<Oksana>
Not lab mice, computer mice :-) But yes, in a lab
<wpwrak>
i would worry a little about the smoking wire, though. that's not a habit usually considered consistent with a long life expectancy
<Oksana>
Computer lab
<Oksana>
Yes, that's piece of wire-isolation burning because of carelessly exposed wires.
<wpwrak>
(bendable device) from apple, we learn :)
<Oksana>
No, Nokia Morph is what I meant.
<Oksana>
It's not smoking all the time, only when I try to 'bend it back into working shape'
<Oksana>
But yes, I have to switch the mouse off when I am not here. Or there will be a fire at _my_ workplace.
<Oksana>
I would pick the mouse apart, and probably shorten the wire to remove (cut off) 'strange' part of the wire altogether, but there are no screws in the mouse. Only plastic bendable hooks?
<wpwrak>
hmm ... i'd get rid of it quickly. that doesn't sounds good at all. may also damage the pc or hub it's connected to.
<Oksana>
Plugged it out of the PC. I value USB ports of the PC too much. Without them, it would be complete black box, unreachable from outside.
<jonwil>
What sort of mouse was it?
<Oksana>
HP mouse.
<jonwil>
in that case just replace it with another mouse
<Oksana>
Usual: two buttons, one scroll wheel.
<jonwil>
mice are cheap
<Oksana>
Seriously, touch-pad integrated inside the computer would have been much better approach. You cannot bend it, break it, or pull wire out of it. Unless you are in the state of anger when you are ready to destroy whole computer.
<jonwil>
Slowly getting closer to a roadmap towards a newer browser engine on the N900
<Oksana>
About mouse: I know a place where I could have caught a mouse, but it's closed. Will not open until next year.
<Oksana>
:-) Nice
<jonwil>
Timeless pointed me at someone over on moznet called romaxa
<jonwil>
haven't actually gotten hold of him yet though :P
<jonwil>
going to point him at the tmo thread and ask if he has anything useful to add
<Oksana>
So he should know fairly well how it works? MicroB and GtkMozEmbed? :-)
<jonwil>
yeah I would imagine
<jonwil>
how much he is willing to share (code, info, thoughts) is another matter
<Oksana>
As much as I can see, there is single plastic thingie which holds two halves of mouse together. I wonder, how should I hold the mouse, to bend the hook and open the mouse, at the same time? And what kind of tool did they expect me to use, to bend the hook?
<jonwil>
why are you wasting time trying to open and repair a cheap crappy HP computer mouse? :P
<Oksana>
Because I do not see any spare mouse close by :P
<Oksana>
Well, I do imagine there is a couple of them in cardboard box in closed room. I would have to ask for access to the room, though; there aren't volunteers available to hold the room open when end of year is approaching :-/
<Oksana>
If you have time to spare, try to use your computer without mouse-touchpad-touchscreen-pointy-hardware. Keyboard-only. With whatever imitation of mouse you have available on your system.
<wpwrak>
hmm, if your year ends around now ... that would be the islamic year then ?
<Oksana>
'End of year' is November+December. And they are already celebrating Christmas here.
kolp has quit [Read error: Connection reset by peer]
b1101 has quit [Ping timeout: 272 seconds]
Kabouik has quit [Ping timeout: 265 seconds]
b1101 has joined #neo900
b1101 has joined #neo900
xes has quit [Quit: Going offline...]
<wpwrak>
Oksana: i got it ! you're living at the north pole. and of course, at this time of the year, everyone is busy with christmas preparations and has no time to grant you access to the local cyber-rodent store
jonwil has quit [Quit: ChatZilla 0.9.91 [SeaMonkey 2.30/20141013232806]]
<Oksana>
Close. Close. But there is no snow, no ice.
norly has quit [Ping timeout: 260 seconds]
<Oksana>
And that's close to _south_ pole.
norly has joined #neo900
<ds2>
you live up in the northpole?
<Oksana>
No, close to _south_ pole. And it does sound (and look) weird when you imagine Santa Claus in a summer heatwave.
<ds2>
do compasses work down there?
<Oksana>
Yes. Not that close to south pole :-) And there is neither ice nor snow.
<Oksana>
Well, except mountains, I expect.
norly has quit [Quit: Leaving.]
