| 00:53.15 | dalias | landley, yes musl is mine. |
| 00:53.57 | dalias | the uuterm svn repo may be broken. i need to move it to git... |
| 01:03.02 | *** join/#edev djerome (~djerome@ip24-251-138-141.ph.ph.cox.net) |
| 01:17.33 | *** join/#edev djerome (~djerome@ip24-251-138-141.ph.ph.cox.net) |
| 04:55.42 | *** join/#edev kergoth (~kergoth@ip24-251-167-38.ph.ph.cox.net) |
| 05:00.34 | *** join/#edev NR (~LessIsMor@2.176.236.115) |
| 05:01.20 | *** join/#edev SterNiX (~LessIsMor@unaffiliated/nu253r/x-0655220) |
| 05:02.00 | *** join/#edev uwe_mobile (~uwe@static.88-198-8-117.clients.your-server.de) |
| 05:37.26 | *** join/#edev edboogie2011 (~Edward_Ki@pool-108-41-60-162.nycmny.fios.verizon.net) |
| 05:37.32 | edboogie2011 | hey guys. does anybody know the correct name for the plastic or felt cushion usually found on the foot of an electronic enclosure? usually they are backed with a self adhesive attachment |
| 05:37.43 | edboogie2011 | anybody know what these darn things are called? |
| 05:42.32 | *** part/#edev edboogie2011 (~Edward_Ki@pool-108-41-60-162.nycmny.fios.verizon.net) |
| 06:07.31 | *** join/#edev JoeLlama (~snork@unaffiliated/joellama) |
| 06:10.19 | *** join/#edev gvandenbosch (~gvandenbo@86.81.67.223) |
| 06:53.46 | *** join/#edev avinashhm (~avinash-h@85.13.70.251) |
| 07:22.31 | *** join/#edev thannoy (~aloiseau@231.39.0.109.rev.sfr.net) |
| 07:25.54 | *** join/#edev florian (~fuchs@Maemo/community/contributor/florian) |
| 09:47.03 | *** join/#edev prpplague (~prpplague@85.13.70.251) |
| 09:47.30 | prpplague | ho ho ho |
| 09:47.35 | prpplague | merry freakin thursday |
| 09:48.38 | wmat | prpplague: !! |
| 09:49.00 | wmat | prpplague: how was the party? |
| 09:49.31 | prpplague | wmat: it was pretty decent |
| 09:49.38 | prpplague | wmat: a little bit unorganized |
| 09:49.56 | prpplague | wmat: got to introduce myself to linus and shake his hand |
| 09:50.00 | prpplague | wmat: no pic :( |
| 09:50.33 | wmat | prpplague: cool |
| 09:51.03 | prpplague | wmat: trying to finish up getting things ready for the hands on class |
| 09:51.05 | wmat | prpplague: i've been battling some kind of Montezuma's revenge, so decided not to stray far from my hotel room's bathroom ;) |
| 09:51.16 | prpplague | wmat: ahh |
| 09:51.30 | wmat | prpplague: have fun with your workshop |
| 09:51.41 | prpplague | not sure if fun is involved |
| 09:53.32 | wmat | prpplague: i can't remember if you have Flyswatters for the prizes as well? I know there's 5 BeaconBoards though. |
| 09:54.20 | prpplague | wmat: yea, 5 flyswatter2's |
| 09:54.29 | prpplague | wmat: but you can't have them until tomorrow |
| 09:54.58 | wmat | prpplague: cool, don't need them until the closing ceremony |
| 09:55.35 | prpplague | ahh |
| 09:55.54 | prpplague | wmat: hehe, well we are giving away 300 beacon boards at the closing ceremony |
| 09:56.13 | prpplague | wmat: so giving them away for the prizes might not be exactly interesting |
| 09:56.55 | wmat | ah, ok, we'll just go with the Flyswatters then |
| 09:57.54 | wmat | nifty case btw |
| 09:58.02 | prpplague | wmat: dandy! |
| 09:58.39 | prpplague | you been to the mall area? |
| 09:58.59 | wmat | prpplague: I walked through it last night between trips to the bathroom ;) |
| 09:59.18 | wmat | prpplague: i have to find gifts for my kids and wife at some point |
| 10:00.14 | prpplague | wmat: same here |
| 10:00.36 | prpplague | wmat: but i also need to find some bags or boxes to put the flyswatter2's in since i didn't bring any |
| 10:01.11 | wmat | prpplague: ask the concierge to ge tthem for you ;) |
| 10:01.36 | prpplague | wmat: hehe good idea |
| 10:59.21 | *** join/#edev gustavoz (~gustavoz@host8.186-125-80.telecom.net.ar) |
| 11:15.37 | *** join/#edev florian (~fuchs@Maemo/community/contributor/florian) |
| 11:47.49 | *** join/#edev JoeLlama (~snork@unaffiliated/joellama) |
