**** BEGIN LOGGING AT Thu May 17 02:59:57 2007
May 17 03:53:27 <billl_> counter
May 17 03:53:27 <aloril> http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 65%) P1: a week, 3 days 08:06:30 (10.338 +-5.0 days) (1201;207)
May 17 03:53:34 <billl_> doh
May 17 03:59:14 <bmidgley> my mokomake got past uboot... i guess it was fixed
May 17 04:33:09 <aloril> (script) wiki RecentChanges: [[Kernel]]
May 17 05:21:09 <aloril> (script) openmoko-community: Werner Almesberger <werner at openmoko.org> Re: Making Neo Brickproof, was comments after reading Wiki
May 17 06:34:09 <aloril> (script) wiki RecentChanges: [[Wishlist:Text_Input]] [[Applications]] [[Wish_List_-_Hardware]] [[Flight_Mode]] [[Buying_Interest_List]]
May 17 07:36:20 <aloril2> Elrond: there were that many mails about unbrick protection by Werner
May 17 07:49:47 <roh> aloril2 there were also a lot of really stupid suggestions about it ;)
May 17 07:55:28 <Kaloz> hmz, dfu-util always fails on the first try, it looses the device after reset
May 17 07:56:26 <Stephmw> yo
May 17 07:56:46 <roh> Kaloz update it. thats a old bug
May 17 07:56:52 <roh> should be gone by now
May 17 07:56:53 <alphaone> Kaloz: Try to get newer uboot and dfu-util
May 17 07:57:46 <Kaloz> roh: well, currently it has the u-boot from april 15.. also, it didn't fail with that dfu-util
May 17 07:57:59 <Kaloz> it does with the one from yesterday
May 17 07:59:03 <Psi> is anyone working on VOIP on the neo, or is it far too soon for that and everyone is working on getting the main apps to work properly :P ?
May 17 07:59:22 <alphaone> Kaloz: Does uboot print something about i2c error?
May 17 07:59:38 <Kaloz> alphaone: yup
May 17 07:59:50 <Kaloz> alphaone: it's the uboot from april, so it does :)
May 17 08:00:01 <roh> Psi the latter. voip is really easy when the rest is working properly
May 17 08:00:08 <Kaloz> anyways, reflashed :) 
May 17 08:00:16 * Kaloz goes to check if the bug is gone or not
May 17 08:00:24 <alphaone> should be
May 17 08:00:36 <Psi> roh, cool
May 17 08:01:15 <Psi> i guess VOIP isnt very useful until the wifi version anyway
May 17 08:01:49 <hrw> really?
May 17 08:01:55 <buz> i  think so
May 17 08:01:55 <hrw> I plan to have BT AP here
May 17 08:02:03 <buz> i mean how many BT AP have you come across?
May 17 08:02:09 <madwoota> 0
May 17 08:02:11 <Psi> well yeah, if you can get the range with BT, then go for it
May 17 08:02:14 <buz> i would say a grand total of 0
May 17 08:02:24 <Psi> i just dont know if BT really has enough bandwidth at range for VOIP around the home
May 17 08:02:25 <madwoota> theres a reason for that :P
May 17 08:02:28 <mjr> clearly VOIP without wifi has some uses, but it's obviously true that wifi would radically increase VOIP's usefulness
May 17 08:02:38 <koen> Psi: 2Mbit/s
May 17 08:02:54 <mjr> and yeah, BT can do VOIP just fine
May 17 08:03:00 <madwoota> bt's bandwidth is ok, the range is crap tho
May 17 08:03:38 <koen> mjr: I'd be surprised if it couldn't, seeing that all headsets for phones use bluetooth :)
May 17 08:04:00 <Psi> yeah, at short range im sure it would work great
May 17 08:05:37 <Psi> i suppose if you lived in a small flat it would work fine, but a large house i dont think would work to well with BT
May 17 08:15:41 <koen> Class 1 does 100 meters
May 17 08:17:10 <sparq> Psi: I'm interested in working on VoIP, thought I cannot boast any actual *work* done ^_^
May 17 08:21:24 <XorA> koen: on a good day with a headwind :-)
May 17 08:26:34 <aloril2> roh: hehe, only triggering for core team and some mailing lists
May 17 08:28:15 <GaKKGaKk> counter
May 17 08:28:15 <aloril> http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 65%) P1: a week, 3 days 03:31:42 (10.147 +-5.0 days) (1202;208)
May 17 08:38:47 <sparq> dddd d
May 17 08:39:35 <Stephmw> sparq: I couldn't have put it better.
May 17 08:42:22 <PBeck> ahoi
May 17 08:42:32 <Vitaly> counter
May 17 08:42:33 <aloril> http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 65%) P1: a week, 3 days 03:17:25 (10.137 +-5.0 days) (1203;208)
May 17 08:51:30 <Elrond> aloril2 - Ahh, okay. :)
May 17 08:58:20 <sparq> Stephmw: sorry, terminal wonkage
May 17 08:59:03 * sparq peers suspiciously at brctl
May 17 09:02:41 <Stephmw> sparq: LOL
May 17 09:02:45 <Stephmw> sparq: misread that
May 17 09:05:28 * Stephmw cleans up the coffe he spluttered
May 17 09:05:30 <Stephmw> *coffee
May 17 09:08:16 <Elrond> How fragile is the power button on the P0?
May 17 09:14:25 <koen> the powerbutton is pretty sturdy
May 17 09:14:35 <koen> the aux button OTOH...
May 17 09:27:56 <simon__> counter?
May 17 09:27:56 <aloril> "/msg aloril counter?" for actual long message, giving short version here: http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 65%) P1: a week, 3 days 02:32:01 (10.106 +-5.0 days) (1204;208)
May 17 09:46:27 <LuitvD> good morning
May 17 09:53:45 <CIA-24> openmoko: 03thomas * r1992 10/trunk/src/target/OM-2007/applications/openmoko-contacts/ (ChangeLog src/contacts-contact-pane.c src/contacts-omoko.h): (log message trimmed)
May 17 09:53:45 <CIA-24> openmoko: Patch by: Neil J Patel <njp@o-hand.com>
May 17 09:53:45 <CIA-24> openmoko: * src/contacts-contact-pane.c:
May 17 09:53:45 <CIA-24> openmoko: * src/contacts-omoko.h:
May 17 09:53:45 <CIA-24> openmoko: - Uses colours from the theme instead of hard-coded values.
May 17 09:53:48 <CIA-24> openmoko: - Correct alignments for the GtkLabel.
May 17 09:53:50 <CIA-24> openmoko: - When a entry is in focus, its combobox will turn to an 'active' state.
May 17 09:57:18 <CIA-24> openmoko: 03thomas * r1993 10/trunk/src/target/OM-2007/artwork/themes/openmoko-standard/gtk-2.0/ (4 files): 
May 17 09:57:18 <CIA-24> openmoko: Patch from: Neil J Patel <njp@o-hand.com>
May 17 09:57:18 <CIA-24> openmoko: * Updates the contactspane rc file so it will use the new fieldentry and
May 17 09:57:18 <CIA-24> openmoko:  fieldcombo-normal images, depending on state.
May 17 09:57:18 <CIA-24> openmoko: * Makes the font size of the fieldcombo the same as everything else in the
May 17 09:57:22 <CIA-24> openmoko:  pane.
May 17 09:57:24 <CIA-24> openmoko: * Makes checkboxes visible when active.
May 17 10:08:09 <CIA-24> openmoko: 03thomas * r1994 10/trunk/src/target/OM-2007/artwork/icons/openmoko-standard/22x22/stock/ (8 files): * Add/rename contacts stock icons
May 17 10:43:54 <CIA-24> openmoko: 03thomas * r1995 10/trunk/src/target/OM-2007/artwork/icons/openmoko-standard/36x36/stock/ (43 files): * Rename stock icons to match moko-stock.h
May 17 11:01:21 <CIA-24> openmoko: 03thomas * r1996 10/trunk/src/target/OM-2007/openmoko-libs/ (ChangeLog libmokoui/moko-stock.c libmokoui/moko-stock.h): 
May 17 11:01:21 <CIA-24> openmoko: * libmokoui/moko-stock.c:
May 17 11:01:21 <CIA-24> openmoko: * libmokoui/moko-stock.h:
May 17 11:01:21 <CIA-24> openmoko: Modify/add stock items
May 17 11:11:43 <CIA-24> openmoko: 03thomas * r1997 10/trunk/src/target/OM-2007/artwork/icons/openmoko-standard/22x22/stock/ (9 files): * Rename some icons to new naming scheme
May 17 11:18:38 <CIA-24> openmoko: 03thomas * r1998 10/trunk/src/target/OM-2007/applications/openmoko-contacts/ (ChangeLog src/contacts-omoko.c src/contacts-omoko.h): 
May 17 11:18:38 <CIA-24> openmoko: * src/contacts-omoko.c: (create_main_window):
May 17 11:18:38 <CIA-24> openmoko: * src/contacts-omoko.h:
May 17 11:18:38 <CIA-24> openmoko: Use moko stock icons
May 17 11:43:51 <CIA-24> openmoko: 03thomas * r1999 10/trunk/src/target/OM-2007/openmoko-libs/ (ChangeLog libmokogsmd/moko-gsmd-connection.h): * libmokogsmd/moko-gsmd-connection.h: Add MokoGsmdConnectionProgress
May 17 11:44:20 <LuitvD> one more thomas :P
May 17 11:44:53 <CIA-24> openmoko: 03thomas * r2000 10/trunk/src/target/OM-2007/applications/openmoko-dialer/src/dialer-window-outgoing.c: openmoko-dialer: use MokoGsmdConnectionProgress
May 17 11:57:41 <robster> hehe
May 17 12:06:50 * arw_hal9000 saluda a todos
May 17 12:13:25 <kiney> counter
May 17 12:13:26 <aloril> http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 65%) P1: a week, 2 days 23:46:32 (9.991 +-5.0 days) (1205;208)
May 17 12:15:18 <nnpiggy> I'm having trouble powering up my neo. Any suggestions for field diagnostics?
