freemangordon | jonwil: maybe if you look what is upstream and compare it with what is in maemo will give us a clue. | 00:00 |
---|---|---|
*** lizardo has quit IRC | 00:00 | |
jonwil | ok | 00:00 |
freemangordon | if it is the same, then iface replacement does not make sense | 00:00 |
DocScrutinizer05 | struct wl1251_magic {int magic1; int magic2; int magic3} wl1251; | 00:00 |
zeq | O_o | 00:01 |
DocScrutinizer05 | hehehe | 00:01 |
freemangordon | DocScrutinizer05: where did you get that from? | 00:01 |
DocScrutinizer05 | made it up | 00:01 |
zeq | :) | 00:01 |
freemangordon | aah, I was thinking that you found it in upstreamed driver :D | 00:01 |
DocScrutinizer05 | but I bet that's about what you'll find in any 'FOSS' driver | 00:01 |
kerio | goddammit my 0.02 were useless :( | 00:02 |
freemangordon | http://www.valot.fi/kalle/tmp/wl12xx/wl12xx-2.6.28-1.patch | 00:03 |
freemangordon | http://markmail.org/thread/xummknezx5r66fha | 00:03 |
zeq | kerio: didn't cover as much as we hoped | 00:03 |
kerio | you guys ended up compromising! | 00:03 |
Estel_ | shhh | 00:03 |
Estel_ | DocScrutinizer, doesn't know yet | 00:03 |
freemangordon | DocScrutinizer05: does not look like that | 00:03 |
Estel_ | enjoy peace before storm | 00:03 |
kerio | now i'll have to flash the compromising kernel and then kernel-power again | 00:04 |
zeq | looks quite well documented ^^^ | 00:04 |
freemangordon | yeah, no magic so far in the patch | 00:04 |
DocScrutinizer05 | meh, which storm. If you think you don't need my advice, I'll simply beat up each single one of you separately if ever there's any problem arising that I could've warned you about | 00:05 |
zeq | sounds fair :) | 00:06 |
DocScrutinizer05 | my spiders already collecting your N900's IMEI, ser#, MAC, ip-route etc | 00:07 |
freemangordon | wl12xx_op_add_interface(struct ieee80211_hw *hw, | 00:07 |
freemangordon | + struct ieee80211_if_init_conf *conf) | 00:07 |
freemangordon | well, that looks pretty much coming from upper layers | 00:08 |
freemangordon | not something specific to wl1251 | 00:08 |
DocScrutinizer05 | and one "thumbs down" from me will probably cost CSSU 80% of users | 00:09 |
DocScrutinizer05 | so you better keep me happy ;-P | 00:09 |
* freemangordon sends DocScrutinizer05 a beer | 00:09 | |
Estel_ | You think that CSSu is used by 5 people only? | 00:09 |
* Estel_ send DocScrutinizer a kiss ;) | 00:10 | |
DocScrutinizer05 | unless you want CSSU for your own proivate idaho, in which case you better fork anyway | 00:10 |
freemangordon | anyway, I am out of cigarettes, so better have some sleep | 00:11 |
Estel_ | :) | 00:11 |
Estel_ | DocScrutinizer, I'm positively surprised by your sense of humour - it's due to vacations? | 00:11 |
zeq | goodnight freemangordon | 00:11 |
Estel_ | I hope that this remakr about thumb down and 805 of cssu users was also part of this nice humour (no irony) | 00:11 |
Estel_ | night, free | 00:12 |
freemangordon | night guys | 00:12 |
Estel_ | s/805/80%/ | 00:12 |
zeq | I'm sure there was irony :) | 00:12 |
Estel_ | no, it's just "people mean well" :D | 00:12 |
Estel_ | I don't suspect DocScrutinizer of being THAT kind of self-convince one. | 00:13 |
Estel_ | and bashing, if problems arise, seems fair to me too | 00:13 |
DocScrutinizer05 | Estel_: you already should've learnt that none of my posts consists of 0% or 100% humor | 00:13 |
jonwil | ok, its definatly zzz time now | 00:13 |
jonwil | Looks like we had a good meeting | 00:13 |
Estel_ | DocScrutinizer, exactly my thoughts :) | 00:13 |
jonwil | and I know now that wl1251-cal and bluetooth-cal stuff | 00:13 |
zeq | yeah, jonwil, what time is it there?!? | 00:13 |
jonwil | stuff IS worth working on | 00:13 |
jonwil | its 5am :P | 00:14 |
Estel_ | ouh | 00:14 |
zeq | yeah, ouch | 00:14 |
zeq | get some good rest | 00:14 |
jonwil | good to be there for the important stuff though | 00:14 |
jonwil | cya | 00:14 |
Estel_ | then I understand Your repeated questions about "I'm i really needed here?" | 00:14 |
zeq | night | 00:14 |
Estel_ | see ya! | 00:14 |
*** jonwil has quit IRC | 00:14 | |
zeq | I'd better be off too.... | 00:14 |
DocScrutinizer05 | and particularly my statement about a *lot* of users pinging me in PM to keep on struggling for a sane CSSU is not any joke | 00:14 |
DocScrutinizer05 | just the 80% is merely made up | 00:15 |
DocScrutinizer05 | maybe it's 4% or 97% | 00:15 |
zeq | Good night DocScrutinizer05, Estel_ | 00:15 |
Estel_ | Good night! | 00:16 |
Estel_ | DocScrutinizer, but they're same users using different names ;P | 00:16 |
DocScrutinizer05 | cya zeq | 00:16 |
DocScrutinizer05 | Estel_: I'm not an idiot, even when you don't believe that | 00:16 |
Estel_ | but, seriously this time, just read backlogs, I think you will be satisfied by outcome too :) | 00:17 |
DocScrutinizer05 | and my post about spiders also wadsn't 100% joke | 00:17 |
Estel_ | especially, that You were sympathetic to fixing pselect bug | 00:17 |
Estel_ | No worries, my cat eats spiders. | 00:17 |
DocScrutinizer05 | pselect bug is a valid reason to get CSSU kernel, IF and ONLY IF we can demonstrate the issue on stock kernel | 00:17 |
Estel_ | well, d-bus seems to use pselect quite often | 00:18 |
Estel_ | which arises both performance concerns, and possibility of race, which should be avoided at all cost | 00:18 |
DocScrutinizer05 | I would believe that in 0.0001 second | 00:18 |
DocScrutinizer05 | d-bus is so fubar ther HAS TO be some massive bug sleeping in it | 00:19 |
merlin1991 | Estel_: Doc is talking about showing the bug, which basically means someone has to write a really nasty testcase that manages to show the race condition | 00:19 |
Estel_ | also, upstream adopted this fix in kernel 2.6.30 for reason, too | 00:19 |
DocScrutinizer05 | merlin1991: exactly | 00:19 |
Estel_ | merlin1991, there was one, zeq presented it to DocScrutinizer, but DocScrutinizer wasn't satisfied by aesthetic side of this | 00:19 |
merlin1991 | basically a lop of some pselect calls that make sense together with a signal sender | 00:19 |
merlin1991 | and ofc madness | 00:19 |
DocScrutinizer05 | almost correct, Estel_ | 00:19 |
Estel_ | also, I still think that in our little world (N900) case, arguing with upstream is quite silly | 00:19 |
DocScrutinizer05 | merlin1991: there's such code already | 00:20 |
merlin1991 | DocScrutinizer05: what's keeping you unconvinced then? | 00:20 |
merlin1991 | btw I made a really nasty mistake here | 00:20 |
DocScrutinizer05 | missing post to ML, showing a scientific review and the results of this test | 00:21 |
merlin1991 | flashed only rootfs when I had a quite different kernel flashed | 00:21 |
kerio | merlin1991: D: | 00:21 |
kerio | well | 00:21 |
kerio | flash the kernel | 00:21 |
Estel_ | DocScrutinizer, last time you were not satisfied by this code displaying stats, and stoppind when encounter bug - you wanted it to display info on encountering bug, instead of stoppiong doing so. IIRC | 00:21 |
Estel_ | that's why I called it aesthetic | 00:21 |
Estel_ | concern. May be wrong, though | 00:21 |
kerio | and then install backupmenu and unpack the backup you have | 00:21 |
merlin1991 | kerio: did it already, but I spend a few minutes wondering why I'm having a silly reboot loop | 00:21 |
kerio | reflash *all* the things! _ò/ | 00:22 |
Estel_ | well, once I recovered old backup, because I realized 5 minutes too late, that hanging on reboot was caused by typo in rcS_late | 00:22 |
DocScrutinizer05 | Estel_: incorrect, I asked to point me to the quote that explains what this code is supposed to do, freemangordon(?) pointed me to comment#6 which provides exactly what I asked for | 00:22 |
*** BCMM has quit IRC | 00:22 | |
Estel_ | DocScrutinizer, I see. so, after all, you're satisfied with it? | 00:23 |
