**** BEGIN LOGGING AT Sat Sep 15 02:59:57 2007 Sep 15 04:39:49 * rwhitby tries a virgin ipkg upgrade to see if gsmd restart in matchbox/session is still required Sep 15 04:43:59 nope, still required Sep 15 04:44:42 * rwhitby ipkg upgrades the bv3 to match Sep 15 05:05:13 rwhitby: do you have a pin# on your SIM? Sep 15 05:05:19 no Sep 15 05:10:26 * rwhitby wants to work out how to get the neo to connect to a bluetooth dund server via ppp Sep 15 05:11:18 hmm - or maybe I just work out how to get bluetooth and pand running on my freecom fsg-3 router ... Sep 15 05:39:39 * rwhitby sets up dbus from scratch on a consumer router ... Sep 15 05:42:06 what's the URL for the wiki page with all the rediculous rumors/ Sep 15 06:08:05 mrcucumber: slashdot? ;-) Sep 15 06:25:28 * rwhitby wonders why depmod -a is not working, and then realises he forgot to flash the .22 kernel on the bv3 Sep 15 06:26:52 moin Sep 15 06:28:42 good morning Sep 15 06:31:22 Question: Can the OpenMoko currently function as a modem? Sep 15 06:31:23 Selenolycus: See http://wiki.openmoko.org/wiki/FAQ (and also SH1, SH1_FAQ, Neo1973, Developer_preview and Neo1973_Hardware#GTA01Bv4 pages) (Hopefully these links answers your question. This autotrigger for new nick and question during first 15 minutes done only once.) Sep 15 06:36:04 Or, rather, will it be possible that it will ever? Sep 15 06:36:54 Selenolycus: there have been recent u-boot checkins which allow gsm passthrough Sep 15 06:37:44 Nice. Sep 15 06:40:56 and of course there will be BT<->GSM (GTA01) and wifi<->GSM (GTA02) gateways eventually Sep 15 06:41:31 I don't see any mention anywhere of the processor the 1973 uses Sep 15 06:44:40 Selenolycus: the wiki links first given to you by the bot above definitely answer that one Sep 15 07:13:42 SCO has filed for Chapter 11 bankruptcy protection http://biz.yahoo.com/ap/070914/sco_group_bankruptcy.html?.v=4 Sep 15 07:16:00 They finally do something rational! Sep 15 07:19:12 In other news, due to a massive shortages chairs have doubled in price in Redmond Sep 15 07:19:38 hehe Sep 15 07:19:59 I'll buy the OpenServer rights... so I can do /something/ to it.. -_- Sep 15 07:20:02 s/a// Sep 15 07:20:02 Linux_Galore meant: In other news, due to massive shortages chairs have doubled in price in Redmond Sep 15 07:26:09 cesarb, mwester: by "connector" I meant the stuff beneath the plastic inside the jack, and yes, I got the events, not just the mike working Sep 15 07:45:51 is the mailinglist down? Sep 15 07:46:54 edistar: apparently, yes. Sep 15 07:47:15 polz: I was wondering why I didn't get any openmoko mails anymore. Sep 15 07:48:16 edistar: I've only just woken up, so I haven't noticed that yet :) Sep 15 07:48:27 hadn't, sorry Sep 15 08:29:47 hi Sep 15 09:06:26 * bedboi is away: sleeping Sep 15 09:37:40 <_Jan_> hello, are anybody here who can talk german? Sep 15 09:37:41 _Jan_: See http://wiki.openmoko.org/wiki/FAQ (and also SH1, SH1_FAQ, Neo1973, Developer_preview and Neo1973_Hardware#GTA01Bv4 pages) (Hopefully these links answers your question. This autotrigger for new nick and question during first 15 minutes done only once.) Sep 15 09:44:50 <_Jan_> Hello, anybody talks german? Sep 15 09:45:52 _Jan_: we are here: #neo1973-germany Sep 15 09:46:04 <_Jan_> Thanks Sep 15 09:46:28 <_Jan_> Which Server? Sep 15 09:46:40 freenode.net Sep 15 09:47:14 <_Jan_> Can't find you :( Sep 15 09:47:39 try irc.freenode.net Sep 15 09:47:50 <_Jan_> I have Sep 15 09:49:07 <_Jan_> Found you Sep 15 10:03:33 why is there still a screenshot of 2007.1 on the front page of the wiki? Sep 15 10:10:16 I think they are also on the openmoko.com Sep 15 10:10:29 maybe because gta01 comes with 2007.1? Sep 15 10:13:17 Maybe because nobodies updated it. Sep 15 10:18:20 btw, was the dialer in 2007.1 any better than it is in .2 ? Sep 15 10:19:37 no Sep 15 10:19:40 I've only used it a couple of times Sep 15 10:19:45 worked for me Sep 15 10:21:46 I want a neo now... :( but 300$ is a bit expensive if a new improved version comes out in december.. Sep 15 10:23:12 edistar: good things come to those who wait Sep 15 10:23:21 and pay $450 Sep 15 10:26:37 and still not be able to see what's on the screen if you try to read it in the sun :> Sep 15 10:36:14 almost watchable full screen video with mplayer ;-) Sep 15 10:36:43 (if you re-code it to 10fps and turn everything else off...) Sep 15 10:37:21 QVGA mode on the lcm works Sep 15 10:37:32 It plays most VCD sized stuff just fine Sep 15 10:37:42 Though it's not properly setup Sep 15 10:37:49 speedevil: really? how do you set it (fbset doesnt offer..) Sep 15 