May 17 12:17:09 <aloril> (script) openmoko-community: Werner Almesberger <werner at openmoko.org> Re: Few comments after reading Wiki
May 17 12:35:09 <aloril> (script) wiki RecentChanges: [[Talk:Hammerhead_protocol]] [[Hammerhead_protocol]]
May 17 12:37:09 <aloril> (script) openmoko-community: Werner Almesberger <werner at openmoko.org> Re: Making Neo Brickproof, was comments after reading Wiki
May 17 12:57:16 <aloril2> nnpiggy: leave it charging for enough long time?
May 17 12:57:56 <nnpiggy> I'm hoping that's all--I had let the battery go completely flat.
May 17 12:58:25 <nnpiggy> I let it charge off of a normal USB port over night and I was able to power it up...
May 17 12:58:43 <nnpiggy> I turned it off and haven't been able to turn it on since. It's charging again...
May 17 12:59:06 <nnpiggy> When it did power up this morning it reported less than 1/2 of a full charge.
May 17 13:00:11 <Stephmw> I would hazard a guess that the power management is foobared and that 'turning it off' just turns the backlight off
May 17 13:06:23 <XorA> take battery out of Neo when it is off
May 17 13:06:32 <XorA> any other ways just invites pain
May 17 13:07:00 <DukeOfURL> http://gizmodo.com/gadgets/cellphones/crazy-video-20-ways-japanese-smartphones-do-it-better-261161.php
May 17 13:07:29 <DukeOfURL> I guess they use a camera for facial recognition/authentication?
May 17 13:10:20 <nnpiggy> Thanks, Stephmw and XorA. That theory has a certain appeal.
May 17 13:11:02 <XorA> nnpiggy: I have 3 neo here, and although power usage goes down with P1 it still drains too much
May 17 13:11:14 <nnpiggy> However, "Turning on" uniformly elicits a reboot. But I could easily believe that we're bleeding power continuously...
May 17 13:11:33 <SpeedEvil> XorA: Do you have an actual draw figure in mA?
May 17 13:11:48 <XorA> SpeedEvil: no, but I know kernel suspend code is a work in progress
May 17 13:11:50 <Stephmw> nnpiggy: noe Neo related, but the last thing to get fixed on all the protos I get at work is power management
May 17 13:11:53 <Stephmw> nnpiggy: *not
May 17 13:12:05 <nnpiggy> S
May 17 13:12:19 <XorA> SpeedEvil: in suspend P1 lasts less than 12 hours with current kernel, how much less I dont know
May 17 13:12:21 <nnpiggy> Stephmw: that's not surprising...
May 17 13:12:23 <Stephmw> I work with my protos tethered - always on AC
May 17 13:12:46 <nnpiggy> I've been observing about 2 hours of ordinary operation (WITHOUT making phone calls).
May 17 13:12:59 <webjames> wow, check out these phones: http://www.symbian.com/j20/phones.asp they have all sorts of weird stuff they can do
May 17 13:13:16 <webjames> http://www.symbian.com/j20/phones.asp
May 17 13:13:24 <nnpiggy> I understand that the 1 May load is supposed to be a lot more power-efficient 
May 17 13:13:33 <webjames> i think we could implement some features like this into the openmoko
May 17 13:15:56 <balrog-kun> 1 May isn't P1?
May 17 13:16:25 <Stephmw> webjames: any specifics? that phone list is just, well, a phone list
May 17 13:16:52 <webjames> Stephmw: have a look at this one for example: http://www.symbian.com/phones/foma_f702id.html
May 17 13:17:38 <Stephmw> webjames: oh, I thought you were talking about something power related :)
May 17 13:18:07 <webjames> nope just some weird features
May 17 13:19:51 <Stephmw> the key word to follow through on mobiles is: integrated
May 17 13:20:22 <Stephmw> everything, every app, every help file, every single thing must look to the user to provide a seamless user experience
May 17 13:20:41 <Stephmw> none of the gnome or kde funk
May 17 13:23:52 <aloril2> what about leaving it connected to USB (in 500mA mode)?
May 17 13:25:29 <DukeOfURL> Wierd NOT.  We certainly need user authentication on our smart phone.  Fingerprint, facial recognition, iris scan, speaker identification, SOMETHING!
May 17 13:26:52 <balrog-kun> password?
May 17 13:27:32 <Clint> faceprint
May 17 13:27:42 <DukeOfURL> We need to be able to identify an individual user,
May 17 13:28:02 <balrog-kun> username + password
May 17 13:28:59 <tholin> Anyone know how legal an open source mobile would be. Would I be able to import it legaly in Sweden?
May 17 13:28:59 <aloril> tholin: See http://wiki.openmoko.org/wiki/FAQ (Hopefully this link answers your question. This autotrigger for new nick and question during first 15 minutes done only once.)
May 17 13:29:10 <balrog-kun> i don't know why you would need all that on a phone but GPE/OPIE already supports that :)
May 17 13:29:24 <DukeOfURL> If I'm going to keep financial information on the phone then password isn't stong enough (shoulder surfing)
May 17 13:30:05 <balrog-kun> fingerprint/facial recognition/speaker identification doesn't help with that
May 17 13:30:08 <DukeOfURL> Besides, I don't want to punch in a password every <n> minutes
May 17 13:32:25 <flyfish30> press information
May 17 13:47:17 <aloril2> P1+ discount ideas: 1) normal discount for those who buy P1 (for example $100-150?)
May 17 13:49:02 <aloril2> 2) of those who buy P1 and actually develop for it provide discount coupons like Nokia did for N770 developers: they got to buy N800 for 100 EUR (IIRC)
May 17 13:50:15 <aloril2> 3) of those who got (P0+)P1 free and did stuff: provide free P1+ too (and maybe add some new people to free program)
May 17 13:51:25 <aloril2> A3
May 17 13:51:53 <aloril2> uups ;-)
May 17 13:53:23 <LuitvD> why P1+ discount?
May 17 13:53:45 * alphaone likes idea 3 :-)
May 17 13:54:04 <LuitvD> thought the discount for early adopters applied to P2
May 17 13:54:17 <LuitvD> alphaone: did you get a P1 device for free?
May 17 13:54:29 <LuitvD> alphaone: and is that 'get' or 'get to use' ?
May 17 13:54:30 <LuitvD>  :P
May 17 13:55:32 <alphaone> LuitvD: I also got a P0 device, but after the P1 arrived I gave it to wansti to test his ringtones.
May 17 13:55:45 <LuitvD> :P
May 17 13:55:55 <LuitvD> i want to have one too for testing
May 17 13:57:42 <aloril2> about Nokia discount: it was 500 coupouns IIRC
May 17 13:59:03 <LuitvD> aloril2: the 2nd and 3rd combined would be nice too
May 17 14:00:40 <aloril2> LuitvD: hardware(P1+)==hardware(P2) so I don't see any difference ;-)
May 17 14:01:04 <LuitvD> those who got free P0 or P1 and really did something for moko (think that's still all P0 and P1 recievers) get a free P1+ to get on with developement, and the ones who bought the P1 and did some good stuff for moko, 50% discount or so
May 17 14:01:20 * aloril2 was actually thinking about all 1-3 options being in effect at same time
May 17 14:01:36 <LuitvD> aloril: yeah, probably... discount was first mentioned before there were plans for P1+ :P
May 17 14:01:50 <LuitvD> aloril2: all three seem fair :P
May 17 14:02:50 <aloril2> oh, and 0) option too: haven't bought P1 or otherwise qualify for 1-3 options: buy P1+ for full price
May 17 14:04:52 <SpeedEvil> I'd like the option to buy upgrade boards too.
May 17 14:05:08 <SpeedEvil> Which are just the motherboard, and no GPS/mobile antenna/case/LCD
May 17 14:05:10 <aloril2> s:otherwise:otherwise doesn't:
May 17 14:05:42 <SpeedEvil> For both upgrades, and standalone stuff
May 17 14:07:59 * Kero justs wants a P1 right now :)  will bother with discounts when p1+ is there.
May 17 14:08:14 <Kero> though I guess discounts encourage ppl to buy P1 already and not wait.
May 17 14:08:22 <SpeedEvil> yeah
May 17 14:08:32 <aloril2> Kero: ditto
May 17 14:09:36 * aloril2 will get P1 ASAP even without any discount for P1+
May 17 14:11:04 <buz> the neo as embedded board would be neat
May 17 14:11:05 <aloril2> without discount I might get P1 and then wait for V2
May 17 14:11:06 <Kero> saw yesterday it's gonna be 64MB flash for P1 and 1GB for P1+ ?
