**** BEGIN LOGGING AT Wed Jul 16 02:59:59 2014 Jul 16 03:12:12 alexanderhiam : arrays work fine now :) [they worked fine yesterday too, just tested them today and confirmed :p ] Jul 16 03:43:20 karki: how do you create an array? Jul 16 03:46:58 alexanderhiam : as of now I'm using 2 predefined arrays to make sure my code is working. Have to ask chris on how he wants the array creation. Jul 16 03:47:07 oh ok Jul 16 03:47:08 forgot to ask him this monday :/ Jul 16 03:47:43 I brought that up in the last meeting, he said the way on botspeak.org is really how it is now, but he's open to suggestions on better ways Jul 16 03:47:54 alexanderhiam : according to you, what would be an ideal way of array declaration? Jul 16 03:49:05 what about 'SET x[n]', where x is the name and n is the length? Jul 16 03:49:10 I was thinking of a separate "DEC" inst for declaration of array and variables......... but I don't really like the idea of another inst :( Jul 16 03:50:10 btw, he also uses the first item in the list to store the length, so arrays take up one more value then their length Jul 16 03:50:32 or "SET x[], n" ? (just to keep it in sync with the fact that SET takes 2 operands) Jul 16 03:50:57 yeah that makes sense Jul 16 03:51:34 thanks for reminding me of the length thing. Jul 16 03:51:42 you'll also want to throw an error if that gets called twice with the same name, you don't want to deal with changing array sizes Jul 16 03:51:54 so arr[0] can't be used? Jul 16 03:51:57 yeah Jul 16 03:52:19 alexanderhiam : all these nitty gritty stuff is going to take a long time :/ Jul 16 03:52:27 arr[0] points to the base address + 1 Jul 16 03:52:55 I'll go with return values and PWM for now, add the size check, error warning etc. for later Jul 16 03:53:14 I also have to add the labels stuff! Jul 16 03:54:15 with the current design you can't allow creating arrays with variables as lengths Jul 16 03:55:20 I'll just ignore the inst, incase we already have the arr var defined. probably will take just 2 extra lines to do that. Jul 16 03:56:12 huh? Jul 16 03:59:02 I mean if x is already defined, then don't execute an other inst which declares x with some new size Jul 16 03:59:19 e.g. "SET x[], 5" Jul 16 03:59:27 oh gotcha Jul 16 03:59:40 creates an array x with 5 (6) slots Jul 16 04:00:08 if I now give "SET x[], 16" -> will be ignored Jul 16 04:00:29 seems reasonable Jul 16 04:00:57 also, Chris said he just ignores instructions that index past the ends of arrays Jul 16 04:01:41 yeah "nops" :) Jul 16 04:01:54 * karki will be back in a while. Jul 16 04:01:57 btw, I'd recommend making your interpreter object oriented so you can buffer instructions that come in one at a time if SCRIPT is given Jul 16 04:02:54 i.e. if your sending a line at a time to execute_instruction() you should be able to send scripts Jul 16 04:10:22 thats already done :) Jul 16 04:10:43 alexanderhiam : are you still looking at legacy code? Jul 16 10:19:19 exit Jul 16 10:23:01 ok Jul 16 10:24:34 good that i do not have in clipboard a big paste cuz putty is a pain with right click, it automatically pastes and clicks enter :) Jul 16 10:53:42 vvu: ping Jul 16 11:34:23 praveendath92: pong Jul 16 11:37:41 I found the reason for cpu soft lock issue. Jul 16 11:37:49 what was it ? Jul 16 11:37:52 Not the prefetch actually. Jul 16 11:38:16 When I comment that function, code enters an infinite while loop. Jul 16 11:38:32 any hint why ? Jul 16 11:39:06 The while loop condition is satisfied always. Jul 16 11:39:23 w/o the compress function. Jul 16 11:39:52 while(end>start) Jul 16 11:40:11 That start value is got close to end and thus loop exit is in the compress function. Jul 16 11:40:33 okok Jul 16 11:42:45 so now some sending is done ? Jul 16 11:42:48 I received a stupid mail from my college asking for my project reports to get my graduation :( Jul 16 11:42:49 I mean, in a specific format. Jul 16 11:43:21 I could work around that. Jul 16 11:44:13 But I didn't clearly get how he is updating start value in the compress function. Jul 16 11:44:31 The paging error is still there w/o hline_trim function. Jul 16 14:18:05 alexanderhiam: ping! Jul 16 14:44:22 I am currently in my dorm and it's a mess as we all have to relocate :/ to a new block. To make matters worse there's limited connectivity Jul 16 15:11:14 jkridner: