**** BEGIN LOGGING AT Thu Mar 05 02:59:59 2015 Mar 05 03:00:04 hmm, so install a console image then use apt-get to build up from there? Mar 05 03:00:27 move the crap to an sd card Mar 05 03:00:34 besides, if you're going to a hackathon you can't really show up with something as ancient as debian stable right? ("why won't this compile? oh right, stone-age make and stone-age gcc") Mar 05 03:00:42 sensibly, with temperence Mar 05 03:00:42 surely you dont have 4gb of shit installed Mar 05 03:01:17 you should be using a custom handrolled yocto image with bleeding edge sources for maximum cred Mar 05 03:01:17 not me, I just installed ninja-ide and it choked Mar 05 03:01:39 where do I get one? Mar 05 03:01:44 hehe Mar 05 03:02:31 I'd personally go for (and in fact went for) a bone-debian-7.8-console image, upgrade it to sid, then install the stuff you want Mar 05 03:03:15 okay, so boot off the microsd, then use the emmc for /opt? Mar 05 03:03:41 and of course the big decision you need to make... kernel 3.8.13 versus 3.19.0 Mar 05 03:03:59 zmatt .. no question .. $. Mar 05 03:04:00 I just use eMMC Mar 05 03:04:01 4* Mar 05 03:04:44 you should be able to install a basic lxde environment Mar 05 03:04:47 I just want to write stuff in qt and have the bbb run it Mar 05 03:05:09 qt is a bit heavy, no ? Mar 05 03:05:42 although I guess most of it is gtk Mar 05 03:05:43 qt on my host laptop, cross compile then run in bbb - geany for touch ups Mar 05 03:06:09 ah at least you have the sense to cross-compile Mar 05 03:06:35 phew! Mar 05 03:06:41 I gave up my life as a rodeo clown for embedded mcu's Mar 05 03:07:06 strictly speaking its not an mcu Mar 05 03:07:20 (I sometimes still forget that... "why the fuck is this taking so long? oh wait, it's a cortex-a8..." ) Mar 05 03:08:17 bwarff: not sure how to unambiguously draw the line between a big mcu and a small soc though Mar 05 03:08:26 mmu Mar 05 03:08:32 hmmm Mar 05 03:08:37 yeah, no pre-emptive kernel, but I'm not good enough on RTOS to use that on the weekend Mar 05 03:08:56 techincally the onspecific difference is the mmu that turns multiple memory modules into flat addressable one Mar 05 03:09:05 twoten: baremetal programming is fun too Mar 05 03:09:10 versus the mcu just having the single chunk of memory Mar 05 03:09:28 I like assembly myself, but tools are hard to find Mar 05 03:10:00 the cortex m7 is a dual channel, 1ghz, mcu .. faster than many of the cpu's but still an mcu. Mar 05 03:10:51 twoten: no need for much asm per se, apart from a really tiny bit of init and some occasional inline asm Mar 05 03:11:28 twoten: although sometimes I do feel the temptation when I see the idiotic crap compilers sometimes produce :P Mar 05 03:11:30 I used to work with an HC11 and wrote tons of assembly code, not any more Mar 05 03:11:45 arm assembly is pleasant enough Mar 05 03:12:13 I did write a fairly big chunk of it for another project (a Forth kernel) Mar 05 03:12:36 I'll chop away any junk I can find on the emmc and hopefully X11 will come back up Mar 05 03:13:32 twoten: just ponder for a moment how _big_ 4 GB is, and how bizarre it is to be able to manage to fill it up when not storing large media files Mar 05 03:13:46 my emmc is only 2 gig Mar 05 03:13:54 2 GB, same thing (mine is too) Mar 05 03:14:01 the debian image nearly fills the 2gig emmc Mar 05 03:14:09 is about 1.7gb from memory Mar 05 03:14:30 yeah I know, I once wrote a predictive ai system in 20K of asm Mar 05 03:14:31 my old 2gig beagles run from sd cards now, i use the emmc for fast storage Mar 05 03:15:15 twoten: my baremetal code also runs from internal SRAM, I never bother initializing the DDR Mar 05 03:15:48 (most of it is on a dm814x though, which have really a lot of internal SRAM) Mar 05 03:16:08 btw, the bone-debian-7.8-console image is only 50 MB Mar 05 03:16:09 ;) Mar 05 03:16:22 I don't want to look like a weenie with an arduino or raspi Mar 05 03:17:07 the only thing potentially interesting about the rpi is that it apparently boots up in secure mode, so you can play with