May 17 14:12:23 <Kero> not that it matters much if there's a microSD slot.
May 17 14:12:27 <antenagora> counter
May 17 14:12:28 <aloril> http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 65%) P1: a week, 2 days 21:47:29 (9.908 +-5.0 days) (1206;208)
May 17 14:12:29 <buz> that rumor has never been confirmed afaik
May 17 14:13:16 <SpeedEvil> I'm almost certain that that was a mistranslation.
May 17 14:13:20 <SpeedEvil> Of gigabit.
May 17 14:13:26 <SpeedEvil> Or 128M.
May 17 14:13:38 <SpeedEvil> Going from 64-128 would make sense.
May 17 14:13:44 <Kero> ok. I'd expect upgrades, but not huge upgrades :)
May 17 14:13:52 <aloril2> 2 Neos and 2 differents users nearby (only one of them needs to be developer too) enables interesting developent possibilities, so I hope it happens often (and discount should make this more common, especially 2 and 3 options)
May 17 14:13:59 <SpeedEvil> However, I don't speak german
May 17 14:17:00 <aloril2> one simple example: a: "Where are you?" b: presses button "Send position" a: looks at screen to see it in map
May 17 14:17:52 <SpeedEvil> There is the 'upload position constantly, encrypted with a key' app too
May 17 14:18:00 <SpeedEvil> So your friends can see where you are.
May 17 14:20:06 <Kero> that's cute when waiting for friend(s) :)
May 17 14:21:59 <pH5> I've uploaded some partial analysis of the hammerhead cold start, links at http://wiki.openmoko.org/wiki/Talk:Hammerhead_protocol.
May 17 14:22:35 <aloril2> yeah, sending position during call requires some kind of send data during call over voice protocol ('modem') and other side to intepret it (for low volume data like positions should be reasonably easy)
May 17 14:23:19 <SpeedEvil> Annoyingly - it's not trivial to do it inaudibly
May 17 14:23:47 <Kero> watermark-style is inaudible.
May 17 14:23:59 <SpeedEvil> And also doesn't get through GSM codecs well
May 17 14:24:03 <XorA> Kero: watermarking GSM isnt trivial
May 17 14:24:18 <SpeedEvil> The basic problem with GSM is that it assumes the input is speech-like
May 17 14:24:30 <SpeedEvil> And all the coding goes to coding speech.
May 17 14:24:42 <SpeedEvil> It is designed to throw away inaudible stuff
May 17 14:24:51 <XorA> SpeedEvil: your best bet would be to just transmit it hidden in a BEEP, then sender can say, sending you my position gimme a sec BEEP, done
May 17 14:24:54 <Kero> makes the problem much harder indeeed.
May 17 14:25:11 * Kero is not an audio expert at all.
May 17 14:25:20 <SpeedEvil> The nice fastish way to transmit data isn't AIUI available free.
May 17 14:25:35 <SpeedEvil> It's basically you take a human throat model, and feed it data
May 17 14:25:42 <SpeedEvil> and then decode at the other end
May 17 14:25:51 <SpeedEvil> you can get some 1.3kbits/sec from this
May 17 14:25:54 <XorA> DTMF and make it a tune
May 17 14:26:38 <mjr> yeah, if you don't mind some beeps, DTMF or similar would be workable...
May 17 14:26:39 <SpeedEvil> Imagine for example the datarate you get if every sylable can be a different pitch, loudness, and you're not constrained by a real human throat.
May 17 14:26:53 * mjr imagines a Klingon
May 17 14:27:25 <pH5> so not VoIP but IPoV?
May 17 14:28:28 <Kero> but why not sending data over gprs? that's not audio, no?
May 17 14:28:34 <mjr> hmh
May 17 14:28:49 <mjr> wonder if you can tell the chip to suspend voice for a while and let gprs through instead...
May 17 14:29:06 <mjr> 'cause, you know, it doesn't do both at the same time, and usually the gprs connection is bumped for voice
May 17 14:29:15 <pH5> and have the neo play a nice standby tune while the voice call is on hold
May 17 14:29:29 <Kero> ah-ah...
May 17 14:29:31 <SpeedEvil> You can't do that AIUI.
May 17 14:29:50 <mjr> SpeedEvil, kinda suspected that... Plus, really, the DTMF approach would probably be faster ;]
May 17 14:29:56 <SpeedEvil> It's not.
May 17 14:30:12 <SpeedEvil> DTMF gets you 20 bits/sec or so
May 17 14:30:21 <SpeedEvil> the 'fake voice' approach gets you 1300
May 17 14:30:38 <aloril2> SpeedEvil: I think it is simple: receiving side delays incoming voice by tiny amount, enoung so that it can notice beep at certain frequency: do 2 things at that point:
May 17 14:30:59 <Kero> wikipedia mentions gprs class A devices which should be able to use gsm and gprs simultaneously, and class B and C which can't
May 17 14:31:09 <SpeedEvil> class A devices have two radios
May 17 14:31:24 <Kero> that's cheating :)
May 17 14:32:15 <aloril2> 1) start decoding data
May 17 14:32:16 <aloril2> 2) at receiving size instead of delivering beeps or whatver encoding is used use text->speech to describe it (position being transmited or N bytes transmitted or something)
May 17 14:32:42 <mjr> SpeedEvil, hmm, come to think of it, I assumed that gprs would have to be set up as well, if it was in suspend it'd probably be faster to switch, yeah
May 17 14:33:07 <SpeedEvil> AIUI, in principle GPRS can come back almost immediately after a call
May 17 14:33:19 <mjr> it can
May 17 14:35:00 <SpeedEvil> I wonder if you can send SMS on a call
May 17 14:35:09 <XorA> SpeedEvil: yes
May 17 14:35:09 <aloril> (script) wiki RecentChanges: [[Talk:Hammerhead_protocol]] [[Press_Coverage]] [[GPS_Navigation]]
May 17 14:35:18 <XorA> SpeedEvil: and I beleive you can GPRS during a call as well
May 17 14:37:50 <aloril2> without delay you would hear short beep and then there is no reason to echo communication to ear unless receiver really wants it instead of something nice ;-)
May 17 14:38:23 <SpeedEvil> You can just use a speech-beep
May 17 14:38:41 <mjr> mmh, yeah, sms while calling would be a natural way to transmit side data if one doesn't mind the possible extra cost
May 17 14:38:45 <SpeedEvil> 'data' in a wierd pronounciation, for example.
May 17 14:38:57 <SpeedEvil> Delays are annoying
May 17 14:39:12 <tholin> How good will the gps reception be. Will it work of the phone is in a pocket? Will if work if the phone is indoors?
May 17 14:39:59 <mjr> and yeah, aloril's note about the receiving Neo just saying that position is being transmitted would be good
May 17 14:42:15 <pH5> tholin: pocket is no problem, indoors depends on the building I'd guess.
May 17 14:42:28 <mjr> According to Sean the chip got a fix indoors when he was demoed it, but of course depends on particular circumstances
May 17 14:42:48 <mjr> info is a bit scant until the thing is properly supported...
May 17 14:43:14 <pH5> mjr: I sometimes get fixes indoors when I cold-booted it outside, so probably AGPS could help here.
May 17 14:43:27 <mjr> in that case maybe yes
May 17 14:43:40 <SpeedEvil> From first principles.
May 17 14:43:49 <tholin> If I decide to get one of these phones I would like to create a position logger to record my whereabouts and draw some nice maps with google maps api. 
May 17 14:43:54 <mjr> pH5, with the Neo and the first binary blob?
May 17 14:44:08 <SpeedEvil> I think it will get positions almost anywhere there is the barest hint of the signal, if it's had a fix within 5 min or so.
May 17 14:44:09 <tholin> But that would be useless if I had to hold it up to get a signal.
May 17 14:44:34 <SpeedEvil> And maybe slightly less sensitive if it's got a complete almanac
May 17 14:44:50 <SpeedEvil> And is started up after being off for a long time
May 17 14:44:58 <SpeedEvil> tholin: http://www.openstreetmap.com
May 17 14:45:47 <pH5> mjr: yes, so this is by no means a good indicator for the gps performance of the final phone.
May 17 14:46:11 <aloril2> SpeedEvil: yeah nice beep is better than delay
May 17 14:47:02 <mjr> pH5, you getting accurate positions? I remember when somebody was first getting fixes here, and he said (rather vaguely) something about it being close but not quite
May 17 14:47:21 <mjr> (and yeah, not final etc)
May 17 14:49:44 <mjr> hm, openstreetmap.com has some Helsinki mapping present; not a lot, but a starting point
May 17 14:50:28 <XorA> mjr: just remove your roof :-)
May 17 14:51:54 <mjr> the javascript map viewer is buggy though, but a zoom change will whack it into sanity
May 17 14:52:26 <SpeedEvil> My local town (40K) was entirely missing.
May 17 14:52:36 <SpeedEvil> Dundee - 140K - had one road.
May 17 14:53:07 <pH5> mjr: accuracy varies. it's a few m off maybe.
May 17 14:53:25 * XorA notes the SpeedEvil is creeping around just to the north of him
May 17 14:54:19 <SpeedEvil> I'm about 5Km NE of Glenrothes
May 17 14:54:21 <mjr> pH5, well that's good enough for me
May 17 14:54:25 <CoreDump|afk> FWIW, I have _not_ been able to get a correct lock w/ my Neo. All tests were a few hundred meters off, even when I was on the top of a building...