Not sure if I'll be able to make it today but the blocking issue for BeagleLogic right now is a little bit of time that has to be paid to some urgent work here. It will take a day or two for things to stabilize Jul 16 15:14:18 KumarAbhishek: will you be able to make up the missed time? Jul 16 15:15:02 jkridner: I should be able to this weekend Jul 16 15:42:24 av500: around here ? Jul 16 15:50:18 karki: I was talking about bs_us_interpreter.py Jul 16 15:52:36 8 minutes til and I only have 1 status report? Jul 16 15:52:41 is that right? Jul 16 15:54:06 jkridner: Mine will be there in 3 mins. Jul 16 15:58:33 greetings guys! Jul 16 15:58:45 alexanderhiam : so am I :p Jul 16 15:58:52 apologies i couldn't make it last week Jul 16 15:58:59 but yeah.... I get what you mean Jul 16 15:59:06 one sec.... Jul 16 15:59:47 if you put it in a class it makes it easy to have the script_mode flag persistent across calls to execute_instruction() Jul 16 16:01:20 meeting time! Jul 16 16:01:40 Hello vmayoral :) Jul 16 16:01:51 so actually i wanted to share a couple of issues that I found over the last days, maybe some of you have also encountered it and found a solution already Jul 16 16:02:16 rseethamraju: you working on the weekly status report? Jul 16 16:02:56 alexanderhiam: yes. Jul 16 16:03:00 First one is an issue with the Language in Debian that doesn't allow to configure locale properly. For now i'm fixing it with this script Jul 16 16:03:01 https://gist.github.com/vmayoral/44d6d9f93ac4504f711b Jul 16 16:03:07 hi all Jul 16 16:03:09 onyone has encountered this in Debian? Jul 16 16:03:11 getting started a bit late... Jul 16 16:03:16 hi jkridner Jul 16 16:03:19 hi Jul 16 16:03:24 still reading the reports that are coming in at the after-last-minute. Jul 16 16:03:41 rseethamraju: ok, just making sure Jul 16 16:03:46 I wrote my report, but forgot to press *post* Jul 16 16:03:52 alexanderhiam: done Jul 16 16:04:25 karki: I agree debugging the PRU is hard. there is some kind of open source debugger and CCS also has PRU debug abilities, but I've been using BeagleLogic for a week or so to try to debug PRU code that updates pins. Jul 16 16:05:02 yeah I was thinking of using BeagleLogic :) Jul 16 16:05:10 rseethamraju: you left of that the eQEP and ADT7310 libraries are fully working now (!) Jul 16 16:05:51 jkridner : thanks for the BBB Jul 16 16:05:52 :) Jul 16 16:05:55 karki: I was thinking that the interpreter could keep count track of the instruction size to keep #s working instead of labels. Just treat it like every instruction had a label that incremented by 1. :-) Jul 16 16:06:06 karki: it arrived?!? Jul 16 16:06:19 yeah..... to my house Jul 16 16:06:28 4 of them........ Jul 16 16:06:35 hi jkridner Just sent you a pm Jul 16 16:06:45 alexanderhiam: right. I forgot. internet’s giving me problems. I’ll push those and PhantSteam before the meeting ends hopefully. Jul 16 16:07:30 vmayoral, karki, rseethamraju, praveendath92, DiegoTc, KumarAbhishek: any issues you'd like to add to today chat outside of your project updates/blockers? Jul 16 16:07:42 no Jul 16 16:07:44 I'd like to share two Jul 16 16:07:50 I have one. Jul 16 16:08:01 Kernel paging error. Jul 16 16:08:01 I have one Jul 16 16:08:15 I'm getting it at sporadic intervals :/ Jul 16 16:08:22 alexanderhiam, av500, mranostay, bradfa, panto, vvu, VoltVisionSteve: same question to you: any topics to add today? Jul 16 16:08:29 I was sick for 3 days. Was on the hospital :( that eas my blocker Jul 16 16:08:41 first one was introduced before: issue with the Language in Debian that doesn't allow to configure locale properly. For now i'm fixing it with this script https://gist.github.com/vmayoral/44d6d9f93ac4504f711b. Seems like an issue with the latest Debian released at debian.beagleboard.org Jul 16 16:08:42 i am at the same level as praveendath92, dunno what is with that error he is getting Jul 16 16:09:00 k, was looking for logistical topics before going around the room. Jul 16 16:09:12 vmayoral: k, starting with your blockers... Jul 16 16:09:35 jkridner, alexanderhiam : we must discuss the JIT with panto! Jul 16 16:09:38 praveendath92: what is the status now? can you send data to the phone ? Jul 16 16:09:42 right Jul 16 16:09:43 jkridner, DiegoTc : I just talked to Oscar.... He said that Paola will send