**** BEGIN LOGGING AT Wed May 21 02:59:58 2014
May 21 03:16:32 <karki_> good morning ^_^
May 21 03:18:43 <karki_> Abhishek_ : I need some help with the PRU, I think it specifically relates to your project.
May 21 03:21:58 * karki_ guesses that Abhishek_ might be sleeping for another two hours.
May 21 03:23:18 <Abhishek_> zzz
May 21 03:27:27 <karki_> hehe
May 21 03:36:06 <Abhishek_> karki: what in the PRU?
May 21 03:44:14 <Abhishek_> *karki_
May 21 03:46:03 <karki_> Abhishek_ : I want to know how you are moving the sampled data from the PRU to your userspace application.
May 21 03:47:10 <Abhishek_> Currently trying two approaches: Writing from PRU to DDR RAM directly or setting up a circular buffer in PRU shared RAM
May 21 03:48:27 <karki_> Hhmm..... I need to implement PRU->remote proc -> userspace in my driver...
May 21 03:48:35 <karki_> was thinking of the best way.
May 21 03:48:41 <Abhishek_> For your application, since you mostly will be moving data from the system RAM to the PRU RAM, your approach may be different, as my approach is mostly to do it as *fast* as possible
May 21 03:49:02 <karki_> then realized you would have done the hard work ;)
May 21 03:49:13 <karki_> mine goes both ways
May 21 03:49:37 <karki_> I need speed for PRU -> userspace
May 21 03:50:08 <karki_> but doesn't need to be as fast as yours
May 21 03:50:22 <karki_> why are you not using the SRAM
May 21 03:50:33 <karki_> wait..... DDR with uio?
May 21 03:50:50 <karki_> wont there be problems with the cache ?
May 21 03:52:36 <Abhishek_> I remember panto telling that, I thought the best way for measuring the impact for my application would be to actually try it and see the results
May 21 03:53:45 <Abhishek_> remoteproc implementation can do things correctly, but I'm going to set up with UIO and move to the remoteproc as the code develops
May 21 04:42:27 <ds2> try doing shared memory first
May 21 04:42:38 <ds2> writing to DDR means you have to allocate DMAable memory
May 21 04:42:42 <ds2> and that can get nasty
May 21 04:42:56 <ds2> esp. if it is userspace accessible
May 21 04:43:31 <Abhishek_> ds2: What if the UIO driver allocates it through extram_pool_sz? I'm not using mmap directly in my code
May 21 04:44:16 <ds2> that can still get tricky to debug
May 21 04:44:46 <Abhishek_> I'm currently trying broadside register access
May 21 04:45:00 <Abhishek_> from one PRU to the other
May 21 04:45:02 <ds2> for you, that is a one way transfer
May 21 04:45:11 <ds2> 2 way xfer is can have other things to watch out for
May 21 04:46:17 <ds2> so basically one PRU acquires and the other acts as a programmable DMA engine?
May 21 04:46:30 <Abhishek_> Currently I'm thinking of an approach where PRU1 does only register->register work, PRU0 handles the dirty work
May 21 04:46:34 <Abhishek_> yes
May 21 04:53:12 <Abhishek_> ds2: Something like this: https://gist.github.com/abhishek-kakkar/aa5f411304a95d1fb28a
May 21 04:53:38 <Abhishek_> If only this can be sustained
May 21 04:58:21 <ds2> what's your current target sample rate?
May 21 05:00:09 <Abhishek_> I'm targeting as fast as possible that can be sustained [100 MHz, 50MHz, 25MHz].
May 21 05:02:09 <ds2> are you doing 8 bits or 16bit of data?
May 21 05:03:06 <Abhishek_> 16bits
May 21 05:03:21 <Abhishek_> There are 12+2 possible HW inputs
May 21 05:37:40 <ds2> hmmmm
May 21 05:37:54 <ds2> it might make it easier if you attempt 8 bit first
May 21 05:38:05 <ds2> cuz with 8 bit, you gain a factor of 4 in speed
May 21 05:38:16 <ds2> the L3 is 32bit wide so...
May 21 05:40:03 * mranostay kicks the 2.0.0 compiler
May 21 05:40:18 <mranostay> Abhishek_: do 8 bit first
May 21 05:40:39 <Abhishek_> k, did you see the gist above?
May 21 05:40:40 <mranostay> be sure they all share the same byte
May 21 05:41:01 <mranostay> so you can do one clock cycle or instruction
May 21 05:41:43 <mranostay> i assume you mapped the registers with XCHNG?
May 21 05:41:59 <Abhishek_> currently doing that part
May 21 05:42:15 <mranostay> btw HUGE stalls if both hit it at the same time
May 21 05:42:29 <mranostay> pru1 has priority over pru0
May 21 05:42:54 <mranostay> you'll want to spinlock till the other pru sends a interrupt
May 21 05:43:11 <ds2> you can change the priority around with a config register
May 21 05:43:16 <Abhishek_> mranostay: https://gist.github.com/abhishek-kakkar/aa5f411304a95d1fb28a
May 21 05:43:17 <Abhishek_> yep
May 21 05:44:34 <mranostay> but better to lock :)
May 21 05:44:51 <Abhishek_> lock with the SLP instruction?
May 21 05:45:22 <mranostay> hmm SLP
May 21 05:45:28 <mranostay> is that a macro for a loop?
May 21 05:45:50 <mranostay> ah he SLP instruction will sleep the PRU, causing it to disable its clock.
May 21 05:46:14 <mranostay> no you want the other PRU to send an interrupt and you spin on the r31 event
May 21 05:47:07 <ds2> something like QBNE $$, r31.tXX
May 21 05:47:20 <ds2> but SLP can be made to work too
May 21 05:47:26 <ds2> slp is like wfi
May 21 05:47:34 <Abhishek_> I thought SLP would be nices
May 21 05:47:36 <Abhishek_> *nicer
May 21 05:47:49 <ds2> looping is easier to debug...
May 21 05:48:03 <Abhishek_> yep
May 21 05:49:10 <Abhishek_> ds2 mranostay: How long before you both leave for tomorrow?
May 21 05:49:38 <ds2> donno
May 21 05:49:43 <ds2> depends on when I doze off
May 21 05:50:56 <mranostay> half hour to a hour i guess
May 21 05:52:12 <Abhishek_> k
May 21 06:39:03 <karki_> panto : where would my sysfs entry be? ( /sys/class/rproc/pru-speak ?)
May 21 06:56:20 <panto> where your platform driver is
May 21 08:14:50 * Abhishek_ finds it interesting to note that the PRU has 16 different kinds on NOPs but wonders why
May 21 08:16:39 <karki_> maybe if you get tired of one way of doing nothing, you can do it another way ;)
May 21 08:17:04 * Abhishek_ *of
May 21 09:48:00 <Abhishek_> vmayoral|pc: Hey. did you investigate the PRU issue?
May 21 09:57:09 <vmayoral|pc> Abhishek_: I didn't have the time yet. Finishing some modifications to the I2CDriver layer
May 21 09:57:12 <vmayoral|pc> Abhishek_: any ideas?
May 21 09:57:41 <vmayoral|pc> Abhishek_: before i look at it in more detail i want to try how the Xenomai-patched kernel responds
May 21 10:02:03 <Abhishek_> k, I saw it in the mailing list, don't have any ideas for now though.
May 21 10:03:36 <Abhishek_> Let me know of any leads on Xenomai.
May 21 12:31:33 <karki_> vmayoral|pc : what exactly is Xenomai ?
May 21 12:32:34 <vmayoral|pc> karki_: a set of patches for the Linux kernel to convert it  in a Real-Time system
May 21 12:32:35 <vmayoral|pc> http://www.xenomai.org/
May 21 12:33:05 <vmayoral|pc> both the RT_PREEMPT and the Xenomai patches convert the Linux kernel into an RT system. They do it in a different way though
May 21 12:33:53 <Abhishek_> Xenomai patch does not exist for 3.x kernels?
