribute to fill the buffers with increasing numbers to check read consistency, and it passes the test. Jun 16 10:13:10 now to do the same with the PRU firmware Jun 16 10:13:13 Abhishek_, that is good Jun 16 10:16:18 what currently needs review in my code is the usage of locks and waitqueues to wake up reads from ISR. Jun 16 10:19:14 k Jun 16 10:19:26 it's on my long long long (long) TODO list Jun 16 10:19:33 but, Monday Jun 16 10:38:43 panto: I get a warning with a backtrace (but no particular message) on kzalloc'ing, is that due to might_sleep? Jun 16 10:43:39 where is that warning generated from? Jun 16 11:49:36 panto: sysfs store (an attribute to allocate memory) **** BEGIN LOGGING AT Mon Jun 16 11:52:42 2014 Jun 16 11:52:48 vvu: which probe is that? Jun 16 11:53:54 i-jet from IAR Systems Jun 16 11:56:17 vvu: interesting. The only probe I have worked with is a ST-LINK probe, and that too in SWD mode only. Jun 16 11:56:53 hi DiegoTc, you around? Jun 16 11:57:36 jkridner: DiegoTc was there in today's meeting. Jun 16 12:06:39 Abhishek_: very sorry to have missed it. :( Jun 16 12:24:11 jkridner: i was at the point of missing it, though it was PM not AM my time :) Jun 16 12:57:50 jkridner: maybe we could have got in more traffic on such a meeting on Fridays, Mondays late night meetings (on your side) would have definitely been a little odd. Jun 16 12:59:44 i am ok with mainly all meetings if they are via Google announced :D Jun 16 13:02:34 vvu: BTW, what is the target chip of your JTAG probe? Jun 16 13:09:41 about that i cannot really talk :) Jun 16 13:10:10 oops, NDA? Jun 16 13:14:39 y Jun 16 13:31:59 Abhishek_: how do you log data ? with printk or dev_dbg ? Jun 16 13:32:38 currently logging with dev_err, Haven't tried dev_dbg Jun 16 13:32:56 *dev_err or dev_info Jun 16 13:33:00 link to how to use it ? Jun 16 13:33:49 didn't really use a link, it's more like dev_info(struct device *dev, "info", ...); Jun 16 13:35:33 for dbg it will show in dmesg or i need to set the level ? Jun 16 13:35:53 yeah I saw just now, for dev_dbg levels have to be set Jun 16 13:36:02 ah that is why it was not working Jun 16 13:36:19 and neither did it for me :) Jun 16 13:42:23 ok good fixed :) Jun 16 13:57:00 alexanderhiam, are you available for a quick phonechat? Jun 16 14:04:38 can I make the kernel release (version string) independent of the commit I am at / or the branch I am at currently? Jun 16 14:13:15 karki: you around? Jun 16 14:13:36 jkridner : very sick Jun 16 14:13:41 :( Jun 16 14:13:47 fever since 5 days Jun 16 14:13:48 bad day to do the hangout with Tufts? Jun 16 14:14:10 yup, later this week? Jun 16 14:14:25 even the Docs don't know whats wrong Jun 16 14:15:11 karki: I might meet with him anyway because we've had so much difficulty putting it together and then do it again next week at the same time? Jun 16 14:15:32 jkridner : sorry about this . yes, sure :) go ahead Jun 16 14:15:44 karki: Get well soon :) Jun 16 14:15:56 get well soon karki :) Jun 16 14:24:38 panto : how do I go about PWM control using PRU? you have manually controlled the pin states in your example, doing that in PRU Speak is going to get very hairy. Jun 16 14:25:00 don't then Jun 16 14:25:08 define a PWM command Jun 16 14:26:21 how would that be implemented? I have to use the PWM subsystem, right? Jun 16 14:27:30 no Jun 16 14:27:53 you can have a small module in the PRU that performs PWM in s/w Jun 16 14:28:40 which has to be triggered by a timer? Jun 16 14:29:36 no Jun 16 14:29:50 you can use the other PRU Jun 16 14:30:02 Abhishek_ is doing the same thing basically Jun 16 14:30:14 have one PRU that is doing the hard real time loop stuff Jun 16 14:30:25 the other soft real-time and handles the comms Jun 16 14:30:41 okay :) Jun 16 14:31:49 I want you guys talking to each other Jun 16 14:58:00 jkridner: botspeak invite to me ? :) Jun 16 14:58:18 just tried to publish it. public. Jun 16 14:58:25 hi panto. Jun 16 14:58:36 jkridner, I've seen it but I can't make Jun 16 14:58:41 k. Jun 16 14:58:43 conflicts with another conf-call Jun 16 14:58:49 *it Jun 16 14:58:58 noon or 1 Eastern makes no difference? Jun 16 14:59:11 next week should be OK Jun 16 14:59:12 * jkridner can't quite figure out the best way to do the invite. Jun 16 14:59:14 k. Jun 16 14:59:20 there is no best way Jun 16 14:59:20 seems like karki can only make it next week. Jun 16 14:59:39 * jkridner wants it to be recurring, but Google Events doesn't seem to enable that. Jun 16 14:59:52 Adding video to the calendar, I'm not sure if it'll record it. :( Jun 16 15:02:33 i'll be around here anyway, gonna tweak on praveen's code Jun 16 15:12:36 any clue where err() is defined for kernel usage ? Jun 16 15:32:18 * Abhishek_ wonders if vmayoral|pc has deployed the PRUs for PPM/PWM usage yet Jun 16 15:55:59 rcn-ee: How do I explain that after replacing the kernel [zImage] on the stock Debian image with a custom built kernel, my BBB on a cold start wants to boot from eMMC instead of the SD card (the LEDs actually light up in sequence twice, even when I press the boot button to boot from the SD card), and when I "touch" the SD card via a fsck and reboots, then it Jun 16 15:55:59 boots correctly? Jun 16 16:01:14 VoltVisionSteve: you still want to chat? Jun 16 16:01:50 sure, can you call me in about 5mins? Thx! Jun 16 16:01:59 sure Jun 16 16:07:50 Abhishek_, that's very odd... I'd like to see the serial boot log (mostly the u-boot stuff) to see how it's interacting in both cases... Jun 16 16:09:23 I don't have a serial cable around. I wonder if it's safe to use my Arduino mega as a VCP in this situation. Jun 16 16:10:46 rcn-ee: How do you prepare the 100MB boot image, would it help to clean and recreate that partition? Jun 16 16:12:46 well it's fat16, so there really isn't anything to clean... it was created by: http://eewiki.net/display/linuxonarm/BeagleBone+Black#BeagleBoneBlack-SetupmicroSD/SDcard Jun 16 16:13:30 for the production images i used 96 instead of 48 as we were shipping a lot more stuff in the fat16 partition (windows drivers/etc) Jun 16 16:26:42 Abhishek_, (which is why it's odd a fsck would fix anything. ;) ) Jun 16 16:29:19 I don't really know what's going on either [unless I hook up a serial cable but as I told earlier, I'm little apprehensive to hook my arduino as a VCP for the BBB], but I found that this fsck thing works, don't know why but it does. I wonder if a sync operation would do the same. Jun 16 16:30:24 Also if I force cut power to the BBB, there's the same issue. fsck sometimes says "recovering journal" for rootfs though, this time Jun 16 16:32:03 anujdeshpande: is the beaglepilot firmware using the PRU for PWM generation? Jun 16 16:32:26 Abhishek_: yes we are using the PRU Jun 16 16:33:17 are you running the testpru application or have you written your own code for the subsystem? Jun 16 16:34:08 started out with testpru, but heading towards our own now. you should talk to sidbh about this though Jun 16 16:34:32 ok, I'll ask him when he's around Jun 16 17:12:57 jkridner, is there a meeting Jun 16 17:13:00 right npw? Jun 16 17:14:29 now? Jun 