10:37:56 I don't know Sep 15 10:38:08 nobodies written up their attempts Sep 15 10:38:25 the other problem is that there is no nice optimised rotator Sep 15 10:38:28 when I checked (~3 weeks ago) stefan said they hadnt implemented it in the driver yet... Sep 15 10:38:41 yeah, I'm rotating in mencoder Sep 15 10:43:34 doesn't the chipset support rotating ? Sep 15 10:43:38 No. Sep 15 10:43:50 but it can scale, right ? Sep 15 10:43:57 ˆNo. Sep 15 10:44:02 the display does Sep 15 10:44:12 oh, OK. Sep 15 11:00:20 I keep getting the error: NOTE: package eds-dbus-1.4.0+svnnow-r6: task do_fetch: failed Sep 15 11:00:41 what do do? it seems to just not be able to download it Sep 15 11:02:51 edistar: what does the svn command look like? i had problems with it always trying to svn co -r 1 ... Sep 15 12:21:27 jpcass: NOTE: Fetch svn://svn.o-hand.com/repos/eds-dbus;module=trunk;proto=http Sep 15 12:21:40 re Sep 15 12:23:06 jpcass: HTTP Anforderung gesendet, warte auf Antwort... 404 Not Found Sep 15 12:24:21 Well, Ijust would like to know, whether openmoko and the neo will support the bluetooth sim access profile to connect to a standard car mobile phone equuipment? Sep 15 12:27:31 matsch: this is unlikely until later models - the SIM is attached directly to the GSM modem chipset and support from this profile will require remoting the sim through the interface to the modem chipset. It will eventually happen, but not likely for a couple of years. Sep 15 12:27:48 matsch: this not official - just IMHO Sep 15 12:28:47 matsch: we also want to remote the entire GSM layer to support UMA Sep 15 12:31:07 I am not familiar with those details, I am afraid, but from the practical point of usage I would really like such a bluethooth connection to work earlier than just in a couple of years Sep 15 12:32:54 It's probably not going to happen in the next model after the december available one Sep 15 12:32:58 matsch: everyone would like to see new stuff appear in products immediately. However, the time from specifications to market is typically 2 to 5 years Sep 15 12:33:24 simply because the GSM chipset is basically a closed 'blob' Sep 15 12:33:54 Making it other than a closed blob is very complex. Sep 15 12:33:55 even if you own it all, the design process for a handset is normally 2 years from start to market Sep 15 12:34:12 and having the bluetooth and SIM setup work together pretty much implies that Sep 15 12:35:30 at least it fits in the general architecture of a GSM7.10 mux based device, so I'm pretty sure it will appear relatively soon. However, it will still take time get it into actual product. Sep 15 12:36:32 It's possible it can be done with a software bridge through the TI chipset? or are you meaning in principle. Sep 15 12:36:48 (without further mods to the software by TI) Sep 15 12:36:56 need TI software Sep 15 12:37:04 but I think FIC has access to that Sep 15 12:37:20 whether its doable depends on internal software architecture. Sep 15 12:37:28 K. Sep 15 12:37:35 in principle its properly layered to do it. Sep 15 12:37:44 but of course the usual race conditions will appear Sep 15 12:38:08 and the bandwidth over the gsm7.10 probably has to be raised Sep 15 12:38:17 which is probably why motorola went USB Sep 15 12:38:30 personally, i think you need better high speed uarts Sep 15 12:38:42 and handle both BT and GSM with serial Sep 15 12:38:57 yeah. Sep 15 12:38:58 due to various unneeded USB complexity suckage Sep 15 12:39:31 BT spec is the right way to do it - protocol that can be transported over HS serial or USB Sep 15 12:39:48 and its already in silicon in high volume for BT Sep 15 12:40:23 at least on ARM7 devices Sep 15 12:40:57 like a $2 CSR chip Sep 15 12:56:41 mickey|bbiab, are you back? Sep 15 13:44:43 is it easy to copy stuff between smart phones (from/to friends phones)? Sep 15 13:47:45 aloril: I'd guess so, since if BT hasn't been locked out by the carrier, you can send things easily via BT OBEX even from a normal phone Sep 15 13:48:29 hmm.. so there is already standard, anybody have practical examples? Sep 15 13:48:39 s/example/experience/ Sep 15 13:48:41 aloril meant: hmm.. so there is already standard, anybody have practical experiences? Sep 15 13:49:20 I must admit thought come from this command at /.: "The iPod touch is the first iPod to have wireless capability, although it doesn't use it to synch with a computer or friends' iPods." Sep 15 13:50:01 aloril: I have been sent background pictures for my phone by a friend, for instance; I also often used to send pictures from my razr to my palm (where I can store it on a SD