May 17 14:54:39 <SpeedEvil> Odd.
May 17 14:54:39 <XorA> SpeedEvil: next to Murrayfield Stadium when Im at work
May 17 14:54:58 <SpeedEvil> You are sure? What have you been comparing to?
May 17 14:55:07 <CoreDump|afk> google maps
May 17 14:55:23 <SpeedEvil> I wonder if it might be geoid problems.
May 17 14:55:27 <SpeedEvil> But that should be under that.
May 17 14:55:30 <mjr> maybe there still is some conversion issue?
May 17 14:55:49 <CoreDump|afk> I hope it is a driver problem with the binary gllin cr*p =)
May 17 14:55:51 <SpeedEvil> Google maps can be wrong
May 17 14:56:09 <CoreDump|afk> SpeedEvil: on two different locations? I don't think so
May 17 14:56:15 <SpeedEvil> And are you sure you diddn't get degree min sec - decimal degree problems?
May 17 14:56:23 <CoreDump|afk> pretty sure
May 17 14:56:35 <Stephmw> pwd
May 17 14:56:39 <Stephmw> hmmm
May 17 14:56:45 * Stephmw changes window
May 17 14:56:53 * CoreDump|afk will do some more testing when he gets a chance
May 17 14:58:06 <SpeedEvil> If you're in an openstreetmapped area, you can use 'gpsbabel' to convert the NMEA logs to GPX, and then load the logs into the JOSM editor, along with the map
May 17 14:58:39 * XorA should play with GPS
May 17 14:58:51 * XorA wonders if P0 had GPS
May 17 14:59:00 <mjr> it did
May 17 14:59:19 <XorA> hmm, I could install it in wifes car and generate Edin maps :-)
May 17 14:59:20 <SpeedEvil> Nobodies got it working though
May 17 14:59:21 <mjr> the daemon reportedly doesn't work with it though due to different power-up-sequence 
May 17 14:59:28 <pH5> XorA: LaF0rge said they had it working even on v2,v3, but I can't get GPS on my Bv3 to work.
May 17 15:00:13 <mjr> hmm, when I get my Neo, have to take up a bicycling hobby. Through all the nearby roads ;)
May 17 15:01:51 <SpeedEvil> Bicycle, with bluetooth headset to note stuff apparently works really well
May 17 15:02:39 <SpeedEvil> 'left foo street' 'left amenity post office'
May 17 15:02:41 <SpeedEvil> ...
May 17 15:03:41 <LuitvD> 'left through wall' 'right towards vault' 
May 17 15:03:44 <LuitvD> ...
May 17 15:03:50 <mjr> :D
May 17 15:04:03 <SpeedEvil> :)
May 17 15:04:22 <SpeedEvil> No officer, I'm not kerb crawling, I'm mapping.
May 17 15:04:27 <LuitvD> hehehehe
May 17 15:04:28 <aloril2> a bit further about: IPoV (or just data over voice): send tone1 at certain frequency (I want t send data) and tone2 (proposed encoding: DTFM, vocoder, etc..) and then wait for reply
May 17 15:05:07 <LuitvD> IPoV? :P
May 17 15:05:12 <LuitvD> damn, who invented that?
May 17 15:05:47 <aloril2> if no reply, then other side is not capable (not Neo nor anything else supporting it)
May 17 15:06:08 <mjr> aloril2, sounds sane
May 17 15:06:53 <SpeedEvil> IMO, tone1 will false trigger _way_ too often
May 17 15:07:00 <SpeedEvil> you want a sample.
May 17 15:07:09 <aloril2> pH5 used IPoV term first
May 17 15:07:11 <SpeedEvil> 'NeoData' in a silly voice
May 17 15:07:43 <guaqua> neodata :)
May 17 15:07:57 <LuitvD> should be openpatented
May 17 15:08:01 <guaqua> maybe frequency human ear can't hear
May 17 15:08:24 <guaqua> i'm guessing there are already ways to transfer data over gsm
May 17 15:08:43 <DukeOfURL> GSM should be able to pass DTMF.  There are 16 combinations of tones.  Maybe with a 'shift' key we could apss data.
May 17 15:08:55 <DukeOfURL> s/apss/pass/
May 17 15:09:24 <mjr> guaqua, SpeedEvil has referred to a vocoder that'd do ~1.3k but proprietary tech
May 17 15:09:42 <apt> DukeOfURL meant: GSM should be able to pass DTMF.  There are 16 combinations of tones.  Maybe with a 'shift' key we could pass data.
May 17 15:09:59 <guaqua> people used to do 9600 kbps over any phone line
May 17 15:09:59 <mjr> DukeOfURL, and yeah, DTMF should work for very small data amounts
May 17 15:10:05 <guaqua> it _should_ be possible
May 17 15:10:07 <SpeedEvil> mjr: More unpublished than proprietory
May 17 15:10:12 <mjr> SpeedEvil, righto
May 17 15:10:13 <DukeOfURL> With mobile-2-mobile calling plan, IPoV would save a ton of money.
May 17 15:10:19 <SpeedEvil> Someone referred to it on this channel earlier
May 17 15:10:25 <SpeedEvil> A couple of months ago
May 17 15:10:26 <DukeOfURL> GSM gateway to PBX, etc.
May 17 15:10:32 <SpeedEvil> I think it was a pdf
May 17 15:10:52 <SpeedEvil> DukeOfURL: It's very marginal for voice though.
May 17 15:11:12 <SpeedEvil> VOIP over IPoV over GSM is just a little silly.
May 17 15:11:14 <mjr> guaqua, the issue is that we can't feed digital data straight to the GSM chip, and the audio data we can feed it will be put through (one of) the GSM codec(s)
May 17 15:11:34 <SpeedEvil> We only have an analog connection in.
May 17 15:11:36 <mjr> guaqua, and the codec will, in worst case, itself have a bitrate of 9k6
May 17 15:11:41 <DukeOfURL> I'm not talking about VoIP--just the ability to send data over a voice channel.
May 17 15:11:56 <SpeedEvil> Yeah
May 17 15:12:20 <mjr> guaqua, so whatever data you want through will have to survive the GSM codec at 9k6bps
May 17 15:12:25 <LuitvD> connect one neo to the internet, and use it with the other?
May 17 15:12:26 <DukeOfURL> Means we could use Cingular pay-as-you-go cards that don't provide data.
May 17 15:12:42 <mjr> completely different from just doing it over "any phone line"
May 17 15:12:59 <guaqua> mjr: okay
May 17 15:13:32 <guaqua> mjr: are you familiar with the sanla technology?
May 17 15:14:44 <mjr> not really, I'm a civil service man :]
May 17 15:15:32 <guaqua> well, it basically does 1200 bps over any line
May 17 15:15:39 <guaqua> and this is any
May 17 15:15:49 <guaqua> i have to dig up the algorithms
May 17 15:16:50 <mjr> again, the GSM voice call is not merely "line", the signal is actively munged in a codec that expects to be compressing human speech and wants to throw everything else away
May 17 15:16:51 <aloril2> another idea: route all your calls trough PC and then PC could decode IPoV -> internet access during call (slow but still..)
May 17 15:17:18 <webjames> i like the SMS idea, if compatible the neo would not load up the sms in message viewer, but onto a map as positioning data
May 17 15:17:38 <mjr> aloril2, yay, but raises the ol' VOIP issues that you really want to use GSM data calls then, and even then adds some latency
May 17 15:17:44 <guaqua> mjr: still, it's a lot better quality than hf frequency transmission from 600 kilometers away at 35W power :)
May 17 15:17:55 <SpeedEvil> SMSs can be _very_ non-realtime though.
May 17 15:18:24 <mjr> SpeedEvil, can. Usually not at least here.
May 17 15:18:31 <webjames> SpeedEvil: i see..
May 17 15:18:35 <XorA> webjames: SMS can be marked as DATA only
May 17 15:18:43 <XorA> webjames: and they wont show up
May 17 15:18:54 <mjr> but I still agree that SMS may be a good side-channel for this kind of thing
May 17 15:18:55 <webjames> XorA, now we just have the latency problem
May 17 15:18:57 <SpeedEvil> Sometimes the things take days. Usually for me minutes
May 17 15:19:58 <webjames> i suppose different places countries people will have different experiences, i would say on the whole in the UK it is fairly instant, but i have had delays of an hour or more
May 17 15:20:15 <XorA> use GPRS, implement IP over DNS (suppliers only seem to charge for http transfer)
May 17 15:20:21 <webjames> which i now see is useless for transmitting real time location data
May 17 15:21:19 <SpeedEvil> If you're doing GPRS, then you can throw your location on a server, and the other party can pick it up after the call
May 17 15:21:25 <DukeOfURL> IP over DNS?  Domain name service?
May 17 15:21:39 <Clint> yes, it's slow as hell
May 17 15:21:40 <webjames> DukeOfURL: must have meant something else?
May 17 15:21:59 <SpeedEvil> much of this stuff works better if there is one nominated server for all neos
May 17 15:23:01 <DukeOfURL> XorA: Do you mean data on port 53?