DocScrutinizer05 | I'd be if I'd see the results posted somewhere | 00:23 |
Estel_ | somewhere = mailing list? Why so? | 00:23 |
merlin1991 | btw DocScrutinizer05 the possible race condition is even on wikipedia :D | 00:24 |
merlin1991 | http://en.wikipedia.org/wiki/Event_loop#Handling_signals | 00:24 |
DocScrutinizer05 | merlin1991: I know | 00:25 |
DocScrutinizer05 | merlin1991: does that wiki also say 2.6.28-omap1 #1 PREEMPT Fri Aug 6 11:50:00 EEST 2010 armv7l unknown does NOT have any fix for it? | 00:26 |
merlin1991 | DocScrutinizer05: that part is what we have zeq for ;) | 00:27 |
DocScrutinizer05 | well, that's what I ask for. DEMONSTRATE we got a problem on stock kernel, so *everybody* noob can read it, and I'll support a fix. No discussion about that | 00:30 |
DocScrutinizer05 | CSSU is about fixing problems | 00:30 |
DocScrutinizer05 | getting a new kernel for 10% performance increase is *creating new* problems, not fixing any | 00:30 |
DocScrutinizer05 | fixing pselect IF WE CAN | 00:30 |
DocScrutinizer05 | DEMONSTRATE | 00:31 |
DocScrutinizer05 | it bites us on stock kernel, is absolutely in line with CSSU (given the negative aspects of bug are way more severe than the negative aspects of the fix) | 00:32 |
merlin1991 | DocScrutinizer05: do you have a link to the sample code? | 00:32 |
DocScrutinizer05 | leem check if I still got the window open | 00:32 |
DocScrutinizer05 | https://bugs.launchpad.net/ubuntu/+source/linux/+bug/319729 | 00:33 |
DocScrutinizer05 | is the comment#6 I mentioned above | 00:33 |
merlin1991 | arf it's 404 | 00:34 |
merlin1991 | hm how does one use the google cache again? | 00:34 |
DocScrutinizer05 | http://people.canonical.com/~scott/childspin.c is dead | 00:34 |
DocScrutinizer05 | :-/ | 00:34 |
kerio | is that like meatspin? | 00:34 |
Estel_ | DocScrutinizer, actually, 10% performance would be worth it :) | 00:35 |
Estel_ | not to mention, that it's not only about it | 00:35 |
DocScrutinizer05 | no, definitely not | 00:35 |
Estel_ | yes, definitelly yes, in device that have such potential, but so scarce resources. | 00:35 |
DocScrutinizer05 | no normal user gives a fuck about 10% performance boost | 00:36 |
kerio | the n900 has no normal users, realistically | 00:36 |
Estel_ | 10% substitutes for 100 mhz oveclock, in most devices, without drawbacks. | 00:36 |
DocScrutinizer05 | kerio: wrong | 00:36 |
Estel_ | normal suers doesn't use cssu | 00:36 |
Estel_ | users, ffs | 00:36 |
Estel_ | what I have with this "sue" thing :p | 00:36 |
DocScrutinizer05 | Estel_: wrong | 00:36 |
Estel_ | DocScrutinizer, I buy n900 from "normal users" few times a month | 00:36 |
DocScrutinizer05 | see? | 00:37 |
Estel_ | they don't have rootsh, cssu, but, suprisingly, the have kernel-power | 00:37 |
merlin1991 | DocScrutinizer05: found the source somewhere else https://launchpadlibrarian.net/34502840/childspin.c | 00:37 |
DocScrutinizer05 | and those are only the ones that sell their device | 00:37 |
Estel_ | DocScrutinizer, because majority of normal users never heard of maemo.org website? | 00:37 |
Estel_ | anyone else isn't normal user anymore | 00:37 |
DocScrutinizer05 | Estel_: CSSU is for normal users BY DEFINITION | 00:37 |
DocScrutinizer05 | it's the founding preamble of CSSU | 00:38 |
Estel_ | so missed target group, as users of custom-installable repo, TMO, and other resources like that isn't normal anymore | 00:38 |
Estel_ | normal guy buy phone to make phone calls and sms'es | 00:38 |
DocScrutinizer05 | your definition of 'normal' is flawed | 00:38 |
Estel_ | Youcould target "normal users" if Nokia would make PR out of cssu, otherwis,e it require custom actions | 00:38 |
Estel_ | to install | 00:38 |
Estel_ | well, normal as in % of n900 owners? | 00:39 |
Estel_ | I suspect no more than 305 of them have been to TMO, even once | 00:39 |
Estel_ | 30% | 00:39 |
Estel_ | s/305/30%/ | 00:39 |
DocScrutinizer05 | I knew about that flawed concept when I seen you thinking that every "normal N900 user" is supposed to read tmo | 00:39 |
Estel_ | TMO OR wiki | 00:39 |
Estel_ | well, you need to know about things from somewhere, yep? | 00:40 |
Estel_ | or mailing list | 00:40 |
Estel_ | any of those | 00:40 |
Estel_ | not necessary all of them | 00:40 |
Estel_ | sadly, people doing so are far from normal phone users, anyway | 00:40 |
DocScrutinizer05 | 90% of N900 (or any other device) users won't give a fuck about wiki, fora, IRC whatever | 00:40 |
Estel_ | and they don't use cssu | 00:40 |
DocScrutinizer05 | CSSU is for them, as well | 00:40 |
Estel_ | cause how they can know about it, without doing so? | 00:40 |
DocScrutinizer05 | that's a good questions, and a problem you should tackle | 00:41 |
Estel_ | also, if they don't give a fuck, but use cssu, they don't give a fuck about kernel used, too | 00:41 |
Estel_ | as long as it works | 00:41 |
DocScrutinizer05 | rather than thinking of CSSU as the geeks' breeding-edge distro | 00:41 |
DocScrutinizer05 | bleeding even | 00:41 |
Estel_ | where we comes to situation, where you're right about beating everyone involved if serious problems arise | 00:41 |
Estel_ | no no | 00:41 |
Estel_ | kernel-power is for bleeding-edge | 00:42 |
Estel_ | no one said kernel in cssu will contain bq2415x_charger and such things, before proven stable | 00:42 |
Estel_ | after proven stable, though, no reasons why not. | 00:42 |
DocScrutinizer05 | Estel_: let's say I managed a product, while you obviously didn't yet | 00:42 |
Estel_ | sure. which doesn't make you better suited to discuss it :) | 00:42 |
Estel_ | it's just equal | 00:42 |
DocScrutinizer05 | and let's assume it's true that I get private mail asking me to not stop with what i'm doing for CSSU | 00:43 |
merlin1991 | hm compiled the sample code, and it's still running | 00:43 |
Estel_ | DocScrutinizer, irrelevant, You imagine how many mails and psots cssu team get about including various features? | 00:43 |
Estel_ | even ones that shouldn't get into cssu? | 00:43 |
Estel_ | usually 1 feature request per day ;P | 00:43 |
Estel_ | negatrons are not better in any way than those people | 00:43 |
Estel_ | cssu need to follow common sense and logic - it works -> it gives benefits - > it doesn't cause regression - > include | 00:44 |
DocScrutinizer05 | well, the difference is, my requests are in line with CSSU foundation manifest, while yours are noobs who didn't get the idea of CSSU | 00:44 |
Estel_ | of course it mean core system components, like kernel | 00:44 |
Estel_ | or toolchain | 00:44 |
Estel_ | DocScrutinizer, sophism and demagogic approach. "Mine are better than yours" | 00:44 |
DocScrutinizer05 | if you want sometjing that CSSU definitely is NOT, you're free to fork | 00:45 |
Estel_ | it's jsut about interpretation of CSSU manifest, and merlin1991 seems to share "common sense" way of looking at it. Last time I checked, he and MohammadAG were maintainers | 00:45 |
merlin1991 | srly Estel_ relax | 00:45 |
Estel_ | same apply in reverse - if You want cssu-lite to contain less, You're free to fork | 00:45 |
Estel_ | merlin1991, why so? i'm relaxed | 00:45 |
DocScrutinizer05 | it's however bad habit to try and redefine a project on the fly, just because you're too lazy to fork | 00:45 |
Estel_ | we're discussing, in civil way, i suppose, no anger here from any party (I hope) for sure not from my side | 00:45 |
DocScrutinizer05 | or have even worse reasons to try and change CSSU to something differnet | 00:46 |
merlin1991 | hm this testhtingy is still running | 00:46 |
Estel_ | DocScrutinizer, the main questions always remain - IMO - a) does it work well b) does it give benefits c) does it cause regression d) how much time we need to work on it | 00:46 |
Estel_ | aka feasiobility | 00:46 |
kerio | i propose changing cssu-thumb in cssu-leet | 00:46 |