card), and palm's OS is the same (almost) as one found on a treo Sep 15 13:50:03 what _does_ it use it for? Sep 15 13:50:06 friends want to share images they have snapped, music they listen to, videos they look at, etc.. Sep 15 13:50:07 aloril: both via bluetooth Sep 15 13:50:20 "Instead, you can use it to browse the Web, watch YouTube videos or download music from a WiFi-specific iTunes Music Store. " Sep 15 13:50:24 aloril: ah, I've done that too (sharing images I've snapped) Sep 15 13:50:41 aloril: this all is why BT was such a must that FIC delayed GTA01 to add BT Sep 15 13:50:45 and its easy as phones are configured? Sep 15 13:51:03 cesarb: But, we can't use the BT for a headset? Sep 15 13:51:58 aloril: very easy. on my razr, I press menu on a picture, chose a menu option badly placed way down on the menu ("Copy"), chose the destination (or the last option, to scan for nearby BT devices), and the receiver just has to accept Sep 15 13:52:14 aloril: on the palm, it's even easier, since the menu option is much less badly placed Sep 15 13:52:36 sagaci2: we can, but I've never used that (I don't have a headset) Sep 15 13:52:44 hmm.. ok, then no extra potential for Neo1973 here (except for using stuff) Sep 15 13:53:36 aloril: the extra potential would be about the same one would have on a smartphone (not a normal phone), where you can (at least on palm) create new bt-using applications (like a palm chat one) Sep 15 13:53:45 cesarb: We've tried with some headsets. Somebody said the driver's not compiled with the option for headsets (something different perhaps). I haven't looked into it myself, just tried to use it. Sep 15 13:54:11 hmmm.. p2p school wide network ;-) Sep 15 13:54:33 Somebody updated the wiki page, I'll give it a go again. Sep 15 13:54:38 aloril: heh, not school wide, but you could share with people nearby... (BT range is very short) Sep 15 13:55:04 yes, but if crowd is enough big, you could forward packets Sep 15 13:56:06 bt only has 3-bit addressing iirc Sep 15 13:56:22 aloril: perhaps. I know BT can do it, but I've never seen it being used to connect to more than one device at a time; besides, it's slow (and would get even slower with forwarding), and you can only connect to a few people at a time (about 5 IIRC) Sep 15 13:56:54 could still be used for free IM Sep 15 13:57:11 aloril: BT was never intended as a general-purpose networking protocol, after all... It's mostly for interconecting your gadgets and interacting directly with other people's nearby devices (the sending pictures scenario) Sep 15 13:57:44 aloril: yeah, it could (hampered by the lack of a standard IM-over-bluetooth scenario, unlike file transfer, which uses IR's OBEX) Sep 15 13:58:37 * cesarb has seen one of those picture-printing kiosks with sockets for about every existing memory card format, CD, USB... just for fun, I did a bluetooth scan, and there it was... Sep 15 14:06:56 what about PAN, isn't that an 1:n network? Sep 15 14:07:41 (i often wonder why companies put wlan in such small devices, where bt would be sufficient...) Sep 15 14:10:20 good morning Sep 15 14:14:43 no big joy on the BT headset. It connected, then later disconnected. Never got the audio working through it. :( Sep 15 14:14:47 stefan|hospital: alphaone: Shoragan: oh, anything important? Sep 15 14:16:20 zecke: He broke his hand boarding... Sep 15 14:16:34 Responses will be slow due to one-hand-typing Sep 15 14:17:11 hmm Sep 15 14:17:19 zecke: I broke my right arm on thursday. Snowboarding Sep 15 14:17:19 zecke: And i'm right-handed :( Sep 15 14:17:51 where do you have snow? Sep 15 14:17:52 oh, get well soon Sep 15 14:17:56 emdete: For some reason, the PAN BT profile is seldom supported by devices. Sep 15 14:18:18 emdete: Even my swiss army knife N95 doesn't. Sep 15 14:18:46 zecke: Thanks, still have some hope for OEDEM Sep 15 14:18:50 DukeOfURL: snow-dome.de Sep 15 14:20:49 Lynet: but linux, as usual, already do support it ;-) Sep 15 14:22:04 cesarb: Yup, which would be a big win for OpenMoko compared to many other phones. :) Sep 15 14:22:09 emdete: they probably put wlan in not because of the efficiency, but because of the sheer amount of wlan APs you can find almost everywhere Sep 15 14:22:30 emdete: PAN client would be next to useless with no BT PAN APs around Sep 15 14:24:17 cesarb: shure... avm supports it alot. but you are right, i don't know any other device but linux-pc to support it out of the box Sep 15 14:28:32 stefan|hospital: is that near Garmish? Sep 15 14:38:27 DukeOfURL: no, near hamburg http://www.snow-dom.de/ Sep 15 14:39:47 1004 users more ;) Sep 