| 00:24.13 | dc_ | very. |
| 00:24.32 | dc_ | pb_: have a look if you can later, just to double check. |
| 00:25.22 | dc_ | I need sleep now. |
| 00:25.22 | dc_ | gotta go.... |
| 00:25.22 | dc_ | out tomorrow. |
| 00:25.22 | dc_ | scary though. |
| 01:50.51 | *** join/#gpe paul (~paul@itg-gw.cr008.cwt.esat.net) |
| 04:01.46 | *** join/#gpe TheMasterMind1 (~aman@h-68-166-68-149.MCLNVA23.covad.net) |
| 05:09.49 | *** join/#gpe walters (walters@verbum.org) |
| 08:29.43 | *** join/#gpe zecke (~ich@pD9E7FD4E.dip.t-dialin.net) |
| 08:49.49 | *** join/#gpe HET2 (~diman@chello080110092229.507.15.vie.surfer.at) |
| 09:24.01 | *** join/#gpe PaxAnima (~petter@v058b.studby.ntnu.no) |
| 09:42.01 | *** join/#gpe pb_ (~pb@pc2-cmbg4-3-cust239.cmbg.cable.ntl.com) |
| 10:00.50 | pb_ | morning |
| 10:02.29 | PaxAnima | morning, pb_ |
| 10:02.50 | zecke | moin |
| 10:10.43 | *** join/#gpe florian (~fuchs@pD9E5F6CD.dip.t-dialin.net) |
| 10:10.51 | florian | morning |
| 10:15.05 | pb_ | hi zecke |
| 10:15.06 | pb_ | hi florian |
| 10:15.14 | pb_ | hi PaxAnima |
| 10:16.55 | pigeon | hi pb_ |
| 10:17.05 | pigeon | hi florian |
| 10:20.49 | pb_ | hey pigeon |
| 10:21.25 | pigeon | hi all gpers. |
| 10:26.02 | florian | do we have a new greader package? |
| 10:28.03 | florian | hmm... we don't... |
| 11:18.01 | *** join/#gpe florian_ (~fuchs@p509296F2.dip.t-dialin.net) |
| 12:24.32 | *** join/#gpe HET2 (~diman@chello080110092229.507.15.vie.surfer.at) |
| 13:36.21 | *** join/#gpe ade|desk (~adrian@unimatrix.ex.ac.uk) |
| 14:00.11 | *** join/#gpe HET2 (~diman@chello080110092229.507.15.vie.surfer.at) |
| 14:51.52 | jblinux | florian left again ? :( |
| 15:19.54 | pb_ | no doubt he'll be back. |
| 15:29.23 | jblinux | pb_: he will? |
| 15:29.36 | pb_ | well, eventually, yeah. |
| 15:29.44 | pb_ | I expect he'll be back later today. |
| 15:31.02 | jblinux | pb_: ok. Thanks for letting me know pb_ . Would really hope he could me with the wireless driver problem... |
| 15:31.18 | pb_ | you could send him email, or leave a message for him with ibot. |
| 15:31.41 | jblinux | pb_: how exactly do I do that? |
| 15:31.55 | pb_ | ibot, botmail for jblinux: like this. |
| 15:32.55 | jblinux | just replace my nick with his, and my message to him to replace "like this"? |
| 15:33.03 | pb_ | yep |
| 15:33.26 | jblinux | pb_: thanks alot |
| 15:34.23 | jblinux | ibot, botmail for florian: |
| 15:44.43 | *** join/#gpe walters (walters@verbum.org) |
| 15:48.21 | jblinux | I got a ibot window pop up, can I just type more messages to him in that? |
| 15:48.39 | moray | almost certainly |
| 15:48.56 | pb_ | that's just a /query with ibot. so yeah, but you need to say "botmail for florian" each time. |
| 15:49.19 | jblinux | ok, thanks. |
| 15:50.59 | pb_ | ibot, botmail for florian: I just tried hostap-cs 0.0.3 on my h3900, and it seems to work okay. gpe-aerial doesn't like it much, though - I get many "SET failed on device wlan0 ; Invalid argument." messages scrolling up the screen. |
| 18:05.19 | *** join/#gpe florian (~fuchs@p509290C6.dip.t-dialin.net) |
| 18:05.23 | florian | re |
| 18:08.43 | florian | jblinux: here i am :-) |
| 18:09.06 | jblinux | florian: Hi florian! |
| 18:09.20 | florian | jblinux: hi, what's up? |
| 18:10.40 | jblinux | florian: been having trouble getting the orinoco silver card to work as what I need. Couldn't get a driver that will do AP scanning, report infomration such as signal strengths |
| 18:12.40 | jblinux | florian: any idea which driver out there I could achieve that? Got a iPAQ 3630, if that helps... |