Estel_ | + minor but important aspects, like maintainability, etc | 00:46 |
Estel_ | every thing discussed today have allo of them fullfiled | 00:47 |
Estel_ | fix included in upstream, well documented, people interested to work on it and maintain it, benefits, no regressions. | 00:47 |
Estel_ | it's about practice > ideology. | 00:47 |
Estel_ | You're free to beat anyone involved if it cause serious problems, but not before, just "to be on safe side" | 00:48 |
DocScrutinizer05 | a) it works well for those who use it b) it has benefits for them, without any changes of the policies c) the point is it mustn't have regressions that are not inevitable and d) it's your time. don't worry about our time! | 00:48 |
Estel_ | merlin1991, about testhingy - no idea, You're better asking zeq about it | 00:48 |
Estel_ | well, future will tell - I'm sure it will end like thumb thing after all, i.e. perfectly doable and benefitable, without regressions | 00:49 |
Estel_ | lets see what future will bring :) | 00:49 |
Estel_ | atfer all, nothing is un-rollback-able | 00:50 |
DocScrutinizer05 | Estel_: that's a BS statement, sorry. CSSU policy and manifest will not 'end' anywhere - it IS | 00:50 |
Estel_ | in worst case scenario, which isn't going to come, in my humble opinion. | 00:50 |
Estel_ | DocScrutinizer, and what you will do? Lie in the path of train as act of protest? last time you've said "well, if You think like that it's ok, i'll beat You if something goes wrong" | 00:51 |
Estel_ | which is fair | 00:51 |
Estel_ | I'm also eager to see what future will bring | 00:51 |
Estel_ | and it seems to be best conclusion | 00:51 |
DocScrutinizer05 | Estel_: you're again short of arguments and thus starting to use your annoying tedious style which I'll ignore like I always do. cya | 00:52 |
Estel_ | no worries, if this makes you happy during this pleasant summer time, we may agree that i'm short of arguments :) | 00:52 |
Estel_ | see ya. | 00:52 |
javispedro | ah, the winds of maemo never change. | 00:54 |
DocScrutinizer05 | javispedro: !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! yeeeeehaaaaaa! hello ol'fart! | 00:55 |
javispedro | :D hello! | 00:55 |
merlin1991 | javispedro: any idea why tracker indexers like the flc ones are missing if you udpate the core tracker package? | 00:55 |
merlin1991 | s/flc/flac/ | 00:56 |
infobot | merlin1991 meant: javispedro: any idea why tracker indexers like the flac ones are missing if you udpate the core tracker package? | 00:56 |
javispedro | you get me thinking when did tracker gain fuzzy logic capabilities... | 00:56 |
javispedro | *got | 00:56 |
javispedro | last time the flac indexers stopped working, they were horribly packaged/optified, linking to non-existing libraries. | 00:57 |
javispedro | since then they have always worked for me. | 00:57 |
javispedro | there was this big libFLAC/pulseaudio/maemo-optify triangle of DOOM. | 00:57 |
merlin1991 | well since we updated tracker in -testing users have been claiming that tracker did not index ogg / flac / whatever fancy extra codec anymore if the support was installed prior to the update | 00:58 |
Estel_ | javispedro, hi there :) abusing Your presence - could You, please, link me to your latest, unreleased in repos version of radio program for N900? | 00:58 |
merlin1991 | playback from filemanager still works though | 00:58 |
javispedro | merlin1991: maybe ABI change? | 00:58 |
merlin1991 | and I found out that by reinstalling ogg-support (it does ogg and flac) tracker suddenly indexed again | 00:58 |
merlin1991 | and abi change, well all we did was apply your patches :D | 00:58 |
javispedro | Estel_: https://github.com/javispedro/cfmradio.git | 00:58 |
Estel_ | javispedro, any concerns if I would put it to the repos, of course crediting You properly? | 00:58 |
Estel_ | thanks | 00:58 |
javispedro | Estel_: not much difference I think from what was packaged | 00:59 |
javispedro | Estel_: most of it was just testing stuff | 00:59 |
Estel_ | I see | 00:59 |
javispedro | (iirc there were some additional nonsense buttons in the appmenu for changing audio routes) | 00:59 |
Estel_ | I remember something about possibility ot define output for audio, and it's volume | 00:59 |
Estel_ | which was lacking in released version | 00:59 |
Estel_ | or smth like that | 01:00 |
javispedro | did not play with volume | 01:00 |
javispedro | but you can trigger the analog bypass | 01:00 |
javispedro | so that radio uses 0% cpu time | 01:00 |
javispedro | but then you might fry your speakers | 01:00 |
javispedro | choice is yours! ;) | 01:00 |
Estel_ | It's possible to use analog bypass with this radio? 0_o? | 01:00 |
Estel_ | awesome | 01:00 |
Estel_ | someone told me long time ago, that it's almsot undoable in N900 :) | 01:00 |
Estel_ | javispedro, we have (currently not included in our upstream) kernel-module for filtering output for speakers | 01:01 |
Estel_ | maybe it would be good reason to revive it | 01:01 |
javispedro | Estel_: analog will bypass that too. | 01:01 |
Estel_ | I see | 01:01 |
Estel_ | any way to put lowpass filter alongside analog output? | 01:01 |
Estel_ | or it's just so 1337 that it will bypass everything? | 01:01 |
Estel_ | stupid android uses analog bypass, and somehow, they don't blow up speakers | 01:02 |
javispedro | there were some talks about implementing the filters on the audio chip itself (forgot the name :/) | 01:02 |
javispedro | maybe that's what the kernel module does? I have no idea which kernel module are you talking about btw. | 01:02 |
Estel_ | javispedro, yeah, isn't that related to kernel module talking with that chip? | 01:02 |
Estel_ | yea, it was this thing | 01:03 |
Estel_ | on audio chip, can't recall name though | 01:03 |
Estel_ | so, analog bypass is available on git versin only, not in the pre-compiled one? | 01:03 |
javispedro | only git | 01:03 |
javispedro | also you need to #define sth, lemme check | 01:03 |
Estel_ | why the hell haven't You released it? It's awesome, for us, radio fans | 01:04 |
Estel_ | :) | 01:04 |
javispedro | https://github.com/javispedro/cfmradio/blob/510f98cf1749ddae72aebf1d9f8561ebff56831f/cfmradio.c#L11 | 01:04 |
javispedro | #define ADV_AUDIO_ROUTING 1 there | 01:05 |
* Estel_ nods | 01:05 | |
Estel_ | is switching this from gui implemented, or is it decided on compile time (bypassing or not) | 01:05 |
javispedro | gui | 01:05 |
Estel_ | great, thanks a lot | 01:05 |
javispedro | hit the menu and you'll see some extra buttons | 01:06 |
Estel_ | And I suppose that you had it ready for ages, just not released? And we struggled with FM Radio for years? You cruel one :) | 01:06 |
javispedro | struggled? | 01:06 |
javispedro | it's just some mixer setting | 01:07 |
Estel_ | it's irritating as hell, the msot popular FM Radio | 01:07 |
Estel_ | when it crashes, N900 microphone stop working | 01:07 |
Estel_ | wise heads were shaking at it, without finding out "why" | 01:07 |
Estel_ | only reboot solve it | 01:07 |
javispedro | same would happen to mine | 01:07 |
Estel_ | OTOh, CFM Radio is great, just lacked few things sitting in your git for ages :) | 01:07 |
Estel_ | no idea why, but CFM radio *never* resulted in such weirdness for me | 01:08 |
javispedro | (but I guess mine crashes slightly less often, being so simple and everything) | 01:08 |
Estel_ | or any other radio program, other than fm radio | 01:08 |
* Estel_ nods | 01:08 | |
*** taziff1 has joined #maemo-ssu | 01:08 | |
javispedro | I prefer the N950/N9 radio hardware, despite the fact you'll never get 0% cpu usage playing there. | 01:09 |
Estel_ | whats ie better there, then? | 01:09 |
javispedro | faster tuning and digital audio output | 01:09 |
javispedro | and you don't even need to turn bt on | 01:09 |