<ds2>
I see
<Oksana>
:-)
che1 has quit [Remote host closed the connection]
che1 has joined #neo900
che1 has quit [Remote host closed the connection]
wicket64 has quit [Remote host closed the connection]
b1101 has quit [Ping timeout: 265 seconds]
nox- has quit [Quit: Leaving]
<Oksana>
Quiet...
<wpwrak>
Oksana: if bored while the rest of us is busy or asleep, you may want to try this: http://gameaboutsquares.com/
<Oksana>
Heh... Comparing two datasets: in one case difference of 10 units, in another case difference of 15 units, and outside source says that the difference is 12 units.
jonwil has joined #neo900
Oksana has quit [Ping timeout: 265 seconds]
<freemangordon>
jonwil: romaxa is the guy behind embedlite, toldya, been there :)
<freemangordon>
for sure he knows (or have forgotten) everything about microb
<jonwil>
definatly worth talking to then if I can get hold of him (how much info/know-how/ideas/code he is willing/able to share is another matter)
<freemangordon>
join #embedlite, he was there last time I checked
<jonwil>
I see him on moznet #mozillians
<jonwil>
which is where timeless told me to look for him
<freemangordon>
jonwil: BTW what happens to -voice? Am I left alone there?
<jonwil>
I will still be helping with -voice where possible
<freemangordon>
ok
<jonwil>
the browser work is a sideline
<freemangordon>
I started some RE job on it
<jonwil>
ok
<freemangordon>
and was able to find all the function names :D
<jonwil>
great
<jonwil>
This is what I have for a few data structures
Oksana has quit [Read error: Connection reset by peer]
<jonwil>
are these names you invented or names that you pulled from somewhere?
<freemangordon>
actually the names are in the symbols file, it is just that IDA can't match those automagically
Oksana has joined #neo900
<jonwil>
k
<freemangordon>
jonwil: see ^^^. if you open only the .so with symbols in IDA, you'll see the names with the function addresses
<jonwil>
anyhow, all that FPU/Neon code is outta my league, I will let you do that and focus what time I have on the browser engine stuff and on those bits of -voice that are not armcc
<freemangordon>
jusa_: please look at ^^^. From memory, was that fexp() (I guess this is fast exponent) done in ASM, or was it in C?
<jonwil>
I think the f probably stands for floating point
<freemangordon>
no, it is integer
<jonwil>
hmmm ok
<freemangordon>
jonwil: my point is - if that was in C, then armcc supports some language extension I've never heard about, like (builtin?) for supporting ARM instructions like qadd etc. (qadd is add with saturation)
<freemangordon>
or it is very smart and can optimize the couple of lines in C needed to implement such addition into a single instruction
<freemangordon>
I am almost sure it was in ASM (no function prologue and epilogue, BEQ instead of SSATNE), but still I am curious :)
b1101 has joined #neo900
<Oksana>
wpwrak: Well, IT brought new mouse. Not sure why it was so easy. Do they have too many mice?
<jonwil>
Most IT departments have extras of all sorts of components so they can replace them easily
<jonwil>
at big companies anyway
<jonwil>
that way people dont have downtime unable to do useful work waiting for someone to get a new part
<jonwil>
whether that be a desk chair or a mouse or a monitor
<Oksana>
Still, too easy.
<Oksana>
And a friend of mine tried to get a second monitor some time ago - they said no, unless absolutely necessary to have two monitors.
<Oksana>
I can understand this, though. Everybody having two monitors is a luxury. Everybody having their broken mouse replaced (whatever the reason of being broken) is a necessity.
<jonwil>
bah, no closer to finding the Gecko mainline codebase that matches microb-engine :(
<jonwil>
hang on, this dump of the maemo fremantle SDK has microb-engine_20100401-1.9.2.orig.tar.gz and microb-engine_20100401-1.9.2-5.2+0m5.diff.gz
<jonwil>
I wonder if the .orig.tar.gz is the original unmodified Mozilla code
<jonwil>
with the .diff.gz being the maemo changes
<jonwil>
its a 3.8mb diff file though
<jonwil>
good news is that I just saw microb-geolocation source in that SDK dump meaning its one less thing we need to care about cloning :)
<Oksana>
Good :-)
<Oksana>
About me using Firefox 17: some less fortunate souls still use Firefox 10.
<jonwil>
ugh, that's almost as bad as someone still using IE6
<Oksana>
I am thinking about a new computer mouse: 1. mechanical, not optical; 2. with scroll ball instead of scroll wheel; 3. for USB port. How likely is its existence?