TrustZone and setup your own secure monitor Mar 05 03:17:26 so if I just plain boot I'm on the emmc and if I hold reset and boot I'm on the microsd - is that correct? Mar 05 03:17:34 no Mar 05 03:17:43 if you hold reset, you'll hold it in reset (surprise!) Mar 05 03:18:03 hold reset and power on Mar 05 03:18:10 if however you hold the sd-boot button at power-on, it'll boot to sd Mar 05 03:18:38 right - I'll get the console image 50meg sounds great Mar 05 03:19:30 specifically, by default the boot order is: eMMC, SD, uart, usb Mar 05 03:19:51 with button S2 held, the boot order is: spi, SD, usb, uart Mar 05 03:20:07 ah sweet Mar 05 03:20:33 so holding S2 without an SD card inserted is also used to boot it into usb mode (e.g. for BBBlfs) Mar 05 03:22:12 and if you hold S2 and also pull sysboot0 (lcd_data0) up to 3.3V it'll try to boot via ethernet (bootp + tftp) Mar 05 03:22:58 I'm using that a lot lately for some baremetal tests, saves me from having to flash anything Mar 05 03:23:31 how long have you worked with bbb? Mar 05 03:25:03 not long, but subarctic (am335x) is architectually closely related to (but much simpler than) centaurus (dm814x / am387x) where I have a few years of experience Mar 05 03:26:05 aha, where do I get the debian console image? Mar 05 03:26:39 https://rcn-ee.net/rootfs/bb.org/testing/*/console/bone-debian-*-console-armhf-*-2gb.img.xz Mar 05 03:26:53 replace * by highest number available Mar 05 03:28:02 ! hey isn't rc-ee a local expert here? Mar 05 03:28:05 note that if you use BBBlfs to flash it (which is the easiest way), beware that the script is quite broken and e.g. assumes the main partition is partition 2 Mar 05 03:28:18 while it's partition 1 for these images Mar 05 03:29:14 so either just use the usb_flasher of BBBlfs to make your bbb appear as an "usb stick" and flash the image manually, or if you use the script use fdisk to fix the partition table afterwards Mar 05 03:29:15 ah yes - I got a page at elinux.org with lots of rcn-ee images Mar 05 03:31:17 also, if you choose to upgrade from debian wheezy to jessie or later, the upgrade will hang after setting up udev -- open a separate terminal and perform (as root): systemctl daemon-reexec .. the apt-get upgrade will then continue normally Mar 05 03:32:10 good to know - well I'm back on hdmi having booted off an sd image - thanks very much Mar 05 03:32:56 also, even though the console image is minimalistic, it still contains some irrelevant crap: bbx15-ducati-firmware-installer tiomapconf mt7601u-modules-* Mar 05 03:33:11 and of course nano (real men use vim) Mar 05 03:33:32 (or emacs I guess, if you're that sort of person) Mar 05 03:33:45 I like kate and geany, and qt! Mar 05 03:34:22 well, nobody's perfect Mar 05 03:35:24 some poor saps think vis studio and basic are all that you need Mar 05 03:38:28 as for kernel 3.8.13 vs 3.19.0 ... use 3.8.13 if you need any of the stuff that's not in mainline yet, or 3.19.0 otherwise ;) Mar 05 03:42:41 Any replicape users out there? I kind of want one, but haven't found anyone using one. Mar 05 03:45:56 whats replicape do Mar 05 03:46:24 makes endless copies of itself obviously Mar 05 03:46:46 soon the whole world will be covered in replicapes Mar 05 03:47:36 bwarff_: 3D Printer/CNC controller. Mar 05 03:48:10 if you make one, you've made 1000 Mar 05 03:48:12 aah, so another mcu-with-motor-drivers Mar 05 03:49:19 Yep. Related other works are Smootheboard (Cortex MCU), RAMBo (Arduino), etc. Mar 05 03:52:11 an mcu for the motor drivers? are the three PWMSS instances on the am335x inadequate? :P Mar 05 03:54:03 zmatt: Yes. :) Mar 05 03:54:14 ok Mar 05 03:54:14 6 pwm is ideal. 