May 17 15:23:44 <XorA> DukeOfURL: DNS compliant requests which happen to carry IP traffic, there is a RFC for it
May 17 15:24:19 <XorA> DukeOfURL: ie, DEADBEEF.spanner.org where DEADBEEF is the packet
May 17 15:24:42 <Clint> it's a wonder more people aren't blocking that sort of thing
May 17 15:25:13 <aloril2> mjr: well I was thinking something similar to SkypeOut (so from Neo viewpoint it would be just normal voice call until IPoV), but latency yeah I guess so
May 17 15:25:21 <webjames> bit of a change of subject, it would be cool if if one is in a wireless signal area one could transmit ones position to a server so ones parents/employer/teacher could see where one is
May 17 15:25:56 <mjr> webjames, old hat, we all want to do that
May 17 15:26:21 <Maulkin> Well, the issue is that it's very hard for them to actually block it due to the way they use acls to privide the pay-for wireless points
May 17 15:26:27 <webjames> mjr, would be cool, then at the end of the day you could map your movements throughout that day
May 17 15:26:31 <Maulkin> (re: IP over DNS)
May 17 15:26:56 <Maulkin> webjames: nice thought, and I've done it :)
May 17 15:27:07 <webjames> Maulkin: nice, nice
May 17 15:27:08 <mjr> webjames, you can collect the data locally too for the latter purpose. And with gprs, you're pretty much always in signal area.
May 17 15:27:08 <Maulkin> admittedly, with a A780 and GPRS
May 17 15:27:37 <Maulkin> I sued it for OpenStreetMap mapping, but as a handy roll-off, you had a "Where is Maulkin" thing
May 17 15:27:41 <Maulkin> *used
May 17 15:27:42 <webjames> mjr yeah, i think thats such a good feature, also would help to buld up openstreet map quickly to
May 17 15:27:52 * Maulkin smiles
May 17 15:27:54 <DukeOfURL> I have built a database of streets for western U.S.  I can do geocoding and reverse geocoding.  Using WebServer, I am drawing location points.
May 17 15:27:55 <Maulkin> great minds etc
May 17 15:28:14 <DukeOfURL> D Is there street information for the UK?
May 17 15:28:22 <Maulkin> DukeOfURL: What do you mean?
May 17 15:28:26 <Maulkin> DukeOfURL: street names?
May 17 15:28:44 <DukeOfURL> line segments with lat/lon points that describe each street.
May 17 15:29:00 <DukeOfURL> with street name, etc.
May 17 15:29:28 <Maulkin> DukeOfURL: Well, it's an interesting one. Most of the mapping data in the UK is Crown Copyright by Ordanace Survey, and they charge a hefty licence fee for it. There is openstreetmap.org who have been doing their own mapping, and that's free (speech and beer)
May 17 15:29:44 <Maulkin> And don't even ask about post codes :)
May 17 15:30:09 <webjames> DukeOfURL: openstreetmap is great, except for postcodes + large unmapped areas
May 17 15:30:30 <guaqua> can someone give some info on this ip over dns?
May 17 15:30:57 <Maulkin> For postcodes, theres... erm...
May 17 15:31:02 <webjames> guaqua: http://thomer.com/howtos/nstx.html
May 17 15:31:04 <Maulkin> freethepostcode.org I think
May 17 15:31:09 <DukeOfURL> forget postcodes
May 17 15:31:33 <DukeOfURL> who here is sending letters?
May 17 15:31:41 <XorA> guaqua: google can
May 17 15:32:13 * k-s[WORK] is away: lunch!
May 17 15:32:24 <DukeOfURL> Were taxpayer funds used to develop the Survey?
May 17 15:32:32 <DukeOfURL> in the UK?
May 17 15:32:39 <Maulkin> DukeOfURL: sort of, yes.
May 17 15:32:55 <webjames> DukeOfURL: a lot of gps software uses postcodes, often to find someone they tell you their postcode
May 17 15:33:03 <DukeOfURL> Well then the information belongs to the taxpayer.
May 17 15:33:12 <Maulkin> Nope
May 17 15:33:19 <Maulkin> It belongs to the crown
May 17 15:33:25 <Maulkin> Which is different :)
May 17 15:34:07 <SpeedEvil> It now doesn't.
May 17 15:34:19 <SpeedEvil> It belongs to the spun-out ordinance survey organisation.
May 17 15:34:27 <Maulkin> yup :)
May 17 15:34:28 <webjames> well if openstreetmap and freethepostcode can be combined that'd be great
May 17 15:34:29 <SpeedEvil> Who is now wholly supported by selling maps
May 17 15:34:45 <Maulkin> Wikipedia explains it best: "Although OS is a government agency it is required to act as a "trading fund" or commercial entity. This means that it is totally self funding from the commercial sale of its data whilst at the same time being the public supplier of geographical maps"
May 17 15:34:58 <Maulkin> So it's sort of odd
May 17 15:35:14 <DukeOfURL> it's fucked
May 17 15:35:19 <Maulkin> The UK doens't have the same "produced by teh government == public domain" thing that the US has
May 17 15:35:50 <mjr> and the US doesn't _really_ have it either...
May 17 15:36:21 <DukeOfURL> well. I have 25M line segments in a database from a government source.
May 17 15:36:30 <DukeOfURL> free
May 17 15:36:43 <SpeedEvil> The tiger one?
May 17 15:36:53 <DukeOfURL> Tiger
May 17 15:37:02 <DukeOfURL> Census bureau
May 17 15:39:06 <webjames> gprs can still be very expensive it would be possible to store the neo instead of downloading them, or getting maps whilst in a wifi zone
May 17 15:40:09 <SpeedEvil> For openstreetmap, the compressed planetary map will fit on a microSD card.
May 17 15:40:18 <SpeedEvil> It would be nice if it diddn't.
May 17 15:42:09 <DukeOfURL> Would it be possible to have transparent WiFi?  Where the user isn't even aware the phone is making a WiFi connection to whatever AP is in range?
May 17 15:42:19 <SpeedEvil> Open APs, sure.
May 17 15:42:23 <SpeedEvil> Legality may vary.
May 17 15:43:09 <LuitvD> lol, would be nice if the map that didn't fit on microSD did actually fit...
May 17 15:43:13 <webjames> SpeedEvil: you are implying that it would be nice if it had more data
May 17 15:43:38 <LuitvD> I guess he is
May 17 15:44:08 <webjames> SpeedEvil: also the more clever features we can say the phone has, like 'updating map information when in wifi range of the country you are in'
May 17 15:44:11 <LuitvD> would be cool to have the global street map on one microSD card
May 17 15:44:30 <SpeedEvil> Well. The nearest city to me had one road.
May 17 15:44:44 <webjames> vectors take up little space
May 17 15:44:49 <SpeedEvil> I added a couple more, but it's still a long way from a useful streetmap in many places.
May 17 15:45:06 <SpeedEvil> But for a streetmap, you've got (ballpark) a vector a person.
May 17 15:45:10 <webjames> SpeedEvil: i named the road i live on, but i still cannot see the name
May 17 15:45:28 <SpeedEvil> the map takes about a week to update
May 17 15:45:34 <webjames> ahh right
May 17 15:45:43 <webjames> i'll look now
May 17 15:46:00 <SpeedEvil> the '+' on the map viewer shows alternate renderings, some of which may update faster
May 17 15:46:58 <aloril2> bad implementation of IPoV could be vulnerable for worms spreading trough GSM network
May 17 15:47:41 <webjames> SpeedEvil: my road is now named!
May 17 15:48:16 <SpeedEvil> :)
May 17 15:49:04 <mjr> aloril2, good luck having the user wait for an interruption in a voice call to transfer a viable worm ;)
May 17 15:50:58 <pH5> mjr: regarding accuracy, http://www.gpsvisualizer.com/display/1179416906-14022-85.178.222.100.html, I did not swim.
May 17 15:54:27 <DukeOfURL> Anyone seen this?  http://rss.slashdot.org/~r/Slashdot/slashdot/~3/117457150/article.pl
May 17 15:55:20 <aloril2> mjr: yeah: "hi foo" beep beep quiet/synth voice saying data being transmitted "what!" click
May 17 15:56:09 <aloril2> example of worm reading address books and trying to automatically spreading
May 17 15:56:37 <aloril2> (still of course good idea to consider security of course)
May 17 15:58:08 * aloril2 thinks that if updating openstreetmap from Neo is made easy and assuming significant % of people enable it and enough Neos sold -> big boost for openstreetmap
May 17 15:59:28 <mjr> yes, that would be nice; Somebody (TM) needs to do a friendly OSM tracking app
May 17 15:59:48 <XorA|gone> mjr: nicely volunteeered
May 17 16:00:25 <mjr> was there something already for Maemo, I wonder. Does the mapper do this sort of thing too?
May 17 16:00:32 <mjr> XorA|gone, or non- :]
May 17 16:01:57 <SpeedEvil> Evne just raw (anonymised) position data thrown to a server and released to OSM after a month would be a help.
May 17 16:02:20 <webjames> i think openstreetmap is such a good idea, and i'm sure with the support of the neo community it could be truly great, we'll just have to make sure it's well documented and it's easy to get the data from the neo to the map
May 17 16:02:55 <mjr> SpeedEvil, yeah I suppose it would. It'd probably be easy to get people to sign up for that.
May 17 16:03:07 <mjr> since they'd have to do almost nothing actively
May 17 16:04:24 <mjr> cooking such data from another user might not be too fun but :]
May 17 16:04:48 <SpeedEvil> From one user, no.