May 21 12:36:43 * Abhishek_ seems to have been confused between Xenomai versions and Linux kernel versions
May 21 12:40:28 <Abhishek_> panto, mranostay, ds2: is the C application fast enough to catch two interrupts to the PRU raised in quick succession?
May 21 12:40:58 <panto> don't mess with RT_PREEMPT
May 21 12:41:04 <panto> it's more trouble than it's worth
May 21 12:41:36 <av500> tglx agrees?
May 21 12:41:53 <Abhishek_> panto: I mean, if I raise two successive PRU0_ARM_INTERRUPTs in quick succession, will I be able to catch both of them from the UIO impl
May 21 12:41:54 <panto> well, we can ask him
May 21 12:42:13 <panto> Abhishek_, if you depend on that your logic is broken
May 21 12:42:24 <vmayoral|pc> panto: then you recommend me to go straight to Xenomai without looking into RT_PREEMPT?
May 21 12:42:32 <panto> don't go to either
May 21 12:42:47 <panto> standard linux is good enough for a few 10s of ms
May 21 12:42:54 <panto> use the PRU for the rest
May 21 12:43:24 <panto> you have to understand that no-one developed all those drivers under RT_PREEMPT
May 21 12:43:39 <panto> every single one of the drivers you use, plus what ever subsystems you use have to be verified under it
May 21 12:44:06 <panto> it might work, or it might not (probably not)
May 21 12:44:12 <vmayoral|pc> makes sense
May 21 12:44:32 <vmayoral|pc> the dicussion we had (https://groups.google.com/forum/#!topic/beaglepilot/7DKcdm0AEPo) pointed out that the issue might be in the capemgr though
May 21 12:44:41 <panto> before you jump into RT patches for linux, evaluate if standard linux + PRU is good enough
May 21 12:44:52 <panto> vmayoral|pc, well, I never tried capemgr under it
May 21 12:45:03 <panto> and since the lock in question seems to be the devtree lock
May 21 12:45:16 <panto> it might even be a problem with the DT locking in general
May 21 12:45:23 <panto> you're on 3.8
May 21 12:45:27 <vmayoral|pc> yes
May 21 12:45:38 <panto> all of the patches for the new overlays is against 3.15
May 21 12:45:45 <panto> or will be
May 21 12:46:19 <Abhishek_> panto: Can you suggest an alternative approach? What I'm trying to debug is: PRU1signals->PRU0 , PRU0 sends an interrupt (so that I know that PRU0 has been signalled), transfers some data into the RAM, and raises an interrupt again.
May 21 12:46:29 <panto> so... if you can't pin-point what exactly the problem is I have no inclination to figure out what's broken on 3.8
May 21 12:46:45 <panto> and broken under PREEMPT_RT
May 21 12:47:26 <panto> mdp & I used to make a good living a few years back figuring out what's broken under random PREEMPT_RT versions
May 21 12:47:41 <panto> and most of the time it wasn't something obvious
May 21 12:48:28 <vmayoral|pc> panto: thanks for your comments, I honestly don't feel really comfortable with PREEMPT_RT and i'm afraid that it could take me too much time to figure out what's wrong. i'll share your thoughts with the group. Everthing's working fine in vanilla kernel
May 21 12:49:02 <vmayoral|pc> and there're plans to move what we need to the PRU.
May 21 12:50:32 <panto> if you don't really really need PREEMPT_RT don't use it unless you're willing to fix each and every possibly affected driver
May 21 12:51:51 <vmayoral|pc> I'm afraid that'll be too much complexity.
May 21 12:52:18 <vmayoral|pc> I appreciate that you shared your thoughts. I'll keep going with vanilla for now.
May 21 12:56:09 <panto> Abhishek_, try to use some kind of ack scheme
May 21 12:56:23 <panto> don't rely on the other end picking up interrupts fast enough
May 21 12:56:28 <panto> this is by definition racy
May 21 13:03:03 <Abhishek_> panto: Do you have a few minutes to take a quick look at the current firmware?
May 21 13:03:29 <Abhishek_> https://github.com/abhishek-kakkar/BeagleLogic/blob/prutest/PRULATest/src/pru0fw.asm
May 21 13:03:35 <Abhishek_> https://github.com/abhishek-kakkar/BeagleLogic/blob/prutest/PRULATest/src/pru1fw.asm
May 21 13:14:34 <karki_> Hi jkridner
May 21 13:15:38 <jkridner> hi karki_
May 21 13:15:43 <karki_> jkridner : Could you give panto the link for the google doc for the student hardware requirements?
May 21 13:15:58 <karki_> if he hasn't already filled it!
May 21 13:16:17 <jkridner> https://docs.google.com/spreadsheet/ccc?key=0AtD7XdBlve3HdFRaU1NEZE40cUgxNFlYTEdjNklYcEE&usp=sharing
May 21 13:16:34 <karki_> panto : could you please fill this up for me?
May 21 13:16:40 <jkridner> I don't see an entry.
May 21 13:17:31 <karki_> meeting in 3 hrs?
May 21 13:17:41 <av500> 2:40
May 21 13:20:47 <vmayoral|pc> jkridner_: did you make up your mind about the "Google three day unconference". I'd be interested in participating
May 21 13:20:57 <vmayoral|pc> ?
May 21 13:21:05 <jkridner> I started typing a response...
May 21 13:21:11 <jkridner> I really appreciate that you brought it up in public...
May 21 13:21:33 <vmayoral|pc> i was unsure if it was the right thing. Glad to hear that
May 21 13:21:42 <jkridner> I guess I was hoping to use the $$$ for myself, but that might be too selfish.
May 21 13:21:50 <Abhishek_> hi jkridner
May 21 13:21:50 <jkridner> hopefully I can get TI to pay.
May 21 13:22:14 <jkridner> let's discuss in the meeting today. I'll try to get TI to pay for me and Google to pay for you + av500.
May 21 13:22:36 <jkridner> I'll need to see if Google will approve a 3rd person going, but I don't think it'll be an issue.
May 21 13:23:10 <jkridner> I think because you are a 2-time Beagle GSoC'r that it'll be understood. Also helps that you jumped on it fast.
May 21 13:24:55 <av500> vmayoral|pc: are you a mentor?
May 21 13:24:58 <av500> or student?
May 21 13:25:02 <vmayoral|pc> a student
May 21 13:26:19 <av500> ok
May 21 13:27:26 <av500> ah yes, students can be delegates too
May 21 13:27:28 <av500> np then
May 21 13:27:41 <vmayoral|pc> yeap
May 21 13:28:08 <av500> they changed that for this year
May 21 13:29:45 <DiegoTc> hey jkridner, tell me you like this: http://diegotc.github.io/bone101/Support/GSOC/BLINKINGTUTORIALS/Blink.html and that the blue color as background is what you don't like  :p (And the View All)
May 21 13:31:12 * DiegoTc started to make a list of things to improve, fonts, color, image borders. Guessing jkridner will mentioned it 
May 21 13:31:49 <jkridner> DiegoTc: I think it is OK, but I think it could look better.
May 21 13:34:26 <DiegoTc> jkridner you are referring to the gist view? I have to define fonts and other things at the moment for showing it. Or you mean the way it's
May 21 13:34:54 <jkridner> overall cleanliness
May 21 13:41:51 <jkridner> DiegoTc: what do other people think of it?
May 21 13:42:11 <DiegoTc> from this one I haven't get feedback
May 21 13:42:58 <DiegoTc> from the last one, I took Steve idea of parallex
May 21 13:43:17 <DiegoTc> if you see, you can use arrows from keyboard or the ones of top
May 21 13:47:57 <Abhishek_> DiegoTc: scrolling with keyboard isn't working on firefox but is fine in chromium
May 21 13:48:30 <DiegoTc> Abhishek_: thanks!