5 is manageable. Mar 05 03:54:25 There needs to be at least 2 ADC. Mar 05 03:54:26 each PWMSS can produce up to 3 pwm signals Mar 05 03:54:56 (and if your needs are simple each timer with an external pin can also produce a PWM signal) Mar 05 03:56:17 Hm. The "fake PWM" would be rubbish. Mar 05 03:56:31 the am335x has 7 adc channels pinned out on the BBB Mar 05 03:56:33 And the replicape doesn't feature an additional MCU. Is that what we're talking about? Mar 05 03:56:52 Yea - the replicape is a cape on the beaglebone black. Mar 05 03:57:12 ah, since bwarff mcu-with-motor-drivers and you said yep Mar 05 03:57:33 And I misread what you said earlier. The PWM on the am335x is just fine. For some reason, I read that as a swizzler and assumed something from another project. :) Mar 05 03:57:56 actually replicape apparently uses PRU to control the steppers Mar 05 03:58:59 Most other projects are just MCU based - Cortex or Arduino. That's what I meant by "yep" to "mcu-with-motor-drivers". Mar 05 03:59:30 an arduino.... -.- Mar 05 03:59:37 The SeeMeCNC Delta printers use replicape. Not sure if anything else "off the shelf" does. Mar 05 04:00:12 zmatt: Yea, Arduinos are the oldest printer boards. RAMPS. Mar 05 04:00:52 I mean, there's so much fancy stuff on the am335x, yet I've even seen people here connect use an external 555-timer connected to gpio Mar 05 04:01:14 Overengineering is fun? :) Mar 05 04:01:34 part of the problem is of course that linux gets in the way of easily playing with the hardware Mar 05 04:02:13 Writing to files is not a fun way to use gpio. Mar 05 04:02:29 I've been thinking it would be really useful to have a generic peripheral driver that just lets you mmap the registers (like /dev/mem), receive interrupts, and setup dma if applicable Mar 05 04:02:39 you can just mmap the gpio peripherals via /dev/mem Mar 05 04:03:03 then reading all 32 pins of one gpio bank is suddenly 1 instruction Mar 05 04:03:08 :) Mar 05 04:03:16 oh, neat trick Mar 05 04:03:35 (1 instruction that will still take quite a few cycles, but still) Mar 05 04:03:49 I've been semi-stuck in libmraa lately, so it doesn't really matter... but still... the thought of files is icky. Mar 05 04:04:11 sysfs is for config, not for any substantial data transfer Mar 05 04:04:19 Exactly. Mar 05 04:04:20 using it for gpio for anything but simple tests is silliness Mar 05 04:06:26 if you need to use a peripheral, just mmap it, as long as you don't need irqs/dma and the peripheral isn't in use by the kernel (or such use doesn't conflict, for example gpio reads are a non-issue) Mar 05 04:06:56 but for example, configuring the direction (input/output) should still go via the kernel to avoid getting into a fight Mar 05 04:07:49 but getting irqs into userspace shouldn't be that hard either Mar 05 04:08:30 and would greatly extend the possibilities of directly working with the peripherals without the complication of writing a kernel driver Mar 05 04:09:10 (it would also make the adc suddenly useful) Mar 05 05:10:12 Good morning all, Am working in Linux-3.8.13 for beagleboneblack. There is no support for power management. I have applied so many patches but still not working. Is there any update or successful patched for PM in Linux-3.8.? Mar 05 06:12:06 Is there any fix for the incorrect powerdown issue mentioned on the wiki where the power light flashes and then nothing happens Mar 05 09:00:43 Can any one support me ? Mar 05 09:02:18 no, but if you tell what the issue is, maybe Mar 05 09:04:26 he needs help with power save Mar 05 09:04:49 I didn't get this to work either Mar 05 11:38:50 Hello! I would like to send request to your sales department! Pleas, tell me email Mar 05 11:45:12 <_av500_> ? Mar 05 11:51:27 av500 its nearly friday Mar 05 12:02:09 Am working in Linux-3.8.13 for beagleboneblack. There is no support for power management. I have applied so many patches but still not working. Is there any update or successful patched for PM in Linux-3.8.? Mar 05 15:11:32 hi Mar 05 15:12:09 type is too small Mar 05 15:53:31 I connected my BBB with logitech c270 webcam and executed a simple program to display captured image. the error is "vidioc_querymenu invalid argument" and "timeout" Mar 05 15:58:39 hello Mar 05 16:04:05 hii Mar 05 16:05:16 anyone online? Mar 05 16:06:22 Guest91537, ask the beagleboard group mailing list. Mar 05 16:14:14 heloooo Mar 05 16:26:50 hello Mar 05 16:27:01 anyone online? Mar 05 