**** BEGIN LOGGING AT Mon Oct 14 02:59:58 2013 Oct 14 03:17:58 wow Oct 14 03:55:46 the convo about frontcam was very inspiring. I think we need to check indepth how it#s handled and what might apps like "video call2 do to the hw config in cam chip Oct 14 03:56:13 yes it's known to need lots of light to work semi-decent Oct 14 03:56:47 it's also known from early N900 days that Nokia shiped several camdriver improvements to main cam but iirc also to frontcam Oct 14 03:57:35 freemangordon: ^^^ you might be interested in those details as of convo above Oct 14 03:58:34 DocScrutinizer05: hmm? Oct 14 03:59:16 abysmal quality of frontcam when using mirror, until you did a video call after which is seems to be much better. Interesting, isn't it? Oct 14 03:59:36 ("mirror" is an app) Oct 14 03:59:43 no, it isn't, it just shows that mirror is of bad quality Oct 14 04:00:21 there is V2L API to control front cam properties, obviously mirror isn't using it Oct 14 04:00:40 :nod: Oct 14 04:01:05 but how would frontcam the "get better" after doing a video call, for mirror app? Oct 14 04:01:23 it is the same story about gtalk video has way better quality than skype Oct 14 04:01:28 wouldn't cam get powered down and reset each time it gets opened? Oct 14 04:01:59 sure, but driver restores what's been last set Oct 14 04:02:02 s/the7then get better... Oct 14 04:02:07 or v2l, not sure which layer Oct 14 04:02:16 ooh Oct 14 04:02:21 interesting Oct 14 04:02:54 so where's the "persistance storage"? Oct 14 04:02:58 the same happens if you use man camera through gstreamer Oct 14 04:03:03 *main Oct 14 04:03:22 weird Oct 14 04:03:35 if you want good quality, you need to open the camera-ui, close it and then use gstreamer Oct 14 04:03:36 kernel module var area? Oct 14 04:03:44 gconf? Oct 14 04:03:53 cam chip registers themselves? Oct 14 04:04:07 DocScrutinizer05: not sure, but I think it is kernel, as it saves the context on suspend and restores it on resume, iirc Oct 14 04:04:28 the context == registers Oct 14 04:04:30 yep, so a modprobe -r should "fix" that, eh? Oct 14 04:04:45 maybe, never tried Oct 14 04:04:59 IOW unloading the kernel driver module should kill qulaity again Oct 14 04:05:11 :nod: Oct 14 04:05:25 we need sb to test this Oct 14 04:05:32 sb? Oct 14 04:05:36 I never even noticed the effect to start with Oct 14 04:05:41 somebody Oct 14 04:05:51 ever used the mirror app? Oct 14 04:05:55 yes Oct 14 04:06:11 and yes the quality is abysmal Oct 14 04:06:14 for sure the quality is bad Oct 14 04:06:31 I never noticed it getting better, not to think of *how* to do that Oct 14 04:06:42 no video calls here Oct 14 04:06:51 make a video call :) Oct 14 04:06:52 aah Oct 14 04:06:59 never managed to do a single video call in my life Oct 14 04:07:21 register 2 google accounts and call your PC from your n900 Oct 14 04:07:29 or the opposite Oct 14 04:07:39 video call that is Oct 14 04:07:49 I tried from one SIP client to another one on same account, but my SIP provider doesn't even support video calls at all Oct 14 04:07:59 gtalk supports Oct 14 04:08:07 I for sure won't register a google account Oct 14 04:08:11 as is skype :P Oct 14 04:08:22 no skype on my devices Oct 14 04:08:42 thus I said "we need somebody..." Oct 14 04:09:06 best somebody who actually noticed the improvement already Oct 14 04:09:09 no options left then, but you can ply a bit with mplayer, afaik it should support camera properties Oct 14 04:09:14 play Oct 14 04:09:29 so you can tweak exposure etc from it Oct 14 04:09:36 sure, I know Oct 14 04:10:02 btw what do you want to "test"? Oct 14 04:10:10 but I'd prefer to get the improvement anf reversal thereof by rmmod approved first Oct 14 04:10:15 ^^^ Oct 14 04:11:06 use mplayer to put brightness/exposure to max, check the quality with mirror, rmmod, modprobe, check