| 18:13.03 | florian | jblinux: hmm, the silver cards should work with orinoco_cs quite well. i'm not shure if it's able to do active scanning... but passive (including monitor) should work quite well. |
| 18:13.42 | *** join/#gpe jg_ (~jg@h005018015b26.ne.client2.attbi.com) |
| 18:14.43 | florian | jblinux: do you only need automatic connection to the next accesspoint or really scan for all available? |
| 18:14.53 | florian | jg_: hello |
| 18:15.09 | jg_ | florian: hi |
| 18:15.17 | jblinux | florian: scan all the avaiables |
| 18:15.36 | jblinux | florian: I was trying to get some that mwvlan could do.... |
| 18:17.04 | florian | jblinux: ok, this i din't ever try - use the orinoco_cs from familiar, this includes the monitor patch. if you want to compile the most recent version you need the monitor patch for it... |
| 18:17.28 | jblinux | florian: orinoco_cs works great for my card. |
| 18:17.40 | jblinux | florian: and it does have monitor mode |
| 18:17.51 | pb_ | hi florian |
| 18:18.07 | florian | hi pb_ |
| 18:18.21 | pb_ | hi jg_ |
| 18:19.17 | florian | pb_: would you like to do a small check of hostap_cs? |
| 18:19.19 | jg_ | pb_: ho. |
| 18:19.27 | florian | jblinux: correct |
| 18:19.31 | pb_ | florian: sure, just tell me what to do. |
| 18:20.52 | florian | pb_: i assume it's the wlan0 device... |
| 18:21.02 | florian | pb_: ifconfig wlan0 up |
| 18:21.19 | pb_ | yup |
| 18:21.52 | florian | pb_: iwconfig wlan0 mode monitor |
| 18:22.07 | florian | pb_: iwconfig wlan0 channel 1 |
| 18:22.51 | florian | pb_: if no error messages so far, repeat last step with other channels |
| 18:22.54 | pb_ | ok |
| 18:23.16 | pb_ | it doesn't like the "channel" command. |
| 18:23.21 | pb_ | ~ # iwconfig wlan0 mode monitor |
| 18:23.23 | pb_ | ~ # iwconfig wlan0 channel 1 |
| 18:23.24 | pb_ | Error for wireless request "Set Frequency" (8B04) : |
| 18:23.24 | pb_ | <PROTECTED> |
| 18:23.24 | pb_ | ~ # |
| 18:23.49 | jblinux | florian: from kismetwireless, it seems I will be able to scan SSIDs around with channel_hopping, but monitor mode doesn't seem to measure signal strengths well (http://www.kismetwireless.net/Forum/General/Messages/1051021867.932186). Any idea how mwvlan's implemenatition is different? |
| 18:24.49 | florian | hmm... this should be the same message that you get running gpe-aerial, right? |
| 18:25.10 | pb_ | yep, looks the same. |
| 18:25.18 | jblinux | florian: gpe-aerial? |
| 18:26.10 | florian | jblinux: a wireless lan lool for gpe. it's designed to find and select a desired wireless lan. |
| 18:29.10 | florian | jblinux: i din't read the whole thread so far, but a part of the info you get from kismt should be from the header added by the card. this contains a signal strength info... let me look how this works in prismstumbler.... |
| 18:29.36 | jblinux | florian: thanks florian. |
| 18:30.04 | florian | pb_: try to switch to ad-hoc mode and try to change channel here... |
| 18:31.27 | florian | jblinux: no problem :-) |
| 18:32.10 | jblinux | florian: also, I've been very interested in mwvlan or mwavelan_cs because there is actually an API that could come with it...it kinda helps alot since I could do scanning and uses those values right as a part of my program. |
| 18:35.35 | florian | jblinux: don't they use the standard linux wireless api? |
| 18:38.43 | jblinux | florian: I don't know about the standard wireless api, but this one is call MAPI so I suppose it's different? (http://www.cs.umd.edu/~moustafa/mapi/mapi.html) |
| 18:42.03 | florian | jblinux: i think this should work "wireless extensions" is what i meant - this is supportet by orinoco_cs. |