us all source files today. He also said that his wife will be glad to make the changes we talked about in the email. He wants DiegoTc to ping him on skype this afternoon so they can walk through the details with his wife. Jul 16 16:09:45 vmayoral: can you file on bugs.elinux.org? Jul 16 16:09:48 * panto pops in for a brief while Jul 16 16:10:00 sorry, I've been under the gun this whole week Jul 16 16:10:24 jkridner: Will submit it sure. Jul 16 16:10:29 vmayoral: sounds like you have a work-around for now... any blockers? Jul 16 16:10:30 panto : you wanted to have a JIT for the pru speak later on, right? Jul 16 16:10:34 karki, I might have some time to talk about it in 12h (my morning) Jul 16 16:10:36 Data is always being sent. I'm testing with a patch to the soft lock Jul 16 16:10:39 karki, yeah Jul 16 16:10:42 jkridner: the second issue could be Jul 16 16:10:43 vmayoral: we'll get it fixed in the upcoming release if you get the bug report in. Jul 16 16:10:54 sorry, just extremely busy with $DAY_JOB this week Jul 16 16:10:56 jkridner: basically, the BBB spuriously shutsdown as if power button is pressed when connecting/disconnecting batteries. Related to http://bugs.elinux.org/issues/85. Jul 16 16:11:20 jkridner, karki: remind me what the JIT concept seemed to be getting in the way of... was it the PWM driver? Jul 16 16:11:26 jkridner alexanderhiam and me wanted to know more about what you had in mind when you ment "JIT" Jul 16 16:11:36 panto : ^ Jul 16 16:11:37 Besides the urgent work, however I did get a functional Socket.IO link between the BBB and the PC. My next goal after I am done with that is to pipe sigrok-cli output to the web page Jul 16 16:11:41 vvu: Problem is my VM in linux is not detecting my tablet well. Jul 16 16:12:37 vmayoral: have you put a scope to the power rail? Jul 16 16:12:48 alexanderhiam : It was mainly cause I didn't know how to use the other PRU cause I did not know what the JIT implementation would look like. Jul 16 16:13:05 right Jul 16 16:13:07 vmayoral: trying to figure out how sensitive it is and if it is really the BBB's fault. might be possible to address with some additional software debounce. Jul 16 16:13:30 jkridner, no topics from me to add Jul 16 16:13:34 jkridner: actually I was hoping to somehow disable the shutdown on software Jul 16 16:13:39 panto : 12h from now seems good :) Jul 16 16:13:40 jkridner: is there a way to do so? Jul 16 16:13:52 vmayoral: yeah, removing it is trivial. Jul 16 16:14:03 just removing a software package.... if I can remember the name. Jul 16 16:14:18 vmayoral, you're getting ground bounce, is it a clean shutdown or instant like a reset button? Jul 16 16:14:27 err, make it 14h :) Jul 16 16:14:36 9am my time Jul 16 16:14:38 bradfa: it's a clean shutdown Jul 16 16:15:04 vmayoral: something with acpi in it Jul 16 16:15:09 panto : k, better for me :) Jul 16 16:15:13 So it's like a spurious power button press? Jul 16 16:15:17 rseethamraju: thanks for updating that, don't want to leave anything off the 'accomplished' list! Jul 16 16:15:17 bradfa: i stored it in a gist some days ago, here it is https://gist.github.com/vmayoral/5102b2ec442308add0fd Jul 16 16:15:18 vmayoral, possibly there's a glitch in the pmic which sets the pwrbutton interrupt, you can disable the shutdown with software Jul 16 16:15:37 vmayoral: try 'apt-get remove acpid' Jul 16 16:15:47 alexanderhiam: sorry :) Jul 16 16:16:03 jkridner: great thanks i'll try removing "acpid" Jul 16 16:16:19 * vmayoral hooking up the board... Jul 16 16:16:22 vmayoral, you can either disable it in your kernel (panto kanged some code I wrote a while back but not sure what's in the evil vendor tree now for this) or you can try to setup your init system to not use what ever button the kernel sends up as the shutdown button Jul 16 16:16:37 vmayoral: though I believe looking at what you are doing to the power rail might solve some problems down the road and that so would looking at better debouncing on the pin. Jul 16 16:16:56 karki, panto, jkridner: that's 2AM for jkridner and me, would it be worth finding a time when we could all be there (and conscious)? Jul 16 16:17:19 I got a baby's visit to the doctor in the afternoon Jul 16 16:17:52 maybe noon my time? Jul 16 16:17:58 alexanderhiam : I don't mind now, but I'm really fine whenever panto wants it, need to get some things sorted out! Jul 16 16:18:22 what about just an hour earlier? 