May 21 13:48:40 <DiegoTc> Abhishek_: do you feel confortable using it?
May 21 13:50:23 <Abhishek_> While going from the 4th to the 5th card, IMO it would be better to have it on the same row rather than the animations of  going down (others may have a different opinion)
May 21 13:52:09 <Abhishek_> the bone outline adjusts well on my 1920x1080 screen, could be made a little bigger
May 21 13:52:36 <DiegoTc> Abhishek_: could you send me a screenshot please :)
May 21 13:57:18 <DiegoTc> praveendath92, vmayoral|pc karki_ hey guys, could you do me a favor and give me feedback for this: http://diegotc.github.io/bone101/Support/GSOC/BLINKINGTUTORIALS/Blink.html
May 21 13:57:29 <DiegoTc> do you like it, feel confortable using it
May 21 13:57:32 <DiegoTc> thanks
May 21 14:00:06 <praveendath92> DiegoTc: It's good. I didn't understand the drastic row shift animation for card 5
May 21 14:00:51 <DiegoTc> ahh i haven't add the css, but the idea is that you connect the BBB, plug the breadboard to it
May 21 14:00:56 <praveendath92> Also, it would be nice if you can add support for swipe events. Check onepage.js on mobile device once.
May 21 14:00:59 <DiegoTc> and you can run the code live
May 21 14:02:34 <praveendath92> No. What I meant was all the cards from 1-4 are animating from left-right or the other way while this one is different because of the functionality it offers?
May 21 14:04:07 <DiegoTc> ahhh, no, I was playing around with it yesteday and like the effect
May 21 14:04:29 <DiegoTc> there's no special reason for having it
May 21 14:05:27 <praveendath92> Oh.
May 21 14:05:29 <Abhishek_> DiegoTc: http://s9.postimg.org/te83t1q2n/scrshot2.png
May 21 14:06:18 <DiegoTc> praveendath92, Abhishek_ I will have all cards in the same row :)
May 21 14:06:25 <DiegoTc> thanks Abhishek_
May 21 14:07:19 <vmayoral|pc> DiegoTc: i support jkridner's comment. Make it cleaner. Use big fonts if possible and responsive
May 21 14:09:13 <Abhishek_> DiegoTc: possible issue in bone layout: http://s17.postimg.org/g878sbn27/scrshot3.png
May 21 14:09:58 <praveendath92> jkridner: Hello.
May 21 14:12:00 * jkridner likes the font size at https://learn.adafruit.com/connecting-a-push-button-to-beaglebone-black/you-will-need
May 21 14:12:14 <jkridner> same with https://learn.adafruit.com/connecting-a-push-button-to-beaglebone-black/writing-a-program
May 21 14:15:15 * praveendath92 is having mild fever. I will try to stick around for the meeting. In case I couldn't I want to inform jkridner about this. 
May 21 14:16:48 <vvu> praveendath92: any luck with the module ?
May 21 14:17:13 <praveendath92> vvu: Actually I figured that the probing and all are working fine.
May 21 14:17:21 <vvu> what was the issue ?
May 21 14:17:54 <praveendath92> It's just that when I ismod the execution stops or whatever after usb_register.
May 21 14:18:08 <praveendath92> However, the driver is being successfully registered.
May 21 14:18:37 <vvu> so not usb_register does not hang anymore ?
May 21 14:18:58 <praveendath92> And also, the prink's of those lines after usb_register are logged when device is first connected.
May 21 14:19:16 <vvu> did u push the code to github ?
May 21 14:19:20 <praveendath92> No. It doesn't. It's just those lines printk are logged late.
May 21 14:19:32 <vvu> oh ok
May 21 14:19:37 <vvu> i think u need to put KERN_ALERT
May 21 14:19:42 <praveendath92> The code looks messy. I will clean up and push it.
May 21 14:20:08 <praveendath92> Was sleeping the whole day due to fever :/
May 21 14:20:20 <vvu> and a \n at the end to get it pushed in dmesg
May 21 14:21:31 <praveendath92> Will do. Was checking with tail -f /var/log/kern.log so didn't have an issue with that.
May 21 14:21:44 <vvu> ok, cya at the meeting if u can make it
May 21 14:22:28 <praveendath92> I will mostly be there.
May 21 14:28:44 <praveendath92> vvu: It worked.
May 21 14:29:19 <praveendath92> The driver registered message also displayed.
May 21 15:09:05 <bradfa> meeting is in 51 minutes, right?
May 21 15:09:14 <Abhishek_> yep
May 21 15:09:25 <bradfa> k, thanks Abhishek_
May 21 15:28:39 <Abhishek_> ds2, mranostay, panto: if I do a XOUT from PRU1, do I need to do a corresponding XIN into PRU0 as well?
May 21 15:34:04 <vvu> praveendath92: so \n u need to get it out if the ring buffer i think
May 21 15:36:11 <praveendath92> Makes sense. It was probably been pushed to the next printk from the driver when \n is not used.
May 21 15:36:47 <vvu> yep but also printk got dold
May 21 15:36:49 <vvu> old
May 21 15:37:02 <vvu> use err smth, i think they have it for the usb core
May 21 15:37:14 <vvu> run checkpatch.pl on ur file
May 21 15:37:28 <vvu> to see if it complies witk lk coding style
May 21 15:44:53 <karki_> jkridner : we have updated our statuses in the google group. So whats todays discussion going to be about?
May 21 15:53:40 <ds2> Abhishek_: not immediately...there are 3 banks of registers
May 21 15:55:57 <jkridner> karki_: make sure everyone has started coding, providing status updates, getting guidance from mentors, have hardware and generally unblocked.
May 21 15:56:47 <Abhishek_> ds2: Section 5.2.4.2 of PRU Ref guide, there's this Device ID 14 which says: "Selects other PRU Core (Direct connection Mode)". What does it mean exactly?
May 21 15:56:58 <Abhishek_> hi cdsteinkuehler
May 21 15:57:30 <cdsteinkuehler> Hello.  The PRU direct connect copies data from one register bank to the other.  Both PRU cores must execute complimentary XIN XOUT instructions
May 21 15:57:54 <cdsteinkuehler> There is a 1024 clock timeout waiting for the other PRU core to execute the complimentary instruction
May 21 15:58:34 <Abhishek_> Okay, so that means I can XIN before I XOUT from the other PRU, and it will block till data is available?
May 21 16:00:05 <cdsteinkuehler> Yes, up to 1024 clocks
May 21 16:00:31 <jkridner> shall we do a quick roll call?
May 21 16:00:37 <ds2> Abhishek_: yes
May 21 16:00:42 <cdsteinkuehler> present
May 21 16:00:52 * Abhishek_ is present
May 21 16:00:58 <vmayoral|pc> same here
May 21 16:01:27 <alexanderhiam> ahoy
May 21 16:01:41 <voltvision2> here (on the road @ a temporary office)
May 21 16:01:51 <bradfa> aloha
May 21 16:02:29 <jkridner> Per status reports, I see something from karki_, praveendath92, DiegoTc, disdi (not present?), Abhishek_, vmayoral|pc
May 21 16:02:31 <cdsteinkuehler> See details in table 4-20 of SPRUH73C
May 21 16:02:41 <jkridner> anyone seen disdi and joel-brb?
May 21 16:03:10 <cdsteinkuehler> I will need to leave early again this week
May 21 16:03:33 <ds2> besides the email complaints? :D
May 21 16:03:46 <disdi_> hi all
May 21 16:03:48 <jkridner> about the missing hardware?
May 21 16:04:07 <jkridner> hi disdi_... board still seems to be at TI Bangalore, but hopefully will arrive soon.