| 18:45.51 | pb_ | florian: ah, selecting monitor mode seems to have crashed the card. |
| 18:45.55 | pb_ | looks like I need to reboot. |
| 18:46.19 | pb_ | not really good news, I guess. |
| 18:46.23 | jblinux | florian: I did iwpriv eth0 and it's got monitor mode....not ones shown (http://www.cs.umd.edu/~moustafa/mwvlan/examples.txt) ap_scan and such |
| 18:46.32 | florian | pb_: the same like with the old driver with my 3800 |
| 18:47.03 | pb_ | florian: ah. |
| 18:47.19 | pb_ | well, sounds like hostap_cs still isn't really usable for scanning, which is a shame. |
| 18:47.28 | *** join/#gpe algogr (~jim@ppp93-234.kavala.access.acn.gr) |
| 18:47.31 | jblinux | pb_: :( |
| 18:48.01 | florian | pb_: yep - some of us... and really more than one should request this... |
| 18:48.23 | pb_ | florian: yeah, I'll send a note to the maintainers. would you like to do the same? |
| 18:48.40 | algogr | Hello, i am new here. Is there any problem with sound on ipaq3950 |
| 18:48.55 | pb_ | algogr: recording doesn't work on h3900, but playback should be ok. |
| 18:49.30 | florian | jblinux: i think they use active scanning which is not supported by much drivers or cards. |
| 18:49.47 | algogr | do you know if they are going to fix it? |
| 18:50.09 | pb_ | algogr: yeah, we know what the fix is now, just a case of finding time to implement it. |
| 18:50.22 | pb_ | you could try Christian Pellegrin's alsa driver, which should work with recording already. |
| 18:50.22 | algogr | thanks |
| 18:50.39 | algogr | where can i find it? |
| 18:51.38 | florian | jblinux: you'll need to collect this data manually i think... what kind of software do you write? |
| 18:52.21 | jblinux | florian: positioning system using wifi... |
| 18:52.53 | jblinux | florian: do you mean it's still doable with orinoco_cs ? |
| 18:52.59 | florian | jblinux: hey, really cool... |
| 18:53.03 | pb_ | algogr: not sure, try google. |
| 18:53.16 | algogr | pb_: Thanks |
| 18:53.26 | jblinux | florian: thanks, but I already stuck at the first stage :( |
| 18:53.32 | florian | jblinux: shure... are you afraid of bad c source code? |
| 18:54.25 | jblinux | florian: I don't mind...as long as it could help :) |
| 18:56.16 | florian | jblinux: then check out apps/prismstumbler from handhelds cvs or grab an old tarball from http://sourceforge.net/projects/prismstumbler |
| 18:56.48 | jblinux | florian: also, there is a pre-compiled mwvlan-modules ipk that is in familiar 0.7. Tried that, didn't work... |
| 18:57.45 | florian | jblinux: the file iface-wlan-14.c will teach you how to handle the card and pcap and analyzer.c how to get the info from these packages. |
| 18:58.05 | florian | jblinux: you'll only need a very small subset of this. |
| 18:58.41 | jblinux | florian: atm, I just need ESSID and signal strength |
| 18:59.42 | florian | jblinux: yep, this is easy - with the struct in analyzer.h you'll be able to read it direct from the package. |
| 19:18.16 | jblinux | florian: for the iface-wlan-14.c....looks as if there isn't much change except replacing wlan0 with my the wireless alias (in my case, eth0) ? |
| 19:20.29 | jblinux | is the header for orinoco supposed to be the same as DLT_PRISM_HEADER? |
| 19:20.40 | florian | jblinux: yep, i think that's it... maybe remove this wlan-ng stuff. |
| 19:21.29 | florian | jblinux: yep, the orinoco driver creates the same header like the prism card... |
| 19:23.24 | jblinux | florian: yeah. the src seems need to be compiled with driver source....what should I do with it? |
| 19:25.29 | florian | jblinux: you'll only need the linux includes unless you are not using wlan-ng |