agaimn Oct 14 04:11:45 that would assume I see an improvement and I'd know to play with mplayer Oct 14 04:12:07 I want the effect as reported by dos getting checked via "rmmod" Oct 14 04:12:13 that's all Oct 14 04:12:36 you'll see improvement, that is for sure, I already tried it an year or so ago :) Oct 14 04:12:55 once we know how that improvement works, we can think of proper initialization on modprobe Oct 14 04:13:09 I even got mirror sources thinking to fix it, but I gave up as there were more interesting things to play with Oct 14 04:14:01 this shouldn't get done on a per-app basis. udev is the right place to invoke a post-modprobe configurator daemon Oct 14 04:14:45 DocScrutinizer05: you can't, setting camera props should be done on as-needed basis Oct 14 04:14:58 this would fix _all_ usecases that forget to do proper init themselves Oct 14 04:15:27 sure, but it can't hurt to have optimized defaults Oct 14 04:15:41 no, you can't know in advance what the light conditions will be when those broken apps are started Oct 14 04:15:59 obviously what modprobe resp kernel driver init gives you is suboptimal Oct 14 04:16:06 the same for the resolution btw Oct 14 04:16:31 the default that driver gives you, is everything set to min Oct 14 04:16:38 inc resolution Oct 14 04:16:50 which is sane IMO Oct 14 04:16:56 so what? since you can't know which light we will have, you configure the module to 1billion lux ambient in kernel init? Oct 14 04:17:06 it is up to the application to decide how to use the camera Oct 14 04:17:19 that's insane approach Oct 14 04:17:41 the option is to implement some "auto mode" in the driver Oct 14 04:17:48 every hardware should come up with semi-usable defaults, not any crap that leaves it semi-broken Oct 14 04:18:31 obviously frontcam startup defaults are unbearable for *every* usecase Oct 14 04:18:40 yep Oct 14 04:18:47 they aren't optimized for *anything* Oct 14 04:19:19 sure they aren't, but again, it is not the driver to guess what application wants to do Oct 14 04:19:26 so no matter for what you optinize them - nightview or flash use - OPTIMIZE THEM for something! Oct 14 04:19:33 it is the application to tell the driver Oct 14 04:19:43 again, that's an insane rationale Oct 14 04:19:55 could be, but this is what I think Oct 14 04:20:19 ok, what are the values you'd pick for defaults of frontcam? Oct 14 04:20:39 your birthday? Oct 14 04:20:50 NaN? Oct 14 04:20:51 everything on min Oct 14 04:21:04 or something that at least *sometimes* works? Oct 14 04:21:13 and this falls in your "sane defaults" category :) Oct 14 04:21:18 why not everything at max? Oct 14 04:21:24 no reason Oct 14 04:21:29 no, it does NOT Oct 14 04:21:36 that is why it is hard to define "sane default values" Oct 14 04:21:59 sane default values are what works best with the most common usecase Oct 14 04:22:17 it's not at all hard to define Oct 14 04:22:17 the most common usecase for the front cam is video calling Oct 14 04:22:19 ;) Oct 14 04:22:28 fine Oct 14 04:23:12 freemangordon, DocScrutinizer05, to add my 2c - the app _needs_ to be changing these as it is environment Dependant, but semi-usable default would be nice (if only for testing) Oct 14 04:23:16 hten make kernel module configue the chip to the values needed for video calling, on module init Oct 14 04:24:13 otherwise I insist in using MY birthday date for register init, not min, max, or freemangordon's birthday ;-P Oct 14 04:25:22 chainsawbike: you can use V2L API to set the camera to whatever you like, V2L is still driver "layer" albeit not in the kernel Oct 14 04:25:45 that's all fine and would be sad if it wasn't like that Oct 14 04:25:59 maybe V2L have a way to set some defaults, dunno Oct 14 04:26:12 but that doesn't change *anything* in the reasoning to get *sane* defaults on module init Oct 14 04:26:56 i don't think you can have sane defaults Oct 14 04:27:11 camera settings depend on the outside world too much Oct 14 04:27:28 sure, we can easily do a helper that gets invoked by udev on modprobe and opens cam for 0.01s with correct parameters, exploiting gstreamer or whatever you see fit Oct 14 04:27:46 kerio: please read backscroll!!!! Oct 14 04:28:09 DocScrutinizer05: and that would make absolutely no difference to now Oct 14 04:28:18 kerio: while you can't have sane defaults maybe, this is no exciuse to have MAD defaults Oct 14 04:28:49 because I am saying that putting everything to min is best for me as I livo on the beach and use the cam only between 10 and 16 Oct 14 04:29:21 even then you can't use that default since it simply doesn't work Oct 14 04:30:10 it's a flaw of kernel module devels who simply didn't think of initializing their values to something semi-decent. Just like with current-max on LP5523 Oct 14 04:30:14 DocScrutinizer05: and what about is someone changes those settings Oct 14 04:30:23 is/if/ Oct 14 04:31:10 then they are changed? Oct 14 04:31:17 don't get you Oct 14 04:31:34 but they are no more those "works for everyone" defaults Oct 14 04:31:46 should we keep them or reset? Oct 14 04:31:51 DOENT work for everyone is what we got now Oct 14 04:32:07 aaah *now* I get it Oct 14 04:32:15 I'm talking about kernel init Oct 14 04:32:24 DocScrutinizer05: see, it is not only the exposure etc, it is also the resolution Oct 14 04:32:34 what is the best resolution? Oct 14 04:32:36 that's why I asked for checing if those persistent settings go away with rmmod Oct 14 04:33:12 when they vanish with rmmod you want have new sane ones with insmod Oct 14 04:33:19 and not only the resolution, but framerate too Oct 14 04:33:29 yesyesyesyes Oct 14 04:33:42 all quite irrelevant Oct 14 04:33:55 every app is free to change that to whatever it needs Oct 14 04:34:05 exactly Oct 14 04:34:13 and not only free to, but have to Oct 14 04:34:15 I#m talking about the initial defaults Oct 14 04:34:19 I know Oct 14 04:34:41 and those HAVE TO be saner than they are now, cause the ones we have now are defunct Oct 14 04:34:49 * chainsawbike would set it to work under the ambient lighting of my workspace by default to make testing easy and call any application broken that did not adjust it for the current conditions Oct 14 04:35:02 chainsawbike: exactly Oct 14 04:35:17 since that's most likely the most common usecase anyway Oct 14 04:35:23 But a camera application that simply starts a stream without giving the user a tool to control the camera is simply broken. IMO Oct 14 04:35:39 so what? Oct 14 04:35:48 and I would call that WFM approach :P Oct 14 04:35:48 yes, it is Oct 14 04:36:17 I would call that a "let's punish them2 approach by the kernel driver devels Oct 14 04:37:16 DocScrutinizer05: if we assume something is broken, I'd rather say it is V2L, not the module Oct 14 04:37:16 "ooh your PSU exploded? too bad, you shouldn't have activated it without proper config of mains power range before you do so" Oct 14 04:37:55 DocScrutinizer05: on the opposite, keeping everything to min is the safest approach Oct 14 04:38:04 in my book every module should come up with sane safe working defaults Oct 14 04:38:22 safe for what, goddamn? Oct 14 04:38:27 for your PSU Oct 14 04:38:36 I took your example Oct 14 04:38:38 safe for making every crappy app fail epically Oct 14 04:39:00 it is not the kernal that should babysit crappy apps Oct 14 04:39:09 kernel even Oct 14 04:39:22 in my PSU example keeping mains voltage to min is a sure way to make the PSU explode ;-P Oct 14 04:40:35 freemangordon: you're arguing on a "let's punish them,CBA to init that to 100 instead of using the compiler default" approach Oct 14 04:41:24 