16:35:43 sometimes Mar 05 16:47:44 ok thanks all. i lost my patient :( Mar 05 16:49:04 Guest91537, see my previous... Mar 05 17:27:21 users, anyone happen to have an unmodified rev C, need a pastebin.com of u-boot if you have a chance, just starting to dl/flash one of mine.. Mar 05 17:27:39 (may 2014 release u-boot) Mar 05 17:31:58 rcn-ee, you mean its not already in revision control somewhere?! :p Mar 05 17:33:20 oh it is.. on the server, just got to download (150kb/s) and then flash microSD, then fflash eMMC.. so it'll take another hour for me. ;) Mar 05 17:34:06 why flash microSD first? Mar 05 17:35:52 (BBBlfs is not all that great from a technical point of view, but it does save time) ... of course with your download speed (I feel for you) it'll be a while before downloading BBBlfs pays off :/ Mar 05 17:36:52 with the speed of the usb interface, it's the same length as flashing the microSD (5 minutes) and flashing the eMMC (5 minutes) Mar 05 17:39:09 are you sure? I always got the impression eMMC was the bottleneck, not USB Mar 05 17:39:32 hmm, you may be right Mar 05 17:39:47 if the image is > 2 GB Mar 05 17:39:49 BBBlfs is just easier for end users. ;) Mar 05 17:41:12 I'm working on something Ethernet-based, but I've been sick all week so it's not been going fast Mar 05 17:41:39 (also, I'm not sure what the people who wrote the the eMMC standard were on, but wtf...) Mar 05 17:41:57 zmatt: what exactly ethernet-based ? Mar 05 17:42:08 i am always trying to improve bbblfs Mar 05 17:43:43 vvu, have you figured how to replace musb ip on the am335x? ;) Mar 05 17:44:12 vvu: all steps actually.... a rescue-loader which allows examining and reflashing a BBB (with authentication) via ethernet, and for dev purposes I'd also serve it via ethernet boot (bootp/tftp) for convenience probably Mar 05 17:47:23 I've already done baremetal IPv6 on a close cousin of the am335x in the past, so I've been focussing on the eMMC part Mar 05 17:49:39 it's terrible what years of backwards compatibility can do to a standard :( Mar 05 17:49:54 that's why they dropped spi on eMMC. ;) Mar 05 17:50:02 the only good thing... Mar 05 17:50:11 (the only good thing that was availble) Mar 05 17:50:39 except spi would get you only a tiny fraction of the performance Mar 05 17:51:06 and the messy parts of the MMC link-layer protocol are handled in hardware anyway Mar 05 17:52:26 although the mmc/sd peripheral also had some undocumented surprises Mar 05 17:53:17 (peripherals which actually change their behaviour depending on whether interrupts are enabled.. how fun) Mar 05 18:03:55 I already realized that the ROM bootloader is probably causing unnecessary data loss on soft resets :/ Mar 05 18:03:59 *also realized Mar 05 18:04:28 yay Mar 05 18:04:39 * nerdboy verifies he actually has a rev c Mar 05 18:05:18 since rather than recovering the state of the (already-initialized) eMMC it issues an eMMC reset, which means any internal programming operation is aborted and any cached writes are discarded Mar 05 18:06:49 (essentially equivalent to power loss) Mar 05 18:18:18 Hi Mar 05 18:19:32 I'd like to use music frome the commercials of BeagleBone Black. Where should I ask about premission? Mar 05 18:25:03 <_av500_> bbb-user: email jason Mar 05 18:25:06 <_av500_> jkridner|work: ^^^ Mar 05 19:15:05 hello folks Mar 05 19:15:48 anyone knows how improve the device tree load time at the BBB? Mar 05 19:17:30 My BBB was taking 14 seconds to configure the peripherals. I want do this faster! Mar 05 19:18:00 Well, skip the overlay, set it in your dts and add the pinmux to u-boot.. Mar 05 19:19:13 oh thanks Mar 05 19:19:15 I've jumped the overlay, I'm setting right on the device tree. Mar 05 19:20:14 but how do I add the pinmux in uboot? Mar 05 19:21:03 trying to use ccsv6 for BBB.. Mar 05 19:21:15 igorTavares, see: http://git.denx.de/?p=u-boot.git;a=blob;f=board/ti/am335x/mux.c;hb=HEAD Mar 05 19:21:36 but when try to install using "./ccsv*** " nothing happens..any help ?? Mar 05 19:23:21 thanks Robert Nelson! Mar 05 19:25:06 geekswine_: you did make the installer executable after downloading? also, if using a 64-bit linux (like everyone does since the last 32-bit intel processor is now I think a decade old) you'll need to install a fuckton of 32-bit libraries since ccs is still 32-bit Mar 05 19:25:32 see also http://processors.wiki.ti.com/index.php/Linux_Host_Support_CCSv6 Mar 05 19:26:07 i have gone through this.. Mar 05 19:26:25 do we need to run another command to make it executable ?? Mar 05 19:26:52 did you get an error? Mar 05 19:27:08 noo..it just doesn't do anything.. Mar 05 19:28:57 you checked the wiki page for any relevant instructions for whichever distro you're using? Mar 05 19:29:42 let me see once again if am i missing anything.. Mar 05 19:30:18 (note that in general, be aware that using CCS requires a certain amount of masochism, as it *will* involve pain :P ) Mar 05 19:31:46 yaa...its very painfull..:P Mar 05 19:32:48 initially tried with eclipse but to install some additional software for starterware its was taking a hell lot of time...:( Mar 05 19:33:23 note btw that starterware is a steaming pile of pig manure Mar 05 19:34:23 so far the code I've examined of it seems to work mostly by accident if at all Mar 05 19:34:53 the uart echo example compiles to a deadloop if optimization is enabled Mar 05 19:34:53 haha.. Mar 05 19:35:38 actually need to write librararies so that it access board without OS.. Mar 05 19:36:13 Robert Nelson, to change the peripheral configured by direct by uboot I'll have to recompile it? Mar 05 19:36:24 you can actually find some examples by me if you search the e2e forum Mar 05 19:37:08 I still intend to put some on github when I find the time :/ Mar 05 19:37:20 can u provide me the link of any example ?? it will be very helpful for me ... Mar 05 19:37:47 will pace my work somewhat faster :) Mar 05 19:39:24 igorTavares, correct Mar 05 19:40:11 igorTavares, btw, once you set the correct pinmux's in u-boot, you can also remove those setting in kernel land, as they should use the u-boot setting and should speed things up even more.. Mar 05 19:41:42 geekswine_: http://e2e.ti.com/support/arm/sitara_arm/f/791/p/389752/1388187 Mar 05 19:42:19 geekswine_: it's a fairly old snapshot of Mar 05 19:42:37 my baremetal code though, I should really cobble together a bit more up-to-date version Mar 05 19:43:43 thanx..:) i'll have a look at it.. Mar 05 19:45:29 the toolchain link in that post is also old, the latest linaro gcc binaries are http://releases.linaro.org/14.11/components/toolchain/binaries Mar 05 19:47:34 and more recent versions of some of my headers are actually in this project -> https://github.com/dutchanddutch/jbang Mar 05 19:47:57 and the .bin file got executed..:D Mar 05 19:48:33 except to run under linux, instead of declaring a peripheral e.g. "extern Prcm prcm;" and assigning its address in the linker script I declare it "extern Prcm &prcm;" and initialize it with some utility code that mmap()s /dev/mem Mar 05 19:49:28 (I'll warn beforehand that my code style is ... an acquired taste probably ;) Mar 05 19:49:56 haha..:P Mar 05 19:50:25 why just we don't use starterware header file directly to write user friendly programs ? Mar 05 19:51:35 because starterware's headers are typically already mutually exclusive with readable code Mar 05 19:52:57 yeah i have observed that... Mar 05 19:54:45 that's the main reason for me to use C++ .. syntax sugar ;) Mar 05 19:56:01 prcm.mod_mmc1.enable(); Mar 05 19:56:01 wait_until( prcm.mod_mmc1.ready() ); Mar 05 19:56:42 instead of long lines screaming at you with long all-uppercase constants Mar 05 19:58:34 its easy to understand too..:) Mar 05 20:01:28 of course as usual with C++, to make the "end user" code look nice sometimes you have to commit unspeakable horrors in the headers that make the magic happen ;P Mar 05 20:02:25 hahaha... Mar 05 20:03:28 to be honest, C and C++ suck for baremetal programming... they're simply not expressive enough to allow you to convey the semantics of memory-mapped I/O to the compiler, or exert