| 19:32.24 | *** join/#gpe nils__ (~nils@kdr.de) |
| 19:33.15 | nils__ | Hey ho, back online. Any news? |
| 19:33.35 | nils__ | The latest image (thanks pb!) seems to be very nice! |
| 19:34.14 | florian | hi nils__ |
| 19:36.51 | nils__ | florian: N'Abend :) |
| 19:37.19 | florian | nils__: Party over?! |
| 19:39.12 | jblinux | florian: just to make sure florian. To make a quick working copy, looks like I could just edit iface-wlan-4.c and do a make -f Makefile.arm ? |
| 19:40.05 | jblinux | what about wlan-ng source that needs to be specified in config.mk? |
| 19:40.42 | nils__ | florian: "Party" is tomorrow; we simply arrived today. We'll be back on sunday night (hopefully). |
| 19:41.15 | *** join/#gpe chrisime (~chrisime@pD9E8CCB5.dip0.t-ipconnect.de) |
| 19:41.27 | florian | jblinux: yep, this should work - maybe look at this file, it contains only some stuff for my broken toolchain |
| 19:42.08 | nils__ | Has dc updated the filemanager yet? |
| 19:42.13 | florian | jblinux: this is only needed for the wlan-ng support which is off by default |
| 19:43.07 | chrisime | hi kids |
| 19:43.15 | florian | hi chrisime |
| 19:44.17 | jblinux | florian: oh good. look at which file? |
| 19:44.18 | nils__ | Hi kid :) |
| 19:44.25 | chrisime | ;) |
| 19:44.41 | chrisime | nils__, *neuen_sharp_c750_ham_will* |
| 19:44.59 | florian | ohoh |
| 19:45.09 | chrisime | *g* |
| 19:45.24 | nils__ | chrisime: Aeh ... new c750? Have a link? |
| 19:45.28 | jblinux | florian: looking at iface-wlan.c, how is it different to iface-wlan-14.c? which one gets used by default? |
| 19:45.55 | chrisime | nils__, jo. somewhere |
| 19:45.57 | chrisime | just google |
| 19:45.59 | florian | jblinux: to enable it? forget it... it's possibly broken |
| 19:46.03 | chrisime | you'll find several |
| 19:46.25 | jblinux | jblinux: u mean iface-wlan.c is broken and not used? |
| 19:46.27 | nils__ | chrisime: What's the big difference to the c700? |
| 19:46.40 | jblinux | whoops, it's to florian |
| 19:46.58 | chrisime | nils__, neue cpu, mehr ram, ... |
| 19:47.21 | florian | jblinux: iface-wlan.c is the old one - it's not needed anymore... i should clean up this code :-/ |
| 19:47.22 | chrisime | nils__, cpu ohne bug |
| 19:48.00 | jblinux | florian: I see :) |
| 19:48.55 | florian | nils__: |
| 19:49.29 | florian | nils__: sorry... any idea about a "Compaq C-Series PDA" ? |
| 19:51.15 | florian | jblinux: for wireless lan monitoring you may use the psfront package in familiar - it provides a frontend for this prismstumbler scanner. |
| 19:52.02 | nils__ | florian: What's that again now? Gee ... one day off and the whole world goes crazy :) |
| 19:52.10 | jblinux | florian: frontend as a GUI? |
| 19:52.33 | *** join/#gpe tuhl (~tuhl@B084b.pppool.de) |
| 19:53.07 | tuhl | hi all |
| 19:53.27 | florian | hi tuhl |
| 19:53.45 | florian | jblinux: yep, a bad but working gtk frontend |
| 19:54.24 | florian | nils__: http://cgi.ebay.de/ws/eBayISAPI.dll?ViewItem&item=3025941798 |
| 19:54.48 | florian | nils__: looks not really fresh.... |
| 19:55.20 | jblinux | florian: looking at Makefile.arm, gcc is used in (cc), that's a bit strange...shouldn't it be arm-linux-gcc? |
| 19:55.34 | chrisime | hi tuhl |
| 19:55.55 | nils__ | florian: Looks like an HP Journada to me! |
| 19:55.58 | tuhl | hi chrisime! |
| 19:56.08 | jblinux | florian: I was thinking Makefile.arm was the Makefile speficially for arm processor compilation..... |
| 19:56.12 | chrisime | jo |
| 19:56.15 | tuhl | chrisime: Bist Du auf dem Linux-Tag? |
| 19:58.05 | chrisime | nein |