May 17 16:05:01 <SpeedEvil> From 10 users who all went down the same road, it gets easier
May 17 16:06:01 <SpeedEvil> All 10 users going down a path at >50MPH, from another designated road, without stopping, and it's probably safe to auto-tag it preliminarily as a proper road of some sort.
May 17 16:06:23 <mjr> *nod*
May 17 16:06:25 <aloril2> it could even be a 'game': be first to map streets: as time goes, you need to go to more and more exotic places
May 17 16:06:42 <so_solid_moo> make it so the user benefits from entering the information
May 17 16:06:45 <mjr> aloril2, oo. Game. Scores. Experience points!
May 17 16:06:56 <aloril2> keep highscore!
May 17 16:06:57 <so_solid_moo> e.g., they identify the street, they then can mark out hangouts and arrange nights out there and stuff
May 17 16:07:19 <mjr> "You have reached level 4 and learn the spell 'Mark public office'"
May 17 16:07:48 * aloril2 thinks he just invented basic idea on how to get a lot mapped for osm fast ^
May 17 16:09:21 <webjames> i like the ideas!
May 17 16:13:21 <aloril2> lets assume % of people updating osm * number of Neos sold >= 10000 within 1E9 people -> every 100,000 person mapping
May 17 16:13:41 <woglinde> *g*
May 17 16:14:31 <aloril2> -> 50 players in Finland competing about highscores hmm..
May 17 16:16:02 <aloril2> of course if its 10x higher including all other mobiles having gps + support for this game -> probably quite fast coverage
May 17 16:16:31 <SpeedEvil> Needs prizes
May 17 16:17:04 <pH5> yeah, like vacations in unmapped areas
May 17 16:19:24 <webjames> pH5 love it
May 17 16:19:47 <webjames> you've won a trip to map the Bahamas 
May 17 16:20:20 <webjames> ..waypoint: beach bar, NOTE: sells nice rum and coke
May 17 16:21:05 <happycube> lol
May 17 16:23:35 <aloril2> good position in highscore list alone is very motivating 'prize' for many people
May 17 16:25:25 <aloril2> there should also be some score for being second/third/etc.. person mapping street (1/nth for example: 1.0 for 1st, 0.5 for 2nd, etc..)=
May 17 16:28:05 <webjames> given this we want quality, not just quantity
May 17 16:31:32 <aloril2> yeah, this get score from bein second, etc.. adds redundancy: game could ask user to confirm earlier street name (or fix it), if you are 10th or so it would just automatically add to score (and add yet another redundant datapoint)
May 17 16:34:09 <aloril> (script) wiki RecentChanges: [[Main_Page/es]] [[FIC/es]]
May 17 16:41:33 <aloril2> risk: cheaters uploading data from commercial database instead of being personally present (but if data is time stamped this can be detected)
May 17 17:10:34 <minime> counter
May 17 17:10:34 <aloril> http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 65%) P1: a week, 2 days 18:49:23 (9.784 +-5.0 days) (1207;208)
May 17 17:12:42 <aloril2> you could even play this game against yourself: have you yet visited all streets in your home city?
May 17 17:12:57 <LuitvD> XD
May 17 17:13:39 <SpeedEvil> That can be surprisingly hard.
May 17 17:13:55 <SpeedEvil> You've got to tag all junctions to make that possible.
May 17 17:14:17 <LuitvD> would be fun to have a program that tells you you're driving differently when going to work and avioding thick traffic :P
May 17 17:14:35 <LuitvD> or a program that recognizes where you're going
May 17 17:18:14 <SpeedEvil> Collaborative traffic monitoring would be nice
May 17 17:18:50 <aloril2> marking public buildings, etc.. should give score, not only from mapping+tagging streets
May 17 17:39:37 * balrog-kun just picked up a grey packet from the post office...
May 17 17:39:48 <Stephmw> tick-tock?
May 17 17:40:12 <balrog-kun> it doesn't tick ;) it's got lots of duct tape on it though
May 17 17:42:07 <balrog-kun> umm it starts looking like there's only duct-tape
May 17 17:49:16 <LuitvD> lol
May 17 17:51:08 <balrog-kun> yay, it's the black one
May 17 17:51:19 <LuitvD> XD
May 17 17:51:31 <LuitvD> black P1 device?
May 17 17:52:10 <balrog-kun> yep
May 17 17:52:39 <LuitvD> why is everyone getting P1 devices... :(
May 17 17:52:52 <balrog-kun> S/N: 000351
May 17 17:53:04 <LuitvD> did you recieve a P0 too?
May 17 17:53:14 <balrog-kun> nope
May 17 17:53:44 <Kaloz> balrog-kun: 321 and 322 here :)
May 17 17:53:57 <balrog-kun> dammit ;)
May 17 17:54:03 <Marex-notebook> ;-)
May 17 17:54:13 <Marex-notebook> 000000 here ;-)
May 17 17:54:15 <Kaloz> wondering if Sean or Harald got 001
May 17 17:54:17 <Kaloz> :P
May 17 17:54:23 <PBeck> re
May 17 17:54:31 <Kaloz> or maybe Mr. mickey ;P
May 17 17:55:12 <balrog-kun> has anyone taken off the screen protector already? :p
May 17 17:55:40 <Marex-notebook> balrog-kun, try taking off the digitizer ... that might be much more fun ;)
May 17 17:56:41 <balrog-kun> for sure
May 17 17:57:34 <Marex-notebook> balrog-kun, ok, try not grooving a hole into the LCD with a screwdriver
May 17 17:57:45 <Marex-notebook> ... I did that accidentally on TT ;)
May 17 17:57:51 * lrak knocks on current-devel
May 17 17:58:04 <Pander> hi all
May 17 17:58:14 <balrog-kun> did it make it look prettier?
May 17 17:58:26 <balrog-kun> hmm the stickers are upside down
May 17 18:02:12 * k-s[AWAY_WORK] is back.
May 17 18:16:20 * CoreDump|afk utters something about "358"
May 17 18:29:00 <roh> heh.. i'll really need to ask where the numbers are started at.. i thought we only sent about 50 bv4 out
May 17 18:29:51 <happycube> there was a second batch of ~150 gta01bv4's i thought
May 17 18:29:59 <happycube> maybe 200 now if the other 50 got fixed?
May 17 18:30:38 <roh> eventually if you add the bv2, 3 and 4 its 350-300
May 17 18:30:45 <happycube> *nod*
May 17 18:30:56 <roh> eh 250-300
May 17 18:31:01 <happycube> how many are going to be in the may ~24 run?
May 17 18:31:42 <happycube> i wonder how gta02 is coming along, too ;)
May 17 18:31:42 <roh> dunno. thats happening in mainland china, i'm in .tw right now
May 17 18:31:45 <happycube> ok
May 17 18:32:09 <tobmaster> hi, are there any chances to get an openmoko phone preordered somewhere?
May 17 18:32:09 <aloril> tobmaster: See http://wiki.openmoko.org/wiki/FAQ (Hopefully this link answers your question. This autotrigger for new nick and question during first 15 minutes done only once.)
May 17 18:32:34 <tobmaster> Neo1973 even
May 17 18:33:37 <roh> happycube when we publicly announce the p1 to be sold, the webshop will be up, the exact price will be announced and the preview about the gta02 will be there, so everyone can decide to either support us by buying a 'release-gta01' or to wait for gta02
May 17 18:33:50 <happycube> sounds good :0
May 17 18:33:50 <happycube> :)
May 17 18:34:09 <aloril> (script) wiki RecentChanges: [[Wish_List]] [[Wishlist:ANARM]]
May 17 18:34:46 <roh> thats also why we currently don't do preorders. you should exactly know what you'd get before placing an order. at least that is what we think thats fair and good.
May 17 18:40:56 <tobmaster> roh: will the gta01 be upgradeable to gta02?
May 17 18:41:12 <tobmaster> or do you expect/know about hardware changes?
May 17 18:43:09 * sannes is so looking forward to ordering one :)
May 17 18:44:27 <CIA-24> openmoko: 03mickey * r2001 10/trunk/src/target/OM-2007/openmoko-libs/libmokogsmd/ (moko-gsmd-connection.c moko-gsmd-connection.h): libmokogsmd: register signal "pin-requested"
May 17 18:44:54 <CIA-24> openmoko: 03mickey * r2002 10/trunk/src/target/OM-2007/applications/openmoko-dialer/ (4 files in 2 dirs): openmoko-dialer: prepare to show PIN dialog on demand
May 17 18:46:43 <roh> tobmaster that information will be released when we start selling the gta01
May 17 18:47:56 <roh> i really cannot comment much further about this
May 17 18:48:35 <tobmaster> roh i understand thanks :)
May 17 18:49:06 <roh> tobmaster take a look at the wiki, a lot if infos are already there
May 17 18:49:13 <roh> s/if/of/
May 17 18:49:13 <apt> roh meant: tobmaster take a look at the wiki, a lot of infos are already there
May 17 19:08:17 <aloril2> tobmaster: gta02==P1+, gta01==P1 at [[Neo1973]] page at wiki (linked from above FAQ page)
May 17 19:12:37 <aloril2> different hardware, there should be discount but no idea what kind or how much
May 17 19:13:05 <aloril2> software you can always upgrade of course
May 17 19:49:08 <tobmaster> thanks guys
May 17 19:52:45 <Bish> thanks for what!?