8 for you panto Jul 16 16:18:26 back again, sorry Jul 16 16:19:02 vmayoral|pc, does your internet have ground bounce too? :) Jul 16 16:19:15 bradfa: seems like :) Jul 16 16:19:18 DiegoTc: sorry to hear about your bad health. please get well and send us a status update when you can. hopefully you can continue working wherever you are at. :-) Jul 16 16:19:28 not sure if you saw: vmayoral, you can either disable it in your kernel (panto kanged some code I wrote a while back but not sure what's in the evil vendor tree now for this) or you can try to setup your init system to not use what ever button the kernel sends up as the shutdown button Jul 16 16:19:42 bradfa: for some reason when a new device gets into the network (BBB) i lose connection :( Jul 16 16:20:05 bradfa: got that, thanks Jul 16 16:20:59 that's all from my side, thanks for addressing it guys Jul 16 16:21:05 we keep moving forward Jul 16 16:21:07 vmayoral, but if acpi is doing this now, not sure how much help I can be, I'm still on old kernel and bbw Jul 16 16:22:02 KumarAbhishek: can you try in the future to send your status report early if you need to not be around for the meeting? will you be sending an updated status soon? Jul 16 16:22:08 thanks vmayoral|pc Jul 16 16:22:28 praveendath92: congrats on getting a real image! Jul 16 16:22:49 we are still having issues with the sensor precision but for now we have enough for a rover Jul 16 16:22:54 jkridner: will do it soon. Jul 16 16:23:00 Siddharth did a nice video on it Jul 16 16:23:04 https://www.youtube.com/watch?v=G4sGv707Q-o&list=PLf4Fnww4KiFeFsGuP-s5HS1iSAtg-5SFH&index=2 Jul 16 16:23:10 planes and copter will come after Jul 16 16:23:14 karki: odd that you got 4 boards. wonder if you got boards meant for other students as well. Jul 16 16:23:32 yeah.... idk really Jul 16 16:24:00 jkridner: Sad part is that is from framebuffer data captured on linux and send to Android :| Jul 16 16:25:09 praveendath92: that sounds like a very odd config.... Jul 16 16:25:39 I'd assumed you were hooking an Android phone up to a Linux box and capturing on the Android phone. Jul 16 16:25:49 praveendath92: I guess I don't understand what you've done. Jul 16 16:25:56 I guess vvu and av500 do Jul 16 16:26:23 cat /dev/fb1 then result just pasted in an imageview in java Jul 16 16:26:57 jkridner: I have some issues with framebuffer driver. Jul 16 16:26:58 I captured a raw frame data through a linux command. Jul 16 16:26:58 I took that data on to android and processed as a bitmap. Jul 16 16:26:58 yes. as vvu said. Jul 16 16:28:01 jkridner: I'm guessing I'm the last one on the list, I will be updating the report right now Jul 16 16:28:29 /dev/fb1 is the device created by your driver that is meant to send data to the Android phone, yes? Jul 16 16:28:40 Yes. jkridner Jul 16 16:28:59 and you are just looking that you are able to put data INTO the buffer? Jul 16 16:29:01 When driver sends the data, it is not is raw form. Jul 16 16:29:17 Compressed and other commands added. Jul 16 16:30:24 is this soft lock-up issue something you have a good handle on solving? Jul 16 16:30:41 I was testing if I can process the raw data sent from driver on Android. Jul 16 16:31:02 That's because of an infinite loop. I will have to check the code in the loop. Jul 16 16:31:33 vvu, av500: are you comfortable with where praveendath92 is at right now and what blocking issues are there? Jul 16 16:31:35 I will be able to do it. Jul 16 16:32:08 jkridner: vvu said he has is heading home. Jul 16 16:32:41 k Jul 16 16:32:47 karki: so, when are we meeting? Jul 16 16:33:13 karki: is the PRU debugging something that needs a better solution? Jul 16 16:34:02 karki, panto, alexanderhiam: when are we getting our heads wrapped around the priorities, including thoughts on JIT, etc.? Jul 16 16:34:53 I'm feeling now like it would be best to ignore JIT and just make the simplest PWM driver you can on the second PRU Jul 16 16:35:36 it's isolated enough that it can be rewritten later (post-GSoC) to add the JIT functionality without much trouble Jul 16 16:35:38 k. hopefully we are going to get a time when we can all chat w/ panto included. Jul 16 16:35:47 please do that Jul 16 16:36:04 I'm swamped