| 19:58.09 | chrisime | ich hab null zeit |
| 19:58.10 | chrisime | wegen uni |
| 19:58.16 | florian | nils__: i mailed him... if he is able to tell me what i need... i'll get... some |
| 19:58.22 | tuhl | chrisime: was macht die gnoppix cd? |
| 19:58.35 | chrisime | amu fragen |
| 19:58.39 | chrisime | der hat server zerschossen, der held |
| 19:59.14 | PaxAnima | warum sprechen wir Detusch in #GPE? ;) |
| 19:59.22 | florian | jblinux: yes, it is - did you get the cvs version? |
| 19:59.24 | florian | :-) |
| 19:59.43 | tuhl | chrisime: wird also nix mit gnoppix ? |
| 20:00.04 | florian | i have no idea... anyone problems with english language? |
| 20:00.36 | chrisime | tuhl, doch, aber amu is ned der schnellste |
| 20:00.47 | chrisime | der penner braucht immer eine aufm arsch ;)) |
| 20:02.00 | tuhl | chrisime: harte worte |
| 20:02.07 | chrisime | hehe |
| 20:02.11 | chrisime | muss scho mal sein |
| 20:02.13 | chrisime | *g* |
| 20:02.46 | jblinux | florian: looks like not :P which one would make myself easier? |
| 20:02.47 | tuhl | nils__: habe einen freien Kernel Mann für Stuttgart gefunden! |
| 20:03.42 | nils__ | tuhl: Cool! Oh great then... Congratulations! |
| 20:04.05 | tuhl | nils__: war recht aufwendig |
| 20:04.41 | tuhl | nils__: how usable is GPE! |
| 20:05.22 | florian | jblinux: the arm makefile from the tar package is maybe quite old... try "make CC=arm-linux-gcc" then |
| 20:08.56 | nils__ | tuhl: Depends on what you want to do :) |
| 20:09.19 | nils__ | tuhl: For quite general PDA use its IMHO quite good! |
| 20:09.26 | chrisime | tuhl, there's still much to do |
| 20:09.29 | chrisime | but it's fun |
| 20:09.46 | nils__ | tuhl: And we have some nice toys that others lack... |
| 20:11.49 | jblinux | florian: no worries, cvs is much nicer. |
| 20:14.38 | florian | jblinux: yeah, the tarball is really old.... |
| 20:15.17 | florian | brb |
| 20:24.45 | nils__ | So... what about our Release? |
| 20:26.34 | nils__ | I have seen that most points from our WiKi page seem to be done, apart from the filemanager and the gpe-beam thing - nothing major IMHO. |
| 20:27.25 | nils__ | nelson: Are you with us? |
| 20:36.31 | *** part/#gpe tuhl (~tuhl@B084b.pppool.de) |
| 20:40.05 | jblinux | florian: compiling complains about gtk++2.0 package not found, can I skip that part. I just want quickly try it on command line. |
| 20:46.44 | *** join/#gpe chrisime__ (~chrisime@pD9E8CC1F.dip0.t-ipconnect.de) |
| 21:00.18 | pb_ | nils__: Yeah, I think most of the important things are under control. Damien looked at the filemanager thing yesterday but wasn't able to reproduce the problem on his desktop; it sounds like it may be hard to fix. |
| 21:00.43 | moray | pb_: do we know if he was trying it under matchbox on his desktop machine? |
| 21:00.44 | chrisime__ | hey dude |
| 21:00.54 | nils__ | pb_: Oh well, then be it without, or? |
| 21:01.08 | pb_ | moray: ah no, probably a normal window manager. |
| 21:01.19 | pb_ | hi chrisime__ |
| 21:01.35 | moray | pb_: yes, I was just wondering if that might make a difference with this kind of issue |
| 21:01.51 | pb_ | nils__: Yeah, I think we can live without a fix for that one. It's annoying, but the filemanager isn't absolutely critical for most people. |
| 21:02.07 | nils__ | pb_: Not yet, yes :) |
| 21:02.29 | nils__ | Has nelson said something about having fetched the new packages? |
| 21:02.55 | pb_ | No. Someone should send him a mail with a list of the packages that need to be pushed from unstable into 0.7. |
| 21:03.27 | nils__ | pb_: We have the list on the WiKi page if this is still up-to-date. |