Estel_ | well, on N900 we also have digital audio output, just not connected anywhere... like fmtx in N950/N9 :P | 01:09 |
Estel_ | I see | 01:09 |
javispedro | I still don't know if fmtx is connected or not on n9. | 01:10 |
Estel_ | reportedly, isn't, but it may be urban legend. On N950, it seems that it isn't for sure | 01:10 |
javispedro | On N950 it's the actual tx hardware missing. | 01:10 |
Estel_ | I wonder, if on N950 it is modable? I wasn't able top trace visible track | 01:10 |
javispedro | the wl127x variant used just doesn't have it. | 01:10 |
Estel_ | hm, reportedly, it was present, but not connected, although, I have no idea how reliable this info is | 01:10 |
Estel_ | so is this whole "unconnected fmtx" a urban myth, then? | 01:11 |
javispedro | no one really seems to know the n950 hardware for certain around. | 01:11 |
Estel_ | need X-ray'ing it :) | 01:11 |
Estel_ | at least for patches | 01:11 |
*** taziff has quit IRC | 01:11 | |
Estel_ | path* | 01:11 |
Estel_ | cooper ones | 01:11 |
javispedro | Estel_: I like to think that if noone made it yet means it is not connected. | 01:11 |
kerio | are the alsa mixers saved somewhere, or can you reset them by rebooting if you fear you screwed up something? | 01:11 |
Estel_ | javispedro, probably yes, and people talking about unconnected fmtx were quite certain about it | 01:12 |
javispedro | kerio: you can definitely mess stuff up to the point you need to remove batteries | 01:12 |
javispedro | I don't remember if maemo saves alsa mixer values though. I think not. | 01:13 |
javispedro | (plus the entire "burn the speakers" thing, of course) | 01:13 |
javispedro | my holidays started, hope to be more often around here. For a few days at least ;P | 01:17 |
DocScrutinizer05 | (there were some talks about implementing the filters on the audio chip itself (forgot the name :/) ) /me and this eastern guy HNZ or whatever | 01:18 |
javispedro | DocScrutinizer05: you know if that's what "the kernel module" mentioned above does? | 01:21 |
DocScrutinizer05 | nfc what's this "the kernel module" mentioned above ;-) | 01:22 |
* DocScrutinizer05 got as few context as javispedro | 01:22 | |
DocScrutinizer05 | javispedro: holidays? congrats, me too :-D | 01:22 |
DocScrutinizer05 | so I will rethink my planning | 01:22 |
DocScrutinizer05 | you being around changes weights between IRC/maemo and going to camp in the woods | 01:23 |
javispedro | heh | 01:23 |
merlin1991 | the gentoo guy dunno his name wrote something for the kernel | 01:23 |
Estel_ | luke_jr | 01:23 |
Estel_ | although, he didn't wrote it | 01:23 |
DocScrutinizer05 | ooh luke | 01:23 |
Estel_ | just send :P | 01:23 |
merlin1991 | or adapt it, or whatever | 01:24 |
Estel_ | yea | 01:24 |
merlin1991 | I think the patch was in kp for some time but got removed again | 01:24 |
Estel_ | it was for a 5 mijnutes in non-repo version of kp, but got kicked out, as he was not interested in maintaining it | 01:24 |
merlin1991 | you'd have to ask Pali for the details | 01:24 |
Estel_ | and fixing to apply only for speaker output, not all output | 01:24 |
javispedro | Support-for-tlv320aic3x-codec-highpass-filter-needed.diff | 01:24 |
javispedro | ^^ found it | 01:24 |
Estel_ | yea, I'm collecting irc log and will pester Pali about usability of properly integrating it | 01:24 |
DocScrutinizer05 | that's it | 01:24 |
Estel_ | yea | 01:24 |
Estel_ | it is it | 01:24 |
DocScrutinizer05 | that's the codec hw filter | 01:25 |
javispedro | so you should fetch the tlv320 datasheet and read whether the filtering applies to the 2nd analog bypass | 01:25 |
DocScrutinizer05 | hehehe | 01:25 |
DocScrutinizer05 | HF | 01:25 |
javispedro | I think that means "probably not" ;) | 01:26 |
DocScrutinizer05 | the datasheet is BS | 01:26 |
DocScrutinizer05 | I got severe headache and left me as wise as before more often than not | 01:26 |
javispedro | https://garage.maemo.org/plugins/ggit/browse.php/?p=kernel-power;a=blob;f=kernel-power-2.6.28/debian/patches/Support-for-tlv320aic3x-codec-highpass-filter-needed.diff;h=962061affdbff324d640a6edf3b5fd17889b74e7;hb=HEAD | 01:27 |
javispedro | from reading the patch, MNZ did it. | 01:27 |
DocScrutinizer05 | so dunno, afaik the filter is a part of digital domain, so if your bypass is mere analog domain, it obviosuly won't get any filtering | 01:27 |
DocScrutinizer05 | MNZ was the name | 01:28 |
DocScrutinizer05 | yep | 01:28 |
DocScrutinizer05 | I guided him | 01:28 |
javispedro | it seems to be right before the DAC, so digital domain. | 01:28 |
DocScrutinizer05 | yep, sure | 01:28 |
javispedro | well, you can't have everything | 01:28 |
javispedro | at least you can now safely use alsa directly | 01:28 |
DocScrutinizer05 | :nod: | 01:29 |
DocScrutinizer05 | if only @¼¼³&§%##!!”ĸµ̣Nokia would bother to tell about the *parameters* of their friggin XPROT | 01:29 |
Estel_ | so, this path doesn't have anything to do with our analog bypass, hm | 01:29 |
Estel_ | no idea to re-implement it for analog too? | 01:30 |
DocScrutinizer05 | not feasible | 01:30 |
DocScrutinizer05 | there are no analog filters | 01:30 |
Estel_ | fun mod - hardware highpass filter connected before speaker's springs | 01:30 |
javispedro | also, the patch will add some new 3d effects to alsamixer | 01:30 |
DocScrutinizer05 | well, THAT was feaqsible, in theory | 01:31 |
Estel_ | 0_o 3d effects is something i react allergically to | 01:31 |
DocScrutinizer05 | so you'll appreciate you now can turn them off ;-P | 01:31 |
javispedro | hah | 01:31 |
javispedro | you could probably implement a hardware equalizer | 01:31 |
Estel_ | DocScrutinizer, thin plate with cooper, to be placed between motherboard and speakers, that lead to one highpass hardware filter :P | 01:31 |
DocScrutinizer05 | javispedro: that's what MNZ actually planned to do | 01:32 |
Estel_ | hardware equalizer, hujh | 01:32 |
javispedro | very nice | 01:32 |
Estel_ | interesting | 01:32 |
javispedro | I used to have a hweq with my creative card | 01:32 |
DocScrutinizer05 | he gave up when he realized complexity to calculate the coefficients of digital filter | 01:32 |
Estel_ | Audigy 2ZS video editor gold, here ;) | 01:32 |
javispedro | yeah, I had 2ZS too | 01:32 |
javispedro | best Linux sound card ever imho, hwmixing, sf2. | 01:33 |
Estel_ | pity that they have castrated X-fi and later ones | 01:33 |
Estel_ | yes | 01:33 |
javispedro | X-Fi is not castrated, driver just does not support anything | 01:33 |
javispedro | the new ones ARE castrated. | 01:33 |
DocScrutinizer05 | this didital filter in codec is a biatch, regarding handling | 01:33 |
Estel_ | this oen i'm talking about is external usb one, works even with N900 via already build in modules (!) | 01:33 |
Estel_ | IOI also have PCI one lying in drawer, also gold | 01:33 |
Estel_ | sad fact - despite the usb one being video editor, I can't seem to utilize video part of it, even on windoze, LOL. Pity, as it's video part was decent | 01:34 |
DocScrutinizer05 | you got like 8 32bit coefficients, and none of them is exactly trivial to calculate | 01:34 |
Estel_ | hardware mpeg2 compression with very high output quality... eh | 01:34 |
Estel_ | javispedro, wasn't X-fi lacking sound processor on card?> | 01:34 |
Estel_ | off-loading it to CPU? | 01:35 |
Estel_ | it was quite... smalll, light (as per weight), and seemed to contain small ammount of components | 01:35 |
javispedro | Estel_: emu20kx ones don't | 01:35 |
Estel_ | or I mistaked it with next one | 01:35 |
Estel_ | I see | 01:35 |
javispedro | on my todo list there's an item "buy titanitum card and figure out how to get emux synthesizer working" =) | 01:36 |
Estel_ | I used to have this big desk thing, from audigy 2zs, from "pro" version, with big jacks... | 01:36 |