| 12:05.02 | *** join/#edev bsdfox (~h36sa@c-67-172-188-250.hsd1.ca.comcast.net) |
| 12:05.02 | *** join/#edev bsdfox (~h36sa@unaffiliated/bsdfox) |
| 12:42.23 | *** join/#edev SterNiX (~LessIsMor@unaffiliated/nu253r/x-0655220) |
| 12:44.17 | *** join/#edev dfarnsworth (~dfarnswor@nat/montavista/x-excymximdaeoeoif) |
| 13:06.09 | *** join/#edev dijenerate (~dijenerat@173.225.251.175) |
| 13:30.54 | *** join/#edev file (crash@conference/astricon/x-ulgemuugbpbgpznw) |
| 13:59.43 | *** join/#edev GPSFan (~kenm@64.92.145.112) |
| 13:59.43 | *** mode/#edev [+o GPSFan] by ChanServ |
| 14:07.44 | *** join/#edev bsdfox (~h36sa@c-67-172-188-250.hsd1.ca.comcast.net) |
| 14:07.44 | *** join/#edev bsdfox (~h36sa@unaffiliated/bsdfox) |
| 14:22.12 | *** join/#edev risca (~risca@f-static-78-70-87-29.business.telia.com) |
| 14:43.40 | *** join/#edev CosmicPenguin (~nobody@soa.codeaurora.org) |
| 14:43.41 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 14:58.15 | *** join/#edev file[rawr] (crash@conference/astricon/x-txbesejajbapnlrh) |
| 15:07.12 | *** join/#edev jayabharath (~jayabhara@nat/ti/x-atavneztzcjzujss) |
| 15:08.21 | *** join/#edev sndcrb (~sndcrb@dsl-hkibrasgw2-fef8de00-126.dhcp.inet.fi) |
| 15:31.56 | *** join/#edev avinashhm (~avinash-h@85.13.70.251) |
| 15:48.22 | *** join/#edev ZeZu (null@c-98-227-57-88.hsd1.in.comcast.net) |
| 16:07.34 | *** join/#edev purl (~purl@rikers.org) |
| 16:07.34 | *** topic/#edev is Discussion of embedded systems (Hardware/Software/Linux/WinCE/Commercial Device Repurposing) : Busybox and Buildroot : http://www.elinux.org |
| 16:22.00 | CosmicPenguin | ELC starts on Valentines day next year - thats problematic |
| 16:37.20 | calculus | wmat: maybe your wife and kids would like some flyswatters and beacon boards ;) |
| 16:45.38 | *** join/#edev file[rawr] (crash@conference/astricon/x-pfnysqogsadiikwl) |
| 17:24.25 | landley | dalias: somebody's been poking at aboriginal's arm system images to port musl to arm, apparently... |
| 17:24.35 | wmat | calculus: that would certainly make things easier for me |
| 17:31.18 | *** join/#edev file (~file@asterisk/developer-and-muffin-lover/file) |
| 17:40.45 | *** join/#edev kergoth (~kergoth@ip24-251-167-38.ph.ph.cox.net) |
| 18:21.27 | *** join/#edev JoeLlama (~snork@cpe-50-113-132-35.socal.res.rr.com) |
| 18:21.29 | *** join/#edev JoeLlama (~snork@unaffiliated/joellama) |
| 18:40.49 | *** join/#edev flyback- (~flyback@c-98-219-134-156.hsd1.pa.comcast.net) |
| 19:02.15 | *** join/#edev araujo (~araujo@190.73.44.29) |
| 19:02.15 | *** join/#edev araujo (~araujo@gentoo/developer/araujo) |
| 19:26.08 | CosmicPenguin | Okay, dumb linux question. it seems to me if I want to do a wait with an optional timeout, I have to do an if (timeout == 0) wait_event() else wait_event_timeout() |
| 19:26.37 | CosmicPenguin | Which double sucks because wait_event() returns 0 on success, and wait_event_timeout() returns the number of jiffies remaining on the timeout - is there a less dumb way to combine the two? |
| 19:55.42 | *** join/#edev CosmicPenguin (~nobody@soa.codeaurora.org) |
| 19:55.42 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 20:17.08 | landley | CosmicPenguin: if you want the timeout to be optional you need to make a wrapper, yes. It's about two lines. |
| 20:17.44 | landley | The kernel guys are big on efficiency. It's a thing. Each call is generally going to be one or the other. |
| 20:18.07 | CosmicPenguin | Done wrappered |
| 20:18.32 | CosmicPenguin | just making sure I wasn't being stupid and there was a magical unicorn that pooped rainbows |
| 20:23.04 | CosmicPenguin | There is a school of thought that said that the timeout shouldn't be optional, but hey, if you want to wait_event_interruptible for ever, who am I to stand in your way? |
| 20:31.15 | landley | And now I have to cope with armv8. |
| 20:31.25 | landley | Sigh. Gonna have to upgrade the toolchain... |