May 21 16:04:19 * jkridner doesn't know how long it takes to ship something in Bangalore...
May 21 16:04:36 <jkridner> apparently disdi_'s professor contact was required for shipping due to some policy.
May 21 16:04:43 <ds2> yep
May 21 16:04:57 <disdi_> jkrinder: yes....
May 21 16:04:58 * karki_ thinks it shouldn't be more than a day
May 21 16:04:58 <DiegoTc> jkridner: you told me last week you were going to see the hardware related to bone101
May 21 16:05:03 <jkridner> rseethamraju: you around?
May 21 16:05:10 <karki_> disdi_ : where are you located?
May 21 16:05:21 <disdi_> karki_: new delhi
May 21 16:05:23 <DiegoTc> I don't need it right now, just a remainder
May 21 16:05:25 <jkridner> DiegoTc: yes. should get there in 3-4 weeks.
May 21 16:05:32 <DiegoTc> ahh ok
May 21 16:05:53 <karki_> disdi_ : Then why get it shipped in B'lore?
May 21 16:06:36 <rseethamraju> ya sorry
May 21 16:06:40 <jkridner> karki_: The TI office is in Bangalore and they are acting as a intermediary.
May 21 16:06:40 <disdi_> jkrinder: I am sorry nothing was in my hands ... it had to go first to TI Banglore office
May 21 16:06:53 <rseethamraju> I dodn’t see the status update thing so I was just putting that up
May 21 16:06:55 <karki_> jkridner : as in my status report, I will be a bit slow this week :(
May 21 16:07:04 <jkridner> rseethamraju: k, thanks. was just about to ask.
May 21 16:07:07 * karki_ hates exams
May 21 16:07:20 * vvu subscribes
May 21 16:07:28 <jkridner> hi vvu
May 21 16:07:28 <disdi_> jkridner : My video can only be made once I have the Hardware... I hope you understand
May 21 16:07:35 <vvu> hello jkridner
May 21 16:08:28 <disdi_>  jkridner : I have been keeping myself busy in all other tasks that do not require the HW as of now
May 21 16:08:31 <jkridner> disdi_: I'd think that you could describe what you are going to do without hardware, but joel-brb should conclude if that is the case or not.
May 21 16:08:50 <jkridner> disdi_: I'd think you'd not have much else to do *but* make the video, since coding can't be tested. :(
May 21 16:08:54 <rseethamraju> done
May 21 16:09:24 <jkridner> disdi_: I'd also think that info like code repositories can be setup.
May 21 16:09:55 <disdi_>  jkridner : there are issues like making the present driver stack SDHCI compliant
May 21 16:10:06 <jkridner> it seems most all projects have code repos now, which is excellent.
May 21 16:10:16 <joel_> jkridner, hardware would be ideal ofcourse
May 21 16:10:20 <joel_> but..
May 21 16:10:21 <vmayoral|pc> It seems many people (specially in India) is having conflicts with exams and GSOC. Google should probably be informed about this guys.
May 21 16:10:26 <jkridner> only 2 have their own wiki pages as of now.
May 21 16:10:37 <joel_> jkridner, I agree disdi_ should set one up
May 21 16:10:44 <jkridner> vmayoral|pc: k, I will pass that on to Carol.
May 21 16:11:04 <disdi_>  jkridner : I have setup a code repo on git hub as of now
May 21 16:11:06 <jkridner> vmayoral|pc, karki_: are they all complete this week?
May 21 16:11:14 <DiegoTc> jkridner: is necessary to setup a wiki?
May 21 16:11:19 <jkridner> disdi_: great. would be good if you can put a link on the wiki page.
May 21 16:11:33 <joel_> jkridner, I doubt any coding can be done immediately but disdi_ can say more... I think he is still finding his way around and such. Probably some time will be spent studying the problems
May 21 16:11:42 <disdi_>  jkridner : Sure I have mentioned it in my weekly report even
May 21 16:11:48 <karki_> jkridner : just a couple more days :-)
May 21 16:12:01 <vmayoral|pc> jkridner: I'm doing a PhD so fortunately no exams (for a lifetime I hope) but i share the pain of the rest.
May 21 16:12:02 <jkridner> DiegoTc: I'd like to have a page with notes and background that you can hack easily, but if the README.md is up-to-date, maybe it just needs to point there.
May 21 16:12:06 <rseethamraju> yes! My exams gto postponed and now they start on June 4th
May 21 16:12:22 <jkridner> rseethamraju: yuck!!
May 21 16:12:25 <karki_> 3 more days!
May 21 16:12:33 <praveendath92> June 4th!?
May 21 16:12:40 <rseethamraju> ya elections
May 21 16:12:40 * Abhishek_ is grateful that his exams ended within one week of acceptance
May 21 16:12:41 <vvu> i had the same last year, exams until 5-6th
May 21 16:12:46 <rseethamraju> and telengana formation
May 21 16:12:47 <karki_> I agree with jkridner rseethamraju !
May 21 16:13:01 <karki_> blast the congress
May 21 16:13:03 <_av500_> hello
May 21 16:13:07 <vvu> hi av500
May 21 16:13:12 <DiegoTc> I'm assuming Google consider that all universities work with the same time period
May 21 16:13:33 <_av500_> they do not
May 21 16:13:42 <vvu> they work just on US basis
May 21 16:13:43 <rseethamraju> karki_: 2 weeks! and I haven’t begun studying yet. exams are so boring. How are you managing?
May 21 16:13:49 <praveendath92> Hello av500.
May 21 16:13:51 <_av500_> gsoc dates are discussed at each summit
May 21 16:13:53 <karki_> DiegoTc : not actually. GSoC period keeps changing every year!
May 21 16:13:53 <vmayoral|pc> vvu: yeap that was my guess as well
May 21 16:13:56 <vvu> in RO for example they start exams @10th june
May 21 16:14:08 <_av500_> its not easy to find a date that fits all
May 21 16:14:14 <_av500_> so, live with it :)
May 21 16:14:27 <jkridner> vvu: I think most students are not in US.
May 21 16:14:38 <karki_> rseethamraju : I have a very high CGPA, so I can let this semester pass by, + I have a job offer from CISCO
May 21 16:14:40 <Abhishek_> stats say 401/1307 from India!
May 21 16:14:43 <vvu> yeah but the schedule they have fits US the most
May 21 16:14:48 <ds2> jkridner: maybe it is just us
May 21 16:14:54 * karki_ does not really care about exams
May 21 16:14:55 <rseethamraju> this yr is a one off. Its beacuse of the elections
May 21 16:15:00 <jkridner> I think _av500_ is right... it is just a pain we must live with and try to still keep to GSoC schedule as much as possible. Life will always get in the way some.
May 21 16:15:30 <praveendath92> Usually GSoC dates go fairly well with most University timetable I think.
May 21 16:15:34 <_av500_> lets move on
May 21 16:15:37 <_av500_> :)
May 21 16:15:37 <jkridner> karki_: won't hurt to still try to make a good grade.
May 21 16:15:38 <praveendath92> yep.
May 21 16:15:39 <rseethamraju> karki_: Wohoo! Good for you.
May 21 16:15:40 <jkridner> ok _av500_
May 21 16:15:50 <jkridner> thanks for the status reports....
May 21 16:16:06 <jkridner> I think disdi_ is the only one fully blocked on hardware, no?
May 21 16:16:13 <ds2> on a different topic - is the BeaglePilot project going to move to proper kernel drivers in the GSoC timeframe?
May 21 16:16:40 <disdi_> jkrinder: i am
May 21 16:17:08 <karki_> jkridner : I'll have to settle with a decent grade this semester :(  It was preparing for GSoC vs reading tasteless stuff, I had to make a choice!