atm Jul 16 16:36:31 panto: you mean ignore JIT? Jul 16 16:36:35 karki: please coordinate with panto and alexanderhiam for a time. I will find a way to join whenever. Jul 16 16:36:43 no, I mean let's talk tomorrow Jul 16 16:36:46 alexanderhiam: I think finding a time when panto is available Jul 16 16:36:47 oh :P Jul 16 16:36:50 :) Jul 16 16:36:51 jkridner alexanderhiam panto : I'm not doing PWM this week, so we have a few days. Jul 16 16:36:56 hello Jul 16 16:36:59 hi mranostay Jul 16 16:37:03 welcome to the fun. Jul 16 16:37:07 panto does 8AM your time work? Jul 16 16:37:13 rseethamraju: waiting around patiently? Jul 16 16:37:25 it does somewhat Jul 16 16:37:38 ping me on the irc window, and I will see it Jul 16 16:38:22 ok, jkridner, can you be in irc at 1AM? Jul 16 16:38:41 yikes. tonight? ok. Jul 16 16:38:42 jkridner: kind of Jul 16 16:38:58 better tonight than that Monday monstrosity. Jul 16 16:39:04 jkridner: it's slightly better than the 1:30 meeting ;) Jul 16 16:39:23 yeah, I'm trying to avoid continuing to be out of sync until Monday Jul 16 16:39:30 rseethamraju: how is PyBBIO coming along? was testing done with OpenCV capture now? Jul 16 16:39:57 no. I’m doing the recording/streaming using gstreamer Jul 16 16:40:00 k Jul 16 16:40:43 saving a picture is complicated in gstreamer so maybe that I’ll do in opencv Jul 16 16:40:52 using you’re program Jul 16 16:41:06 * karki wonders how heavy PyBBIO will become......... Jul 16 16:41:43 karki: alexanderhiam is going to release it on pip in a few days Jul 16 16:41:58 karki: a lot of the libraries it uses are already installed on the newer debian images Jul 16 16:42:06 cool Jul 16 16:43:08 * jkridner doesn't understand the significance of running https://github.com/sparkfun/phant Jul 16 16:43:37 jkridner: people love IoT Jul 16 16:43:58 but a node.js toolkit rather than python? Jul 16 16:44:05 jkridner & VoltVisionSteve I just updated the status, sorry for the delay Jul 16 16:44:32 jkridner: rseethamraju is only writing a client for it, it posts data to a remote phant server Jul 16 16:44:50 by default to the data.sparkfun.com server Jul 16 16:45:10 or ppl can install it on their bbb Jul 16 16:45:16 DiegoTc: thanks Jul 16 16:45:47 not just some python code to push to https://data.sparkfun.com/streams/, but instead using the phant node.js lib to do the pushing? Jul 16 16:46:41 jkridner: the python code is always just interfacing to HTTP API Jul 16 16:46:52 alexanderhiam: I'm told what would make a cool demo is using Google Analytics for data logging. Jul 16 16:47:07 the library is used to push/log data to the node.js phant server installed on the bbb Jul 16 16:47:24 ah. Jul 16 16:47:28 jkridner: how would that work exactly? Jul 16 16:47:59 alexanderhiam: they publish an API. you'd need a google analytics account (free for base features). Jul 16 16:48:21 so you can push data yourself? Jul 16 16:48:36 I like the idea of using the bone for the server.... I didn't understand that was what you were doing. Jul 16 16:49:01 so google analytics is another IoT library? Jul 16 16:49:02 you are enabling pushing data to http://data.sparkfun.com or to a BeagleBone using the same simple interface (telnet) Jul 16 16:49:05 like phant? Jul 16 16:49:16 rseethamraju: kinda. Jul 16 16:49:28 oh Jul 16 16:49:29 jkridner: that's up to the user, if they've installed the phant server on their bone they can use localhost instead of data.sparkfun.com Jul 16 16:49:48 rseethamraju: it is typically used to track web page traffic, but you can log anything. Jul 16 16:50:42 alexanderhiam: I like your approach. didn't know about it. thanks for the education. you'll definitely need to do some explaining of the new features when this is all done. Jul 16 16:51:09 seems like lots of time is getting lost in the image capture stuff. Jul 16 16:51:22 ya. We’ll do a wiki. Jul 16 16:51:43 any plan to get around the image capture/streaming issues? Jul 16 16:51:56 using openCV Jul 16 16:52:11 jkridner: yup, gstreamer is huge and complicated. I feel we've been keeping on track with other stuff in parallel though Jul 16 16:52:26 you’re capture program works so we’ll use that Jul 16 16:52:37 k. Jul 16 16:53:04 rseethamraju: wrapped up? Jul 16 16:53:08 cv2 is preinstalled so there won’t be a problem Jul 16 16:53:25 yes Jul 16 16:53:48 DiegoTc: it seems the issues are all that you just haven't had time to work on things since you've been sick. you starting to feel better? Jul 16 16:53:58 karki: fyi: running a script now that sets DIO[0] low then high without any delay then loops, and it's toggling at about 300kHz Jul 16 16:54:03 jkridner: yes Jul 16 16:54:29 alexanderhiam : hmm thats bad :/ Jul 16 16:54:31 Start working 120% to recover time Jul 16 16:54:39 alexanderhiam: I wonder if compiler optimization flags are even set? Jul 16 16:54:40 jkridner: did you saw the create? Jul 16 16:54:44 is that Ok for you Jul 16 16:54:58 no, i haven't set compiler optimizations Jul 16 16:54:59 the only thing I left was the CSS line of the editor Jul 16 16:55:12 karki: judging y the pulse widths there's about a 1us overhead in the looping Jul 16 16:55:14 DiegoTc: yes. Jul 16 16:55:21 DiegoTc, are you ok pinging Oscar this afternoon to talk about the graphic design mods his wife will make? Jul 16 16:55:30 (on skype) Jul 16 16:55:37 VoltVisionSteve: yes I will ping him today Jul 16 16:55:46 DiegoTc: I noticed if not logged in it takes me to the listing. Jul 16 16:55:53 alexanderhiam : thats too much. 1000 inst? Jul 16 16:56:01 DiegoTc: the toolbar on top looks odd to me. Jul 16 16:56:41 jkridner: yes that way we will have a better control from who creates tutorials Jul 16 16:56:44 DiegoTc: I like the pulldown for the code type selection. Jul 16 16:57:21 jkridner: which toolbar? Jul 16 16:57:48 DiegoTc: above the cards. Jul 16 16:58:19 odd, works on my mac. Jul 16 16:58:30 On my Windows machine, they are all squashed. Jul 16 16:58:35 both running Chrome. Jul 16 16:58:44 karki: https://gist.github.com/alexanderhiam/bfc43b44b1109c1012ee Jul 16 16:59:09 <_av500_> jkridner: sorry, in a train with crap net Jul 16 16:59:13 karki: 2.2us positive width, 1.12us negative width Jul 16 16:59:15 <_av500_> (unexpectedly) Jul 16 16:59:17 could you send screenshot jkridner please. Currently I'm using windows Jul 16 16:59:22 DiegoTc: will you be adding an "edit" (existing content) soon? Jul 16 16:59:28 yes Jul 16 16:59:52 alexanderhiam : thats weird; i *need* and oscilloscope :/ Jul 16 16:59:59 but is the create OK, for you, you wouldn't ask for drastical changes next week Jul 16 17:00:43 karki: I can help with scoping, if you send me tasks I can send you captures back Jul 16 17:01:44 http://imgur.com/mzuVTZe Jul 16 17:01:50 alexanderhiam : I'll look into it once I finish the implementation of all the instruction for the pru interpreter. then we will know the final speed. Then we'll optimize from there. Jul 16 17:02:49 DiegoTc: nothing drastic. VoltVisionSteve? Jul 16 17:03:02 jkridner: I'm asking because my plan is to have the edit with the same interface Jul 16 17:04:00 alexanderhiam : the instructions are in memory (DDR), thats the problem. Jul 16 17:04:14 right Jul 16 17:04:30 DiegoTc: I'm not sure that "Card 2" and "Card 3" need to be there by default. Jul 16 17:04:52 Heres what it looks like for me on Win8/Chrome... http://imgur.com/17dwUfO Jul 16 17:05:26 alexanderhiam : but it still does not explain why it is so slow, I was expecting it to be a few MHz Jul 16 17:05:33 DiegoTc: is the autosave there? Jul 16 17:05:38 yes! Jul 16 17:05:48 even on initial create? Jul 16 17:06:01 I got an "Error creating the tutorial" Jul 16 17:06:15 shall we wrap up the official meeting? Jul 16 17:06:21 * jkridner raises gavel Jul 16 17:06:34 *slam* Jul 16 17:06:48 DiegoTc: are you still using css to make the board outline? Jul 16 17:06:54 alexanderhiam: yes Jul 16 17:07:09 DiegoTc: any idea why I might get that error? Jul 16 17:07:23 I'm skeptical that will be consistent across browsers and viewports... Jul 16 17:07:31 jkridner: not really, did you wrote something? Jul 16 17:07:37 or just where changing cards? Jul 16 17:08:00 I added a card, changed the names of the cards, added some content and hit "Save" Jul 16 17:09:02 jkridner: this is your gist: https://gist.github.com/jadonk/40d5b1696ded55360c2b Jul 16 17:09:05 Me too... I just edited the 3 default cards, then hit save ... this is the error... http://imgur.com/D3pjjGh Jul 16 17:10:11 http://imgur.com/7WZPlaF Jul 16 17:12:17 jkridner & VoltVisionSteve could you send me a screenshot but seeing the debugger console on chrome Jul 16 17:13:02 DiegoTc: if you're interested