May 17 19:54:19 <LuitvD> thank goodness, microsoft can still do something right :P
May 17 19:55:14 <LuitvD> though it's not really usefull in correctly designed finger-touch-GUIs, Microsoft's Shift is quite a nice feature
May 17 19:59:37 <SpeedEvil> last recentchanges
May 17 19:59:52 <SpeedEvil> oops
May 17 20:04:07 <DukeOfURL> LuitvD: wonder if there will be licensing issues with this...
May 17 20:04:58 <LuitvD> DukeOfURL: I bet it's Windows Mobile exclusive...
May 17 20:05:27 <DukeOfURL> I mean could we write our own without license issues.
May 17 20:05:52 <LuitvD> hmm, why would you? 
May 17 20:06:31 <DukeOfURL> for using pen-based apps with finger touch
May 17 20:06:44 <LuitvD> ah well...
May 17 20:07:14 <LuitvD> I guess they already patented the technology
May 17 20:07:47 * SuN is in Europe and consequently does'nt care :D
May 17 20:07:54 <SuN> doesn't either.
May 17 20:08:49 <Bish> i would like to buy my first free mobilephone :/
May 17 20:15:58 <raynet> we can always disable the touch interface and make it voice controlled
May 17 20:17:12 <noidd> raynet: you involved with UK raynet org?
May 17 20:18:16 <raynet> noidd: no.. i am named after a star in StarControl I
May 17 20:18:21 <raynet> StarCOntrol II that is
May 17 20:18:41 <LuitvD> I wonder if laptop-touchpad-style controlled cursor on a mobile device would work...
May 17 20:18:54 <LuitvD> just using the touch screen in a touchpad-like way...
May 17 20:19:07 <guaqua> the joystick on my nokia works
May 17 20:19:17 <guaqua> i wouldn't mind that kind of an interface in a mobile
May 17 20:19:39 <LuitvD> wouldn't mind, or wouldn't use?
May 17 20:20:30 <LuitvD> maybe for precise clicking hitting a button and a cursor appears, and the touchscreen becomes a touchpad... think it'll work...
May 17 20:20:30 <guaqua> i like it, that's what i meant
May 17 20:21:00 <LuitvD> wonder wether it's possible to hack it into openmoko...
May 17 20:21:11 <LuitvD> touchscreen < > touchpad switcher
May 17 20:21:55 <LuitvD> especially for Neo1973's high DPI screen...
May 17 20:22:30 * LuitvD is off for a drink downstairs
May 17 20:39:38 <Bish> lol the text input ideas a pretty cool xD
May 17 20:41:04 <xFallenAngel> a bit off topic, but does anyone know technical details of the HDLC protocol?
May 17 20:41:39 <DukeOfURL> flag (0111111) data CRC flag
May 17 20:41:49 <DukeOfURL> s/0111111/01111110/
May 17 20:41:50 <apt> DukeOfURL meant: flag (01111110) data CRC flag
May 17 20:42:23 <DukeOfURL> USART chips will do this low-level stuff for you
May 17 20:42:55 <DukeOfURL> constant flags between frames
May 17 20:44:13 <DukeOfURL> HDLC is a knock off of IBM's SDLC protocol--the lowest layer of SNA
May 17 20:45:08 <xFallenAngel> I'm not quite clear in which order I and RR frames are to be sent
May 17 20:45:38 <xFallenAngel> i.e If I have a window size 3, the sender sends 3 I frames. What does the reciever send back?
May 17 20:45:43 <xFallenAngel> and when?
May 17 20:45:51 <DukeOfURL> I is an information (data) frame, RR is "receiver ready" response
May 17 20:46:21 <DukeOfURL> RR contains the acknowledged packet number
May 17 20:46:40 <DukeOfURL> window size is typically 7
May 17 20:47:04 <xFallenAngel> so the reciever sends a RR frame for each I frame it recieves?
May 17 20:47:49 <DukeOfURL> no, when it is ready to send, it sends the highest acknowledged frame number
May 17 20:48:18 <DukeOfURL> both ends can be sending I frames simultaneously
May 17 20:48:59 <DukeOfURL> (synchronous circuits are typically full duplex)
May 17 20:49:52 <DukeOfURL> HDLC is a standard--you should be able to get the specs.  "Red Book" comes to mind.
May 17 20:51:27 <DukeOfURL> if window=3 then A can send 3 frames to B before B sends an RR
May 17 20:52:11 <xFallenAngel> so B definitly sends an RR after three frames. What happens if the third frame is lost in transfer?
May 17 20:52:22 <DukeOfURL> timeout, resend
May 17 20:52:42 <DukeOfURL> or B acks the second frame and then A retransmits #3
May 17 20:53:12 <DukeOfURL> this goes back 30 years for me--you really need the spec.
May 17 20:53:54 <xFallenAngel> I'll see what I can find out of them. Thanks so far, you helped me a lot :)
May 17 21:04:44 <xFallenAngel> argh it costs 216 CHF
May 17 21:20:44 <Bish> ausser schweiz xFallenAngel ?
May 17 21:21:02 <xFallenAngel> nee, mein nachname ist kewisch :)
May 17 21:21:07 <xFallenAngel> aber deutschland
May 17 21:21:16 <Bish> hier auch ;)
May 17 21:21:38 <xFallenAngel> auch was?
May 17 21:21:56 <Bish> jop jop aber ich glaub uns ist es nit gestattet hier was anderes als englisch zu reden :P also qry
May 17 21:22:16 * CoreDump|afk kindly reminds the audience that is is an English channel ;)
May 17 21:22:29 <Bish> i just did before CoreDump|afk :P
May 17 21:22:43 * guaqua thinks CoreDump is afk
May 17 21:22:51 <guaqua> so he can't really think in irc ;)
May 17 21:23:00 <Bish> xD 
May 17 21:23:18 * CoreDump|home is typing via a brain<->PC interface. hah! take this!
May 17 21:24:05 <Bish> is this like hexinput -_- ?
May 17 21:24:35 <CoreDump|home> no, just my usual excuse for chatting while "afk"
May 17 21:24:37 <CoreDump|home> ;)
May 17 21:26:50 <Sufflope> counter
May 17 21:26:51 <aloril> http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 65%) P1: a week, 2 days 14:33:07 (9.606 +-5.0 days) (1208;208)
May 17 21:42:02 <Bish> lolz .. dasher is fun
May 17 22:34:09 <aloril> (script) wiki RecentChanges: [[Wishlist:Text_Input]] [[Wish_List]]
May 17 22:43:08 <Stephmw> are there any raw performance numbers for the graphics hardware?
May 17 22:50:05 <Bish> http://wiki.openmoko.org hardware Stephmw ?
May 17 22:59:32 <Stephmw> Bish: I'm looking there as well :)
May 17 23:00:33 <Stephmw> looking at the wishlist, people are wanting accelerater gfx, so I'm guessing it's probably not terribly fast
May 17 23:01:27 <balrog-kun> it's only a dumb framebuffer, it's as fast as RAM
May 17 23:02:11 <Stephmw> cheers
May 17 23:02:18 * Stephmw ticks a box on his todo list
May 17 23:02:56 <LuitvD> :)
May 17 23:06:56 <Bish> rrrrrrrrright.. but it will be faster Stephmw... i hope
May 17 23:07:25 <Stephmw> *accelerated even
May 17 23:07:48 <Stephmw> Bish: aye, it's no big problem
May 17 23:07:51 <Elrond> Any hints on how to mark new bad blocks on flash?
May 17 23:08:12 <balrog-kun> you can add an accelerator, yes, it's a bit difficult if it's not inside the CPU, so they would probably rather change the CPU
May 17 23:08:22 <balrog-kun> currently the only graphics hardware is really the physical LCD and the dma channel that connects it to RAM, but the LCD and the dma channel don't affect system speed or system running and are faster than a human eye
May 17 23:08:59 <Bish> :/ i really would like to buy the final version of oopenmoko :P
May 17 23:09:02 <Stephmw> balrog-kun: I was wondering about raw data throughput for say, full screen wipes, etc
May 17 23:09:04 <Bish> im in love with that project
May 17 23:10:06 <balrog-kun> Stephmw: a full screen wipe takes same time as writing 480*640*2 bytes into RAM
May 17 23:10:45 <SpeedEvil> The DMA channel does affect performance.
May 17 23:11:15 <SpeedEvil> It is pulling 600K*60 - or 40M/second from main RAM
May 17 23:11:22 <SpeedEvil> this reduces RAM bandwidth
May 17 23:11:34 <SpeedEvil> Not to mention hurting power-wise.
May 17 23:11:52 <balrog-kun> afaik part of the bandwidth is reserved for the LCD DMA channel and for the 4 remaining DMA channels
May 17 23:12:17 <SpeedEvil> For many apps. Simply going to 240*320 mode will give quite enough performance.
May 17 23:13:06 * balrog-kun also doesn't think a gfx accelerator would be of much use
May 17 23:13:26 <balrog-kun> and certainly a pain to drive
May 17 23:14:55 <Elrond> It might help. But I doubt it.
May 17 23:16:02 <sparq> It's possible that using a CPU with a graphics-oriented extended instruction set would be nearly as good as a low end graphics accelerator
May 17 23:16:25 <sparq> Does ARM have anything like MMX/SSE?