<Oksana>
I wish I could write IE6-theme for Firefox-latest. Not enough time, though.
<Oksana>
The parcel with chargers not arrived yet. Are post-men expecting to bring it after Christmas holidays? As belated gift? Not funny.
<Oksana>
It's within the same city, for the last 5 days, if tracking site is to be believed.
freemangordon_ has joined #neo900
<Oksana>
When it arrives, it will certainly be Unexpected.
<Oksana>
I will put two or three batteries to charge, all at the same time.
<Oksana>
Good bye. Going to get home, and sleep. Or read fiction instead of sleep. Depends on circumstances.
<Oksana>
Or maybe, edit bibliography styles.
<Oksana>
Good bye.
<jonwil>
mechanical mice have gone the way of the dodo
<jonwil>
and good riddance I say
<jonwil>
a GOOD optical mouse with a decent laser sensor is far better than any mechanical ball mouse
<freemangordon_>
jonwil: what about starting the integration of the voice module files from nemo/meego etc into fremantle branch while I am on ASM/NEON stuff?
<jonwil>
I dont think whats in nemo is any help to us, its whats in meego and in harmattan that would help us if anything
<jonwil>
nemo is probably just using harmattan or meego bits
<freemangordon_>
jonwil: yeah
<freemangordon_>
fortunately there is lots of such bits in meego/harm that are 100% or similar matches to fremantle
<jonwil>
yeah
<jonwil>
I was planning to do more RE of fremantle -voice module next actually
<freemangordon_>
great
<jonwil>
that is, take bits of Meego code and RE them back to match Fremantle
<freemangordon_>
I took the rest I needed after that EAP stuff, so I am back in the business too :)
<jonwil>
biggest job is finishing RE of massive userdata struct
<jonwil>
as you can tell from the earlier paste, its got a LOT of unknowns left in it
<freemangordon_>
yeah
<jonwil>
it doesn't help that a number of the more interesting functions in -voice cause hex-rays to crash with the dreaded "internal error" :(
<jonwil>
the one that suggests you send your db to hexrays
<freemangordon_>
yeah, I know it
sixwheeledbeast has joined #neo900
sixwheeledbeast has left #neo900 [#neo900]
sixwheeledbeast has joined #neo900
<jonwil>
freemangordon: I have attempted to give names to all the functions in -voice, so far I got most of them, the main ones I cant get are ones in eap where the armcc stuff has changed too much between Fremantle and Harmattan
<jonwil>
also some strange stuff where on Fremantle Generic_Init calls various functions with unknown names but in the same place on Harmattan I see things like VCMPE.F64 D0, D8, VMRS APSR_nzcv, FPSCR and VMOV R2, R3, D0
<jonwil>
almost looks like this is somehow connected to the wierd stuff they used on Fremantle to do FP without using FPU instructions
<freemangordon_>
seems like inlined
<jonwil>
yeah some of it might be inlined
<jonwil>
anyhow, figuring out this whole mess is going to be a LOT of work :P
<freemangordon_>
yeah
kolp has joined #neo900
<jonwil>
lets hope jusa_ can throw us bits of info to help
norly has joined #neo900
Kabouik has joined #neo900
sixwheeledbeast has quit [Ping timeout: 250 seconds]
sixwheeledbeast has joined #neo900
sixwheeledbeast has quit [Ping timeout: 244 seconds]
che1 has joined #neo900
norly has quit [Quit: Leaving.]
arcean has joined #neo900
sixwheeledbeast has joined #neo900
sixwheeledbeast has quit [Ping timeout: 256 seconds]
che1 has quit [Ping timeout: 250 seconds]
<Oksana>
For robocat, thank you. Beautiful. Mechanical mice may be less precise, but I am worried about bright 'laser' rays of optical mouse reaching human eyes, and potentially harming them. Infrared is worse: you do not even see it as it blinds your eyes.