I made a couple quick SVGs you could use for the outlines instead of drawing them with css. That way you could also use a :hover to do some snazzy drop shadowing when you mouse over a card (http://www.w3schools.com/svg/svg_feoffset.asp) Jul 16 17:13:10 error from github is 442 (unprocessable entity) Jul 16 17:13:34 like this...? http://imgur.com/GhAHLTP Jul 16 17:14:18 alexanderhiam: let's ask jkridner what he thinks about that Jul 16 17:14:42 do all browsers support SVG now? Jul 16 17:15:29 the 422 error is cause of the json send to github. Wondering what you are adding. Jul 16 17:15:49 jkridner: Note: Internet Explorer and Safari do not support SVG filters yet! Jul 16 17:16:17 I just added text on the 3 different cards... I made sure the code text was a comment in python. Jul 16 17:17:01 alexanderhiam: can you make a fork and show us what it looks like to help us decide? Jul 16 17:17:28 * jkridner doesn't really know all of the SVG vs. CSS tradeoffs. Jul 16 17:17:44 I'd like to be able to easily embed decks on any web page. Jul 16 17:18:02 jkridner: ok, I'll do that soon. I think the support is about the same as it is for the fancy CSS drawing (CSS3+HTML5) Jul 16 17:19:30 VoltVisionSteve: could you try it again Jul 16 17:19:37 and let me know if you get the error again Jul 16 17:20:39 DiegoTc, when I hit "Home" nothing happens, so I dont know how to "get back to create"...I will close the window and re-open it now. Jul 16 17:21:37 jkridner: http://caniuse.com/svg Jul 16 17:23:38 DiegoTc, I tried it again with a fresh create and still got an error ... http://imgur.com/jbL25GJ Jul 16 17:24:11 mmm VoltVisionSteve would you be on todays meeting with Oscar Jul 16 17:24:12 ? Jul 16 17:24:18 that way I can see what you do Jul 16 17:24:32 I'm not sure why you're having that error. I'm not getting it Jul 16 17:25:15 DiegoTc, I have to leave here to go to customer in about 1hr....but I can do screenshare with you right now (after this meeting). Jul 16 17:25:50 or could you save a small video Jul 16 17:26:07 if that's not bad Jul 16 17:27:02 I dont have any screencapture/screencast software...my trial version of Camtasia Studio expired. Jul 16 17:27:19 I can do it with my phone I guess. Jul 16 17:27:58 DiegoTc: I did nothing but type "Dummy" into the preview card and hit "Save" and still got the same error. Jul 16 17:28:36 I agree that "Home" should take you back to the listing, not the Create page. Jul 16 17:28:53 jkridner: it works for me: http://diegotc.github.io/bone101/Support/GSOC/app/views/tutorial.html?gistid=802f77db2107eeecd381 Jul 16 17:29:50 DiegoTc, I agree. All I am doing it typing text into the 3 cards and hitting save. Jul 16 17:30:30 * DiegoTc doesn't knows whats going on. I will try on different computers right now Jul 16 17:31:18 This is the link: http://diegotc.github.io/bone101/Support/GSOC/app/views/create.html Jul 16 17:34:02 DiegoTc, I just started a new one...I only typed text into Card1 (nothing else)...hit save... got error... Jul 16 17:34:03 http://imgur.com/qNFVyGK Jul 16 17:35:43 DiegoTc, the first time you hit save it gives a 404 error... the second and third times it gives a 422 error. Jul 16 17:39:51 DiegoTc, does this help looking @ the network tab of the Chrome Develope Tools? http://imgur.com/FXoB2ce Jul 16 18:03:39 DiegoTc: I'd recommend leaving all html out of the gists. IMO the gists should only contain content, then the javascript code that fetches them should put that content into html elements Jul 16 18:04:39 imagine if you had a whole bunch of live gists then realized you wanted another
or something, currently you'd have to go through and change every single gist Jul 16 18:08:30 DiegoTc: you could instead have files like 'description.txt', 'card1.txt', etc. you could also have user defined code files, e.g. 'my_example.js', then reference them in the card files with something like [code src='my_example.js'] Jul 16 18:09:34 then that file would be inserted there by the javascript code Jul 16 18:10:08 templating is good Jul 16 18:10:54 alexanderhiam: I am doing that Jul 16 18:11:09 I'm getting the content of each html file Jul 16 18:11:45 and displaying it on the layout. That's why the editor has the toolbar on top, WYIIWYG Jul 16 18:12:35 but it's saved in the gist with html elements Jul 16 18:12:40 I don't quite get your example of having whole bunch of live gists then realized you wanted another