| 20:31.33 | CosmicPenguin | 7 is so 2011 |
| 20:31.58 | landley | It's still 2011. |
| 20:32.06 | CosmicPenguin | Thats part of the joke |
| 20:32.19 | landley | (It's actually thumb2 I really miss.) |
| 20:32.29 | landley | You can build a usable kernel with thumb2. Can't with thumb1. |
| 20:35.32 | landley | http://www.marketwatch.com/story/appliedmicro-showcases-worlds-first-64-bit-arm-v8-core-2011-10-27?sf2435627=1 |
| 20:36.28 | landley | http://www.arm.com/about/newsroom/arm-discloses-technical-details-of-the-next-version-of-the-arm-architecture.php |
| 20:43.59 | roxfan | cool stuff |
| 20:58.29 | SpeedEvil | wonders when the first phone with >4G RAM will hit. |
| 20:59.30 | MonMotha | I'd say there's no need for it, but at the rate of bloat in software, probably soon |
| 21:01.42 | SpeedEvil | Indeed. |
| 21:02.16 | roxfan | armv7 can do >4gb already with PAE |
| 21:02.19 | SpeedEvil | It would be interesting to plot CPU time vs indirection level. |
| 21:02.35 | SpeedEvil | So hand-written ASM code would be 1. |
| 21:02.43 | SpeedEvil | C might be 3 |
| 21:02.56 | SpeedEvil | stdlib - 4 |
| 21:02.58 | SpeedEvil | ... |
| 21:03.12 | SpeedEvil | And z80 code running on a javascript emulator about 14 |
| 21:07.25 | MonMotha | SpeedEvil: so if I'm running something written in BASIC for a TRS-80 on an emulator written in Javascript that's running in my web browser on .NET/Mono on Linux, what level would that be? |
| 21:07.30 | MonMotha | NaN? |
| 21:08.03 | MonMotha | or maybe COBOL running on an S/360 emulator (which is actually quite common) |
| 21:10.53 | *** join/#edev ZeZu (null@c-98-227-57-88.hsd1.in.comcast.net) |
| 21:11.32 | SpeedEvil | :) |
| 21:11.58 | SpeedEvil | The above plot over time, I mean. |
| 21:12.09 | SpeedEvil | The levels of indirection hve gone way up over time. |
| 21:12.52 | MonMotha | definitely |
| 21:22.53 | landley | roxfan: and DOS could do more than 640k with extended/expanded/extruded memory. |
| 21:22.59 | landley | The 8080 could do > 64k with bank switching. |
| 21:23.30 | landley | SpeedEvil: the invention of the MMU was probably the big jump. The rest is details. |
| 21:24.52 | landley | MonMotha: you'll probably want to throw http://bellard.org/jslinux/ in there. |
| 21:25.35 | SpeedEvil | Bellard is one of the people who I wish I could be as awesome as. |
| 21:26.03 | landley | SpeedEvil: someday, I hope to set up another combinatoin linux expo and science fiction convention (ala penguicon and linucon) and invite him as a tech-side guest. |
| 21:26.46 | SpeedEvil | Him and (for other reasons) Jeff Bezos. |
| 21:26.58 | MonMotha | landley: that's...terribly awesome |
| 21:27.14 | SpeedEvil | He's just got so many awesome projects. |
| 21:27.37 | SpeedEvil | From tccboot, to digital-video out of a VGA card. |
| 21:28.35 | SpeedEvil | (Bezos - for spacex) |
| 21:28.42 | MonMotha | oh, VGA cards are really useful as essentially just high speed DACs with a few timing limitations |
| 21:28.53 | MonMotha | I've seen people dump ATSC and DVB out of them at RF, for example |
| 21:29.03 | SpeedEvil | Developing a human-capable rocket on around a tenth of what NASA would charge. |
| 21:30.06 | SpeedEvil | I noted elsewhere that for the price of the proposed NASA rocket - 30 billion in the next decade for 2 launches, - you can launch 40000 tons into orbit on the upcoming SpaceX rocket. |
| 21:32.05 | SpeedEvil | Oops - Elon Musk. |
| 21:32.08 | SpeedEvil | is tired. |
| 21:34.39 | landley | MonMotha: Bezos' space stuff is fun but John Karmack (armadillo aerospace) did it first. |
| 21:34.52 | landley | Nah, it's Robert Young (Red Hat, then Lulu) who'd be the real catch... |
| 21:35.20 | landley | SpeedEvil: developing a human-capable rocket on 1/10th of what nasa would charge is not a serious trick, actually. |
| 21:35.35 | SpeedEvil | It depends. |
| 21:35.40 | SpeedEvil | Technically, no. |