sixwheeledbeast has joined #neo900
<jonwil>
I have been using optical mice for years and as long as you dont stare into the mouse you should be perfectly fine
<DocScrutinizer05>
those mice don't use collimated lasers
sixwheeledbeast has quit [Ping timeout: 272 seconds]
<jonwil>
If staring at the laser of an optical laser mouse was dangerous enough to hurt your eyes, it would have the same "warning laser device" sticker as dvd players and stuff
<DocScrutinizer05>
and power level most likely eye safe even for directly looking into the mouse optics from 5mm distance
che1 has joined #neo900
<Oksana>
I know, but still... Dust-gathering mechanical mouse is a pleasant antique, akin to a paper book, when compared against shining optical mouse and bright display. And about harm to eyes: imagine a kid staring into the bright light, holding it as close to the eye as possible, for a long time , fascinated with the bright light and its optics.
sixwheeledbeast has joined #neo900
<bencoh>
Oksana: imagine a kid staring at the sun at midday ;)
<DocScrutinizer05>
why would anybody stare into a apparently completely black optical window of a mouse?
<DocScrutinizer05>
from 5mm distance
<bencoh>
DocScrutinizer05: because "shiiiinyyyyy" :p
<DocScrutinizer05>
I don't even see any Laser Class warning on my Logitech laser mouse. I doubt they'd dare to sell something in Swiss or Germany without proper labeling
<DocScrutinizer05>
bencoh: there's nothing shiny in a infrared laser
<bencoh>
oh, infrared, true
<bencoh>
(but on the internet, nobody knows you're a dog!)
<DocScrutinizer05>
http://de.wikipedia.org/wiki/Lasermaus >>... Es wird auch kein konzentrierter Lichtstrahl verwendet...<< >>Der verwendete Laser ist so schwach, dass er keine Gefahr für das menschliche Auge darstellt<<
<DocScrutinizer05>
interesting background: http://de.wikipedia.org/wiki/Speckle and the RF-analogon glitter. You can see glitter effect when your RF signal amplitude is fading in and out by up to a few dozen dB when moving your phone only 50cm or one meter
<DocScrutinizer05>
glitter is annoying since it's not sttaic, every change in "speed of light" in air will change the glitter pattern. This can already happen by air warming up a few fractions of a centigrade
<DocScrutinizer05>
static*
Kabouik has quit [Ping timeout: 265 seconds]
<DocScrutinizer05>
means that this one sweet spot on your desk where the phone has decent signal will move rather quickly to somewhere else
che1 has quit [Ping timeout: 258 seconds]
Kabouik has joined #neo900
sixwheeledbeast has quit [Ping timeout: 272 seconds]
sixwheeledbeast has joined #neo900
sixwheeledbeast has quit [Ping timeout: 245 seconds]
sixwheeledbeast has joined #neo900
Svetlana has quit [Quit: re-connecting]
kolp has quit [Remote host closed the connection]
<jonwil>
Looks like updating the browser engine may be a lot harder than it looks and may in fact be too much work for the few Maemo developers that are left
<freemangordon_>
correct :)
Svetlana has joined #neo900
che1 has joined #neo900
* jonwil
has hit a blank trying to reverse engineer more of -voice
<jonwil>
going to stop that for now and come back later when maybe I can figure more out :)
<freemangordon_>
jonwil: please share what you have done so far
<DocScrutinizer05>
suggestion : focus on fleshing out all cmtspeech stuff and replace it with sth that simply feeds audio to a ALSA HW:1.0 card
<jonwil>
and yes I think for the short term we can pretend parts of aep are "do nothing" functions
<DocScrutinizer05>
cmtspeech is an abomination of a nightmare, doing timing corrections via timing adj messages from ISI
<jonwil>
That said if we plan to run our pulseaudio-nokia clones on N900 as well as Neo900, we will need to RE the CMT bits at some point
<jonwil>
RE using the meego stuff as a base that is
<DocScrutinizer05>
yep
<DocScrutinizer05>
just saying that for Neo900 cmtspeech is obsolete
<DocScrutinizer05>
we didn't test PxS8 modem PCM/IIS audio interface in detail yet, but according to TRM it's capable of doing master as well as slave, and in both modes there's no further restrictions on the amount of syncing done, which meant the modem kind of "resamples" internally and you don't need to worry about keeping correct pace in sync with GSM 20ms timeslots, unlike BB% N900 modem which is using GSM timeslots as "master clock" for audio
<DocScrutinizer05>
BB5*
<DocScrutinizer05>
PCM with a clock either provided by modem, or provided by CPU and ~8kHz will do
<DocScrutinizer05>
plain vanilla ALSA soundcard
<DocScrutinizer05>
just saying that when you don't RE anything except this particular detail to call ALSA snd_pcm_*() instead of cmtspeech statically linked functions, possibly by patching the binary blob, we already had solved the most critical part of audio problem in Neo900 voice calls
<jonwil>
patching out cmtspeech in the binary isn't as simple as you imply but yes in theory we could replace the various calls to the cmt libraries with calls to something else if we wanted to
jonwil has quit [Remote host closed the connection]
mvaenskae has joined #neo900
<DocScrutinizer51>
citydruck says 'building such lightspreader is absolutely possible, though the machine will run a while. precision they plan for is <0.1mm
sixwheeledbeast has quit [Ping timeout: 245 seconds]
sixwheeledbeast has joined #neo900
arcean has quit [Ping timeout: 250 seconds]
sixwheeledbeast has quit [Ping timeout: 240 seconds]
<DocScrutinizer51>
Oksana: would you want to create a outline drawing pdf from domesheet-bottom?