May 21 16:17:14 <rseethamraju> I have a bbb
May 21 16:17:39 <vmayoral|pc> ds2: it's not in the initial plan
May 21 16:17:40 <praveendath92> I have one too.
May 21 16:17:55 * Abhishek_ got one very early
May 21 16:18:02 <vmayoral|pc> ds2: still, there's alreday somebody working on it and hopefully if i have the time i can push forward as well
May 21 16:18:29 <karki_> jkridner : when is TRE coming out?
May 21 16:18:38 <jkridner> karki_: beta just started....
May 21 16:18:42 * vvu big question over here
May 21 16:18:42 <ds2> vmayoral|pc: gotcha... then for practical purposes, consider me off the mentoring team for the project
May 21 16:18:50 <jkridner> karki_: if you have a specific task, I might be able to get you on the beta list.
May 21 16:19:07 <karki_> by task, you mean project, right?
May 21 16:19:14 <jkridner> vmayoral|pc: we should have a schedule, no?
May 21 16:19:19 <jkridner> where is tridge?
May 21 16:19:42 <jkridner> I know this is a bad time slot for tridge, but we need someone accountable for the direction of the project.
May 21 16:19:43 <vmayoral|pc> jkridner: it's 2am now in Canberra, i think he's asleep
May 21 16:20:14 <jkridner> it might be fine if ds2 drops off, but I think his concerns should be addressed. the inputs are valuable.
May 21 16:20:23 <vmayoral|pc> jkridner: i can talk to him and find a way so that he provides feedback
May 21 16:20:59 <vmayoral|pc> ds2: i had a quick talk to koen and he suggested to go for iio kernel drivers. Would you agree?
May 21 16:21:32 <vmayoral|pc> ds2: my initial idea was to finish userspace drivers asap and then code the kernel ones as far as possible
May 21 16:21:37 <jkridner> vmayoral|pc: the idea of having real Linux drivers can be a bit philosophical, but real, practical arguments can be made regarding the approach and the intended audience, including consideration for long-term support and a growing user base.
May 21 16:22:05 <ds2> vmayoral|pc: that is tolerable but keep in mind, IIO is not rock solid yet so if you want to leverage older code, you will have a very hard time
May 21 16:22:27 <jkridner> karki_: yeah, project, assumedly associated with your gsoc project or something else with high visibility like that.
May 21 16:22:41 <ds2> vmayoral|pc: if you care enough about timing to look at the PRU... userspace is absolutely positively downright a waste of time
May 21 16:23:09 <jkridner> ds2: but we want to advance the state of Linux, no? given it is a bit difficult, but I'd think this is an excellent use-case for IIO.
May 21 16:23:29 <vmayoral|pc> ds2: tridge suggested to start with userspace drivers to have a Minimum Working Prototype of BeaglePilot
May 21 16:23:37 <vmayoral|pc> and then construct from there
May 21 16:23:57 <vmayoral|pc> anyhow, the userspace drivers are pretty much coded already (have been doing my homework during the bonding period)
May 21 16:23:58 * karki_ thinks about an extension for PRUSpeak........ maybe AtmelSpeak ? jkridner ;)
May 21 16:24:10 <ds2> jkridner: trying to be practical on it... there is a trade off with what we can do in this 8 weeks of GSoC... a lot of those sensors out there have non IIO drivers already... getting that going is quick...getting an IIO driver can be a GSoC project in itself
May 21 16:24:39 <jkridner> karki_: BotSpeak already exists for Arduino Uno.
May 21 16:24:46 * praveendath92 just added all Beagle GSoC repos to watchlist.
May 21 16:24:51 <jkridner> karki_: I'd think you'd be using it as a reference.
May 21 16:25:06 <ds2> vmayoral|pc: that's why I think it is best if userspace is the path, that i drop out. I cannot agree with doing any of that in userspace beyond possibily validating hw
May 21 16:25:56 <vmayoral|pc> ds2: userspace is drivers are pretty much completed (although need to test them though) https://github.com/BeaglePilot/ardupilot/commits/master
May 21 16:26:16 <ds2> vmayoral|pc: unless the 9 axis part really truely uses a different gyro then the other Invensense parts, there are drivers out in the while already.
May 21 16:26:39 <jkridner> ds2, vmayoral|pc: shall we wrap up this meeting so you can have that one?
May 21 16:26:56 <ds2> jkridner: sure, I'll follow up later.
May 21 16:27:01 <ds2> sorry about the attempted hijacking.
May 21 16:27:10 <karki_> alexanderhiam : How do you see PyBBIO on the TRE? What other cool features do you think can be added? I was thinking of a PyBBIO support for the atmel chip........
May 21 16:27:17 <jkridner> _av500_: any other business from you or topics you feel I'm not bringing to the fore?
May 21 16:28:07 <ds2> jkridner: should we go through the expectations of these mtgs?
May 21 16:28:13 <ds2> (in general)
May 21 16:28:15 <jkridner> karki_: I'm sure there must be Python support for Firmata. I'd think you couldn't have direct Python support.
May 21 16:28:23 <jkridner> ds2: yes, please.
May 21 16:28:35 <alexanderhiam> karki_: atm I really don't get the tre, seems like it's basically replacing the PRU with an Arduino... what am I missing?
May 21 16:28:44 <jkridner> ds2: right now, I'm primarily looking to hear the way students and mentors are interacting...
May 21 16:28:47 <rseethamraju> karki_: that would be really cool. alexanderhiam said his plan was to add support for more than just the bbb
May 21 16:29:19 <jkridner> wanting to know that status reports are getting done so that everyone can review on-going work and confirming mentors are actually reviewing and guiding the status.
May 21 16:29:35 <cdsteinkuehler> I also do not "get" the tre...
May 21 16:29:36 <jkridner> ds2: also looking to see what blocking issues exist that I need to correct.
May 21 16:29:39 <karki_> jkridner : yes, there is PyFirmata!
May 21 16:29:58 <ds2> jkridner: okay
May 21 16:30:21 <jkridner> cdsteinkuehler: I think you have to look at the Tre as a bit of a teaching platform in some ways more so than a prototyping tool.
May 21 16:30:25 <Abhishek_> ds2, cdsteinkuehler, mranostay: I'm sending you a mail shortly, there's something curious going on with the Xfer instructions
May 21 16:30:32 <_av500_> jkridner: no, seems fine for me
May 21 16:30:34 <ds2> just wanted to add - any problems that may effect progress should be brought up as early as possible. please do not spend a week or two spinning your wheels
May 21 16:30:43 <ds2> GSoC is short. there is no time to spin wheels
May 21 16:30:44 <_av500_> as usual, if you are strudent or mentor and something is amiss, shout out!
May 21 16:30:51 <_av500_> student*
May 21 16:30:57 <_av500_> and what ds2 said :)
May 21 16:31:09 <_av500_> mentors, hound your students
May 21 16:31:13 <jkridner> cdsteinkuehler: it is useful for prototyping, but under constraints most useful to people very new to electronics.
May 21 16:31:22 <_av500_> lightly
May 21 16:31:30 <jkridner> ds2: +++++++++++++
May 21 16:31:49 <vmayoral|pc> regarding issues, i found some conflicts between capemgr and the PREEMPT_RT-patched kernel
May 21 16:31:56 <vmayoral|pc> talked with panto about it
May 21 16:32:18 <jkridner> even seemingly very small issues can destroy your whole project if not worked to resolution quickly.
May 21 16:32:55 <vmayoral|pc> he suggested thtat the effort of fixing (every possible driver in the ) PREEMPT_RT might be to much, so i'd take it into account and focus on the objectives i set
May 21 16:33:40 <cdsteinkuehler> I have to sign off...  I'll review the issue with xfer instructions in a couple hours when I return
May 21 16:34:03 <ds2> vmayoral|pc: PREEMPT_RT can expose a lot of bugs :(
May 21 16:34:11 * jkridner should really be taking mentor roll call more so than student roll call to make sure mentors are interacting/pushing/helping students as much as possible to help avoid them failing.