15 14:43:01 is there any new info concerning the AT&T SIM cards? Sep 15 14:43:25 the ones that won't operate in the Neo? Sep 15 15:20:41 hmm..i didnt get any mails from openmoko devel lists.. Sep 15 15:22:11 They're just resting. Sep 15 15:23:38 :P all right if you say so Sep 15 15:39:10 i wonder why everyone wants diferent backs, but no-one has suggested a modular back Sep 15 15:40:45 oh dang, anyone getting webkit-gtk built?, compile error here Sep 15 15:44:33 Sup3rkiddo: have you accidently tried building as root, and then resumed the build as a user? Webkit bombed here after I made that mistake.. Sep 15 15:50:35 * bedboi is back (gone 06:44:05) Sep 15 15:54:55 nope Sep 15 16:06:36 Sup3rkiddo: svn: Syntax error in revision argument '25582-r25571 Sep 15 16:06:54 mwester, yes!, wow, atleast i am not the only one..yay Sep 15 16:46:28 hey all Sep 15 16:52:52 re Sep 15 16:53:48 mwester: yes, if you edit /openembedded/conf/distro/include/sane-srcrev.inc you can correct to "25582" Sep 15 16:55:36 can anyone help me out with a little serial port problem? Sep 15 16:56:14 Im running the GPS location sharing project, on projects.openmoko.org Sep 15 16:56:27 but im a java programmer.. .so not really deep into linux Sep 15 16:56:48 now im trying to test the connection to my Garmin Sep 15 16:57:04 but i don't know how to reach it Sep 15 16:57:22 im using a com port Sep 15 16:57:37 the mobo offers only one, so i guess its com0 Sep 15 16:58:53 neo serial ports are /dev/ttySAC[0|1] Sep 15 16:59:09 ttySAC0 is used for communicating to GSM modem... Sep 15 16:59:32 can someone help me? I don't know what the error is exactly in the following: http://nopaste.org/p/aylBoNJdlb Sep 15 17:00:23 dont know what is exposed on the debug board... Sep 15 17:00:46 edistar: your build is issuing fetch command: svn co -r {now} http://svn.o-han...... Sep 15 17:00:58 -r must have a number as an argument... Sep 15 17:01:31 hmm... is anyone else having problems with building webkit-gtk ? Sep 15 17:02:01 polz: yes, there is an incorrect entry in /openembedded/conf/distro/include/sane-srcrev.inc Sep 15 17:02:08 v Sep 15 17:02:11 edit /openembedded/conf/distro/include/sane-srcrev.inc you can correct to "25582" Sep 15 17:02:15 jpcass: so what do I need to change? Sep 15 17:02:16 sorry CTRL-v Sep 15 17:03:03 edit the webkit like in that inc file, it should just ="25582" Sep 15 17:03:10 s/like/line/ Sep 15 17:03:12 jpcass meant: edit the webkit line in that inc file, it should just ="25582" Sep 15 17:03:24 jpcass: how did you find _that_ ? Sep 15 17:03:37 polz: quite some grepping... Sep 15 17:03:57 * polz only checked the .bb file for webkit-gtk Sep 15 17:04:15 jpcass: thanks :) Sep 15 17:04:35 jpcass: how do I tell the make to add a number? :P Sep 15 17:04:38 polz, i do Sep 15 17:04:51 polz: youre welcome ;-) Sep 15 17:05:09 jpcass, thanks man...welcome me too..:P Sep 15 17:05:13 wheee, the lists seem to be working again Sep 15 17:05:24 * Sup3rkiddo goes to fix the issue Sep 15 17:05:50 jpcass: plz help... been trying for about 2 days Sep 15 17:06:30 edistar: not sure about eds-dbus ... I'll try to have a look... in general you are looking at the SRC_URI or SVNREV variables in .bb files... Sep 15 17:07:42 jpcass: I'll need to learn bitbake generally then? Sep 15 17:09:25 edistar: look in packages/eds/eds-dbus_svn.bb : SRC_URI doesnt specify a specific revision... Sep 15 17:09:40 jpcass: what number do I need to add? Sep 15 17:10:00 but PV is set to svn${SRCDATE} ... Sep 15 17:10:26 i think that's where your -r {now} is coming from ... Sep 15 17:10:40 dont know still looking... Sep 15 17:10:53 jpcass did you update your mtn recently? i see a diff here in the commitlog which says: Sep 15 17:10:55 -SRCREV_pn-webkit-gtk ?= "25507" Sep 15 17:10:56 +SRCREV_pn-webkit-gtk ?= "25582-r25571" Sep 15 17:11:16 in sane-srvrevs.inc Sep 15 17:11:41 jpcass webkit has multiple components of which webkit and webcore are taken to form the rev Sep 15 17:12:10 roh: yes my mtn is quite recent - all entries in sane-srcrevs must be numbers they are used directly by svn as in : svn co -r {number from sane-srcrevs} Sep 15 17:12:48 roh: sane-srcrevs is automatically generated, perhaps there is a bug in the logic? Sep 15 17:12:58 jpcass the comment is: sane-srcrevs: fix webkit entry, it uses SRCREV_FORMAT, which the generator script doesn't pickup Sep 15 17:13:15 ask koen, he checked that in, so he knows the logic there Sep 15 17:13:26 changing the line in .inc works here... Sep 15 17:15:17 edistar: for a quick fix you could try changing svn://svn.o-hand.com/repos/${PN};module=trunk;proto=http