Estel_ | but 2zs video editor is even better, all kind of audio output and imputs, including optical ones and midi. + built in powered 4 port usb hub :P | 01:37 |
Estel_ | javispedro, :) | 01:37 |
Estel_ | isn't it better to just use old good 2zs? | 01:37 |
Estel_ | what titanium offer over it? | 01:37 |
javispedro | pciexpress | 01:37 |
Estel_ | yes, and it's required for? | 01:37 |
merlin1991 | low latency? | 01:38 |
javispedro | in two years you won't be able to find motherboards with pci | 01:38 |
merlin1991 | and that :D | 01:38 |
Estel_ | merlin1991, would need implementation in card for that | 01:38 |
javispedro | _good_ ones I mean | 01:38 |
Estel_ | javispedro, pci to usb adaptors? | 01:38 |
Estel_ | + DIY case | 01:38 |
Estel_ | and You have desktop external card :) | 01:38 |
javispedro | pci to usb adaptor? :P | 01:38 |
Estel_ | may be BS, heard somewhere about it | 01:38 |
Estel_ | anyway, it's good argument - I have notebook acting as desktop, it was another reason why audigy 2zs video editor (it's USB) | 01:39 |
Estel_ | there was also some kind of notebook-friendly one, powered from USDB, but it was fubar - advertised as 24 bit one, it did internal downsampling to 16 bi for processing, then, upsampling to 24 bit on exit | 01:39 |
javispedro | also, pci 2zs usually die of the "crackling" problem | 01:39 |
Estel_ | Creative never was very reliable partner :P | 01:39 |
javispedro | they are harder to get every day :( | 01:40 |
Estel_ | good to know, my gold one bought for 30 dollars may be worth more now :p | 01:40 |
Estel_ | but this usb external one (requiring another power source from mains) zs video editor is really state of art | 01:41 |
Estel_ | for both audio and video. If it would only work for me re video part... :P | 01:41 |
Estel_ | never needed to use it, so noticed that video isn't working 2 years after purchase | 01:41 |
Estel_ | oddly, video processing part is visible from operating system as working ok, it just refuses to work in practice | 01:41 |
Estel_ | every video capturing thing just hangs out, when card's imput is selected | 01:42 |
Estel_ | odd as hell | 01:42 |
Estel_ | and of course only windoze drivers for video part, say hi to creative | 01:42 |
Estel_ | notg popular one enough to have linux Re drivers | 01:42 |
Estel_ | although I was surprised as hell, when noticing that I may connect it to N900 via hostmode, and enjoy 7.1 audio output (5.1 in practice) | 01:43 |
Estel_ | only through mplayer, though, of course | 01:43 |
Estel_ | say hi to N900's pulseaudio :P | 01:43 |
Estel_ | (every attempt to redirect mafw output to usb sound card failed msierably) javispedro, You even were in this thread | 01:44 |
javispedro | I remember the last consensus was "kernel bug" | 01:44 |
Estel_ | wut? | 01:46 |
Estel_ | either I remember it wrong way, or it was something different - I would be pestering Pali to fix it in KP | 01:46 |
javispedro | meh, my memory is quite hazy | 01:47 |
Estel_ | javispedro, http://http://talk.maemo.org/showthread.php?t=83270 | 01:51 |
Estel_ | consensus was that You haven't had USB card/speaker :) | 01:52 |
Estel_ | and were wondering about emulating it | 01:52 |
javispedro | and I don't, you got me interested in the video capture + audio card thing though | 01:52 |
Estel_ | see: http://talk.maemo.org/showpost.php?p=1191590&postcount=77 | 01:52 |
Estel_ | and this: | 01:52 |
Estel_ | http://talk.maemo.org/showpost.php?p=1191583&postcount=75 | 01:52 |
Estel_ | hm, audio and video thing>? ncie to hear, i though you've bored You :) | 01:53 |
Estel_ | well, it is | 01:53 |
Estel_ | lemme search | 01:53 |
Estel_ | http://www.google.pl/search?q=audigy+2zs+video+editor&hl=pl&prmd=imvns&tbm=isch&tbo=u&source=univ&sa=X&ei=3wQbUPXUN62N4gTZ24H4Dg&ved=0CGUQsAQ&biw=1540&bih=773 | 01:53 |
Estel_ | it's lovely external USB card, this time not fake'ing anything, and being real 24bit/96khz one | 01:54 |
Estel_ | (up to, you may use real-life values of 16bit 48 khz too) | 01:54 |
*** DocScrutinizer05 has quit IRC | 01:54 | |
*** DocScrutinizer05 has joined #maemo-ssu | 01:55 | |
Estel_ | it have hardware control over imput level via gauge, and gauge for outpout, hoever, the latter is mixed with OS control | 01:55 |
Estel_ | inas and outs visible in photos | 01:55 |
Estel_ | + 4HUB usb 2.0 active hub (to the point of requiring external power to have even only hub working) | 01:55 |
javispedro | are you sure it is a emu10k1 card? | 01:56 |
javispedro | which ALSA driver are you using? | 01:56 |
Estel_ | erm, in fact I'm not sure at all, but for sure it works with alsa... | 01:56 |
Estel_ | hmmm | 01:56 |
Estel_ | which ones we have in N900? :D | 01:56 |
Estel_ | that's the oldest I've tried | 01:56 |
javispedro | just lsmod and grep for emu10k1 | 01:57 |
Estel_ | gimme a second, I'm on windoze actually (i know, shame...), it should work when i connect it to n900 too? | 01:57 |
Estel_ | (emu10k1) | 01:57 |
Estel_ | (not card itself) | 01:57 |
javispedro | http://www.alsa-project.org/main/index.php/Matrix:Vendor-Creative_Labs | 01:58 |
javispedro | they don't seem to know anything about it | 01:58 |
javispedro | I bet it is probably not emu10k1 and just USB sound card ;P | 01:59 |
javispedro | also, kernel bug here: http://talk.maemo.org/showpost.php?p=1193473&postcount=80 | 02:00 |
Estel_ | javispedro, tried it and no emu10k1 | 02:04 |
Estel_ | what drawbacks it does mean? | 02:04 |
javispedro | no hw equalizer | 02:04 |
javispedro | and no hw mixing I think | 02:04 |
Estel_ | strange, I can definitely mix in hw | 02:04 |
Estel_ | maybe it require custom drivers, though | 02:04 |
javispedro | not dmix? :) | 02:04 |
Estel_ | definitely not dmix | 02:04 |
*** DocScrutinizer05 has quit IRC | 02:05 | |
Estel_ | no much experience with using it under linux, sadly, for any advanced things, thanks to it being very niche | 02:05 |
Estel_ | so under linux I use it for... well, normal audio output :P | 02:05 |
Estel_ | under windoze it definitely have hardware mixing and hardware equalizer, but it's thing of propertiary drivers, probably | 02:05 |
Estel_ | because, upon connection it's recognized (under linux) ass usb hub first, then, card is connected to internal ports... | 02:06 |
Estel_ | on windows, You see more components | 02:06 |
Estel_ | i.e. more devices detected | 02:06 |
Estel_ | probably, under linux it works as normal USB card | 02:06 |
Estel_ | so no rela 5.1 for me, then, from N900? :) | 02:06 |
Estel_ | continuation - I think that under windoze another device is detected, as one talking with hardware mjxing parts | 02:07 |
Estel_ | mixing* | 02:07 |
Estel_ | probably even Hw equalizer, although I don't use equalizers at all, so hard to tell | 02:07 |
Estel_ | definitely no video device detected under linux, too | 02:07 |
javispedro | I am reading the alsa snd-usb driver and they only do 1 substream, so no hw mixing | 02:07 |
* RST38h moos at javispedro evilly | 02:07 | |
* Estel_ nods at javispedro | 02:08 | |
javispedro | moo RST38h | 02:08 |
Estel_ | pity :D | 02:08 |
Estel_ | hey, so no multi-channel recording too? | 02:08 |
Estel_ | :((( | 02:08 |
javispedro | it might have multichannel recording, dunno about that | 02:08 |
Estel_ | it's pity that RE things (as under windozxe it works, definitely) isn't ferasible | 02:08 |
Estel_ | feasible, lol | 02:09 |
Estel_ | sorry for typos | 02:09 |
* RST38h runs out with a pitchfork | 02:09 | |
javispedro | Estel_: CT is not very friendly any longer either | 02:10 |
Estel_ | I just suspect that my video processing part is faulty, which is strange, as it's bundled on 1 board, and as said, drivers report hardware detected properly (on windoze) | 02:10 |
javispedro | previously they would even submit research papers frm time to time. | 02:10 |
Estel_ | Creative was friendly anytime?:p | 02:10 |
Estel_ | ah, I see | 02:10 |
