**** BEGIN LOGGING AT Tue May 26 02:59:58 2015 May 26 09:52:14 gm May 26 15:00:17 Abhishek_: dts works now. used rnc-ee's dtb-rebuilder repo May 26 15:01:19 https://github.com/RobertCNelson/dtb-rebuilder/tree/3.14-ti May 26 16:47:09 hi jkridner May 26 16:47:41 jkridner: was debugging the display cape today, apparently the LED backlight is causing too much ripple in the power supply May 26 16:51:41 scope shows me 1.5Vpp! May 26 16:55:14 yikes May 26 16:56:26 when I disable the backlight, the display was apparently stable (as seen under a flashlight) May 26 16:56:45 fixing it right now May 26 16:58:58 *fingers crossed* May 26 17:26:34 * nerdboy strying into mips64 territory... May 26 17:26:42 *straying even May 26 17:27:28 jkridner: can you send two PNI rm3100 eval boards to neemo ? May 26 17:39:46 no improvment in display so far May 26 17:40:13 nerdboy: what is a PNI RM3100 eval board? May 26 17:40:53 vvu: any chance we can add the dependent image builds to builds.beagleboard.org? May 26 17:41:03 (if they aren't build by travis-ci) May 26 17:41:27 arianepaola_: coding start OK? May 26 17:42:01 julianduque, nodebotanist: you seen DiegoTc or ehab lately? May 26 17:42:06 hello jkridner May 26 17:42:25 yes jkridner May 26 17:43:05 great. May 26 17:43:11 I'm trying to make sure I May 26 17:43:23 am logged in noon-1pm US Eastern time every working day. May 26 17:43:45 so, be sure to look for me around now or a couple hours earlier every day if you are looking to chat. May 26 17:44:04 * jkridner doesn't know how you and moto-timo have broken down the development schedule... May 26 17:44:09 great, thank you. ok May 26 17:44:16 I suspect you have some kind of milestone every week or so. May 26 17:45:04 jkridner: yes I will review the milestones with moto-timo May 26 17:46:07 jkridner moto-timo ah I receive the Beagle Board May 26 17:48:02 Abhishek_: no luck with the blog draft on beagleboard.org? May 26 17:48:23 jkridner: the only hi-res magnetometer available May 26 17:48:26 http://www.pnicorp.com/products/geomagnetic-sensors May 26 17:48:55 yeah, that's where google brought me. May 26 17:49:08 for some (stupid) reason their previous one is discontinued May 26 17:49:13 nerdboy: address and phone number? May 26 17:49:23 nerdboy: I assume this is absolutely required for the project, right? May 26 17:49:25 you sent him a beagleboard... May 26 17:49:41 yes it is, and i don't have his address May 26 17:50:21 need two sensors for noise/data measurements May 26 17:51:11 wait, i think you sent bbb to his address in croatia May 26 17:51:42 neemo: you don't have a school address in china? May 26 17:52:42 jkridner: it gets charged back to google, or does it come out of org's hide? May 26 17:53:25 it's all of $139 plus tax/shipping... May 26 17:53:46 er, 138 technically... May 26 17:55:09 nerdboy: it comes out of the org's hide, but google gives a $500 donation for each student mentored, so even thought the org technically pays for it, it makes sense to pick up any hardware expenses for the students that are required. May 26 17:56:14 well, technically we need imu also but i think he has one May 26 17:56:22 nerdboy: I do May 26 17:56:27 I have two in fact :) May 26 17:56:56 * neemo needs a sec to scroll up through the chat May 26 17:57:15 okay, so first cut is imu processing on arm side and fast/hi-res magnetic data in PRUs May 26 17:57:34 what is neemo's name? May 26 17:57:46 Niko Visnjic May 26 17:57:54 hi neemo... May 26 17:58:03 hi jkridner May 26 17:58:04 thought you were out of Croatia for the moment... May 26 17:58:09 same address? May 26 17:58:09 I am May 26 17:58:22 to your mom or so? May 26 17:58:23 nope, if you're sending the PNI sensors May 26 17:58:25 send to china May 26 17:58:33 Im here until mid July May 26 17:58:44 