Sep 15 17:15:28 to svn://svn.o-hand.com/repos/${PN};module=trunk;proto=http;rev=HEAD in the .bb file Sep 15 17:15:39 might work ;-) - no guarantees! Sep 15 17:23:51 stefan|hospital: *sigh* sorry about that, i knew how you were looking forward to the vacation w/ Nicole Sep 15 17:24:01 stefan|hospital: get well soon!!! Sep 15 17:26:39 ouf. strongly seconded. fast healing, stefan|hospital! Sep 15 17:39:25 mickey|dinner, pH5: Thanks Sep 15 17:39:54 edistar: it should be picking up a date from openembedded/conf/distro/include/sane-srcdates.inc (eds-dbus = "20070704") Sep 15 17:40:12 edistar: for some reason I dont know that's not getting included... Sep 15 17:41:02 edistar: if you explictly set SRCDATE = "20070704" in the eds-dbus_svn.bb file it should download the right revision (safer than HEAD) Sep 15 18:10:56 <_buz> nokia wants to do away with competing flash formats, by introducing yet another one http://feeds.arstechnica.com/~r/arstechnica/BAaf/~3/156903198/20070915-nokia-at-work-an-yet-another-removable-flash-memory-format.html Sep 15 18:11:06 Haha. Sep 15 18:11:10 * SpeedEvil stabs. Sep 15 18:11:11 * SpeedEvil stabs. Sep 15 18:11:21 * SpeedEvil sighs. Sep 15 18:11:59 What makes otherwise sane people wake up in the morning and think we need a new compression, or flash format. Sep 15 18:20:38 that's easy Sep 15 18:20:44 nokia would own it and therefore make more $ Sep 15 18:23:46 <_buz> sony tried that and it pretty much didnt work Sep 15 18:23:50 <_buz> as did fujitsu Sep 15 18:25:27 It's too late for them; the multitude of SD/MMC variants seem to already have won Sep 15 18:26:31 The only way it can suceed is if they manage to turn it into something backed by a coalition of manufacturers... Sep 15 18:26:55 umm guys any news on the new devices shipment? Sep 15 18:27:02 berkus: no. Sep 15 18:27:08 :( Sep 15 18:29:14 openmoko: 03shoragan * r2973 10/trunk/src/target/ (12 files in 2 dirs): Change the filename to something similar to the existing ones. Sep 15 18:33:37 cesarb: that article says that nokia is partnering with samsung and sony ericsson and several chipmakers. That doesn't count as a coalition? Sep 15 18:35:02 Jiten: in the best slashdot tradition, I wrote that first and then went to read the article Sep 15 18:35:14 though, it seems they're planning on making it several times faster than anything on the market right now Sep 15 18:35:17 (of course, the *real* slashdot tradition would be to simply never actually read the article) Sep 15 18:35:58 * Jiten rejects traditions that make no sense. Sep 15 18:37:06 considering the features they're planning, I wouldn't be surprised if it becomes standard for PCs too. Sep 15 18:37:26 it sounds like it'd be faster than anything on the market right now Sep 15 18:38:06 90 minute hd movie in a few seconds sounds quite a bit faster than current hard drives Sep 15 18:38:15 openmoko: 03mickey * r2974 10/trunk/src/target/OM-2007.2/artwork/sounds/ringtone_classy.wav: artwork/sounds: add classy ringtone Sep 15 18:40:07 haha Sep 15 18:40:14 that's a nice ringtone, mickeyl Sep 15 18:40:22 :) Sep 15 18:40:29 I love it Sep 15 18:42:52 Jiten: there's a small problem with something which will only be released on 2009 being "faster than everything on the market" Sep 15 18:43:11 Jiten: you are comparing something from 2009 with the common ones from 2007 Sep 15 18:43:25 Jiten: it's not that surprising that the 2009 one is faster than the 2007 one Sep 15 18:43:44 mickeyl: can we http://en.pastebin.ca/698774 to not include the gps panel applet on gps-less devices by default? Sep 15 18:45:13 cesarb: should go without saying that I'm comparing it to current situation :P Sep 15 18:46:09 though, wouldn't be surprising to have seriously fast flash drives for SATA soonish. Sep 15 18:46:13 Jiten: yeah, it being faster than anything on the market now isn't that big of a deal for something which doesn't even exist yet. Most important, will it be faster than anything on the market when it actually gets to be part of the market? Sep 15 18:46:30 What's the upper speed limit on SD anyway? Sep 15 18:46:42 Bandwidth over the connector Sep 15 18:46:53 which has poorly defined inpedance at high speeds. Sep 15 18:47:04 Jiten: making a fast SATA flash drive isn't even that hard - just use several chips and read from them in paralell, instant xN speedup, repeat as often as you want Sep 15 18:47:08 It's already up to 50Mhz at 4 bit wide Sep 15 18:47:39 I wouldn't be astonished at it getting to 150Mhz - more might be problematic. Sep 15 18:47:41 Jiten: the SATA interface is already quite fast, and