Estel_ | I remember them as asshats from the beginning, though. | 02:10 |
javispedro | well, I might be confusing EMU with CT ;P | 02:10 |
Estel_ | I wonder what bite them | 02:10 |
*** DocScrutinizer05 has joined #maemo-ssu | 02:19 | |
*** taziff1 has quit IRC | 02:33 | |
*** taziff has joined #maemo-ssu | 02:35 | |
*** taziff has quit IRC | 02:38 | |
*** taziff has joined #maemo-ssu | 02:38 | |
*** kolp has quit IRC | 02:45 | |
*** javispedro has quit IRC | 03:19 | |
*** toxaris has quit IRC | 03:48 | |
*** arcean has quit IRC | 03:58 | |
*** nox- has quit IRC | 04:08 | |
*** LaoLang_cool has joined #maemo-ssu | 05:06 | |
*** LaoLang_cool has quit IRC | 05:09 | |
*** LaoLang_cool has joined #maemo-ssu | 05:11 | |
*** LaoLang_cool has quit IRC | 05:15 | |
*** LaoLang_cool has joined #maemo-ssu | 05:18 | |
*** LaoLang_cool has joined #maemo-ssu | 05:19 | |
*** amiconn_ has joined #maemo-ssu | 05:28 | |
*** amiconn has quit IRC | 05:28 | |
*** amiconn_ is now known as amiconn | 05:28 | |
*** LaoLang_cool has quit IRC | 05:54 | |
*** LaoLang_cool has joined #maemo-ssu | 05:54 | |
*** LaoLang_cool has quit IRC | 05:58 | |
*** LaoLang_cool has joined #maemo-ssu | 05:59 | |
*** LaoLang_cool has quit IRC | 06:02 | |
*** LaoLang_cool has joined #maemo-ssu | 06:04 | |
*** LaoLang_cool has joined #maemo-ssu | 06:04 | |
*** LaoLang_cool has quit IRC | 06:06 | |
*** LaoLang_cool has joined #maemo-ssu | 06:08 | |
*** LaoLang_cool has quit IRC | 06:16 | |
*** LaoLang_cool has joined #maemo-ssu | 06:18 | |
*** LaoLang_cool has quit IRC | 06:53 | |
*** jonwil has joined #maemo-ssu | 07:06 | |
*** LaoLang_cool has joined #maemo-ssu | 07:13 | |
*** LaoLang_cool has quit IRC | 07:17 | |
*** LaoLang_cool has joined #maemo-ssu | 07:19 | |
*** taziff has quit IRC | 07:26 | |
*** taziff has joined #maemo-ssu | 07:28 | |
*** LaoLang_cool has quit IRC | 07:44 | |
*** LaoLang_cool has joined #maemo-ssu | 07:50 | |
freemangordon | merlin1991: DocScrutinizer05: childspin just hanged here | 08:10 |
freemangordon | on device, running HAM update in foreground | 08:11 |
freemangordon | merlin1991: the very reliable way to make it lock is: run childspin -> opeh HAM -> click on menu -> application catalogs -> rotate the device | 08:17 |
freemangordon | usually it hangs on the first rotation, but may take 2 or 3 | 08:17 |
freemangordon | kernel-cssu3, not stock kernel, so BTB clear may affect it. | 08:32 |
freemangordon | though I doubt | 08:32 |
*** amiconn has quit IRC | 09:16 | |
*** amiconn has joined #maemo-ssu | 09:17 | |
*** _rd has joined #maemo-ssu | 09:21 | |
*** Pali has joined #maemo-ssu | 09:23 | |
Pali | jonwil, netlink is standard way how to communicate between userspace and kernel... it is socket orientated communication, so you need to create "protocol" how are data sent. and nokia created a lot of non standard protocols (e.g how to send NVS wifi data to kernel driver...) | 09:28 |
jonwil | yeah thats what I thought | 09:29 |
jonwil | Is there a better way to do nvs->kernel? | 09:29 |
Pali | wl1271 (*71*) has NVS data in firmware file | 09:29 |
Pali | so it is loaded as firmware data | 09:30 |
jonwil | is that the way the N900 should go? | 09:30 |
jonwil | or should we stick with WL1251_NL_CMD_NVS_PUSH? | 09:30 |
Pali | I think that wl1251 driver can use this same way too | 09:30 |
Pali | wl12XX share more code in kernel | 09:31 |
*** _rd has quit IRC | 09:31 | |
Pali | look here: http://linuxwireless.org/en/users/Drivers/wl12xx/calibrator | 09:32 |
Pali | this is tool which create NVS data as firmware file | 09:32 |
Pali | this also add mac address into NVS data | 09:32 |
Pali | and then kernel can use it when module wl1251 is loaded | 09:32 |
Pali | the right way should be: when updating kernel driver, postinst script will call that tool (maybe patched to read data from CAL) and it will create proper NVS firmware file | 09:34 |
Pali | and then wl1251-cal can be deleted, because kernel have all data in firmware file | 09:34 |
Pali | and for regional wifi settings (e.g. enabled channels): for linux kernel is cfg80211 driver which regulate it | 09:35 |
Pali | from userspace you can change country with program "iw" | 09:35 |
jonwil | ok, well what wl1251-cal does is this: | 09:36 |
jonwil | firstly is asks cell network for country code | 09:37 |
jonwil | then it compares it against stored list to determine if its "FCC country" | 09:37 |
jonwil | then it reads NVS from CAL | 09:37 |
jonwil | makes 4 changes to NVS data depending on whether its FCC country or not | 09:38 |
jonwil | and pushes that to driver via WL1251_NL_CMD_NVS_PUSH | 09:38 |
jonwil | then it reads MAC address from CAL | 09:39 |
jonwil | MAC address is sent to driver via SIOCSIFHWADDR | 09:40 |
jonwil | it also uses NL80211_ATTR_REG_ALPHA2 to set country code | 09:40 |
jonwil | right now my plan is to clone wl1251-cal except that it wont use systeminfo and it will pull country value from CAL (same place as sysinfo-tool gets /certs/ccc/pp/wlan-channel from) | 09:41 |
jonwil | just like I did with bluetooth-cal | 09:42 |
jonwil | I think best answer is to:1.Replace non-standard bluetooth hwaddr interface with something standard | 09:43 |
jonwil | 2. | 09:43 |
jonwil | 2. | 09:43 |
jonwil | wait ignore that :P | 09:43 |
jonwil | 2.Write tool to read NVS from CAL and apply country change based on value of wlan-channel | 09:44 |
jonwil | 3.Make wl1251 driver use stored firmware nvs | 09:44 |
jonwil | i.e. tool at #2 would produce identical NVS to wl1251-cal | 09:44 |
jonwil | then 4.Write new tool to read bluetooth and wifi MAC plus country code and send all 3 to driver | 09:45 |
jonwil | drivers | 09:45 |
jonwil | but yeah I will write the wl1251-cal clone | 09:45 |
jonwil | and others can go from there | 09:45 |
jonwil | will also write the libcal clone | 09:46 |
jonwil | at some point | 09:46 |
*** BCMM has joined #maemo-ssu | 09:49 | |
jonwil | the goal is to replace wl1251-cal and bluetooth-sysinfo (and possibly libcal) on Fremantle and also to replace sysinfod-rx51, libcxal-rx-51-bin, libppu-bin, libwl1251-bin and wl1251-cal-bin on MeeGo & friends | 09:50 |
Pali | jownil, for setting country code, you can use standard iw tool: "iw reg set US" (change US) | 09:54 |
Pali | jonwil, look at linux macchanger program | 09:54 |
Pali | how it set mac address and check if same interface is also for bluetooth | 09:54 |
jonwil | ok, well I will clone the bits and someone else can find a way to make them do things the Right Way(tm) | 09:55 |
Pali | maybe asking on linux bluez mailinglist how to change mac address of bluetooth is good idea | 09:55 |
freemangordon | hcitool? | 09:56 |
Pali | maybe, hcitool can change mac address, I do not know | 10:02 |
Pali | or hciconfig... | 10:02 |
Pali | freemangordon, for vkb see: https://maemo.org/downloads/product/Maemo5/ukeyboard/ and http://repo.or.cz/w/ukeyboard.git | 10:03 |
freemangordon | hmm, not sure whether it was hcitool or hciconfig :D | 10:03 |
Pali | this tool can create binary vkb layout | 10:03 |
Pali | there are both: hcitool and hciconfig | 10:04 |
freemangordon | Pali: I know ukeyboard as I am using it, the tools there is for maem4, iirc jonwil fixed it for fremantle | 10:04 |
Pali | ukeyboard working on fremantle too | 10:04 |
Pali | I'm using it for Slovak qwerty layout | 10:04 |
jonwil | You want http://talk.maemo.org/showthread.php?t=79899 | 10:05 |
Pali | http://maemo.org/packages/view/ukeyboard/ | 10:05 |
jonwil | Thats the best way to do the vkb on N900 | 10:05 |
freemangordon | I know, I am using it for bulgarian :) | 10:05 |
freemangordon | Pali: I was talking about RE of .vkb files | 10:05 |
Pali | jonwil, it that your code included in ukeyboard? | 10:05 |
jonwil | no | 10:05 |
jonwil | I have nothing to do with ukeyboard | 10:06 |