k, I'm going to need a new address then. May 26 17:58:47 I'll send you the address and all May 26 17:58:48 yeah sure May 26 17:59:12 panto-cloud: not sure if you see these, but Greece customs is still complaining about your board shipment. May 26 18:03:29 jkridner: Can you print out or C/P the address May 26 18:03:30 ? May 26 18:08:20 jkridner: International Exchange Center Hotel, Gubei Road 620, ChangNing District, Shanghai, 200336 China 上海 上海市 长宁区 仙霞新村街道 上海 古北路 620 号, 国际交流中心大酒店 (Niko Visnjic (Room D406) 收) 手机号码 18616352346 May 26 18:08:54 k. Don't guess you saw my private message, but it wasn't this address anyway. May 26 18:08:56 hope you can copy this, or that you like writing chinese characters :) May 26 18:09:19 yeah, that was the Croatian one May 26 18:09:23 I can type a bit of Kanji, but I'm not sure Fedex will take it. May 26 18:09:38 btw, the BBB arrived and is waiting for me there, thanks :) May 26 18:10:15 yeah, well, I guess Fedex knows it's way around Shanghai May 26 18:10:15 or so I hope May 26 18:10:15 the chinese hanzi are useful for the local small delivery guys May 26 18:10:48 I'm sure Fedex will call me, either for delivery or for customs (not sure how the chinese will react to that :/) May 26 18:11:26 can't get those on alibaba or whatever? May 26 18:11:40 not the test modules May 26 18:12:00 As I mentioned, I can get the asic and the sensors apiece here without an issue May 26 18:12:12 I'll have to bake them on a board myself though May 26 18:12:15 * nerdboy tries newer kernel headers on funky laptop May 26 18:12:46 nerdboy: We can still opt for some more available sensor if that makes sense May 26 18:12:47 I'm just not sure what alernatives we have, apart from Honeywell stuff May 26 18:13:14 just go with those May 26 18:13:37 i think there's plenty of software/integration stuff to do... May 26 18:13:38 is 18616352346 the phone number? May 26 18:13:58 chinese zip code? May 26 18:15:04 上海 is the province. May 26 18:15:21 So I thought 200336 was the zip code. May 26 18:15:22 nerdboy: for example http://item.taobao.com/item.htm?spm=2013.1.20141002.4.EdAmOf&scm=1007.10009.6098.i9920063933&id=10217452757&pvid=c9a48e13-2bf8-4a13-9214-6115faab2f10 May 26 18:15:43 neemo: is 18616352346 the phone number? May 26 18:16:01 * moto-timo catches up May 26 18:16:35 nerdboy: 1 euro for a the HMC5883L, I'll test with that one as well I guess May 26 18:16:42 jkridner: yeah, that's the phone May 26 18:17:01 jkridner: Adding bulk capacitors (10uF ceramic) close to LED driver pads does nothing to reduce the noise. Any hints? May 26 18:17:11 *LED driver supply pin May 26 18:17:13 I've written 手机号码 18616352346 which is phone nr *** in chinese May 26 18:17:54 jkridner: the zip for Shanghai is 200000, my Uni says they're local one is 200336, but you can put 200000 I think May 26 18:18:01 jkridner: btw, Do you have any specific pointers on which the blog post should be about? May 26 18:18:07 neemo: k. I'm not sure where to enter some of the hanzi. I'll try to add it all to the notes. May 26 18:18:37 jkridner: If they have an extra details form, you could copy paste it there I guess May 26 18:18:59 jkridner: ehab has been asking about a commit question. I haven't had that git experience that never thought about it. "Question"Mentioning the issue number is the commits is a good way to make it easy to map the commits to its issue, the only problem with it is that it will missed up when merging to the main repo in the end so should I keep mentioning it? May 26 18:19:15 but fedex should be able to find me easily, I just put it there for good measure :) May 26 18:19:39 Abhishek_: imagine people just wanting to know what a PRU is, why they should be excited about what it can do, how do they see it for themselves and how do