May 21 16:34:33 <vmayoral|pc> ds2: yes. And it also seems that Xenomai is no better, right?
May 21 16:34:37 <jkridner> DiegoTc: and just because I'm running the meetings doesn't mean you can't say "hey, I need more of your time/feedback".
May 21 16:35:03 <vmayoral|pc> ds2: still i want to compile the Xenomai-patched one and give it a try
May 21 16:35:13 <jkridner> Xenomai is not Linux?
May 21 16:35:22 <vmayoral|pc> it's Linux
May 21 16:35:25 <ds2> vmayoral|pc: I think Xenomai basically virtualizes Linux so the issues will be different there
May 21 16:35:35 <jkridner> it's on-top of Linux, but really isn't Linux.
May 21 16:36:09 <vmayoral|pc> my bad then
May 21 16:36:18 <vmayoral|pc> ds2: let's hope you are right
May 21 16:36:25 <DiegoTc> jkridner hahaha. I was going to tell you that. I'm guessing voltivision2 hasn't seen the new update
May 21 16:36:42 <karki_> jkridner : me, panto and ds2 were discussing about the interface for PRU speak yesterday. Like, how much would we be pre-processing the bot speak before the PRU reads etc. I would like to discuss about it with you too before I make things concrete. I hope you will be free tomorrow. ( I have to study for my exam now :(  )
May 21 16:36:49 <DiegoTc> neither David
May 21 16:36:56 * jkridner needs to run to another meeting
May 21 16:37:29 <vmayoral|pc> jkridner: anyhow, all good by my side. Good progress and the group for BeaglePilot is great
May 21 16:37:50 <vmayoral|pc> jkridner: i'm really happy and you can hope results (videos) soo
May 21 16:37:51 <vmayoral|pc> n
May 21 16:37:52 <jkridner> is everyone happy that we can follow-on with chatting and I'll be back to attention here in about 30 minutes.
May 21 16:37:52 <ds2> karki_: are you familiar at all with yacc/bison/lex/etc?
May 21 16:38:01 <jkridner> DiegoTc: you available in about 30 minutes?
May 21 16:38:05 <DiegoTc> no problem for me
May 21 16:38:06 <DiegoTc> yes
May 21 16:38:13 <vmayoral|pc> ds2: i'll be taking in good consideration your comments :)
May 21 16:38:51 <_av500_> praveendath92: jkridner: i am in china next week
May 21 16:38:55 <_av500_> IRC should be possible
May 21 16:39:00 <voltvision2> DiegoTc, I checked for updates this morning, but didnt see any.  I will drive back to my office now (30mins) and I will be available in 30mins also.
May 21 16:39:01 <jkridner> DiegoTc: we'll try to have a good chat at that time.
May 21 16:39:11 <ds2> vmayoral|pc: the reason I am set against userland drivers, esp. for your sensors, is the overall timing is downright horrible. even in the kernel, the timing is somewhat iffy
May 21 16:39:12 <karki_> ds2_ : yes I have worked with yacc/lex/antlr
May 21 16:39:18 <praveendath92> That works.
May 21 16:39:22 <jkridner> also, anyone want to raise any objections to having vmayoral|pc come out to the Mentor Summit?
May 21 16:39:29 <ds2> karki_: so tokenizing it shouldn't be too painful?
May 21 16:39:47 <praveendath92> av500:  jkridner: I will be in US 1st week of next month.
May 21 16:39:53 <ds2> _av500_: china?! OMG... you getting shipped to support the factory? ;)
May 21 16:39:59 <jkridner> praveendath92: where?
May 21 16:40:06 <karki_> ds2 : not painful at all! just want to make sure jkridner is with us.
May 21 16:40:07 <praveendath92> San Francisco.
May 21 16:40:13 <praveendath92> I read about the unconference.
May 21 16:40:16 <_av500_> ds2: nah
May 21 16:40:18 <praveendath92> San Jose?
May 21 16:40:21 <_av500_> shopping :)
May 21 16:40:36 <Abhishek_> _av500_ HQB, Shenzhen?
May 21 16:40:46 <karki_> in case he has some inputs that will prove useful. after all he has been with the botspeak project the longest!
May 21 16:41:51 <vmayoral|pc> ds2: fine then. There's somebody pushing the kernel drivers within BeaglePilot. I'd share your concerns and we will try to speed up the kernel ones. As you say there're already some (e.g. MPUxxxx) so shouldn't be that much work.
May 21 16:42:52 <ds2> vmayoral|pc: yes... the MPU drivers are really a Gyro driver + accel driver
May 21 16:43:07 <ds2> look for the ones under the older name made by these folks
May 21 16:43:31 <ds2> itg3200, mpu3500, mpu6500, mpu6000
May 21 16:43:53 <vmayoral|pc> great, thanks
May 21 16:43:57 <ds2> vmayoral|pc: ping me if you can't find them
May 21 16:44:05 <vmayoral|pc> actually, i was just thinking
May 21 16:44:13 <ds2> the accel and mag should be easy to tack on
May 21 16:44:34 <Abhishek_> vmayoral|pc: Look for MPU6050 drivers as well
May 21 16:44:35 <vmayoral|pc> that to have ardupilot code running in the BB(B) we just need the IMU (accel and mag should be enough) and the barometer
May 21 16:44:54 <jkridner> we seem to have gone all over the place, so I will close up the official meeting and close with individuals.
May 21 16:44:54 <ds2> which barometer are you using?
May 21 16:44:57 <ds2> bmp85's?
May 21 16:45:08 * jkridner sounds gavel
May 21 16:45:17 <vmayoral|pc> ds2: right now there's an SPI MS5611 in the PXF (PixFire Cape) and in Erle
May 21 16:45:31 <vmayoral|pc> you guys now about that one?
May 21 16:45:57 <Abhishek_> vmayoral|pc : are you aware of the FreeIMU arduino library?
May 21 16:45:57 <vmayoral|pc> (we also had in the I2C bus there's a design problem in both boards)
May 21 16:46:12 * ds2 looks for who makes the MS5611
May 21 16:46:29 <ds2> vmayoral|pc: you guys just need 3 axis + barometer, right?
May 21 16:46:30 <vmayoral|pc> Abhishek_: yes, thanks unfortunately FreeIMU doesn't provide kernel drivers
May 21 16:46:50 <vmayoral|pc> yes, that's the minimum to run ArduPilot
May 21 16:47:07 <vmayoral|pc> i could hack the barometer but we have the sensors so i'll be silly
May 21 16:47:17 <ds2> I have a tiny board that has all those
May 21 16:47:29 <vmayoral|pc> that's really good, you can follow up then
May 21 16:47:32 <ds2> 9 axis + barometer all on a tiny i2c interfaced board...but i2c might have too much jitter
May 21 16:47:44 <ds2> not finding much on the MS5611
May 21 16:47:55 <ds2> n/m
May 21 16:47:58 <ds2> found datasheet
May 21 16:47:59 <Abhishek_> ds2: L3G4200+ADXL345+BMP085 one?
May 21 16:48:09 <vmayoral|pc> ds2: i feared so, mmm yeap I2C is troublesome
May 21 16:48:16 <ds2> Abhishek_: no... MPU6500, BMP180
May 21 16:48:22 <ds2> HMC5883
May 21 16:48:29 <Abhishek_> k
May 21 16:48:54 <ds2> bmp085's are huge
May 21 16:49:16 <vmayoral|pc> ds2: anyhow, there're a couple of boards set up
May 21 16:49:19 <vmayoral|pc> with all the hardware
May 21 16:49:42 <vmayoral|pc> i could talk to tridge and you could have access to them
May 21 16:49:54 <ds2> is the MS5611 a BMP085 clone?