freemangordon | there is a python/perrl/whatever script flying over the inet | 10:06 |
freemangordon | jonwil: ^^^ | 10:06 |
jonwil | my stuff in that tmo thread contains all the info one needs on the vkb format | 10:06 |
freemangordon | jonwil: I remember I sent that to you, and you fixed it for fremantle .vkb format | 10:06 |
jonwil | including a fremantle-compatible version of the aformentioned perl script | 10:06 |
freemangordon | jonwil: :nod: | 10:06 |
jonwil | yeah | 10:06 |
Pali | I will ask roman if he used your code or not | 10:07 |
jonwil | so yeah that stuff contains decode script, vkb format doc plus code to generate vkb file using libimlayout | 10:07 |
Pali | we do not need more implementations of vkb format... | 10:07 |
freemangordon | Pali: vkbrenderer.so is the one responsible of reading that format and actually displaying the UI | 10:08 |
jonwil | libimlayout is what parses the vkb file format | 10:08 |
jonwil | vkbrenderer is what displays the results recieved from libimlayout | 10:08 |
freemangordon | though not sure if it receiveswhat to display from upper/lower layers | 10:09 |
freemangordon | jonwil: :nod: | 10:09 |
zeq | merlin1991: I should have mentioned having HAM running is a very good way of putting enough load on the system to open up the race window. | 10:09 |
zeq | morning guys :) | 10:10 |
freemangordon | zeq: morning | 10:10 |
Pali | ukeyboard has vkb-format-v3 | 10:10 |
freemangordon | zeq: but not enough, it is the rotation that triggers the bug in 90% of the cases | 10:10 |
jonwil | bah, maemo.org is being annoying | 10:10 |
freemangordon | Pali: so? | 10:11 |
freemangordon | jonwil: don't you like the pink nailpolish? | 10:11 |
freemangordon | it is lumia after all, what color do you expect :D | 10:11 |
zeq | freemangordon: ok, but the point is running the test on an idle system minimizes the race window. | 10:12 |
jonwil | ukeyboard is not using my code as far as I can see | 10:12 |
freemangordon | zeq: agree | 10:13 |
zeq | it probably depends on time spent in kernel(maybe rotation triggers some long latency kernel code?) | 10:13 |
freemangordon | zeq: maybe | 10:13 |
zeq | not that it really matters :) | 10:14 |
freemangordon | but it really does nto matter | 10:14 |
freemangordon | yeah | 10:14 |
Pali | I can write Roman (now maintainer of ukeyboard) to update code | 10:14 |
freemangordon | Pali: to add new keyboards or what? | 10:14 |
Pali | ukeyboard has rewritten control panel aplet | 10:14 |
Pali | to update documentation | 10:15 |
Pali | jonwil has some more info about format | 10:15 |
freemangordon | BTW ukeyboard has an annoying bug (from user POV) | 10:15 |
freemangordon | once installed there are no more messages "input layout has been switched to ..." | 10:16 |
jonwil | my gen_vkb script is as close as its possible to get to the way Nokia does it | 10:16 |
jonwil | not script, binary | 10:16 |
jonwil | I found gen_vkb binary for older Maemo, reverse engineered it and then exended it for Fremantle | 10:16 |
jonwil | Nokia obviously has gen_vkb for Fremantle internally | 10:16 |
jonwil | real one that is | 10:17 |
freemangordon | maybe I missed the point, but what is the relation between vkbrenderer and ukeyboard/gen_vkb? | 10:17 |
freemangordon | anyway, I gtg, see ya | 10:19 |
*** Pali has quit IRC | 10:20 | |
jonwil | basically gen_vkb generates a .vkb file of the keyboard layout | 10:21 |
jonwil | then libimlayout0 parses that and hands the data off to vkbrenderer | 10:21 |
jonwil | which then renders it | 10:21 |
*** luf has joined #maemo-ssu | 10:53 | |
*** zeq has quit IRC | 10:55 | |
*** zeq has joined #maemo-ssu | 11:08 | |
*** Sv has joined #maemo-ssu | 11:08 | |
*** toxaris has joined #maemo-ssu | 11:11 | |
*** LaoLang_cool has quit IRC | 11:44 | |
*** jonwil has quit IRC | 11:48 | |
kerio | idea: rotation lock in the powerkey menu, and out of the status menu | 11:59 |
*** kolp has joined #maemo-ssu | 12:07 | |
*** kolp has quit IRC | 12:14 | |
*** arcean has joined #maemo-ssu | 13:09 | |
*** andre__ has joined #maemo-ssu | 13:25 | |
*** andre__ has joined #maemo-ssu | 13:25 | |
*** lizardo has joined #maemo-ssu | 13:35 | |
*** Estel_ has quit IRC | 13:51 | |
*** DocScrutinizer06 has joined #maemo-ssu | 13:53 | |
*** EdLin has quit IRC | 13:53 | |
*** infobot has quit IRC | 13:53 | |
*** DocScrutinizer05 has quit IRC | 13:53 | |
*** DocScrutinizer06 is now known as DocScrutinizer05 | 13:54 | |
*** EdLin has joined #maemo-ssu | 13:54 | |
*** DocScrutinizer05 has quit IRC | 13:55 | |
*** DocScrutinizer05 has joined #maemo-ssu | 13:55 | |
DocScrutinizer05 | kerio: ++ | 13:57 |
*** BCMM has quit IRC | 14:00 | |
*** DocScrutinizer has quit IRC | 14:09 | |
*** DocScrutinizer has joined #maemo-ssu | 14:12 | |
*** Sv has left #maemo-ssu | 14:35 | |
*** toxaris has quit IRC | 14:40 | |
*** jonwil has joined #maemo-ssu | 15:18 | |
kerio | hm, for what? lost the backscroll | 15:33 |
DocScrutinizer05 | [2012-08-03 10:59:55] <kerio> idea: rotation lock in the powerkey menu, and out of the status menu | 15:35 |
kerio | oh yeah | 15:35 |
kerio | kerio++ | 15:35 |
*** zeq has quit IRC | 16:13 | |
*** ekze has joined #maemo-ssu | 16:13 | |
*** zeq1 has joined #maemo-ssu | 16:33 | |
*** DocScrutinizer has quit IRC | 17:08 | |
*** DocScrutinizer has joined #maemo-ssu | 17:08 | |
*** DocScrutinizer05 has quit IRC | 17:09 | |
*** DocScrutinizer06 has joined #maemo-ssu | 17:09 | |
*** DocScrutinizer06 is now known as DocScrutinizer05 | 17:09 | |
*** BCMM has joined #maemo-ssu | 17:16 | |
*** luf has quit IRC | 17:33 | |
*** infobot has joined #maemo-ssu | 18:17 | |
*** ChanServ sets mode: +v infobot | 18:17 | |
*** BCMM has quit IRC | 18:35 | |
*** BCMM has joined #maemo-ssu | 18:38 | |
*** NIN101 has joined #maemo-ssu | 19:04 | |
*** jon_y_ has joined #maemo-ssu | 19:05 | |
*** jon_y has quit IRC | 19:05 | |
*** jon_y_ is now known as jon_y | 19:05 | |
*** zeq1 has quit IRC | 19:49 | |
*** zeq1 has joined #maemo-ssu | 19:49 | |
kerio | ok, i have the operator name problem again | 20:09 |
kerio | jonwil: *poke?* | 20:10 |
kerio | i can test the bugfix i guess | 20:10 |
kerio | DocScrutinizer05: you too | 20:15 |
*** DocScrutinizer has quit IRC | 20:15 | |
*** DocScrutinizer has joined #maemo-ssu | 20:15 | |
*** DocScrutinizer has quit IRC | 20:15 | |
*** DocScrutinizer has joined #maemo-ssu | 20:15 | |
*** DocScrutinizer has quit IRC | 20:16 | |
*** DocScrutinizer has joined #maemo-ssu | 20:16 | |
DocScrutinizer05 | eh? | 20:18 |
kerio | DocScrutinizer05: i'm ready for more cbsms tests | 20:18 |
kerio | with the fixed version | 20:18 |
DocScrutinizer05 | mhm | 20:18 |
kerio | that i don't know how to install and/or compile | 20:18 |
DocScrutinizer05 | alas i'm not familiar with the actual implementation, just contributed some architecture | 20:18 |
kerio | k | 20:19 |
kerio | should i ask jonwil then? (was it him?) | 20:19 |
*** javispedro has joined #maemo-ssu | 20:19 | |
DocScrutinizer05 | probably for specifics about the current binary/pkg you better ask pali or jonwil | 20:19 |
DocScrutinizer05 | :nod: | 20:19 |
DocScrutinizer05 | jonwil: | 20:19 |
kerio | jonwil! | 20:20 |
kerio | why pali? :o | 20:20 |
kerio | he's everywhere | 20:20 |
DocScrutinizer05 | he did some fixes to the last version before the assumed jonwil bleeding edge one | 20:21 |
DocScrutinizer05 | since jonwil isn't really available that frequently | 20:21 |
DocScrutinizer05 | though he's logged in right now, so odds are he'll answer in a minute | 20:22 |
DocScrutinizer05 | kerio: for now you could provide apt-cache policy | 20:22 |
DocScrutinizer05 | for that thing | 20:22 |
DocScrutinizer05 | so we know what you're using and what we're talking bout | 20:23 |