thy get started trying to learn. May 26 18:19:56 http://beagleboard.org/pru is the best intro I have right now and it is pretty bad. May 26 18:19:59 jkridner: got my last mail? May 26 18:20:31 ebadawy: will look at it after finishing this shipment. May 26 18:20:49 jkridner: okay May 26 18:21:05 jkridner: that was a joke -> large number of digits... May 26 18:21:20 ebadawy: besides the git questions May 26 18:21:24 do you have another type of question May 26 18:21:24 ? May 26 18:21:42 DiegoTc: nop May 26 18:22:08 * nerdboy wanders off to try "gap analysis"... May 26 18:22:24 DiegoTc: actualy there is! what about the BBUI time shifting? May 26 18:23:14 how like that? May 26 18:23:48 neemo: check out the image in the message I sent you privately. May 26 18:25:24 DiegoTc: I mean to reorganize the time line and make it after the midterm ,, as I'm not sure I could finish it in two weeks May 26 18:25:25 ebadawy: copying files to simplify the commit should be OK, but be sure to annotate DiegoTc as the author as to not take credit for his work. May 26 18:25:53 ebadawy: also, try to simplify whatever you can regarding his work. I think DiegoTc committed quite a bit that can be trimmed up a bit. May 26 18:25:53 jkridner: how could I do that? May 26 18:26:08 ebadawy: oh, I thought you were suggesting it. May 26 18:26:09 git preserves author May 26 18:26:29 nerdboy: the issue I have is that ebadawy has some very complex commits.... May 26 18:26:30 jkrinder: I mean about DiegoTc credit May 26 18:26:40 make patch, reapply and sign-off May 26 18:27:09 *maybe more than one patch May 26 18:27:36 nerdboy: and not sure if DiegoTc has time to simplify the patches himself. May 26 18:27:50 either that or get DiegoTc to sign off on appropriate May 26 18:28:23 nerdboy: mmm how you do that? May 26 18:28:31 just make final commit(s) with both and commit message May 26 18:28:36 I will be happy with a md file saying i did it May 26 18:29:08 one person makes original commit, make git patch, re-apply with other person sign-off May 26 18:29:25 * DiegoTc hates git conflicts lol May 26 18:29:32 could do annotated/signed commit May 26 18:30:13 most important is probably to apply patches with both names/emails and good commit message May 26 18:30:28 anything fancier is bonus May 26 18:30:52 nerdbody: do u consider that we are both working on a separated repos? May 26 18:30:55 I guess sound quality issues caused some students to pull their videos ? Now we only have 1. May 26 18:31:11 ebadawy: thus the git pathc tools May 26 18:31:55 you fiddle, make commits, then make patch, send to DiegoTc who can apply with signoff May 26 18:32:11 many parts are edible... May 26 18:32:34 nerdboy: will I guess the DiegoTc is the one who should do that if I am getting u right May 26 18:33:11 either way, doesn't really matter as long as you capture the right patches with good commit message May 26 18:34:13 nerdboy: yea, I meant that it is his work so he can make the right commit message, and the credit goes to him May 26 18:34:58 moto-timo: nerdboy jkridner : about the quality issues, is it OK if I submit the video before next wednesday (or sooner, depending when I get the mic cable)? May 26 18:35:25 yes, i'm also saying you can make last commit with both names/emails and explain May 26 18:35:31 I feel it doesn't do the project any good to submit an intro video with crappy audio :/ May 26 18:36:00 use phone maybe? May 26 18:36:28 iphone makes good video, also android ... May 26 18:36:33 nerdboy: oh ok ,, but that is a good suggestions anyway, thanks May 26 18:36:41 I'd even settle for computer synthesized voice :) May 26 18:36:45 DiegoTc: what u say? May 26 18:36:59 ebadawy: I will work on that today May 26 18:37:07 but I have to read again this conversation May 26 