LVDS can be made to run quite fast Sep 15 18:47:41 cesarb: that's exactly why I'm wondering about not seeing any. Sep 15 18:47:48 Jiten: price Sep 15 18:48:02 Jiten: SSDs are still very expensive Sep 15 18:48:32 SpeedEvil: I suppose the SD spec sets an upper limit somewhere. Sep 15 18:48:36 They have been 'going to replace rotating disks real soon now' since I had a 500M hard drive. Sep 15 18:48:37 Jiten: using more chips in parallel would also mean you would have to make a higher-capacity disk (or use lower-capacity chips), thus increasing the price even more Sep 15 18:48:46 Lynet: no. Sep 15 18:48:58 Lynet: as faster transfers would nescessitate a faster spec Sep 15 18:49:03 and newer spec Sep 15 18:49:08 well, there are 64GB SSDs on the market already... rather expensive though. Sep 15 18:49:11 SpeedEvil: they probably would, in a ideal world, end up being a new step on the memory hierarchy Sep 15 18:49:25 I read about a 416GB one too. not yet available though Sep 15 18:49:36 SpeedEvil: registers>cache>cache>cache>memory>SSD>HDD>tape>internet Sep 15 18:49:36 Flash fast enough for XIP - excecute in place would be nice Sep 15 18:49:45 but that requires it to be _very_ fast. Sep 15 18:50:17 as it means you don't need to copy applications into RAM, before running. Sep 15 18:51:07 SpeedEvil: isn't it more convenient to just copy everything into RAM? there is not much need for XIP anymore since large enough memories are now cheap enough... Sep 15 18:51:17 SpeedEvil: or are you worried about the power usage for the refresh? Sep 15 18:51:19 To an extent, yes. Sep 15 18:51:36 It's an 'it would be nice' feature, rather than required. Sep 15 18:52:25 For example, in principle, all applications running do not need any space for in-memory code. Sep 15 18:52:32 they only use data memory. Sep 15 18:52:36 which can be a big help. Sep 15 18:52:48 hmmh... if you compare current flash to older ram chips, how does the speed match? Sep 15 18:53:06 quite well on read. Sep 15 18:53:12 writes still take ages Sep 15 18:53:23 charge pumping and all that fun Sep 15 18:54:09 so XIP with the temporary parts allocated from RAM might be doable if you're not going for high speed device? Sep 15 18:54:41 Not really - for XIP, you want the flash to be as fast as 'normal' RAM Sep 15 18:54:57 how much slower is flash for reading? Sep 15 18:55:10 compared to ... DDR2-667 I guess Sep 15 18:55:46 slower? Sep 15 18:55:52 that ram is way fast Sep 15 18:56:17 slower is the assumption. I just want to know how much slower :) Sep 15 18:56:31 that depends! Sep 15 18:56:38 it's really too general a question Sep 15 18:56:50 since i could take forty seven zillion 8 bit eeproms and make a faster "ram" Sep 15 18:57:30 but you wouldn't really consider it faster Sep 15 18:57:54 ...well, technically, it would be Sep 15 18:58:00 but you'd have to be batshit insane to use it Sep 15 18:58:28 and it would be mega $ Sep 15 18:58:51 * anrp makes a ram out of latching relays to prove a point of some sort Sep 15 18:59:15 ok lets just compare reading and assume the flash should be the same amount and cost about the same as ddr2-667. Sep 15 18:59:35 specific enough? Sep 15 18:59:50 The problem is that what's available does not equal what's possible Sep 15 19:00:13 as for obvious reasons, most flash is engineered to go into applications like the last hundred million chips that were sold. Sep 15 19:04:15 ... Sep 15 19:04:24 whoever redesigned samsung's website needs to be shot Sep 15 19:05:52 anrp: deeply and fully agree Sep 15 19:06:05 anrp: flash-laden monstrosity Sep 15 19:06:10 anrp: impossible to find anything Sep 15 19:06:26 anrp: AND they broken all the older links from pretty much the whole of the world web Sep 15 19:06:34 yeah, no kidding Sep 15 19:06:50 anrp: these kinds of site were very common ten years ago Sep 15 19:07:08 anrp: back when using JAVA for navigation buttons was somehow considered acceptable Sep 15 19:07:14 Jiten: some of the fastest NOR flash (better kind) available is about equiv to PC-133 Sep 15 19:08:47 ok, so in embedded environment that sounds about fast enough for XIP Sep 15 19:09:11 that stuff is also seriously expensive... Sep 15 19:09:18 nand is what is usually meant when people say flash Sep 15 19:10:24 an example nand speed is around 40mb/s, with 4kb blocks (peak) Sep 15 19:10:52 the blocks is what kills XIP though Sep 15 19:11:05 since a block device and a ram are fundamentally different Sep 15 19:11:11 SpeedEvil: Regarding SD clock, I just checked the spec: "According to the Physical Layer Specification, the maximum SD Clock Sep 15 19:11:12 