16 17:22:56 * vvu closes the day of work! Jun 16 17:23:23 party time :) Jun 16 17:26:53 looks like PRU C Compiler -O3 just messed up my downcalls Jun 16 17:42:39 Abhishek_: that could happen :) Jun 16 17:42:46 iirc I've only used -O2 Jun 16 17:43:07 why does melange suck so bad? Jun 16 17:43:31 Abhishek_: you have your test program posted btw? Jun 16 17:43:52 yep, everything I built is on github Jun 16 17:44:30 mranostay: the PRU test firmware you need? Jun 16 17:48:57 mranostay: I'm currently testing the kernel driver using just dd, should have a proper test app maybe by thursday [if I don't make it in the meeting] Jun 16 17:51:56 so which meeting changed times again? Jun 16 17:53:20 mranostay: the Wednesday meetings are still the same time, there's a new meeting Tuesdays for the PRU Speak project Jun 16 17:56:00 panto, karki: https://www.youtube.com/watch?v=bRaAq6nbC8k Jun 16 17:57:22 DiegoTc: you there? Jun 16 17:57:37 jkridner, YES Jun 16 17:57:45 k. Jun 16 17:57:49 * jkridner pulls up latest e-mails Jun 16 17:58:28 jkridner, we do a hangout or IRC? Jun 16 17:58:58 my preference is IRC, but up to you. Jun 16 17:59:18 let's talk here Jun 16 18:01:34 DiegoTc: did you get in touch with VoltVisionSteve's contacts? Jun 16 18:02:01 * jkridner looks at https://gist.github.com/anonymous/0df6c75ad2194bcd6d27 Jun 16 18:02:14 today I have meeting with them, I'm confirming the hour with VoltVisionSteve (I will be on all day long) Jun 16 18:02:38 jkridner, that was the first demo. This is the current: https://gist.github.com/anonymous/018902cbb045c36a9993 Jun 16 18:02:40 I am pinging them again right now. Jun 16 18:03:06 empty cards appear because I didn't do copy paste to the tutorial, just testing. Jun 16 18:03:20 does https://gist.github.com/anonymous/018902cbb045c36a9993 have any graphics? most cards seem empty Jun 16 18:03:23 k Jun 16 18:04:03 I have 2 issues right now Jun 16 18:05:21 First one autentication. I can't create the cards with the user id. The API returns me error 404 or 403, according to docs, this is because I' m assigning wrong the token Jun 16 18:05:53 jkridner, want to talk of this one first? or I give the second issue? Jun 16 18:06:24 do you have a link to the code that is making the error? Jun 16 18:07:14 I uploaded the code but I erased the part of the error Jun 16 18:07:19 I will pas it right now Jun 16 18:09:44 jkridner: http://pastebin.com/kyqtCQ8P (Currently I gave all in one file) Jun 16 18:10:07 on this method: app.get('/auth/github/callback',function (req, res) Jun 16 18:11:31 jkridner, the link I just send you shouldn't have something: this is the newest one: http://pastebin.com/gNbng2c8 Jun 16 18:12:52 when I do the github.authenticate({type: "oauth", token: accessToken}); Jun 16 18:13:34 it seems the token is being assign, but at the moment the post is call and the create gist starts to work I receive the error 404 or 403 Jun 16 18:18:58 is there no way to use a session login instead of a token? Jun 16 18:20:08 request username and password? Jun 16 18:20:47 expect user to be logged into github Jun 16 18:21:50 jkridner, not really according to the doc: however, to read or write gists on a user’s behalf the gist OAuth scope is required.https://developer.github.com/v3/gists/#authentication Jun 16 18:22:12 so https://developer.github.com/v3/gists/#create-a-gist is just a single post... Jun 16 18:22:21 yes Jun 16 18:22:22 are you saying that the headers must contain the token? Jun 16 18:23:04 yeap Jun 16 18:23:25 https://developer.github.com/v3/gists/#authentication