18:37:14 I'm not a git guru :) May 26 18:37:28 DiegoTc: cool May 26 18:37:44 let me know if you get into trouble... should not be hard to patch from one repo to another May 26 18:37:56 I'm sure it won't be a big issue :) May 26 18:38:03 * nerdboy is plenty experienced unhorking developer repos May 26 18:38:52 not sure i will claim "git guru" but i can usually beat into submission... May 26 18:39:27 just don't use goofy gui tools with crazy reverse-merge or whatever May 26 18:39:52 stackoverflow has many many git answers May 26 18:41:00 (some) gui tools are nice for visualizing but always do commands from git cli May 26 18:41:03 jkridner DiegoTc nerdboy, I have exam tomorrow, have to go now, I'll keep up with u tomorrow. May 26 18:41:51 * nerdboy will admit to using rabbitvcs sometimes for simple stuff May 26 18:42:04 but that's it May 26 18:42:09 nerdboy: yikes! May 26 18:42:20 * moto-timo is always happier with cli results May 26 18:42:23 * jkridner refuses to use anything bug git now. May 26 18:42:31 s/bug/but/ May 26 18:42:40 * moto-timo is forced to use Subversion at work May 26 18:43:12 even worse, I maintain the server May 26 18:43:14 lol May 26 18:43:16 *rabbitvcs as nautilus git front-end May 26 18:43:39 * nerdboy gave up on cvs and subversion mostly May 26 18:43:53 only when i have to touch legacy repos... May 26 18:44:01 it's better than VisualSourceSafe May 26 18:44:23 it works for simple stuff May 26 18:44:47 plus it inherits your desktop credentials May 26 18:45:17 http://i.imgur.com/4amRqLi.png May 26 18:45:27 moto-timo: nerdboy ^^ May 26 18:45:34 comments? May 26 18:45:35 :) May 26 18:45:56 where's his spacesuit? May 26 18:46:11 yeah, with tail enclosure May 26 18:46:27 neem0: so very Laika May 26 18:46:29 ha, it was going over the budget May 26 18:46:40 neem0: thumbs up :) May 26 18:46:43 do you have any idea how much a spacesuit with a tail costs? :D May 26 18:48:47 I think the Chinese gov't can afford it May 26 18:48:50 good thing there's clip art for a astronaut's helmet May 26 18:48:59 a = an May 26 18:49:43 though, someone with a bit more designer's touch should probably fix it up a bit more May 26 18:49:47 ;) May 26 18:50:58 helmet is good, just add puffy marshmallow suit... May 26 18:52:17 http://coloringpagesjos.net/wp-content/uploads/2015/23487-astronaut-suit-coloring-page.gif ? :D May 26 18:52:57 I'll look into it May 26 18:53:04 this is version 0.2 May 26 18:58:57 now you just need a box of crayons :) May 26 18:59:19 although, clearly Boris had a tail-ectomy May 26 19:09:49 a shame it is May 26 19:10:03 off to a small brake, than back to work (brb) May 26 19:37:58 moto-timo: any thoughts on why I get the --api_key error at https://travis-ci.org/beagleboard/beaglebone-getting-started ? May 26 19:38:13 moto-timo, arianepaola: are you working on an automated build/release? May 26 19:38:59 jkridner, arianepaola: vvu and I have been. I would suggest we continue down that path :) May 26 19:39:26 moto-timo: including adding an nw.js autobuild? May 26 19:40:02 jkridner: including nw.js :) May 26 19:42:22 jkridner: first guess is that $GITHUBDEPLOY needs to be defined as an environment variable May 26 19:42:33 did that. May 26 19:42:57 and then tell travis.yml about it... let me grab the syntax May 26 19:43:49 env: May 26 19:43:50 without the $. May 26 19:43:51 global: May 26 19:44:12 - secure: May 26 19:44:23 https://github.com/beagleboard/beaglebone-getting-started/blob/master/.travis.yml#L14 May 26 19:44:37 hmmmm.... I don't know about this secure hash. May 26 19:44:59 right. there's a command line tool to generate that May 26 19:45:11 let me google it. I didn't make a gist :( May 26 19:46:11 from within the project directory May 26 19:46:45 "travis encrypt GITHUBDEPLOY=" May 26 19:47:01 gem