| 21:04.19 | pb_ | Ah, yeah, I think that list is fairly accurate. We might want to add greader and the updated gpe-apps2. |
| 21:04.47 | nils__ | Good point! |
| 21:05.16 | pb_ | florian, any word from Greg on an eta for the updated greader ipk? |
| 21:08.22 | pb_ | moray: it'd certainly be worth re-testing with matchbox on the host. It's also possible that the ipks are just outdated and need rebuilding, but I don't think so in this case. |
| 21:17.23 | *** join/#gpe herzi (~herzi@kiwi.mediascape.de) |
| 21:20.25 | florian | re |
| 21:20.32 | florian | jblinux: yep, shure |
| 21:21.23 | florian | pb_: nope, nothing... maybe i should ask him |
| 21:22.35 | nils__ | OK then. I have just updated the Wiki page with the two added items greader and gpe-apps2. |
| 21:22.42 | nils__ | Who is going to inform Rus? |
| 21:25.23 | *** join/#gpe paul_ (~paul@s0.alphyra.bbnplanet.net) |
| 21:34.03 | nils__ | Oh well then... I'll quickly write an email to Rus pointing him to our page... |
| 21:35.07 | florian | OK, i mailed Greg |
| 21:41.02 | nils__ | florian: About what? That you changed the icon for him? |
| 21:43.35 | florian | nils__: yep, two days ago about that - he answerd that he'd create a fixed package.. .but no pachage so far. So i maild him again... |
| 21:46.09 | nils__ | florian: Pity... I just emailed Rus and pointed him to our Wiki page. I hope everything works out! |
| 21:47.13 | florian | yep true... |
| 21:47.27 | pb_ | nils__: great, thanks. |
| 21:47.32 | pb_ | florian: ok, cool |
| 21:47.34 | florian | oh i should better whatch what i write... |
| 21:50.44 | nils__ | OK. Will there be someone around tomorrow so that Rus could ask if something's broken? I doubt but just in case... |
| 22:08.10 | florian | looks like kernel downgrading is still difficult... |
| 22:08.22 | nils__ | :( |
| 22:09.16 | nils__ | Apropos kernel... any news on the h38xx h39xx power consumption issue? |
| 22:11.35 | pb_ | yeah, I'll probably be around tomorrow. |
| 22:11.43 | pb_ | no news on the power management thing as far as I know. |
| 22:11.51 | nils__ | pb_: Great, thanks! |
| 22:11.57 | pb_ | It seems to be intermittent, which makes it harder to debug. |
| 22:12.24 | nils__ | pb_: Don't know if Floprian mentioned already, he put his 3870 on a AC/DC adaptor with an amp-meter... |
| 22:12.42 | nils__ | pb_: Looked pretty desastrous :( |
| 22:13.08 | pb_ | Yeah, he did mention that. It's hard to know how accurate external power measurements are, though it did seem rather high. |
| 22:13.15 | nils__ | pb_: Peaks of up to 90mA in suspend mode, even with periods >> 2 secs of that drainage... |
| 22:13.15 | *** join/#gpe AntiProxy (Admin@AntiProxy.com) |
| 22:13.20 | pb_ | We should find out how to measure the current internally. |
| 22:13.26 | AntiProxy | hi everyone |
| 22:13.35 | pb_ | nils__: Oh, really? I hadn't heard it was that bad. |
| 22:13.42 | nils__ | Simply cut off the battery and replace with power source? |
| 22:14.08 | florian | Let's try to find out monday... |
| 22:14.13 | pb_ | nils__: apparently there is a zero-ohm shunt on the motherboard that you can replace with a current sense resistor. |
| 22:14.34 | florian | I'm not really afraid of opening my iPaq :-) |
| 22:15.35 | florian | pb_: Any idea where exactly this is? |
| 22:16.30 | pb_ | florian: Not really, but one of the CRL guys could tell you. Maybe jg knows, otherwise ask Jamey or Andy on Monday. |
| 22:18.38 | florian | pb_: ok, i'll do that.. this can't be too hard to find out otherwise... |
| 22:19.15 | pb_ | Okay, cool. |