jonwil | I am here but busy trying to debug with gdb on my phone keyboard (cant use ssh because wl1251-cal work requires wifi to be offline when running the binary) | 20:24 |
jonwil | the good news is that the mac address seems to be being set correctly | 20:24 |
jonwil | ioctl() isn't returning an error | 20:25 |
kerio | what's wl1251-cal? | 20:25 |
kerio | DocScrutinizer05: k, one sec | 20:26 |
kerio | Installed: 0.2 | 20:27 |
jonwil | wl1251-cal handles settings for the wl1251 wifi chip | 20:27 |
kerio | for operator-name-cbs-widget | 20:27 |
jonwil | i.e. setting mac address | 20:27 |
kerio | oh, in the proper way? | 20:27 |
jonwil | and dealing with regulatory domain | 20:27 |
kerio | instead of using the l33t drivers | 20:27 |
kerio | i see | 20:27 |
kerio | i'd like to be in japan, yes | 20:28 |
jonwil | wl1251-cal is closed source from nokia, I am currently doing something that does the same thing as wl1251-cal from nokia except open | 20:28 |
jonwil | open source | 20:28 |
jonwil | then someone else can make it use better interfaces and not the nokia-specific stuff | 20:28 |
DocScrutinizer05 | jonwil: nice stuff, anyway could you spare a few minutes to direct kerio towards properly exploiting the bug just showing up again in his operator-name-cbs-widget? We eventually like to get that fixed, and he's willing to help and just can reproduce the issue | 20:31 |
kerio | "sure sure, but fix this bug first" | 20:32 |
*** Jade has joined #maemo-ssu | 20:36 | |
*** zeq1 has quit IRC | 20:49 | |
*** zeq1 has joined #maemo-ssu | 21:05 | |
*** Woody14619 has joined #maemo-ssu | 21:07 | |
*** Woody14619 has joined #maemo-ssu | 21:07 | |
*** ruskie has quit IRC | 21:16 | |
*** ThreeM has quit IRC | 21:20 | |
*** ThreeM has joined #maemo-ssu | 21:21 | |
*** ShadowJK has quit IRC | 21:25 | |
*** FireFly has quit IRC | 21:32 | |
*** ThreeM has quit IRC | 21:32 | |
*** ChanServ has quit IRC | 21:32 | |
*** NIN101 has quit IRC | 21:32 | |
*** jonwil has quit IRC | 21:32 | |
*** EdLin has quit IRC | 21:32 | |
*** freemangordon has quit IRC | 21:32 | |
*** Woody14619 has quit IRC | 21:32 | |
*** zeq1 has quit IRC | 21:32 | |
*** infobot has quit IRC | 21:32 | |
*** andre__ has quit IRC | 21:32 | |
*** arcean has quit IRC | 21:32 | |
*** RST38h has quit IRC | 21:32 | |
*** kerio has quit IRC | 21:32 | |
*** jon_y has quit IRC | 21:32 | |
*** Jaffa has quit IRC | 21:32 | |
*** macmaN has quit IRC | 21:32 | |
*** Sc0rpius has quit IRC | 21:32 | |
*** tadzik has quit IRC | 21:32 | |
*** Sicelo has quit IRC | 21:32 | |
*** DocScrutinizer05 has quit IRC | 21:32 | |
*** lizardo has quit IRC | 21:32 | |
*** chem|st has quit IRC | 21:32 | |
*** DocScrutinizer51 has quit IRC | 21:32 | |
*** DocScrutinizer51 has joined #maemo-ssu | 21:37 | |
*** chem|st has joined #maemo-ssu | 21:37 | |
*** lizardo has joined #maemo-ssu | 21:37 | |
*** DocScrutinizer05 has joined #maemo-ssu | 21:37 | |
*** ShadowJK_ has joined #maemo-ssu | 21:37 | |
*** FireFly has joined #maemo-ssu | 21:37 | |
*** ThreeM has joined #maemo-ssu | 21:37 | |
*** Woody14619 has joined #maemo-ssu | 21:37 | |
*** jon_y has joined #maemo-ssu | 21:37 | |
*** NIN101 has joined #maemo-ssu | 21:37 | |
*** infobot has joined #maemo-ssu | 21:37 | |
*** jonwil has joined #maemo-ssu | 21:37 | |
*** EdLin has joined #maemo-ssu | 21:37 | |
*** andre__ has joined #maemo-ssu | 21:37 | |
*** arcean has joined #maemo-ssu | 21:37 | |
*** freemangordon has joined #maemo-ssu | 21:37 | |
*** Jaffa has joined #maemo-ssu | 21:37 | |
*** ChanServ has joined #maemo-ssu | 21:37 | |
*** kerio has joined #maemo-ssu | 21:37 | |
*** RST38h has joined #maemo-ssu | 21:37 | |
*** sendak.freenode.net sets mode: +vo infobot ChanServ | 21:37 | |
*** BCMM has quit IRC | 21:38 | |
*** ruskie has joined #maemo-ssu | 21:46 | |
*** macmaN has joined #maemo-ssu | 21:51 | |
*** Sc0rpius has joined #maemo-ssu | 21:51 | |
*** tadzik has joined #maemo-ssu | 21:51 | |
*** Sicelo has joined #maemo-ssu | 21:51 | |
*** arcean has quit IRC | 22:05 | |
*** zeq has joined #maemo-ssu | 22:06 | |
*** arcean has joined #maemo-ssu | 22:06 | |
*** ivgalvez-N900 has joined #maemo-ssu | 22:09 | |
*** ivgalvez-N900 has left #maemo-ssu | 22:09 | |
*** trumee has quit IRC | 22:15 | |
*** trumee has joined #maemo-ssu | 22:16 | |
*** NIN101 has quit IRC | 22:18 | |
jonwil | ok, wl1251-cal is DONE :) | 22:19 |
kerio | :D | 22:20 |
kerio | should cssu allow manual configuration of regulatory domains' | 22:20 |
kerio | ? | 22:20 |
jonwil | thats up to CSSU | 22:20 |
kerio | i wasn't necessarily asking you c: | 22:23 |
*** BCMM has joined #maemo-ssu | 22:24 | |
jonwil | :) | 22:26 |
*** _rd has joined #maemo-ssu | 22:27 | |
*** _rd has quit IRC | 22:29 | |
kerio | no, :ↄ | 22:29 |
*** NIN101 has joined #maemo-ssu | 22:29 | |
*** javispedro has quit IRC | 22:38 | |
zeq | jonwil: you mean via iw? | 22:46 |
zeq | sorry, *kerio* | 22:46 |
kerio | zeq: yeah | 22:48 |
kerio | maybe | 22:48 |
kerio | but it would be better if it was automatically set to JP | 22:49 |
kerio | or, rather, configurable in a plaintext file | 22:49 |
kerio | is CAL signed? | 22:49 |
kerio | ugh, why can't everything be like fmtxd :( | 22:55 |
kerio | i mean, i suppose it kinda is, you can just use the backported drivers, but they have a crappy powersaving | 22:55 |
*** jonwil has quit IRC | 23:03 | |
*** BCMM has quit IRC | 23:08 | |
*** infobot has quit IRC | 23:11 | |
*** infobot has joined #maemo-ssu | 23:32 | |
*** ChanServ sets mode: +v infobot | 23:32 | |
DocScrutinizer05 | ABI: Application Binary Interface - the interface on binary level used between binaries to talk to each other. ABI compatibility means binary A doesn't notice *any* change in the interface of binary B it connects to | 23:44 |
DocScrutinizer05 | compare API compatibility which defines a compliance of abstract definition of a interface as used by compilers at build time, so a *new* binary A though built from same sources would still connect without problems to the possibly changed ABI of binary B since both use the same API still | 23:46 |
DocScrutinizer05 | to connect to the funny discussion 24h ago here, which managed so elegantly to declare my 2h in garden at BBQ as a 1 week non-availability, by sending out invitations in obscure channels and 5 minutes in advance, and also spreading FUD about my availability | 23:48 |
DocScrutinizer05 | I'm frankly assuming and stating that several attendees of yesterday's discussion did this on pretty purpose, to bully me out | 23:50 |
* kerio grabs some popcorn | 23:51 | |
DocScrutinizer05 | while the experise and less sneaky half knew about my position already and also knew it's in line with theirs | 23:52 |
DocScrutinizer05 | ooh, btw, estel_ is on holiday for the next 24 months, don't bother to invite or ask him about anything, we already know what he will say anyway (and we know we don't give a *sh*t*) | 23:54 |
DocScrutinizer05 | assumptions assumptions assumptions. And FUD. Is all of that always just in best intent? I doubt it | 23:55 |
DocScrutinizer05 | >>problem with 3rd party modules was *always* purely theoretical one. I.e. "there may exist, in some ancient vault, a module compiled in 3000 before christ..." | 23:56 |
DocScrutinizer05 | >>(from my limited knowledge common folk POV, feel free to correct) | 23:57 |
DocScrutinizer05 | >>i.e. such situation is purely theoretical, just for sake of doing so, no practical problems risk, ever. 0% chance. | 23:57 |
DocScrutinizer05 | >>(from my limited knowledge common folk POV, feel free to correct) | 23:57 |
DocScrutinizer05 | WTF?! | 23:58 |
DocScrutinizer05 | thanks a lot really for this elaborated expertise | 23:58 |
DocScrutinizer05 | now we're all a lot wiser | 23:58 |
*** lizardo has quit IRC | 23:59 | |
*** BCMM has joined #maemo-ssu | 23:59 |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!