hmmm, I recon I could clone libdevlock if it wasnt for the totally confusing function that actually stores the devicelock in CAL :P Oct 14 04:41:28 keeping everything at 0 on module init has ZILCH rationale nor benefit Oct 14 04:41:43 DocScrutinizer51: if we apply "sane defaults" to the audio for example, what would be it? Oct 14 04:41:52 0 Oct 14 04:42:11 except for one default volume path Oct 14 04:42:19 exactly the same for the video Oct 14 04:42:24 no Oct 14 04:42:33 since video is a sensor, no actor Oct 14 04:42:50 what about microphone? Oct 14 04:43:09 should it be disabled or enabled? with what gain? Oct 14 04:43:38 obviously should be set to a sane boost level and gain to semidecently handel a 60dBm Oct 14 04:44:05 though that's ALSA stuff, not kernel module. And ALSA DOES that already Oct 14 04:44:23 there is no "sane level" for sensors which vastly depend on the outside world conditions Oct 14 04:44:23 it restores last settings on boot time Oct 14 04:44:50 freemangordon: stop telling nonsense, of course there are settings that are more sane than others Oct 14 04:45:11 WFM? Oct 14 04:45:48 when there is no "sane level2 in your book, then why arguing against my request to change those settings from 0 to maybe 100 or whatever I may come up with? Oct 14 04:46:05 freemangordon, yes it is very dependant on external factors but what is the harm in setting it up to work somewhere by default ( a dev workspace for example) Oct 14 04:46:18 exactly Oct 14 04:46:45 there *is* something like a "standard" environment Oct 14 04:47:13 my point is - you can change the defaults to whatever you like (in the kernel), but broken apps are just that - broken apps, and it is not the kernel that sould fix them Oct 14 04:47:16 and a sensor should work OOTB for such standard environment Oct 14 04:47:35 SO WHAT?????? Oct 14 04:47:46 "punsih them!!"? Oct 14 04:47:55 DocScrutinizer05: what is the standard environment for a handheld device? Oct 14 04:48:12 earth, 12 noon, aequator Oct 14 04:48:13 DocScrutinizer05: exactly, punish them don't tolerate lazy developers Oct 14 04:48:15 FFS Oct 14 04:48:32 >.-( Oct 14 04:49:00 comeon, setting camera properties is like 50 lines of code Oct 14 04:49:19 it's NOT "inside an atomic explosion" or "11km deep in marian trench" Oct 14 04:49:38 SO WHAT?????????????????????????? Oct 14 04:49:46 * chainsawbike likes a dev workspace so you get something resembling a picture not garbage when you first init the camera, rather than wondering wtf the garbage you are getting is, and wasting time realizing it really is the camera output Oct 14 04:50:08 chainsawbike: ack Oct 14 04:51:22 freemangordon: and you wonder why lots of app devels and users think kernel devels are mad and assholes? Oct 14 04:51:30 guys, mirror is not even setting the resolution, what "sane defaults" we talk about here Oct 14 04:51:54 I'm NOT talking about mirror app, at least Oct 14 04:51:58 freemangordon, i consider mirror _broken_ Oct 14 04:52:12 forget that friggin app Oct 14 04:52:28 I want a sensor to work OOTB, no matter how poorly Oct 14 04:52:43 and that is what you have Oct 14 04:52:49 it makes no sense to initialize it to a state where it doesn't work at all Oct 14 04:52:55 NO Oct 14 04:53:10 definizely not true for frontcam Oct 14 04:53:25 it works OOTB and if there is enough light around, you'll even see an image of youself Oct 14 04:53:35 bullshit Oct 14 04:54:00 meh, i'll patch that shite myself and upstream it. Don't need you for that Oct 14 04:54:12 DocScrutinizer05, enough light... Oct 14 04:56:06 setting everything to zero is definitely the most idiotic setting you could find for cam Oct 14 04:56:23 DocScrutinizer05: try it (upstreaming) but I bet you'll get exactly the same responce as mine :) Oct 14 04:56:37 even setting everything to max