May 17 23:16:42 <Stephmw> SpeedEvil: doing a full screen blit is a worst-case scenario for what I have in mind
May 17 23:16:52 <balrog-kun> it might help but not in this kind of scenario i think
May 17 23:16:56 <Stephmw> SpeedEvil: dirtyrects and other optimisations would come into it
May 17 23:16:58 <balrog-kun> sparq: generic ARM nope
May 17 23:17:03 <SpeedEvil> Yeah
May 17 23:17:08 <sparq> that's what I figured
May 17 23:17:12 <balrog-kun> sparq: there's iwMMXt but S3C doesn't have it
May 17 23:18:04 <SpeedEvil> A really simple 2D accellerator would be handy generally.
May 17 23:18:21 <balrog-kun> IMHO the only scenario where an accelerator would make much difference is with these accelerators that hold the whole 3d vertex arrays in their own memory and can make opeations on the whole array and render it automatically
May 17 23:18:24 <sparq> balrog-kun: Hmm... if they bump the CPU speed for the consumer version, does anyone know what CPU they would use? Just the same one, but with a higher clock frequency?
May 17 23:18:35 <balrog-kun> all 2D operations on a 640x480 probably don't take so long
May 17 23:18:48 <SpeedEvil> Except for smooth scroll
May 17 23:19:14 <sparq> and possibly anti-aliasing
May 17 23:19:16 <SpeedEvil> That sucks huge amounts of CPU if you do it that wya
May 17 23:19:26 <SpeedEvil> There is no need for anti-aliasing
May 17 23:19:31 <SpeedEvil> That's done in hardware.
May 17 23:19:36 <LuitvD> :)
May 17 23:19:36 <balrog-kun> sparq: afaik GTA02 is an s3c2440 at 400 MHz, but OMAPs are considered for after the first public release
May 17 23:19:36 <sparq> Oh?
May 17 23:19:52 <SpeedEvil> Unless you have the screen at 10cm with a magnifying glass of course
May 17 23:19:52 <LuitvD> SpeedEvil: that's done in your eye :P
May 17 23:20:07 <LuitvD> not really 'hard'ware
May 17 23:20:13 <SpeedEvil> Wetware
May 17 23:20:17 <LuitvD> XD
May 17 23:20:33 <balrog-kun> SpeedEvil: a smooth scroll is a serie of memcpy()s and i don't think it takes really long at 640x480x2 bytes
May 17 23:20:50 <sparq> Anti-aliasing actually does make a difference for readability even when you can't actually see the pixels.
May 17 23:20:55 <SpeedEvil> No, but it sucks a whole lot of CPU.
May 17 23:21:03 <Stephmw> SpeedEvil: doesn't everyone use the screen at 10cm with a magnifying glass?
May 17 23:21:37 <SpeedEvil> sparq: Given that most 'phone' fonts are going to be 30*20 or so... I don't think antialiasing is a real problem.
May 17 23:22:00 <Stephmw> that's something else. phone fonts are typically hand-optimised for small sizes
May 17 23:22:06 <SpeedEvil> The case where it would actually be useful - smaller fonts - is going to be quite rare.
May 17 23:22:07 <Stephmw> rarely auto-hinted
May 17 23:22:14 <balrog-kun> in fact in the pictures Koen posted on flickr he forgot to expose the included magnifying glass :)
May 17 23:22:37 <SpeedEvil> Handy tip. Get a 2.8" rectangle of paper.
May 17 23:22:42 <sparq> SpeedEvil: Maybe not for the occasional use one would expect for a phone, but I had to learn the hard way that well hinted fonts have a huge impact on my attention span and eye fatigue
May 17 23:22:47 <SpeedEvil> Now set a window to it.
May 17 23:22:59 <SpeedEvil> back off to normal phone viewing distance.
May 17 23:23:33 <SpeedEvil> sparq: On what displays?
May 17 23:23:43 <sparq> SpeedEvil: Pretty much any display, or even printed.
May 17 23:24:01 <SpeedEvil> This display has 9 times the dots per square inch of almost any desktop display.
May 17 23:25:02 <sparq> SpeedEvil: It's subtle, and might not even affect most people, but I notice it even at 3600 dpi, or whatever the effective resolution for trade paperback books is
May 17 23:25:17 <SpeedEvil> Ah. You have golden eyes.
May 17 23:25:27 <sparq> SpeedEvil: nope. Dyslexia.
May 17 23:25:45 <sparq> SpeedEvil: I can't actually see the hinting, but I get a headache when it's not there
May 17 23:26:28 <SpeedEvil> I'd like to see actual tests - I have doubts that you'd reliably be able to tell at 300dpi at normal phone distances.
May 17 23:26:59 <SpeedEvil> As most human eyes simply don't have the resolution to pick up if it's aliased or not.
May 17 23:27:04 <sparq> I guess I must sort of see it, but not well enough to describe it.
May 17 23:27:56 * sparq shrugs
May 17 23:28:00 <sparq> maybe not
May 17 23:29:48 <balrog-kun> i rarely see non antialiased fonts anywhere on computers these days
May 17 23:30:00 <balrog-kun> which is cool
May 17 23:30:02 <happycube> i see badly antialiased fonts on windows all the time though
May 17 23:30:31 <happycube> i played with a vista rc and i was like "the font's *still* suck?  damn."
May 17 23:31:03 <LuitvD> balrog-kun: why the double negative? :P non anti-aliased is called: aliased ;)
May 17 23:33:20 <sparq> Hmm... I guess if they are hand-tweaked bitmaped fonts, that would be ideal. No nasty aliasing, and we can save power by not calculating Bezier curves.
May 17 23:33:24 <balrog-kun> aliased makes me think of shell aliases or something :)
May 17 23:33:54 <balrog-kun> LuitvD: actually a digital image is never 100% aliased
May 17 23:34:08 <balrog-kun> anti-aliasing just reduces the effect
May 17 23:34:21 <SpeedEvil> It can be completely aliased.
May 17 23:34:30 <LuitvD> 8-bit :D
May 17 23:34:33 <balrog-kun> not with square pixels
May 17 23:34:33 <SpeedEvil> Just alias it over the whole screen.
May 17 23:34:42 <SpeedEvil> Not very useful of course.
May 17 23:34:52 <LuitvD> lol
May 17 23:35:00 <LuitvD> anyways, I'm gone...
May 17 23:35:03 <LuitvD> need some rest
May 17 23:35:22 <LuitvD> it's half past one already... :S
May 17 23:35:56 * Stephmw waves
May 17 23:42:09 <Elrond> Any P0 owner still awake?
May 17 23:42:51 <Elrond> Or someone any idea where in /sys I should "echo 0" to shut the phone down _immediately_?
May 17 23:51:47 * balrog-kun doesn't know, but one of the poweroff switches should be able to do that in a cross-platform way
May 17 23:51:59 <balrog-kun> poweroff -n was it?
May 17 23:52:20 <balrog-kun> or -f
May 17 23:54:36 <Elrond> balrog-kun - The fs was completely dead.
May 17 23:55:44 <Elrond> balrog-kun - I even scp'd halt.sysvinit to /tmp (which is on tmpfs) in the hope, that libc6 would be enoughly paged in. To no avail. I finally removed the battery.
May 17 23:58:00 <rwhitby> morning
May 17 23:59:19 <balrog-kun> Elrond: ouch
May 18 00:00:14 <CIA-24> openmoko: 03alphaone * r2003 10/trunk/src/target/OM-2007/applications/openmoko-footer/ (5 files in 2 dirs): 
May 18 00:00:14 <CIA-24> openmoko: * src/Makefile.am, src/taskitem.c, src/taskitem.h, src/taskmanager.c:
May 18 00:00:14 <CIA-24> openmoko: Add support for taskitems (consisting of an image and a label)
May 18 00:00:17 <balrog-kun> well, i know how to suspend Linux in a cross-platform manner with only "echo" and /sys
May 18 00:04:08 <Elrond> balrog-kun - Oh, how?
May 18 00:10:23 <Elrond> (anyway. I need to go to bed. I'll read the answers tomorrow.)
May 18 00:27:59 <balrog-kun> oops
May 18 00:28:11 <balrog-kun> Elrond: echo standby > /sys/power/state
May 18 00:28:24 <balrog-kun> this is immediate, and "standby" is equivalend with "mem"
May 18 00:28:41 <balrog-kun> but i don't know if there's a similar thing for completely powering off
May 18 00:35:09 <aloril> (script) wiki RecentChanges: [[User:Minime]] [[Messages]] [[Dialer]] [[Bluetooth_powered_Multi-SIM_support]] [[Wish_List]] [[Wishlist:Tagging]] [[Talk:Messages]] [[Wishlist:Text_Input]] [[Buying_Interest_List]]
May 18 02:11:55 <CIA-24> openmoko: 03laforge * r2004 10/trunk/src/target/kernel/patches/gta01-pcf50606.patch: fix initialization order: first register input device, then enable interrupt. Otherwise the interrupt might come in too early and cause kernel oops. (Matt Hsu)
May 18 02:26:17 <CIA-24> openmoko: 03laforge * r2005 10/trunk/src/target/u-boot/patches/uboot-20070516-mfsl.patch: this patch is now in u-boot mainline
May 18 02:53:09 <aloril> (script) wiki RecentChanges: [[User:Rich]]
**** ENDING LOGGING AT Fri May 18 02:59:56 2007