install travis May 26 19:47:03 first May 26 19:47:39 ugh... May 26 19:47:48 I'm trying to avoid setting up a local build environment... May 26 19:47:54 so I generated a token on github. May 26 19:48:05 Settings/Personal access tokens May 26 19:48:07 i don't have the right npm/node/nw.js etc. easily. May 26 19:48:39 I generated the token... May 26 19:48:49 and that is what I assigned to $GITHUBDEPLOY May 26 19:49:29 but travis doesn't know about it without the env: \n global: \n GITHUB= May 26 19:49:46 or (to keep your private key private) the -secure: option May 26 19:51:16 I've sent you my notes that I sent to vvu and arianepaola May 26 19:51:51 the travis cli actually inserts the secure token into your .travis.yml May 26 19:51:53 jkridner: Do you have any ideas/leads as to what could be causing the power supply noise due to the LED driver? May 26 19:52:11 added several capacitors, doesn't seem to affect the shaky display May 26 19:52:44 did you examine the current requirement? should you be using a series resistor? May 26 19:52:55 jkridner: unrelated... do you have a fixed IP address at home or are you using dynDNS or similar? May 26 19:53:12 ? May 26 19:53:24 neither... May 26 19:53:28 it changes often... May 26 19:53:44 I try not to need to log into home as that creates more security holes to worry about... May 26 19:53:45 the ubuntu machine too? May 26 19:53:47 jkridner: I'm using a 2R2, Vref=200mV gives ILed~100mA May 26 19:53:47 though it is often nice. May 26 19:54:07 moto-timo: which Ubuntu machine? the one I'm using right now for beagleboard.org? May 26 19:54:11 moto-timo: it is up on AWS. May 26 19:54:19 ah May 26 19:54:56 I have to update AWS with my IP address because I have the X15 asking beagleboard.org for build requests. May 26 19:55:02 I was debating about putting a mac mini on line vs. a VMS in the cloud May 26 19:55:22 so, it is certainly a bit of a security hole, but meh. May 26 19:55:33 ok May 26 19:55:37 I can't get a reasonable uplink speed to do any hosting. :( May 26 19:55:53 I've considered taking a board to a hosting site. May 26 19:55:59 charter usually throttles me down if I do :( May 26 19:56:18 kickstarter opportunity: boot farm for DYI May 26 19:56:20 :) May 26 19:57:10 * Abhishek_ has no clue so far what's wrong with the LED boost driver May 26 19:58:58 s/DYI/DIY May 26 20:00:08 Abhishek_: seems like rcn-ee would be a good person to ask? May 26 20:01:06 moto-timo: It is a hardware issue May 26 20:01:37 moto-timo: LCD is working fine without the backlight on, as soon as I power the LED boost converter, the screen messes up May 26 20:05:38 hrm... gut feel is floating ground or some induced frequency issue... but I'm not an EE May 26 20:08:27 moto-timo: I guess I could try to do it your way, but it seems http://docs.travis-ci.com/user/environment-variables/#Using-Settings says I can set the variable from the web interface... May 26 20:08:33 I just don't get why it isn't set. May 26 20:11:37 jkridner: forking and trying to see if I can figure it out May 26 20:34:09 jkridner: same result... digging deeper May 26 20:44:08 jkridner: https://gist.github.com/douglasduteil/5525750 May 26 20:44:30 seems like we still need to include the encrypted env global May 26 20:53:13 * moto-timo quits to install ruby on cygwin May 26 21:31:41 now that was a dependency nightmare May 26 21:35:59 cygwin is full of pain May 26 21:36:10 i tried to compile openocd in cygwin...never again May 26 21:37:44 *sigh* but at work it's all Windows May 26 21:38:08 I suppose I should run a VM May 26 21:38:19 * moto-timo says to himself for the 100th time May 26 21:39:25 moto-timo, sneak in a 2nd laptop and a usb-eth adapter, enable connection sharing. ;) May 26 21:39:37 win-win! May 26 21:40:15 rcn-ee: not a bad