sensitivity is better than that Oct 14 04:57:09 though... who knows Oct 14 04:57:18 * jonwil wonders if reverse engineering libdevlock and devlocktool is worth the effort... Oct 14 04:58:22 freemangordon: I bet against that, since I can support my point using the cam manuf's suggested range of opperation values from datasheet, and setting values for the module somewhwre to the arithmetic or geometric middle of those is for sure a more sound approach than setting everything to zero Oct 14 04:58:31 probably not :) Oct 14 04:58:49 DocScrutinizer05: ok Oct 14 04:59:10 what about resolution and framerate? Oct 14 04:59:31 you said yourself it's usually used for video Oct 14 04:59:53 frontcam supports 180x120, 320x240 and 640x480 Oct 14 05:00:02 not sure about the framerates Oct 14 05:00:05 that makes VGA 25fps in my book, nit 0 fps and 0 resolution Oct 14 05:00:49 not sure n900 can handle VGA@25fps video streaming Oct 14 05:00:50 and of course you configure sensors to their full capability, not to sth downsized Oct 14 05:01:38 skype uses 180x120, gtalk 320x240 iirc Oct 14 05:01:41 when N900 can't support 25fps then you obviously set it to the max it *can* support Oct 14 05:02:03 and ofc it depends on what the other side can handle Oct 14 05:02:29 and iirc ususlly video call is @ 12fps Oct 14 05:02:35 the sensor shall work by the most simple means, like "cat $input >$output" Oct 14 05:03:05 and it should provide the best it can, in that mode Oct 14 05:03:39 this is a general rule for all sensors Oct 14 05:04:12 so max resolution, max sampling rate, max sensitivity and so on Oct 14 05:04:13 DocScrutinizer05: maybe you should check what ISP in OMAP is Oct 14 05:05:09 maybe I should stop arguinfg with somebody who thinks "punish them" is a good approach since it doesn't require the kernel devel to muse about sane defaults Oct 14 05:05:15 it is not only the sensor (and its module), there is dedicated HW who does exactly what you need, it even gives yout the histogram Oct 14 05:06:04 and I couldn't care less about that last point of yours since it won't change my notion on this Oct 14 05:06:19 ok, lets agree to disagree :) Oct 14 05:06:38 setting everything to zero just because compiler initializes vars to zero is a poor and silly approach Oct 14 05:06:55 pathetic approach even Oct 14 05:08:04 as soon as you start to think about that, you easily find ~65534 better values to initialize than just zero Oct 14 05:10:23 and probably any optiomum is anywhere between 500 and 30000 Oct 14 05:11:21 definitely zero is no optimum, not even a lame compromise, not even any rationale at all to support that approach. Except "let's punish them" Oct 14 05:13:17 what is the default baudrate of an UART? zero? Oct 14 05:14:07 9600 iirc, but it is not set by the kernel UART driver iirc Oct 14 05:14:36 sure you *should* use stty to configure the 8n1 and 9600 or 19200 or 64k or whatever you want. but ZERO for sure is the most idiotic value for UART baudrate to have as default. Same for camera Oct 14 05:15:26 DocScrutinizer05: it is not UART driver that sets the default, get my point? Oct 14 05:15:55 afaik it is tty driver Oct 14 05:16:10 uhuh, so who else is setting uart to a particular speed? is it random? Oct 14 05:16:41 there MUST be some speed the UART is set to when it gets loaded Oct 14 05:17:04 and I doubt you even *can* set that to zero Oct 14 05:17:06 it is passed through cmdline Oct 14 05:17:36 nonsense, you *can* pass that via cmdline. That's not any default though Oct 14 05:17:50 meh, I'm tired of this discussion Oct 14 05:26:36 (just on a sidenote: for UARTs there seem to be ~4 "sane" defaults: 300, 1200. 