| 21:36.12 | SpeedEvil | Getting the money together, and actually doing it, in the face of 'but it's expensive because NASA says it's expensive, so it must be expensive' is harder. |
| 21:36.37 | SpeedEvil | NASA got poisoned by apollo. |
| 21:36.53 | SpeedEvil | 'waste anything but time' is something you never, ever want to get into a beaurocracies heart. |
| 21:37.35 | landley | No, the space shuttle project specifically was a giant boondoggle: http://blogs.discovermagazine.com/80beats/2011/07/06/why-the-space-shuttle-aka-the-flying-brickyard-deserves-to-be-retired/ |
| 21:37.41 | SpeedEvil | I know. |
| 21:37.49 | SpeedEvil | I've been arguing that for more than a decade. |
| 21:38.09 | landley | And don't get me started about Space Station Vodka. (Hijack nasa's budget to prevent the Russian nuclear/ICBM engineers from selling their skills to iran by feeding them make-work...) |
| 21:38.12 | SpeedEvil | NASA has in the past been fundamentally incapable of reducing the cost of space. |
| 21:38.31 | SpeedEvil | Their lat serious attempt - X33/venturestar - used three seperate untried technologies on one vehicle. |
| 21:38.53 | SpeedEvil | It was perhaps a sane attempt at reducing the cost of space launch - in four generations time. |
| 21:40.00 | landley | Nasa is a giant bureaucracy. Kennedy pushed through a mongolian hordes approach the same way D-Day worked in World War II, but with Kennedy dead Johnson and Nixon just let it slide, carter had an oil embargo and recession to deal with, and then Regan was a wind-up toy with about 3 coherent thoughts and a bunch of cue cards (before the altzheimers that made the iran contra "I don't remember" thing much less of a lie than we thought at the time). |
| 21:40.31 | SpeedEvil | Indeed - it needs a coherent mission and leadership. |
| 21:40.33 | landley | Bush I put Dan Quayle in charge of NASA. |
| 21:40.42 | SpeedEvil | But it's been mostly welfare. |
| 21:40.45 | SpeedEvil | Rather than space. |
| 21:40.46 | landley | SpeedEvil: it needs to be flushed and restarted from scratch. |
| 21:41.18 | SpeedEvil | If you're asked to build rockets by NASA, you'd much rather build 2 rockets, rather than 100 rockets for the same money. |
| 21:45.03 | landley | SpeedEvil: I think the most damning indictment of Nasa has to be http://xkcd.com/893/ |
| 21:46.15 | SpeedEvil | Indeed. |
| 21:46.49 | *** join/#edev Matth_FA (~Matth_FA@mtl93-14-78-241-192-235.fbx.proxad.net) |
| 21:47.08 | SpeedEvil | It basically needs to be focussed on how to do space. Not how to act as welfare for boeing/locmart |
| 21:51.34 | landley | moblin, maemo, limo, meego, yocto, tizen... I'M CONFUSED... |
| 21:51.58 | SpeedEvil | You missed mer, and nemo |
| 21:52.20 | landley | Did the Linux Foundation officially endorse those too? |
| 21:52.27 | SpeedEvil | No. |
| 21:52.44 | SpeedEvil | Have all those been endorsed by LF? |
| 21:53.20 | landley | Moblin and maemo merged to form meego, which was LF official. Limo was a LF project. Yocto is a LF project. Tizen is a LF project. |
| 21:53.36 | landley | Pretty much each of them is "the official embedded distro we want you to all switch to". |
| 21:54.25 | landley | I'm not sure anymore when I'm mocking the linux foundation and when I'm explicitly stating their official position. |
| 21:54.27 | MonMotha | are there any "embedded distros" that are actually really maintained and simultaneously suitable for embedded use? |
| 21:54.42 | landley | Depends on your definition of "suitable". |
| 21:55.05 | MonMotha | last I checked, you could use Montavista, which is what all the vendors seem to use, but it's ancient; you could use Angstrom if you feel like messing with OE, and it's somewhat up to date, or you could just install Debian, which is hardly embedded |
| 21:55.22 | landley | I've gotten sucked into the funtoo core team. drobbins is rewriting portage in bash, and I'm bootstrapping it under my aboriginal linux project. |