sufficient control over the representation of structures, so I often end up in fights with the compiler Mar 05 20:03:34 but I don't know any better alternative Mar 05 20:04:12 haha..so who wins finally ?? Mar 05 20:04:34 you win some, you lose some :P Mar 05 20:05:30 my approach to things also changes over time, as I discover new ways to solve something, or new problems with the way I used to do things Mar 05 20:05:31 yeah..that's the nature's law ..:D Mar 05 20:05:54 hmm, I should probably fix some food... Mar 05 20:11:15 Robert Nelso I could understand! Thanks, helped me a lot! Mar 05 23:49:25 WHATS WRONG WITH ALL CAPS zmatt? I always thought it was the result of someone being lazy. Mar 06 00:29:43 GenTooMan: NOTHING WRONG WITH, EVERYONE KNOWS IT MAKES TEXT SO MUCH EASIER TO READ Mar 06 00:29:51 ;) Mar 06 00:35:18 OR AN OLD FORTRAN GEEK Mar 06 00:36:49 EQUIVALENCE foo bar <= WTF were they thinking?!? Mar 06 00:36:54 modern fortran is quite okay though Mar 06 00:57:32 Fortran 99 not your grandpa's fortran! Mar 06 00:58:02 I mean fortran 2008 Mar 06 00:58:46 Details. I suppose 55 year old languages need updates too? Mar 06 00:58:57 there's no f99 btw, you might mean f95 Mar 06 01:00:04 Don't use it just taking a random stab at what Might be a crude number. It's kind of weird though they are using dates for everything since windows 95 Mar 06 01:00:45 I'm pretty sure Fortran 77 predates Windows 95 ;-) Mar 06 01:01:19 (Fortran 90 too for that matter) Mar 06 01:03:08 You are too much into fortran for my own good. ;) Mar 06 01:03:40 hehe, not really Mar 06 01:04:46 but I've had some contact with it, and I was kind of amused that it's still very much a living language (there's even a 2015 revision upcoming) Mar 06 01:05:10 even support object-oriented programming and stuff nowadays Mar 06 01:05:14 *supports Mar 06 01:05:48 Hopefully not as badly as Perl does. Mar 06 01:12:13 hey, I happen to like Perl Mar 06 01:13:12 "ooh there is a key i havent used" perl design philosophy Mar 06 01:17:41 I like Perl Pre 5 at 5 it was OOP? or OOPs? I hated the OOP part. 4 was great used it to generate postscript programs and all sorts of strange stuff. Mar 06 01:18:24 Everyone has a favorite language I guess :D Mar 06 01:19:37 perl is a toolbox not a language .. lol :p Mar 06 01:19:41 perl is fine aslong you use it all the time. Mar 06 01:20:02 perl is never fine. :P Mar 06 01:20:04 coming in blind after monthsof disuse its a nightmare of squiggle squark Mar 06 01:20:06 and yet I use it too often :p Mar 06 01:20:31 not by choice, I admit Mar 06 01:21:27 my choice of language is based on the test of not looking at the codebase for 6 months and then getting a panic call at midnight when your drunk.. perl fails that test. Mar 06 01:23:21 lol .. good call bwarff Mar 06 01:26:10 maybe you should try using comments? ;) Mar 06 01:26:27 oo harsh .. Mar 06 01:26:55 I've found Perl to be very powerful for the right purpose. If used for the wrong purpose it sucks. :D Mar 06 01:27:25 any language sucks when used for the wrong purposes Mar 06 01:27:33 ofc Mar 06 01:27:36 (in case of php, that would be any purpose whatsoever) Mar 06 01:28:32 lol .. see, everyone got their favs/hates :p Mar 06 01:28:49 as long as we can all agree emacs sucks it's all good Mar 06 01:28:51 Yes I used Perl to convert assembly languages between 2 processors with 99% functional code. It would have taken too long by hand for the number of project conversions. Mar 06 01:28:53 actually I need to correct myself Mar 06 01:29:06 PHP is an excellent choice for demonstrating poor programming techniques Mar 06 01:29:18 zmatt .. but php is all OVER the web :p Mar 06 01:29:39 like it or not .. seems to do a job Mar 06 01:29:58 and plenty of harm in the process Mar 06 01:30:00 I would nominate Javascript for bad programming practice exacerbator. (sound dirty). Mar 06 01:30:23 given that it actively hinders rather than encourages secure programming techniques Mar 06 01:30:41 I used PHP ... once. Mar 06 01:30:59 like, in Perl you'd get shot down if you