frequency is 25 MHz in normal speed mode and 50MHz in high speed mode, Sep 15 19:11:12 and shall never exceed this limi Sep 15 19:11:29 gaaaah, stupid copy-paste. Sep 15 19:11:38 mm delicious copypasta Sep 15 19:11:41 * anrp coughs Sep 15 19:11:54 Lynet: however, it does not specify what may occur in future standards. Sep 15 19:12:14 Lynet: as long as it is backwards compatible with this version, it's not a problem. Sep 15 19:12:16 the almighty wiki has a pretty good table comparing different consumer flash techs Sep 15 19:12:45 http://en.wikipedia.org/wiki/Secure_Digital_card#Different_types_of_MMC.2FSD_cards Sep 15 19:13:27 absolute max of any of them is SD at 200mbit -> 25mbyte/sec Sep 15 19:13:35 SpeedEvil: True. So it really depends on how far the physical spec can be pushed before changing connector. Sep 15 19:13:38 but they're all block devices, again Sep 15 19:14:02 er, sorry Sep 15 19:14:07 mmcplus is 416mbit Sep 15 19:15:09 Lynet: it's a tradeoff - speed, power, complexity. In the extreme case, you could go to ADSL-like signalling, rather than simple binary, and get a factor of lots speedup. Sep 15 19:15:10 Which makes me wonder if Nokia et al wants to own the new standard, or whether they have identified limits in the SD standard that can't be fixed by a backwards-compatible rev of the spec. Sep 15 19:16:01 its about the almighty dollar... Sep 15 19:16:07 there isn't a good technical reason Sep 15 19:16:24 there's a good legal/financial reason (if they can get it going in the market) Sep 15 21:16:34 hmm Sep 15 21:28:22 how does one run bootchart? Sep 15 21:28:40 one asks! Sep 15 21:28:47 ...obviously, I have no idea Sep 15 21:29:04 ah, someone is awake :-) Sep 15 21:29:05 sorry ~_~ Sep 15 21:29:17 pretty dead here on a saturday eveing, eh? Sep 15 21:29:28 it's like people have a life Sep 15 21:29:30 or something Sep 15 21:29:32 shocking Sep 15 21:30:53 we're just faking Sep 15 21:30:56 nah, I disagree, now is the time the life goes to bed and sleeps Sep 15 21:31:20 excuse me, i have to go back to my imaginary party Sep 15 21:32:09 don't drink too many imaginary drinks! Sep 15 21:32:57 hmm, I should see that Tasks has a icon, no? does not get one with me... Sep 15 21:34:06 hhf423: too late, i am already imaginary drunk! so drunk that i've started imagining things. Sep 15 21:34:44 aevin: to understand recursion, we first need to understand recursion Sep 15 21:35:02 that's not recursion, that's a catch-22 Sep 15 21:35:54 oh, sco still has about 10m usd left? Sep 15 21:36:16 this is so sad, sco once was a good company Sep 15 21:37:06 morning Sep 15 21:37:18 hi rwhitby Sep 15 21:37:25 Greetings Sep 15 21:37:35 "Nothing to be done" Sep 15 21:37:47 thats a sad thing to hear from ipkg upgrade Sep 15 21:38:05 morning Sep 15 21:51:08 anyone here managed to get gllin running in chroot in 2007.2? Sep 15 21:59:49 Will it blend? http://www.youtube.com/watch?v=7F_ZPhf1Bgw&eurl=http%3A%2F%2Fblog%2Ewired%2Ecom%2Fsterling%2F2007%2F09%2Fthe%2Dapple%2Diphon%2Ehtml# Sep 15 22:00:15 i'd ask if the neo would blend but there in such short supply that it would be in bad taste Sep 15 22:00:32 anrp: yes, at least for now Sep 15 22:01:06 but the neo has less metal than the iphone, so this is basically no question Sep 15 22:02:26 but, lithium! Sep 15 22:03:50 huh, iphone also has a LI battery, no? Sep 15 22:05:37 practically everything that has batteries does Sep 15 22:05:48 well all portable consumer applications Sep 15 22:05:57 stationary industrial is owned by lead acid Sep 15 22:06:39 hahaha, blending glowsticks... Sep 15 22:07:24 super bright Sep 15 22:15:12 heat glowstick to boiling, then break it. Sep 15 22:15:17 _really_ bright Sep 15 22:15:44 or just microwave it Sep 15 22:16:05 it will get too bright if you microwave it though ~_~ Sep 15 22:54:14 roh: about? Sep 15 23:08:01 about this long, I'd say Sep 15 23:09:44 Just heard from someone, that the iphone has an "operating temperature from 0-35 \degree\celcius. Does anyone know the relavant data for the neo? Sep 15 23:12:33 * SpeedEvil considers getting out the oven and the freezer. Sep 15 23:14:55 SpeedEvil - _That_ wasn't the intention of my question. :) Sep 15 23:17:54 I would expect it to more or less work over the same sort of range. Sep 15 23:18:37 Yeah, me too. Sep 15 23:20:45 Only up to 35°C? Isn't that a bit too low? Sep 15 23:21:01 Ambient temperatures sometimes get up to 40°C around here... Sep 15 23:21:55 35 is very low Sep 15 23:22:03 even cheap commercial stuff is usually 0-60 Sep 15 23:22:20 high end stuff is -40-125 D: Sep 15 23:22:41 Consider that 35C is the