says anonymous gists can be created without a token Jun 16 18:23:56 yes, all the gist I have been created are anonymous Jun 16 18:24:05 I found this example: https://github.com/ajaxorg/node-github/blob/master/test/oauth.js Jun 16 18:24:26 it works, but I can't implement it :( it gives me the error I told you Jun 16 18:27:40 but the code you posted creates a token Jun 16 18:27:48 on the server side. Jun 16 18:28:40 yes Jun 16 18:28:43 it does Jun 16 18:28:52 i have debug it, see it Jun 16 18:30:25 that's the part is getting me crazy Jun 16 18:30:55 sorry... I'm got here with things already going on Jun 16 18:31:21 but if you can get a Token and if you're using Ajax, you must include the header with the token in the next request Jun 16 18:31:30 Authorization: token TOKEN-YOU-GOT Jun 16 18:33:08 mmm, I see adao Jun 16 18:34:47 I will try that adao Jun 16 18:38:26 jkridner, still here? Second issue is this one: Images Jun 16 18:38:52 yeah, trying to think about how to do an XHR instead of a request from a server Jun 16 18:39:03 http://jsfiddle.net/XWDRf/ won't work due to cross site stuff Jun 16 18:40:09 jkridner, second is images, have tried the flickr API () Jun 16 18:40:20 and? Jun 16 18:40:29 I'm planning to modified this editor: https://github.com/DiegoTc/bootstrap-wysiwyg Jun 16 18:41:44 to include the options for uploading images, in flickr instead of data Image base 64 Jun 16 18:42:23 this is a demo: http://mindmup.github.io/bootstrap-wysiwyg/ Jun 16 18:44:32 jkridner: a few days back I was looking for large local storage on the host [PC side] and found this: https://github.com/tantaman/LargeLocalStorage . Do you know of any alternatives off-hand? Jun 16 18:44:57 I thought gist was capable of saving the images this way: Jun 16 18:47:46 do you think this is a good idea? Or I go another way? Jun 16 18:47:59 panto: Can reloading the PRU firmware through the sysfs load attribute be corrupted due to the traces of previous firmware? Jun 16 18:51:40 DiegoTc: still debugging http://jsfiddle.net/XWDRf/6/ Jun 16 18:52:13 DiegoTc: URL encoding seems reasonable to me... assuming images are small enough. Jun 16 18:56:58 DiegoTc, I just confirmed with the Guatemalan professor that we will talk @ 3pm Guatemala/Honduras time, so thats 2hrs from now. Is that OK? Thx! Jun 16 18:57:09 on skype Jun 16 18:57:57 DiegoTc, send me a skype message sometime before 3pm your time so I have you ready to bring into the talk. Thx Jun 16 18:57:59 ! Jun 16 18:57:59 DiegoTc: http://jsfiddle.net/XWDRf/8/ seems to work find for creating an anonymous gist Jun 16 18:58:05 DiegoTc: guess that wasn't your challenge... Jun 16 18:58:16 seeing if I can add in a username without adding code to server. Jun 16 18:58:55 yes, I'm seeing thar Jun 16 18:59:35 this is the moduele I'm using: https://github.com/ajaxorg/node-github server side Jun 16 19:01:22 jkridner, I will call you on hagout right now Jun 16 19:01:33 tell you something, before I leave right now Jun 16 19:01:39 k Jun 16 19:30:03 DiegoTc: http://jsfiddle.net/jkridner/XWDRf/9/ storegs image content in github using base64 encoding Jun 16 19:49:37 jkridner, just saw it! Cool!!! Jun 16 19:49:47 you mean http://jsfiddle.net/jkridner/XWDRf/11/ ? Jun 16 19:50:03 working on showing you how to do in-place editing. Jun 16 19:50:32 currently it creates a new fiddle using gist, but doesn't copy over the javascript portion. Jun 16 19:51:29 jkridner, so we forget about showing the content on the gist: http://diegotc.github.io/bone101/Support/GSOC/BLINKINGTUTORIALS/Blink.html Jun 