interpolate a variable into an SQL query instead of using a placeholder Mar 06 01:32:05 in PHP land it's still standard practice, and it actually seems non-trivial to do otherwise Mar 06 01:32:56 every language has its quirks though .. and I prefer a good typed language to perl Mar 06 01:34:02 er, MySQLi is defacto php land. Mar 06 01:34:34 Unless you're in one of those super high tech shops that use pdo. Mar 06 01:34:46 (oh, and rather funny, PHP's comparison operators aren't transitive... in fact you can find a,b,c such that a tamarin: and mysqli::query / mysqli_query doesn't support placeholders for example :) Mar 06 01:37:19 I use SQLite not very muchh but it's a bit better written just my opinion AND it doesn't have dangling attachments since it can be used in comercial products. Mar 06 01:37:26 http://php.net/manual/en/mysqli-stmt.bind-param.php Mar 06 01:37:50 1.4 Mar 06 01:37:55 d'oh wrong keybiard Mar 06 01:38:38 I hates it when me keybeard goes eh wure! Sorry I am making fun of south carolina accents I am ashamed. Mar 06 01:39:46 tamarin: yes, which requires at least 3 steps to execute one statement Mar 06 01:40:06 zmatt: ... and? Mar 06 01:40:08 which means that, people being lazy, they'll take the easy route and interpolate Mar 06 01:40:08 hm Mar 06 01:40:26 zmatt: No. It's PHP. They'll just make random functions. :P Mar 06 01:41:04 php and mysql deserve eachother Mar 06 01:41:20 both have nasty implicit behaviorus retards like and coders hate Mar 06 01:41:59 javascript has fun bits too Mar 06 01:42:12 yes, more implicit garbage in that Mar 06 01:42:26 I don't node what you mean. Mar 06 01:42:27 plus, thorough use of UTF-16, always a delight Mar 06 01:42:42 anyting that swallows errors and ignores types Mar 06 01:43:00 seen this talk? https://www.destroyallsoftware.com/talks/wat Mar 06 01:43:39 it's actually non-trivial why JS exhibits the behaviour they show there :) Mar 06 01:43:49 *he shows there Mar 06 01:50:52 (with the {} + cases javascript's implicit semicolon kicking in, thinking the {} is an empty block, hence the + becomes a prefix operator to numerify the right hand side) Mar 06 01:51:01 *kicks in Mar 06 01:51:40 twitch Mar 06 01:52:08 and of course [] stringifies to "" which numifies to 0 Mar 06 01:52:22 while +"[object Object]" ends up NaN Mar 06 01:53:03 wonderful language Mar 06 01:53:15 if "0" + 0 doesnt error, walk away slowly. Mar 06 01:53:46 no sudden movement, just smile and nod and backaway Mar 06 01:54:16 a colleague of mine has a book on the shelf "Javascript: The Good Parts" ... I couldn't help but notice it was a rather small, thin book Mar 06 01:54:32 6 pages. Mar 06 01:57:03 bwarff: I suppose you also don't like that in Perl you can make values whose numerical value and string representation are unrelated? like $! contains errno when used as number and strerror(errno) when used as string :) Mar 06 01:57:27 (and that's even without operator overloading being involved) Mar 06 01:58:35 Hi every one Mar 06 01:59:02 $_-> ~ is enough to hurt before i worry about behaviour Mar 06 01:59:11 can any body tell me where is the DTS files in new BBB located? Mar 06 02:00:19 imanpakii, /boot/dtbs/$(uname -r)/ Mar 06 02:08:13 Well people were complaining about C and NULL pointer assignment the other day. I asked them if they had considered what you could do for the alternative, that turned into "C++" solutions (I wanted to smack them). Mar 06 02:09:36 I have a simple solution to null pointers. Mar 06 02:13:52 the simple solution to pointer problems is anyone that gets confused by pointers should be using python/c# or one of the other higher level langs that hides them Mar 06 02:13:54 I suppose I could make a macro called DULL? :D Mar 06 02:14:23 Those people have to ride in cars instead of Drive them ;)\ Mar 06 02:14:37 bwarff: That was my argument. Use a language that does not have them. Mar 06 02:14:45 GenTooMan: Harsh. Mar 06 02:15:25 the analogy is building a car versus driving it i think :) Mar 06 02:15:30 the enduser just rides. Mar 06 02:15:56 not everyone