May 21 16:50:03 <praveendath92> av500: jkridner: I'm signing off.
May 21 16:50:15 <rseethamraju> I’m using the MPU6050 fro my project
May 21 16:50:15 <ds2> if so, there is a kernel driver for the BMP085 already
May 21 16:51:00 <ds2> n/m...the address map don't match up
May 21 16:51:10 <ds2> but it shouldn't be too hard to modify the bmp085 driver for this
May 21 16:51:14 * jkridner waves at praveendath92
May 21 16:51:29 <Abhishek_> I've used the IMUs this march, but it was with an arduino. ADXL345, L3G4200, and MPU6050
May 21 16:51:31 <vmayoral|pc> ds2: yeap, can i get a link to the BMP085 kernel driver?
May 21 16:51:31 <ds2> rseethamraju: what are you using that for?
May 21 16:51:43 <praveendath92> Good night/day :)
May 21 16:51:50 <vmayoral|pc> praveendath92: good night
May 21 16:51:56 <Abhishek_> G'night praveendath92
May 21 16:52:09 <ds2> vmayoral|pc: in drivers/misc/bmp085.c
May 21 16:52:44 <vmayoral|pc> so it's already merged into the kernel, great. Thanks!
May 21 16:53:06 <rseethamraju> ds2: I’m using it to write a library for PyBBIO
May 21 16:53:33 <ds2> vmayoral|pc: do you have MPU_INT wired to the BBB?
May 21 16:53:40 <ds2> if so, the driver is so easy
May 21 16:53:46 <disdi_> jkridner: need to leave..... thank for your time
May 21 16:53:50 * vmayoral|pc checking the schematics
May 21 16:53:57 * jkridner waves at disdi_
May 21 16:54:24 * jkridner reminds everyone that the gavel means you aren't required to stick around any longer, but conversations may continue.
May 21 16:54:34 <ds2> IIO takes a little more work but you can bang out a working input driver in about 2 hours or less if the interrupt line is wired up
May 21 16:54:46 <vmayoral|pc> yes, it's wired up
May 21 16:54:51 <ds2> nice
May 21 16:55:53 <ds2> the driver just has to - enable the device, configure range, enable interrupt...then on each interrupt, burst read the registers
May 21 16:56:05 <ds2> shove the data out the pipe. and you are done
May 21 16:56:11 <ds2> this gives you decent timings
May 21 16:56:41 <vmayoral|pc> ds2: can you guess what are those good timings?
May 21 16:56:52 <ds2> vmayoral|pc: what rate do you need them at?
May 21 16:56:55 <vmayoral|pc> ds2: might be handy https://github.com/diydrones/PXF/tree/master/PXF_CAPE_PCB
May 21 16:57:20 <ds2> 100Hz? 1000Hz?
May 21 16:57:35 <ds2> 100Hz should be pretty much right on the spot
May 21 16:57:49 <ds2> as in if you did a timing analysis, you get a single good spike
May 21 16:57:58 <vmayoral|pc> nah the loop runs at 400 Hz
May 21 16:58:04 <vmayoral|pc> we are sampling the gyros at 800Hz
May 21 16:58:40 <vmayoral|pc> @tridge pointed out some timing limits http://uav.tridgell.net/LCA2014/AP_Linux.pdf
May 21 16:59:44 <ds2> 800Hz suggests you should use the fifos
May 21 17:00:05 <vmayoral|pc> the userspace drivers are doing so
May 21 17:00:49 <vmayoral|pc> although i have to say that the tests that i've done didn't reach 800Hz
May 21 17:01:13 <vmayoral|pc> (with the userspace drivers)
May 21 17:01:32 <vmayoral|pc> ds2: it'll be really nice though to compare the performance
May 21 17:01:44 <ds2> the fifos aren't deep enough for reliable userspace drivers
May 21 17:01:51 <ds2> is this why you are looking at PREEMPT_RT?
May 21 17:02:44 <vmayoral|pc> I believe so.
May 21 17:03:04 <vmayoral|pc> @tridge made some tests and realized that using PREEMPT_RT there where many less misses
May 21 17:03:13 <ds2> sigh
May 21 17:04:07 <vmayoral|pc> it's highlighted in the "Scheduling issues" slide (http://uav.tridgell.net/LCA2014/AP_Linux.pdf)
May 21 17:05:50 <ds2> is there a diagram of the overall current design?
May 21 17:06:05 <ds2> like pointing out where the userspace parts are
May 21 17:06:37 <vmayoral|pc> there's no diagram that i know but it's a good idea
May 21 17:25:10 <VoltVisionSteve> DiegoTc, jkridner : I'm back @ my office now.
May 21 17:25:26 <jkridner> k, I'm off my call too... just listening to the videos now.
May 21 17:27:36 <jkridner> DiegoTc: you still around?
May 21 17:28:03 <jkridner> VoltVisionSteve: did you have thoughts about the way the cards look?
May 21 17:28:41 <jkridner> personally, I feel they are somehow unprofessional looking, but it is bad that I'm only really suitable to mentor on the back-end side of HTML/JS things. :(
May 21 17:29:30 <jkridner> VoltVisionSteve: do you think we can bring in an HTML/CSS designer who can help DiegoTc learn to knock-out the appearance and get moving a bit quicker?
May 21 17:30:18 <VoltVisionSteve> Can I get to the most recent cards from here https://github.com/DiegoTc/bone101?
May 21 17:30:33 <VoltVisionSteve> ...or is it a private link?
May 21 17:30:33 <jkridner> VoltVisionSteve: also, I think the /#/Card1, /#/Card2 links are needed to make sure you can link to a particular card.  Also, I don't think view-all should be a unique URL, but also like /#/ViewAll or similar.
May 21 17:31:18 <vmayoral|pc> ok everyone, I'm off for today.
May 21 17:31:27 <jkridner> VoltVisionSteve: it is public, but that only has the source. to view it, you need to use the URL http://diegotc.github.io/bone101
May 21 17:31:28 <DiegoTc> jkridner, VoltVisionSteve I'm here
May 21 17:31:40 <VoltVisionSteve> is this it? http://diegotc.github.io/bone101/Support/GSOC/template.html
May 21 17:31:44 * vmayoral|pc waves
May 21 17:31:48 <jkridner> VoltVisionSteve: I believe http://diegotc.github.io/bone101/Support/GSOC/template.html is the latest.
May 21 17:32:03 <jkridner> VoltVisionSteve: yes, I think so.
May 21 17:32:17 <DiegoTc> yes it's
May 21 17:33:32 <VoltVisionSteve> jkridner, I know lots of "backend guys", but not any "front end guys".  I have a meeting with a team of Guatemalan University people today.  I can ask to see if we could get some front-end support.
May 21 17:34:09 <VoltVisionSteve> I agree...there is something not right about the looks yet.
May 21 17:35:12 <VoltVisionSteve> ...maybe its the white/black?  would CSS shading help?
May 21 17:35:40 <jkridner> VoltVisionSteve: I had some people at Fluent offering to help, but haven't kept my cards organized.
May 21 17:36:08 <DiegoTc> VoltVisionSteve: you are referring to this one right now: http://diegotc.github.io/bone101/Support/GSOC/template.html
May 21 17:37:04 <VoltVisionSteve> DiegoTc, yes.  There is still something not professional looking about the White/Black part.
May 21 17:38:29 <VoltVisionFrench> ...sorry I might have missed something...I was booted out of IRC for a moment...
May 21 17:39:03 <VoltVisionFrench> DiegoTc, you did a good job with the BBB outline....but it still needs some "graphic designer magic".
May 21 17:39:18 <Abhishek_> VoltVisionFrench: no, yours was the last message.