9600, 19200. No kernel driver or whoever will initialize a UART to 120, 600, 38400 or whatever) Oct 14 05:27:43 these values have historical and practical reasons, and they have sound rationale why they are used Oct 14 05:29:36 for every sensor you can find similar reasoning why you should use a certain (range of) values for module init default Oct 14 05:30:08 I honestly doubt zero for all values of a camera has *any* such sound rationale Oct 14 05:31:47 e.g stabdard PAL or NTSC or VGA would be a sane fallback and default Oct 14 05:33:12 and a exposure that somehow meets a normal daylight outside situation Oct 14 05:34:20 a less crappy chip would automatically set those values on reset Oct 14 05:34:42 in taht case the kernel module usually wants to just keep whatever the chip came up with Oct 14 05:35:22 in other cases manufactuers clearly state defaults Oct 14 05:36:33 when they don't then they state recommended operating conditions which usually are ranges and you want to pick a value for defaults that is "middle" of the range by whatever algo you use to find that middle Oct 14 05:37:01 for other sensors you want them to come up with maximum sensitivity Oct 14 05:37:43 actors you usually make come up with actual operation set to inactive, and limits set to those the hw manufacturer specified Oct 14 05:38:20 but you ALWAYS can find a better and less embarrassing (for your reputation as developr) value for defaults, than simply "all zero" Oct 14 05:40:49 there are situations where you don't want to come up with any sane working default at all - e.g. firmware for WLAN chip. But that's another topic not applicable for simple settings like exposure and ISO and framerate/resolution of a camera Oct 14 05:49:23 btw I never looked in V4L2 or the camdriver kernel API, but I could figure there are mandatory parameters you _must_ pass to an open(). When e.g resolution was one of those mandatory parameters then any default for resolution is moot anyway. Oct 14 05:50:29 you don't have a default for filename on an open() either Oct 14 06:50:51 bah this is annoying Oct 14 07:20:08 hm? Oct 14 07:23:15 morning Oct 14 07:44:55 trying to figure out the libdevlock change_devlock_code function and its so far eluded my ability to reverse engineer it (even with an x86 binary to play with) Oct 14 07:45:14 its doing something really funky for generating the salt and I cant work it out Oct 14 08:02:03 haha Oct 14 08:03:36 you don't need to understand how it generates salt. You just need to generate own salt and understand how salt is added to the hash, so you can verify existing hashes incl salt Oct 14 08:04:09 since jack-the-ripper can crack the hash, I guess salt is added in a pertty standard way Oct 14 08:04:23 err johm-the-ripper Oct 14 08:04:29 meh! Oct 14 08:08:38 have you already found the if (hash == "12345") then return 0; ? Oct 14 08:09:43 funny enough the original default hash for default lockcode 12345 is also 12345 ;-P Oct 14 08:10:34 yeah, ignore salt generation. A counter is just as salty as a something random Oct 14 08:11:40 for device lockcode for sure Oct 14 08:11:57 it's a hoax in the end, anyway Oct 14 08:12:43 And as the salt must be stored in the clear, it adds very little security to what's intrinsically a small brute force problem Oct 14 08:19:33 yep. Salt only got invented to make rainbow table attacks harder Oct 14 08:20:10 and maybe to hide password changes Oct 14 08:20:56 in the sense that you could use same password but still a different hash each day Oct 14 08:25:16 A splendid wonderful fine Morning (UGT) on this gorgeous Boomtime! Oct 14 08:27:41 boomtime? Oct 14 08:29:19 Discordian Weekday Oct 14 08:29:29 Read http://principiadiscordia.com Oct 14 08:43:19 Salt predates rainbow tables by decades Oct 14 08:44:21 splendid. Oct 14 08:45:08 * sono would rather go outside and play sink Oct 14 08:45:23 * sono eyes the pond Oct 14 08:45:45 * sono eyes the headachy embedded devices from hell