| 21:55.27 | SpeedEvil | Slackware! |
| 21:55.34 | landley | montavista got eaten by somebody, I forget who. |
| 21:55.43 | MonMotha | I guess my definition of "suitable" is "highly modular" and having "minimal runtime overhead" |
| 21:55.52 | landley | Lots of people are using arch these days, I need to look at pacman. But that's pretty fragmented once you get off x86... |
| 21:55.54 | MonMotha | i.e some systems need X or similar; some don't |
| 21:55.55 | roxfan | windriver iirc |
| 21:56.06 | *** join/#edev ZeZu (null@c-98-227-57-88.hsd1.in.comcast.net) |
| 21:56.16 | landley | MonMotha: I built x.org from source following the blfs instructions, and added fvwm and rxvt. |
| 21:56.31 | landley | roxfan: no, montavista. |
| 21:56.46 | roxfan | i mean windriver ate them |
| 21:56.50 | MonMotha | I just remember messing with buildroot years ago, getting totally frustrated that X didn't work (since I did need it) and moving on to OE and tearing my hair out over that |
| 21:56.58 | landley | http://mvista.com/Cavium/Cavium-acquires-MontaVista.php |
| 21:57.07 | roxfan | oh hm |
| 21:57.25 | MonMotha | every time I see windriver, I think of the old Jungo "WinDriver" that was some awful attempt at USB to Userspace access |
| 21:57.32 | landley | MonMotha: http://landley.net/aboriginal/control-images |
| 21:57.43 | landley | Want me to do an x11 bootstrap on top of linux from scratch? |
| 21:57.52 | landley | I can probably be done this weekend. |
| 21:58.01 | landley | Well, xvnc anyway. |
| 21:59.32 | MonMotha | landley: this was years ago. I'm sure things are quite different, now. |
| 21:59.43 | MonMotha | unfortunately I've been kinda out of the loop during that time in terms of "real" embedded Linux |
| 22:00.01 | MonMotha | all my embedded work has either been too small for Linux or really "just a PC" e.g. Atom SBC |
| 22:00.43 | *** join/#edev SterNiX (~LessIsMor@unaffiliated/nu253r/x-0655220) |
| 22:01.14 | landley | Yeah, the "crippled linux" market's kinda been sucked up by Android recently. *shrug* |
| 22:01.35 | landley | I can wait that out, I expect. They're already failing to scale their integration efforts. |
| 22:01.53 | SpeedEvil | And android has the huge problem of google. |
| 22:01.59 | landley | It's hilarious to watch Google fail to understand open source development. They're reproducing the NT development cycle.... |
| 22:02.18 | SpeedEvil | As in you can't get the shiny user-required google binaries unless you bend over _just_ right. |
| 22:02.58 | landley | Google's problem is slightly more complicated (and way deeper) than that. Steve Yegge's google+ post was _one_ tip of that iceberg... |
| 22:03.02 | MonMotha | landley: what gets me is how woefully out of date the actual Linux derived (kernel + minimal userspace) parts of Android seem to be |
| 22:03.18 | MonMotha | a lot has changed since, what are they still on, 2.6.23? |
| 22:03.36 | SpeedEvil | And yes - that's not the only problem. |
| 22:04.06 | MonMotha | actually, I guess my phone claims 2.6.35.7, it's quite hacked, so who knows how "official" that is |
| 22:04.29 | landley | MonMotha: http://linuxmafia.com/faq/Licensing_and_Law/forking.html |
| 22:04.36 | landley | Re-read it and ponder Google's actions in light of it. |
| 22:05.01 | landley | (Also note that the reason distributed source control ala git is awesome is it lets people marshall patch _sets_ between trees, so integration of short-term forks scales.) |
| 22:05.26 | landley | Also watch the time-based releases video: http://video.google.com/videoplay?docid=-5503858974016723264 |
| 22:05.47 | MonMotha | landley: still waiting on that first link... |
| 22:05.55 | MonMotha | there we go |
| 22:06.00 | landley | One advantage of time based releases: forced resync points so you don't regularly have to merge trees that diverged more than a year ago. |