Jul 16 18:12:54 I'm saying the gists should only contain text content Jul 16 18:14:07 alexanderhiam: could you be more specific on that. Still don't get you why Jul 16 18:14:53 for instance iyour CARD_1_IN_1.html file has . What if you decide later that you want to change that? You'd have to go in and edit the gists themselves Jul 16 18:16:14 another example: if you have 100 gists that all have the content wrapped in a and you realize that you really need to have it wrapped in a

you would currently have to go through and edit all 100 gists Jul 16 18:16:52 whereas if the gists only contained text content, then you'd only have to change the one line in the javascript code that puts the content into a Jul 16 18:16:59 alexanderhiam: not, you will click on the edit option and the content will load on the summernote editor Jul 16 18:17:12 the span and everything is because of the sumernote editor Jul 16 18:17:41 ah, gotcha. Do you see the problem there though? Jul 16 18:18:00 Ithe problem on using the summernote editor? Jul 16 18:18:12 or the problem you were trying to explain me at first? Jul 16 18:18:48 on using summernote I don't see the problem. Jul 16 18:19:01 Could you explain that to me Jul 16 18:21:27 alexanderhiam: ? Jul 16 18:22:03 hmm... perhaps it can be worked with. It's just really hard to style the cards when you don't have control over the html inside them Jul 16 18:22:41 e.g. making the content fit within the small preview cards is really hard when everything is wrapped in a variable number of divs and spans that don't have classes Jul 16 18:26:05 alexanderhiam: that's why jkridner have the idea of using the small card for preview. http://screencast.com/t/VaKvZy282F Jul 16 18:26:20 if it doesn't fit there, forget it will fit Jul 16 18:26:40 that makes snese Jul 16 18:27:00 there's also no need to have 'Overview' on all the cards Jul 16 18:27:41 ahh yes, that was the first tutorials I copy paste Jul 16 18:28:07 alexanderhiam: could you try to create one, and tell me if you''re getting an error Jul 16 18:28:28 jkridner & VoltVisionSteve did, and just added hello to the preview card and click save and error Jul 16 18:28:39 I can't reply that :( Jul 16 18:30:58 DiegoTc: error for me too Jul 16 18:31:59 mmm I got it, cookies where the issue Jul 16 19:18:32 hmmmm Jul 16 19:18:54 all's quiet and well on the GSoC front I see Jul 16 20:36:32 alexanderhiam: ping =) Jul 16 20:36:41 pong Jul 16 20:37:14 you mean print message in the example right? Jul 16 20:37:27 yeah Jul 16 20:37:34 to make it clear what the numbers are Jul 16 20:38:26 ok Jul 16 21:19:44 alexanderhiam: done. I pushed the PhantStream library and the other changes to my experimental branch Jul 17 00:05:12 hi jkridner Jul 17 00:05:37 hi DiegoTc Jul 17 00:06:17 jkridner, I open a post on the list for helping testing: https://groups.google.com/forum/#!topic/beagleboard-gsoc/e-u1KOIlei8 Jul 17 00:06:49 I'm trying on a different computer right now and it works. I can't get the error you and Steve have Jul 17 00:06:57 this is a video: http://screencast.com/t/btfR3UQw Jul 17 00:07:01 saw that Jul 17 00:07:18 I thought you knew it was cookie related? Jul 17 00:08:18 I supposed it was that, but not sure. You're still getting the error Jul 17 00:08:19 ? Jul 17 00:08:38 FYI, it is odd that the preview card shows up so big in the card view... I think the preview card shouldn't even show up except in the list view and editor. Jul 17 00:09:14 yup Jul 17 00:09:23 back in a bit Jul 17 00:09:24 yes I have to fix that Jul 17 00:33:44 I don't have time to debug the issue right now, but it is there. I did a cache flush to make sure it wasn't anything old sitting around. Jul 17 00:35:14 jkridner, you're still getting the error then? Jul 17 00:35:22 yes Jul 17 02:37:32 alexanderhiam : optimizations have to be my top priority :p [I ran out of text space in the PRU] Jul 17 02:48:02 oh no! Jul 17 02:58:18 alexanderhiam : then I realized that the current linker file allowed only 4KB of space, I increased that to 8KB. Now it works fine Jul 17 02:58:44 I have to abstract all the repeating stuff into functions...... Jul 17 02:58:54 also use -o3 while compiling **** ENDING LOGGING AT Thu Jul 17 02:59:59 2014