case temp Sep 15 23:23:06 apple would be conservative in the published specs Sep 15 23:23:17 But on reflection, I think I've had my neo up to 50C or so. Sep 15 23:23:34 (it was in the black case, on doing gps+strace, in full sun) Sep 15 23:23:47 Maybe 60C Sep 15 23:24:10 surely it has a temperature sensor? Sep 15 23:24:22 the battery has a temp sensor. Sep 15 23:24:34 well, the battery is the only component that needs one Sep 15 23:24:42 I think the samsung SoC should run fine at 100°C Sep 15 23:24:50 SpeedEvil: yeah, the case temp will be at least the ambient temperature... Sep 15 23:24:59 thats easy to test, just stick it in a pot of boiling water! Sep 15 23:25:01 ... er Sep 15 23:25:01 Oh, we have the data sheet for the CPU. ;o) Sep 15 23:25:16 SpeedEvil: there's no way the case temp will be lower than the ambient temp for more than a few minutes Sep 15 23:25:29 Well, yes. Sep 15 23:25:34 Elrond: where is it? :) Sep 15 23:25:38 <- lazy Sep 15 23:25:46 thomasg - Linked from the wiki. Sep 15 23:25:46 Elrond: the '02 CPU? Sep 15 23:25:57 you won't find it on samsung's website Sep 15 23:26:07 cesarb - Ahh, no. Only the 01 cpu. Sep 15 23:26:23 Elrond: the '01 CPU has been on the wiki since forever Sep 15 23:26:36 k, for the other lazy guys: http://www.samsung.com/global/business/semiconductor/productRightMenuDown.do?doc_file=26537 Sep 15 23:27:02 Elrond: for a time (after samsung's site was worsened) the link was broken, but some kind people on this channel tracked its new place down Sep 15 23:27:20 cesarb - Ahh. :-)) Sep 15 23:28:43 ok, its -65 to +150 °C Sep 15 23:28:53 (only for the SoC) Sep 15 23:29:24 I suspect some of the plastics will melt at 150C Sep 15 23:29:32 SpeedEvil: yes, I think so, too Sep 15 23:29:39 And that's presumably not 150C operating. Sep 15 23:30:00 it is, but "absolute maximum rating" :) Sep 15 23:30:15 Absolute maximums do not guarantee correct operation. Sep 15 23:30:20 "recommended operation conditions": -40 to +85 °C Sep 15 23:30:25 absolute maximum is 'won't die' Sep 15 23:30:30 it does not mean it will function Sep 15 23:30:31 yes, that's it Sep 15 23:30:40 absolute maximum is 'won't die immediately' Sep 15 23:30:42 thomasg: that's the temp on the chip's outside Sep 15 23:30:58 thomasg: you also have to factor in the fact that the case is not a very good thermal conductor... Sep 15 23:31:08 So call it 10C to the board, and 10C to the case, and we're down to 65C Sep 15 23:31:21 thomasg: and do some fancy calculations with the wattage expended, thermal conductivity, and so on Sep 15 23:32:00 thomasg: or, much easier, use something like cpuburn while measuring the temps, with the case closed and a controlled exterior temp ;-) Sep 15 23:32:31 .oO(put it in the dsc. ;o) ) Sep 15 23:32:31 SpeedEvil: well, as long as it can deal with more than 40°C ambient temp... Sep 15 23:33:05 I guess, the battery really is the most critical part? Sep 15 23:33:17 cesarb: we are talking about the very efficient ARM architecture, I think it will not be possible to create temperatures like 80°C with the input power on the cpu Sep 15 23:35:07 thomasg: true... the battery would probably be the greatest problem (with the charger on, it can get warm... if the CPU is also warming things up, the temps might get even higher) Sep 15 23:35:46 Elrond: well, with the battery, the problem isn't actually it not working anymore, if you get what I mean... Sep 15 23:37:19 the battery is definitely the biggest problem here but the biggest problem is not the heat while charging, it's the sun (if you have it in your car e.g.) Sep 15 23:37:23 cesarb - I think, I get, what you mean... and that really is "The Problem", I guess. Sep 15 23:37:57 thomasg: I thought most people have already figured out you never leave LCDs on the sun Sep 15 23:38:04 When it gets too hot while charging, I guess, the internal safety circuits will stop the charging? Sep 15 23:38:30 Elrond: I'd expect them to utterly cut off the battery from the external contacts Sep 15 23:38:32 cesarb: in my car I have a black navigation device with a battery and a lcd :) Sep 15 23:38:41 Elrond: so it wouldn't charge nor discharge Sep 15 23:39:05 I guess I had it at about 60°C this summer, but this isn't really a problem for the lcd Sep 15 23:40:33 cesarb - Right. Sep 15 23:43:09 Elrond: but, I don't trust these protection devices Sep 15 23:43:20 Elrond: since they are much like rarely-used code Sep 15 23:43:34 Elrond: you can't tell if they actually work... and they probably won't Sep 15 23:44:20 Yeah. Sep 16 02:43:49 hi all **** ENDING LOGGING AT Sun Sep 16 02:59:56 2007