| 22:10.47 | MonMotha | correct me if I'm wrong, but hasn't mainline Linux picked up several of the "improvements" from the Android fork that people generally liked? |
| 22:11.02 | MonMotha | they're just not merging the ugly hackish warts that Google introduced to make Android work |
| 22:11.44 | landley | MonMotha: the linux development community can loot android for ideas. |
| 22:11.46 | landley | And has. |
| 22:12.17 | landley | Android has more of a problem backporting Linux stuff because Linux moves faster than android. |
| 22:12.43 | landley | (Linux has quarterly releases with millions of lines of new code in each: android's lucky to get out a new version a year and has its hands full integrating driver code from TI and such.) |
| 22:12.44 | MonMotha | right |
| 22:12.51 | MonMotha | that's basically the conclusion of the first section of that article |
| 22:13.05 | landley | What android has is windows-style distribution agreements. It comes preinstalled on millions of units of hardware every quarter. |
| 22:13.28 | SpeedEvil | I wonder. |
| 22:13.37 | SpeedEvil | How much of that is 'bad device vendor' |
| 22:13.40 | landley | But most of those don't ship an actual google release, they ship extensively vendor modified snapshot of some random version skew. |
| 22:13.52 | SpeedEvil | And how much is that google won't approve the new version on that hardware. |
| 22:14.00 | landley | SpeedEvil: a lot, but Android is giving them an easy out so they never have to face tough decisions. |
| 22:14.14 | landley | Google's going "You want to be stupid? Here!" |
| 22:14.19 | SpeedEvil | Because if you have a phone coming out, and google won't grant you the permission to use the latest release - what can you do? |
| 22:14.41 | landley | Enabling bad behavior for short-term advantage, that's almost certainly going to blow up horribly long-term. |
| 22:14.43 | SpeedEvil | Or if you want to upgrade devices in the field |
| 22:15.07 | landley | I gotta wonder home... |
| 22:15.19 | SpeedEvil | waves. |
| 22:16.03 | MonMotha | the extensive vendor (and especially carrier) modifications just seem to annoy most android users, too |
| 22:16.09 | MonMotha | I kinda wonder why they all persist |
| 22:16.28 | MonMotha | the carriers I can understand to some degree. They are addicted to control and lock-in |
| 22:16.36 | MonMotha | the vendor stuff though I don't really get |
| 22:17.40 | SpeedEvil | Because managment doesn't actually use phones. |
| 22:17.53 | SpeedEvil | 'we want our phones to be better!' |
| 22:17.59 | SpeedEvil | 'put some shiny shit on it'. |
| 22:18.31 | MonMotha | can it really all be attributed to incompetent management/marketing? |
| 22:18.56 | SpeedEvil | I think assuming competance from vendors is going a bit far. |
| 22:19.16 | MonMotha | well, I'll admit that I frequently see rather good hardware hobbled by awful software |
| 22:19.25 | MonMotha | not just in the android world |
| 22:19.29 | SpeedEvil | Way too often. |
| 22:19.37 | SpeedEvil | looks at the device in his lap. |
| 22:19.55 | MonMotha | (I say this as a hardware engineer) |
| 22:20.03 | MonMotha | ((who also does software)) |
| 22:21.19 | SpeedEvil | N9/N950. Maemo 6/harmattan. The security framework is in principle interesting. It has hooks into various bits of the kernel, and if an app tries to do something it's not authorised to do, it fails. The security tokens are generated automatically at compile time, and can be easily reviewed. |
| 22:21.22 | SpeedEvil | But. |
| 22:21.48 | SpeedEvil | You for example can never ever get a security token to tell the device daemon to flash the indicator light. |
| 22:22.03 | SpeedEvil | And you don't actually need a token at all to open the camera, and start streaming to china. |
| 22:22.39 | SpeedEvil | In principle, the subsystem could be really useful for a community repo. |
| 22:22.47 | SpeedEvil | As it can vastly simplify review. |