idea May 26 21:40:28 plus, no sneaking required May 26 21:40:46 if your really sneaking, you can sneak a mini-itx case in too with a nice quad/octo core. ;) May 26 21:40:48 our IT manager gives me a fair amount of wiggle room May 26 21:41:03 or buy another NUC May 26 21:41:06 I love my NUC May 26 21:41:17 meh, nobody listens to the IT guy May 26 21:41:36 yeah they are awesome, have one on top of my mini-itx... and it runs my git server at home.. May 26 21:42:48 my OpenEmbedded/Yocto Project build system is now a NUC May 26 21:43:01 with the added benefit that I can boot the image I just built for it May 26 21:43:22 this reminds me to get some work done... May 26 21:43:41 oh yeah... work May 26 21:43:56 * vvu was kept busy by his grading assignemnts kinda-of-a-job and exams May 26 21:44:02 * moto-timo goes afk for 5 minutes May 26 21:44:08 * moto-timo hates grading May 26 21:44:18 worst part of teaching, EVER May 26 21:44:42 oh well May 26 21:49:00 moto-time: let the grade be determined by a RNG May 26 21:53:35 ds2: ever saw a LED boost converter generate ripple on the source power supply? May 26 22:07:57 ds2: great idea May 26 22:08:06 luckily I'm done with that May 26 22:16:03 moto-timo: i still have a set of final exams May 26 22:16:26 vvu: be strong and drink strong coffee May 26 22:16:30 :) May 26 22:21:41 yeah... May 26 22:39:16 grading isn't *that* bad, but it's kinda easy to give yourself extra work... May 26 22:39:37 I used to have 600 exams to grade May 26 22:39:40 that is BAD May 26 22:40:07 all at once? that's *too* many... May 26 22:40:15 yep May 26 22:40:24 even with a grader it was days of work May 26 22:41:09 one class of 30-ish usually took multiple evenings to grade full exam May 26 22:43:20 either way, challenging problems/essays take effort to grade May 26 22:43:32 indeed May 26 22:43:44 even automation can't do the hard stuff May 26 22:44:01 and if's not hard/challenging, why bother? May 26 22:44:03 I did some scantrons, but... you can't do it all that way May 26 22:44:11 scantrons suck May 26 22:44:37 plus I had 6 versions of the test... big lecture hall taking the exam May 26 22:44:46 only in dire situations, like when grades are due the day after the last final May 26 22:45:00 and *you* get the last final... May 26 22:45:24 got that right, brother May 26 22:45:31 only happened to me once May 26 22:45:38 but I did not sleep May 26 22:45:40 at all May 26 22:47:12 * nerdboy should start the mr_science-school-of-not-f*cking-around May 26 22:57:35 Abhishek_: yes if it is not bypassed enough May 26 22:58:24 LCD backlight converters draw quite a bit of current May 26 22:58:50 so any impedence on the power line will turn those current pulses into ripple May 26 23:04:50 so larger bypass caps? May 26 23:35:23 jkridner: even with the ssh key, you definitely need to encrypt it an insert it into the .travis.yml May 26 23:35:37 jkridner: so far I still can't get it to work May 26 23:35:50 jkridner: see https://github.com/veewee-community/veewee-push/blob/486102e6f508214b04414074c921475e5943f682/.travis.yml#L13 May 26 23:36:10 jkridner: May 26 23:36:20 jkdridner: May 26 23:36:24 dang it May 26 23:36:44 jkridner: https://gist.github.com/pghalliday/240fe740d523dad21d3f May 26 23:37:14 jkridner: which all seems much harder than the OATH token way May 26 23:52:48 jkridner: with OATH token https://github.com/moto-timo/beaglebone-getting-started/releases/tag/1.1.1.3 May 26 23:52:53 shazaam! May 26 23:54:52 jkridner: so the reason for the error was that GITHUBDEPLOY was never defined... so api-key pointed to NULL May 26 23:59:23 not necessarily larger May 27 00:05:39 * moto-timo 's head explodes and he goes home May 27 00:05:58 might have to do with not eating lunch **** ENDING LOGGING AT Wed May 27 02:59:59 2015