should be building memory mangement and datastructures the way c makes you Mar 06 02:16:01 ^^^ Mar 06 02:16:27 And not everyone who can wants to. Mar 06 02:16:31 <- Mar 06 02:16:39 Well ... and not everyone who wants too can? :D Mar 06 02:16:59 Sure. Mar 06 02:18:02 ive seen a shitload of bad linked lists in c code, c#/pythons highly tuned dynamic arrays eat them alive Mar 06 02:18:29 alot of people program c that really dont have the baseline skills to replicate those bits Mar 06 02:22:01 I am not in favor of bad comparisons. Very few people can write good code. Regardless of the language. Knowing the language helps a lot and knowing it's weaknesses is also good. Sadly I make lots of typos instead of logical errors. :d Mar 06 02:25:05 there is no winner in that argument anyway.. we would get nothing done if everyone handrolled baremetal os's and all the support libs Mar 06 02:28:26 Hmm well someone was 'advertising' a baremetal library system that could quickly target new SoC's didn't hear much after a month though. Mar 06 02:30:29 I've noticed people are greedy and tend to think with the cell phone philosophy with software "I bought a cell phone for $30 why can't you make one for that?" Mar 06 02:31:48 well cheap phones is why we now having rocking little SOC arm boards for stuff all $ Mar 06 02:36:49 Those phones aren't that cheap, no cell phone costs $30, even now. Kind of what I was getting at you either pay up front or pay a bit at a time so you don't notice. The initial money you pay is just technically ernest money. The same goes with software. Integrating everything together reduces interconnect cost but it still isn't free. The actual cost of the phones are in the park of $60 and the "free" phones are just taken Mar 06 02:36:49 out of the monthly contract (think installments). Mar 06 02:38:05 I always hate malls where people attack you to sell you cell phone contracts (sigh). Mar 06 02:38:48 Anyhow their is nothing free. IT takes time, where you choose to do what is important is all I'm saying. Mar 06 02:40:10 So Beagle Boards :D By the way this chat forum is a godsend when I can't connect the dots on the wiki. Thanks for pointers people. Mar 06 02:42:25 well I shelled out $300+ on my last new phone retail .. Mar 06 02:42:31 and that was an Old model Mar 06 02:43:48 apple or samsung? :D Mar 06 02:43:55 aren't iPhones somewhere nearer $600-700 - that's just ridiculous Mar 06 02:44:01 samsung ofc :p Mar 06 02:46:05 Yes apple phones are expensive but they do more testing and spend more engineering $ than typical (LG for example) Samsung spends more engineering $ as well. It takes time to make quality products, Apple takes more care in their products having learned that with the Apple 3 netwon and the Lisa. Mar 06 02:46:55 Hey! Not sure if this is the right place to ask...but does anyone know how to restart the service responsible for analog inputs on the beaglebone? Mar 06 02:47:10 an inputs Service eh? sounds novel .. Mar 06 02:48:48 I keep getting an IOError (Errno 11) and I can only continue to read the analog inputs if I reset the board (power off and back on). I was hoping to find whatever script is responsible for this in order to simulate a reboot. Mar 06 02:50:24 Evan__ what are yo using to read your analog inputs? Mar 06 02:51:07 I'm using the Adafruit BBIO library, the program is in python. Mar 06 02:54:15 ok Mar 06 02:54:49 The error says that resource is temporarily unavailable if that helps... Mar 06 02:56:05 that's EAGAIN... that's normally not an error at all Mar 06 02:56:18 at least I think that's EAGAIN ? Mar 06 02:56:32 Can you explain what EAGAIN is? Mar 06 02:56:37 I gotta eat .. bbiaf :) Mar 06 02:58:11 Evan__: on a non-blocking file descriptor it indicates read/write (whichever gave the error) is currently not possible Mar 06 02:58:22 it's given when a blocking file descriptor would block Mar 06 02:58:41 so it's accessing via /dev the A2D channels? Mar 06 02:58:49 can you poll/select it before using it **** ENDING LOGGING AT Fri Mar 06 02:59:58 2015