xes has joined #neo900
sixwheeledbeast has joined #neo900
<DocScrutinizer51>
a pity snaptron doesn't offer RH15xxx oblong domes
<DocScrutinizer51>
RK
<DocScrutinizer51>
so we will go for P15150
<DocScrutinizer51>
on RK15xxx/P15150 domes + Flexcon PM200 adhesive, spacer 0.1mm
<DocScrutinizer51>
Y axis venting between 3 domees
sixwheeledbeast has quit [Ping timeout: 244 seconds]
sixwheeledbeast has joined #neo900
<wpwrak>
ah yes, 3 in a single line should work nicely
<wpwrak>
did snaptron comment on the dome size ? i would have hoped that they may have something nicer and smaller up their sleeve ...
<DocScrutinizer51>
not asked them yet
sixwheeledbeast has quit [Ping timeout: 245 seconds]
<wpwrak>
ah, so there's still hope :)
<DocScrutinizer51>
I'll ask them for a PeelnPlug dome array with exact shape we need, and RK15xxx or P15150
<DocScrutinizer51>
as said above
<DocScrutinizer51>
on RK15xxx/P15150 domes + Flexcon PM200 adhesive, spacer 0.1mm
<DocScrutinizer51>
then add our lightspreader courtesy Citydruck on top
<DocScrutinizer51>
yet to be found: extremely thin silver coated adhesive
mvaenskae has quit [Quit: Lost terminal]
<DocScrutinizer51>
rescue blankets come to mind
<DocScrutinizer51>
rhough I'd be surprised if 3M doesn't have some sticky tape of exactly the type we need
sixwheeledbeast has quit [Ping timeout: 260 seconds]
<Pali>
pulseaudio is 10 years old!
sixwheeledbeast has joined #neo900
Kabouik has joined #neo900
<DocScrutinizer05>
well... :-x
<DocScrutinizer05>
condolences?
<DocScrutinizer05>
tell it about facts of life?
<Pali>
what is needed to 1) tell developers to develop applications which depends on pulseaudio and 2) tell users to not install and use pulseaudio?
<Pali>
what are fundamental features which are present in pulseaudio and *not* in alsa which cause that pulseaudio is preinstalled and used more?
<DocScrutinizer05>
switch audio routing (aka stack) during playback/record
<Pali>
I know about these features *not* supported by alsa: 1) bluetooth (bluez5) support 2) per application volume settings 3) switching sound card input/ouput without application interrupt/support
<Pali>
it is all? or something I missed?
<DocScrutinizer05>
2) incorrect
<DocScrutinizer05>
1) and 3) is it
<Pali>
2) is really possible?
<Pali>
can I configure (in asoundrc) that application cannot change global mixer level, but set it locally only?
<Pali>
so we just need 1) and 3) and alsa win over PA!
<DocScrutinizer05>
it's damn simple, just using pcm.vol {} definition in (e.g.) .asoundrc
<DocScrutinizer05>
yes
<Pali>
and we need 4)
<DocScrutinizer05>
where I consider 3) a regression
<Pali>
and more important, need some "perfect" asoundrc or something which will autogenerate it
<DocScrutinizer05>
I think a correct system archiotecture would NOT switch to a new audio sink/source while app doesn't even notice
<Pali>
automatically not if user does not configure/enable it
<DocScrutinizer05>
anyway ALSA is probably easily patched to do that switching trick
<Pali>
but my idea is to have policy file where will be written order of cards priority and option to enable/disable switch based on card availablility
<Pali>
do you know how to easily patch alsa to support that switching?