May 21 17:39:52 <DiegoTc> jkridner I will appreciate if you could help me contact the web designer. I don't have a lot of experiencing with good design.
May 21 17:40:18 <DiegoTc> For me it looks good. But I know it's missing something
May 21 17:40:35 <DiegoTc> VoltVisionFrench: what about this: http://diegotc.github.io/bone101/Support/GSOC/BLINKINGTUTORIALS/Blink.html
May 21 17:41:21 <alexanderhiam> DiegoTc: remove the 'background-color: #FFFFFF;' from .container{} in the css, looks a lot better imo
May 21 17:42:25 <alexanderhiam> i.e. shows the body's #fcf7ed background
May 21 17:42:41 * jkridner just sent a note to David Tran with the hope he might be helpful.
May 21 17:43:07 <VoltVisionFrench> DiegoTc, that blue is kinda out of place isnt it?  What about the Bone Color scheme?
May 21 17:44:07 <DiegoTc> ahh yes the blue is out of it
May 21 17:44:19 <DiegoTc> I got sleep yesterday :(
May 21 17:44:43 <DiegoTc> I did use the parallex idea you gave me
May 21 17:45:30 <VoltVisionFrench> DiegoTc, also the navigation buttons on the top-right.  The colors are strange.  Also, it makes me think what will happen when there is 20cards instead of 5.
May 21 17:45:36 <jkridner> VoltVisionFrench, DiegoTc: example of David Tran's work: http://unsignedzero.github.io/PreZen/
May 21 17:45:39 <jkridner> meh. :(
May 21 17:46:09 <alexanderhiam> DiegoTc: would also be nice to use a .bonecard:hover{} on that main page to make it a bit more dynamic (show a shadow for the highlighted card or something)
May 21 17:47:06 <VoltVisionFrench> jkridner, Confused about that PreZen thing...trying to understand it ....
May 21 17:47:29 * alexanderhiam hopes you guys don't mind him butting in :P
May 21 17:48:08 <DiegoTc> alexanderhiam: you're being of great help !
May 21 17:48:35 <alexanderhiam> cool
May 21 17:49:02 <jkridner> alexanderhiam: yeah, your inputs are welcome.
May 21 17:49:16 <jkridner> VoltVisionFrench: background for that at https://github.com/unsignedzero/PreZen
May 21 17:49:35 <alexanderhiam> I've done enough HTML/CSS to be good at quickly getting stuff to decent looking so I can move on to something more fun ;)
May 21 17:49:59 * jkridner notes that PreZen thing has a problem with going off of my screen.
May 21 17:54:50 <DiegoTc> alexanderhiam: I just remove the  'background-color: #FFFFFF;' from .container{}
May 21 17:54:57 <DiegoTc> looks a little better
May 21 17:55:13 <alexanderhiam> yeah I think so
May 21 17:55:59 <DiegoTc> VoltVisionFrench & jkridner I will continue working on the other parts of the project
May 21 17:56:31 <VoltVisionFrench> DiegoTc, I just refreshed and dont see anything changed.
May 21 17:58:34 <DiegoTc> want to have the platform working, and fixed the design during the way
May 21 17:58:48 <DiegoTc> that finishing the platform at the end
May 21 17:59:10 <VoltVisionFrench> alexanderhiam, your input is very appreciated.  Thx!  How do you think the "card outlines" can be "better"?  CSS shading?
May 21 18:00:00 <alexanderhiam> a thinner outline maybe? I tend to not like the look of outlines thicker than a pixel or two
May 21 18:11:22 <VoltVisionSteve> I am playing around with options....glad I took the online class for the Chrome Developer Tools!...but I am no graphic designer...
May 21 18:13:24 <VoltVisionSteve> DiegoTc, do you have any Graphic Designer friends?
May 21 18:15:48 <DiegoTc> not really. I will talk with the ones I know
May 21 18:15:54 <DiegoTc> to see if they can help
May 21 18:20:48 <VoltVisionSteve> DiegoTc, in the top-right corner, what happens if a tutorial has 20+ cards?  This page...
May 21 18:20:49 <VoltVisionSteve> http://diegotc.github.io/bone101/Support/GSOC/BLINKINGTUTORIALS/Blink.html
May 21 18:21:29 * Abhishek_ is now a little more clear with broadside PRU->PRU transfers
May 21 18:22:56 <DiegoTc> VoltVisionSteve: That's a good question. It's going to be horrible. But what about if you want to know how many cards are in the tutorial. I don't how to define this :(
May 21 18:24:56 <alexanderhiam> DiegoTc: what about a (bootstrap?) navbar at the top with Home, Previous, Next, and a dropdown with all the cards?
May 21 18:27:18 <alexanderhiam> if you went with bootstrap you'
May 21 18:27:34 <alexanderhiam> (oops)...d also get built-in mobil support
May 21 18:27:44 <alexanderhiam> mobile*
May 21 18:30:24 <alexanderhiam> (http://getbootstrap.com/examples/navbar-fixed-top/)
May 21 18:32:07 <VoltVisionSteve> alexanderhiam, nice!  I do like it.  DiegoTc, what do you think about that above bootstrap example?
May 21 18:32:17 <VoltVisionSteve> (for the navbar)
May 21 18:35:55 <DiegoTc> VoltVisionSteve: working on it
May 21 18:36:05 <DiegoTc> Going to have lunch
May 21 18:37:34 <DiegoTc> thanks alexanderhiam
May 21 18:37:51 <alexanderhiam> no prob
May 21 18:39:48 <VoltVisionSteve> DiegoTc, sounds good.  Have a great day!  I will be driving for 6 hours tomorrow, so maybe we could Skype?  I will check my email periodically through the day.
May 21 19:08:04 <VoltVisionSteve> jkridner, I'll be in Cleveland for the TI Tech Summit tomorrow...you goin?  I know youve been busy with Maker Faire!!
May 21 19:08:27 <jkridner> VoltVisionSteve: after all, no. :(
May 21 19:08:40 <jkridner> wife scheduled stuff for me.
May 21 19:09:08 <Abhishek_> jkridner: How was Maker Faire?
May 21 19:09:09 <VoltVisionSteve> ok, I'm meeting with the Guatemalan team now and I think I might have found some front-end help....I'll let you know...
May 21 19:09:17 <jkridner> Abhishek_: awesome!
May 21 19:09:43 <jkridner> don't know when I'll have a chance to create a blog post.
May 21 19:12:29 * Abhishek_ will wait for jkridner's post
May 21 19:13:14 <jkridner> might be waiting a while. :-)
May 21 19:14:06 <Abhishek_> np, but I hope it would be up before the next Maker Faire ;)
May 21 19:20:19 <Abhishek_> mranostay, ds2: Ideas? https://github.com/abhishek-kakkar/BeagleLogic/commit/b65da0578cc44f4a929e5c1d6a1399f33e643fac
May 21 19:22:08 <mranostay> writing to the other pru's memory while it is reading will causes a stall
May 21 19:22:44 <Abhishek_> how does that happen here?
May 21 19:24:16 <Abhishek_> ah, gotcha
May 21 19:26:39 * Abhishek_ needs to go to the bed now
May 21 19:26:54 <mranostay> you need to halt the other halt or be sure it locks
May 21 19:27:33 <Abhishek_> mranostay: The issue was that I had loaded the firmware into the mem and then writing to the PRU RAM. It had to break
May 21 19:28:31 <Abhishek_> Now the mem is loaded before fw is loaded into the PRU, and it's fine
May 21 19:30:14 <Abhishek_> works correct
May 21 19:37:20 <Abhishek_> thanks for the heads up!
May 21 20:24:54 <Abhishek_> Good night. ds2, mranostay: shot a mail.
**** BEGIN LOGGING AT Thu May 22 02:26:05 2014
**** ENDING LOGGING AT Thu May 22 02:59:58 2014