next to his desk Oct 14 08:45:48 yea, sink sounds good Oct 14 08:48:46 * sono hands Apic a PIC32 based industrial thing with an oldschool LCD. a JTAG cable dangles from it. Oct 14 08:48:52 * sono nods at the pond Oct 14 08:49:11 * sono grabs something else that looks like the unholy union of an acer netbook and an ipad Oct 14 08:49:41 * sono yells something undiscernable and PULLs Oct 14 08:49:47 Thank You very much! Oct 14 08:49:49 *splash* Oct 14 08:50:09 Shift to the Left, shift to the Right, push up, pull down, Byte, Byte, Byte! Oct 14 08:50:18 =) Oct 14 08:50:25 \o\ /o/ ^o^ ,o, o< o< o< Oct 14 09:06:41 Drink all the booze, hack all the stuff? Oct 14 09:09:47 Do You know Operation Swordfish, the Movie? Oct 14 09:09:57 Log on, hack in, go anywhere, steal everything! Oct 14 09:12:09 What a horrid movie that was. Oct 14 09:12:23 ;-) Oct 14 09:12:32 I liked it. Oct 14 09:20:46 ._." there was simply too much blockbuster crap compared to what infosec was all about. Oct 14 09:21:03 popular movies are generally quite horrid. if some random persons likes a movie, that actually increases the odds that it's quite meh. Oct 14 09:21:27 random persons on average are quite meh. Oct 14 09:21:34 or one could simply say, Oct 14 09:21:36 law of averages Oct 14 09:21:40 and be done Oct 14 09:21:48 It's more apparent if one has underlying knowledge of what the theme of the movie is trying to portray. Oct 14 09:22:19 i haven't seen it, but i can imagine. Oct 14 09:22:28 did they create a gui in visual basic? Oct 14 09:23:26 I read in reader's digest awhile ago of some guy whom collect's spy paraphernalia. When asked how would the James Bond do in real world, he quoted from a ex KGB spy stating the real James Bond would not last in the spy world for more than 4 minutes or something. Oct 14 09:23:56 I actually don't want to think about swordfish, I could imagine swordfish would be interesting is if I was stoned. Oct 14 09:24:06 surely he meant "before being fired" Oct 14 09:24:42 No, probably before James Bond dies because he would be outsmarted by real spies probably. Oct 14 09:24:51 well if you're stoned, gravity is a much better movie to begin with Oct 14 09:24:57 real spies don't go around killing each other Oct 14 09:25:04 Exactly. Oct 14 09:25:18 Well I haven't watched gravity yet anyway. Oct 14 09:25:23 not as a day to day practice anyway. Oct 14 09:25:29 i don't doubt it has happened in the past. Oct 14 09:25:53 The ex-KGB guy quoted that when spies pulls out their gun, intelligence ends. Oct 14 09:25:53 Gravity wasn't bad actually Oct 14 09:26:02 but there are some good boooks about the history of the services if you're in for a long flight or something. Oct 14 09:26:09 no need to rely on holywood. Oct 14 09:26:45 Actually it's a little appealing if movies were to be based on more factual stuff rather than sci-fi crap and overexaggeration. Oct 14 09:27:30 Kids these days playing with real pentesting tools used by professionals thinking they are the bomb but have no actual clue let alone the common sense when it comes to proper pentesting. Oct 14 09:27:35 gravity is a lot of action, and of course it's highly unrealistic as far as the entire story goes. but it somehow manages to take you along with it, and manages to rarely be outrageously impossible. Oct 14 09:27:40 so as far as blockbusters go, it is excellent Oct 14 09:28:04 and it has all the magic Oct 14 09:28:17 and a good arrangement, too Oct 14 09:28:21 and acting! Oct 14 09:28:22 Like Harry Potter? Oct 14 09:28:25 idk Oct 14 09:28:27 ._. Oct 14 09:29:08 magic = mood establishment etc pp Oct 14 09:29:28 i'm sure there are filmologists who can explain all the magic. Oct 14 09:29:34 i am not one. Oct 14 09:29:51 Kind of like creative thinking, but in hollywood sense it would be probably more