<DocScrutinizer05>
actually when you _know_ the cards to switch between, ALSA can (and always could) do it
<Pali>
also if sound card disappear (e.g it is usb card which was unplugged or bluetooth headset which was disconnected)
<DocScrutinizer05>
that's a tad cumbersome
<DocScrutinizer05>
DISappear is no problem, but a NEW card is a problem
<DocScrutinizer05>
since your audio stack needs to get set up in a way that needs knowledge of all possible alternative cards to serve
<DocScrutinizer05>
adding a new card to an existing stack is a problem
<DocScrutinizer05>
but then, when do you ever do *that*?
<Pali>
when I turn on my bluetooth headset
<DocScrutinizer05>
even a USB card you plug in will most probably have been plugged before
<DocScrutinizer05>
no, when you turn on your BT headset, that's not a NEW card
<DocScrutinizer05>
it's already known to system from time when you paired the headset
<Pali>
mplayer is already running (audio+video) on speakers and I need to switch it to bluetooth headset without restarting mplayer (live streaming)
<ds2>
why not have the app goto a virtual mixer device
<ds2>
and have things routed underneath there?
<ds2>
IIRC, the OMAP4 ABE code does something along those lines
<Pali>
ds2: what do you mean?
<Pali>
with virtual mixer device?
<ds2>
app sends to FOO audio sink
<ds2>
FOO audio sink is really a virtual device that can switch between different real sinks
<ds2>
app never sees it
<DocScrutinizer05>
you simply set up an arbitrary number of parallel cards, each one with its own volume, and the app distributes/shares audio stream to all of them, even when some are not even plugged ATM
<ds2>
you can expose the ccotnrols for which one through the mixer
<Pali>
rigth, I have idea to implement something like this
<DocScrutinizer05>
obviously you canot do that for a card you never seen before
<Pali>
if that virtual mixer plugin can handle new card appeared (e.g via udev or other) then it can register it
<DocScrutinizer05>
*sigh*
<Pali>
and probably also switch to it, not?
<DocScrutinizer05>
why?
<DocScrutinizer05>
what for?
<DocScrutinizer05>
will you buy a new BT headset and pair it to your phone, **while doing a phonecall**?
<DocScrutinizer05>
I think that's a bit far fetched requirement specs
Oxyd76 has quit [Remote host closed the connection]
<DocScrutinizer05>
generally the set of possible destinations is known at stack init time
Oxyd76 has joined #neo900
<wpwrak>
sure, all the time :) "i can't hear you. hang on ! i'll get a new headset." -- upwardly mobile yuppie-nerd barges into the next gadget shop and gets a new critter
<Pali>
I think it is silly to restart app if I connected totally new device (sound card/bluetooth card)
<DocScrutinizer05>
why restart? the app only needs to close and reopen the audio
<Pali>
and if I configure (somehow) that I want to switch audio output to *any* connected bluetooth headset it should connect to it... even if is totally new
<DocScrutinizer05>
and that's actually what an app should do to switch audio device anyway
<Pali>
and if application does not support reopen audio?
<Pali>
(e.g aplay)
<DocScrutinizer05>
WUT?
<DocScrutinizer05>
and app that doesn't ever close audio?
<DocScrutinizer05>
an*
<Pali>
I need to restart aplay application if I want to switch to new card, or not?
<DocScrutinizer05>
yeah, exactly :-P "wait, lemme try connecting that 128,99USD BT one to the aplay that's just playing back that 10 hours of interrogation protocol"
<DocScrutinizer05>
aplay hardly counds as "application", it's a cmdline binary
<DocScrutinizer05>
counts*
<Pali>
it is easy cmdline binary which play sound from stdin... this is easy to use in shell scripts
<DocScrutinizer05>
so what?
<Pali>
so is there a way how too switch output from aplay to new connected card?
<DocScrutinizer05>
it's outright silly to demand for on-the-fly switching of audio card for an established audio stack
<Pali>
but it is needed
<DocScrutinizer05>
and no, you can only switch to those audio ALSA cards for which you loaded drivers
<DocScrutinizer05>
AHA
<DocScrutinizer05>
when? why? what for?
<DocScrutinizer05>
tell me about a use case!