| 22:20.13 | nils__ | The more important part is to get to know more about the insides of the machine and which part especially might be guilty. Just finding that there is a drainage is only half the truth. |
| 22:20.40 | pb_ | Maybe I should add an option to the PM code to have it behave the same with AC on and off. That might make it easier to see what's going on, without the periodic wake-ups and LED confusing the issue. |
| 22:20.43 | nils__ | And concerning hardware specs and docs it seems to be quite difficult to get hold of. |
| 22:21.20 | pb_ | nils__: I think most of the specs are available if you ask for them. I didn't have too much trouble getting hold of h3900 and h5450 documentation. |
| 22:21.45 | nils__ | pb_: Oh really? That would be great! |
| 22:22.12 | pb_ | unfortunately, I don't have any data on the ASIC pair in the h3800, which I suspect is where the problem lies. |
| 22:22.33 | nils__ | Apparently only the 38x and 39x seem to be affected by this problem :( My old 3600 behaves quite OK though it also sucks ~30mA in standby. |
| 22:23.05 | nils__ | pb_: You think it's the ASIC? I more thought on the SDRAM side. I know that this can suck a lot! |
| 22:23.08 | pb_ | Yes, indeed. |
| 22:23.44 | nils__ | pb_: I cannoot image how the ASIC could draw that much power. Though I have to admit that I have no clear idea what this beast is used for... |
| 22:24.21 | nils__ | pb_: s/image/imagine :) |
| 22:24.23 | pb_ | nils__: well, the SDRAM is handled identically on h3600 and h3800, whereas it differs between h3800 and h3900. Also, when the CPU goes into suspend, the SDRAM controller is automatically stopped. The only likely failure mode there would be the SDRAM not entering self-refresh, which would have the opposite effect: low power consumption but loss of DRAM contents. |
| 22:24.56 | nils__ | pb_: True... hmmm ... |
| 22:25.01 | herzi | nils__: are you going to the linuxtag in karlsruhe? |
| 22:25.11 | nils__ | herzi: Sure! |
| 22:25.28 | pb_ | nils__: could be that it is failing to drive all signals to the right state when entering suspend. If one of the ASIC's outputs remains high, but the circuitry it is connected to is powered down, the output transistor will try to power the connected device through its input protection diodes. |
| 22:25.30 | herzi | nils__: will there be a gpe booth? |
| 22:26.21 | nils__ | nils__: I don't know yet. I will try to ask Joey and others about it. I hope it is not too late to get a corner! |
| 22:26.23 | pb_ | nils__: conversely, if an input on a device that remains powered is allowed to float, it can oscillate and cause the device to draw excessive current. |
| 22:27.31 | florian | pb_: i'm playing with your new hostap_cs... look like it oopses after a failing to init the card... |
| 22:28.04 | nils__ | pb_: All I know is that this effect does not show imediately. It seems to be connected to the battery level somehow. A fully charged device does keep up until it reaches a certain drain level. After that it looses power very rapidly. |
| 22:28.54 | nils__ | pb_: A little bit of random is in there too because it does not happen all the time so floating pins might be worth closer inspection. |
| 22:44.29 | florian | pb_: hostap_cs dies after refusing to init the card. there was the usual "ignore-cis-vcc=1" setting missing. |
| 22:44.52 | pb_ | florian: aha. |
| 22:46.34 | florian | pb_: This setting should be a part of the module package like for orinoco_cs... |
| 22:52.19 | *** part/#gpe AntiProxy (Admin@AntiProxy.com) |
| 22:58.01 | florian | night all |
| 23:52.30 | *** part/#gpe nils__ (~nils@kdr.de) |