closely related into over-exaggeration of stated facts. Oct 14 09:30:12 Neither am I, I just get annoyed thinking the likes of swordfish. Oct 14 09:30:39 If someone were to tie swordfish with movies about IT security I'd lart them for that. Oct 14 09:31:09 * sono points at oreos "he knows how to lart" Oct 14 09:31:17 so oldschool Oct 14 09:31:42 * sono is young enough to only know the term from reading the jargon file when he was 16. Oct 14 09:32:37 and i'm 37! Oct 14 09:33:04 well i am also contiental so maybe the slang is just different Oct 14 09:33:13 but i don't think i have heard someone use the term in 5 years or so Oct 14 09:40:21 sono: I'm 42, and I'm old enough to have made it into the jargon file ;) (check http://catb.org/jargon/html/meaning-of-hack.html, towards the bottom) Oct 14 09:41:57 * psycho_oreos is 29 but doesn't care about using old terminologies. :p Oct 14 10:45:15 ~lart sono Oct 14 10:45:15 * infobot lowers sono's priority Oct 14 10:46:54 I've fallen into the habit of always doing an ls -lart when I don Oct 14 10:47:23 t have any reason not to Oct 14 11:41:43 hi all Oct 14 11:41:45 can I use backupmenu with microsd ext2? Oct 14 11:41:47 I want to backup my maemo Oct 14 11:44:34 Dunno, just try ;-) Oct 14 11:44:41 Not much can go wrong Oct 14 14:24:08 ~lart sono Oct 14 14:24:08 * infobot acting on orders from an unspecified client drags sono into court suing for $200 million Oct 14 14:24:31 who is sono? Oct 14 14:24:52 [2013-10-14 11:31:42] * sono is young enough to only know the term from reading the jargon file when he was 16. Oct 14 14:25:54 ooh, merlin1991 beaten me to it ;-P Oct 14 15:07:26 Veggen: vegard? =) hey that's new. i didn't even know about it. Oct 14 15:07:30 neat Oct 14 15:08:13 guys meybe I will install the archlinux system on my n900 Oct 14 15:08:26 Veggen: does this mean you lost 5 pidgeons? =D Oct 14 15:08:37 Zammy: I doubt you can do that and have a working device Oct 14 15:08:48 Not if you want to make phone calls, use data, send SMS etc Oct 14 15:08:48 jonwil: nice constraint there. Oct 14 15:09:00 depends on the definition of "working" :) Oct 14 15:09:31 the I mean I am going to install in my next future the ArchLinuxARM-N900-2013.02-rootfs.tar.gz Oct 14 15:09:53 I doubt ArchLinux has working battery charging either Oct 14 15:10:00 jonwil: it has Oct 14 15:10:06 pali's drivers Oct 14 15:10:08 ok Oct 14 15:10:21 afaik :) Oct 14 15:10:29 but I must buy first a microsd to it Oct 14 15:10:30 for it Oct 14 15:11:10 ofc testing before using it definately Oct 14 15:12:09 is it faster right? Oct 14 15:12:58 and I mind to use it mainly for command lines usage... that's pretty cool Oct 14 15:13:30 god reverse engineering hildon status bar gadgets is hard, especially when they use like half of GTK and have a class size of 0x1A0 and a private class size of 0x3C Oct 14 15:14:47 anyway I did not find any video around googling about this archlinux on n900 Oct 14 15:15:49 oh wait the 0x1A0 is misleading since the class is entirely composed of a HDStatusMenuItem structure Oct 14 15:16:25 just some in which you can see the archlinux os for computers emulated like debian img... nothing about this ArchLinuxARM-N900 Oct 14 15:16:58 still, 0x3C of private class data is going to be annoying Oct 14 15:20:45 jonwil: (0xA10 size) - most probably the first member variable is some GObjectClass descendant type Oct 14 15:20:57 *0x1A0 Oct 14 15:20:58 yes it is Oct 14 15:21:04 Its a HDStatusMenuItem Oct 14 15:21:14 :nod: Oct 14 15:21:39 which leaves you woth only a couple of members to find Oct 14 15:21:44 *with Oct 14 15:22:28 jonwil: is ther x86 binary? Oct 14 15:22:33 or only ARM Oct 14 15:22:44 no x86 unfortunately Oct 14 15:22:55 yeah, thats bad Oct 14 15:35:55 anyone know what /dev/ttyGS* might be? Oct 14 15:37:20