<Pali>
aplay plays some sound... and I want to listen it. but I do not want to disturb everybody near me so I switch playback to (bluetooth) headset
<DocScrutinizer05>
even yur standard mp3 player shall close audio device when hitting "pause" (NOT "stop"), and re-initialize the audio stack - possibly with new card - when you unpause
<DocScrutinizer05>
again! *sigh* you won't pair a new device you just bought
<DocScrutinizer05>
you may switch to the BT device you got yeaterday and used before
<DocScrutinizer05>
since such BT device is *known* by your system and thus ALSA could load a driver for it, even when it's not switched on
<Pali>
ok and what for applications without pause (e.g live stream playing where pause is not possible)?
<DocScrutinizer05>
and for BT it's yet another story, since all BT looks same for ALSA
<Pali>
for BT we need new plugin (that old not working anymore)
<Pali>
but once I see fixed all other alsa problems I will write that BT plugin (dbus API)
<DocScrutinizer05>
ok, and you think you want to run into a shop and buy a new device and pair it to your system **while hearing sound** on such non-pausable app?
<DocScrutinizer05>
that's no valid use case
<DocScrutinizer05>
I even doubt PA can do that
<Pali>
no I need to fix scenario when I'm playing non-pausable audio and my card disappear from system
<DocScrutinizer05>
so WAHT? I told you 15 times now this is no issue with ALSA, it been able to handle that since ages
<Pali>
ok, ok
<Pali>
how can I set cards priotity?
<Pali>
e.g. when I have 3 cards (some are connected, some not) and once card with higer priority appear I want to switch to it
<DocScrutinizer05>
ALSA wil _not_ allow you to plug in a JAKUHASHU USB mixer and download the JAKUHASHU kernel soundcard driver and modprobe it and switch to that thing **while not stopping the audio**. And *only* this is what you cannot do
<DocScrutinizer05>
well, how about udev?
<Pali>
modprobe is done by udev (or something else)
<Pali>
so once card is inserted after few seconds appear in system
<Pali>
I just want to switch audio output once that card is ready in system
<ds2>
how does it disappear?
<DocScrutinizer05>
well, modprobe is not what we want to do here, since the driver already needs to be loaded or the audio stack cannot use it as one of several alternative concurrent audio sinks
<Pali>
ds2: plug/unplug usb sound card or turn on/off bluetooth headset
<DocScrutinizer05>
and per default ALSA will output sound to all concurrent audio cards in parallel
<DocScrutinizer05>
even to those which are not phyically powered up, nevertheless their driver must be loaded and must be resilient to physical card not present
wazrus has joined #neo900
<DocScrutinizer05>
if you want to "switch", you simply mute all cards "not in use", and you unmute a "new" card and mute the "old" one when you switch
<DocScrutinizer05>
Plugin: Multiple streams to One
<DocScrutinizer05>
first you create one 8 channel virtual PCM out of 4 stereo PCM, then you play mono sound to that 8 channel virtual PCM and define a conversion that equally distributes 100% of original mono audio data to each of the 8 slaves
wazrus has quit [Quit: Lost terminal]
<DocScrutinizer05>
or you actually write a plugin that opens its own downstream audio stack (acts like an "app") and handles the whole stuff completely hidden from the real app upstream
<DocScrutinizer05>
file plugin would be a relatively easy startpoint for that
<DocScrutinizer05>
did I already mention that ALSA has *abysmal* documentation? ;-)
<DocScrutinizer05>
Pali: on my alsa volume patch: twnkle uses a key in config file like "dev_ringtone=alsa:default" which means "use ALSA sound, open ALSA 'device' ``default´´". ``default´´ is user configurable string, name of ALSA device to open and use. My "patch" replaces this by device ``vol:TwinkleRinger,"default" ´´
<DocScrutinizer05>
in .asoundrc I define what's ``vol´´ plugin
<Oksana>
But since I do not know how to make a PDF, it will be done in SVG.
<DocScrutinizer05>
so a new volume slider appears in whatever_your_mixer_app with name of slider (args.0) = TwinkleRinger and salve of the vol plugin (args.1) = whatever been used before (usually ``default´´)
<Oksana>
Is it acceptable?
<DocScrutinizer05>
Oksana: yes, the task still pending. Alas it seems SVG is not one of the formats snaptron likes for file format