16 19:51:59 why do we need to show the content on gist.github.com as long as it is saved there? Jun 16 19:52:15 but, we will take the info and display aside, just like we do on jsfiddle Jun 16 19:52:40 aside? Jun 16 19:53:10 I don't think you want to use the github embed... it is easy to fetch the content inside the gist and display it as we please. Jun 16 19:53:26 perhaps displaying markdown is too difficult? Jun 16 19:53:51 if you are using an authoring tool that creates html, it seems to me that supporing markdown isn't needed. Jun 16 19:53:58 s/supporing/supporting/ Jun 16 19:54:47 that's the word I was looking embedding will not be necesary anymore Jun 16 20:08:06 jkridner, still getting the error 400 when trying to create the gist with your user idd Jun 16 20:54:19 VoltVisionSteve, give me some minutes restarting computer, can found skype for the Ubuntu version I'm using Jun 16 20:54:22 going to windows Jun 16 23:01:39 jkridner, I have it! Jun 16 23:01:42 server side Jun 16 23:02:11 jkridner, https://gist.github.com/DiegoTc/fa48439ca292a3352476 Jun 16 23:08:17 jkridner, just created: https://gist.github.com/DiegoTc/0a32949f5e24a253bd9b Jun 17 00:32:26 http://jsfiddle.net/gh/gist/jquery/1.9.1/d9a3c5251cd41c2696b4/ Jun 17 00:32:31 DiegoTc: didn't see what you did. Jun 17 00:32:49 DiegoTc: http://jsfiddle.net/gh/gist/jquery/1.9.1/d9a3c5251cd41c2696b4/ has a pure client-side ability to create anonymous gists with an image Jun 17 00:33:35 jkridner, the link I send you was the one of a gist created with my account Jun 17 00:33:49 k, but how do you do that for every user? Jun 17 00:34:02 how do you avoid using secrets? with code in github, you can't use secrets. Jun 17 00:35:30 DiegoTc: http://jsfiddle.net/jkridner/44KQP/show/ is an example of a native view that is forkable. not sure how to do the same off of github, but could do so with the API. Jun 17 00:36:44 jkridner, I uploaded the code: https://github.com/DiegoTc/bone101/tree/nodejs/Support/GSOC Jun 17 00:36:58 it's server side Jun 17 00:37:22 again, you have a "secret" in a public repo. that can't work. Jun 17 00:38:11 you need to work without using a secret. and, ideally, there shouldn't be any need for special server code. Jun 17 00:38:39 av500: ping Jun 17 00:41:42 jkridner, I understand you, but reading the documentation I will need to use oauth https://developer.github.com/v3/oauth/#web-application-flow Jun 17 00:42:05 in order to? Jun 17 00:42:19 just because you want the gists to show up under a user account? Jun 17 00:42:24 yes Jun 17 00:42:53 the idea was that every user have its gist on their account Jun 17 00:43:39 for me it's easier to do it anonymous, I don't complicate Jun 17 00:43:53 but I'm asking because of what we talk at the beggining Jun 17 00:44:39 i could go to beagleboard account on gist.github.com and fork the gist Jun 17 00:45:13 jkridner, So I continue with the anonymous way then? Jun 17 00:45:47 should be possible to do oauth within the browser... Jun 17 00:45:55 I'm looking for examples or to figure it out. Jun 17 00:45:59 looking at http://jsfiddle.net/willjcksn/2hb6B/ right now Jun 17 00:47:48 jkridner, I'm leaving right now. Coming back in 2 hours. Jun 17 00:47:56 k. Jun 17 00:48:49 jkridner, http://jsfiddle.net/willjcksn/2hb6B/ that example works for you? Jun 17 00:48:56 no Jun 17 00:49:07 but, I'm looking at http://oauth.io right now Jun 17 00:49:46 if it's possible to make it run on the client side, it will be almost the same logic I used on the server **** ENDING LOGGING AT Tue Jun 17 02:59:58 2014