Log from #neuros IRC channel

nerochiaro: if you were here, JohnDesigner and I would be kicked back in our chairs at the Neuros lounge dreaming of better design methods and models.

[01:41]

gremlin[it]: congratulations :)

[01:41]

gremlin[it]: just in time for the demo

[01:41]

<gremlin[it]>

sigle applications (imageplayer, xmms, player) work but desktop application raise a 'Illegal Istruction' exception/signal

[01:42]

crweb: agreed

[01:42]

<crweb>

the real question is

[01:42]

<crweb>

would the Neuros lounge allow beer during work hours...

[01:42]

crweb: if you are not a wimp and can handle it, why not ?

[01:43]

<crweb>

count me out then

[01:43]

<gremlin[it]>

yes just in time ... ig good enought to show as a WIP project ... and the multimedia features are working

[01:43]

crweb: there is beer in the office fridge

[01:43]

<gremlin[it]>

now i have to find a TV to attach on neuros :)

[01:43]

<gremlin[it]>

hi crweb !

[01:43]

<crweb>

gremlin[it]: thats always the hardest part

[01:43]

crweb: nothing wrong with that

[01:43]

gremlin[it]: that SIGILL is evil, i wouldn't know where it comes from

[01:44]

gremlin[it]: maybe mixing compilers ?

[01:44]

* crweb is an alcohol panzy

[01:45]

<gremlin[it]>

maybe i know ... if i use your libz dropbear don't work cause a missin __imul3 (something like) simbol ...

[01:45]

<crweb>

probably has something to do with not even tasting alcohol beaverage until 26

[01:46]

wow

[01:46]

gremlin[it]: really smells like different compiler issues, to me

[01:47]

crweb: puritan family or something ?

[01:47]

<crweb>

nerochiaro: i was busy

[01:47]

<gremlin[it]>

so i use OE libz ... so probably some other lib have trouble ... maybe related with softfloat or about ...

[01:47]

<crweb>

JohnDesigner: you could.. use svg?

[01:48]

crweb: I can export .svg from illustrator. Can that be imbued somehow into QT drawn version?

[01:48]

<crweb>

JohnDesigner: svg is a vector image. it would scale perfectly

[01:48]

* JohnDesigner is suddenly hopeful.

[01:48]

<crweb>

JohnDesigner: we can try it. Dunno how heavy svg's are on the system

[01:48]

gremlin[it]: good guess. it's probably the different configuration of libc ?

[01:49]

<crweb>

nerochiaro: never understood drinking. you know, busy with pursuit of knowledge, fine arts (music), and addicted to online gaming and programming

[01:49]

crweb: i vote for giving it a shot

[01:49]

<crweb>

plus terrified of my adictive personality

[01:49]

crweb: but couldn't QT use the SVG as "instruction" to make its own lite version?

[01:49]

crweb: that is the weirdest thing i've ever heard related to drinking. kudos to you, really

[01:50]

<crweb>

JohnDesigner: thats kind of what happens. the slow part is the reading of the file and conversion

[01:50]

<gremlin[it]>

nerochiaro, but i'm using your compiler ... maybe different libc :)

[01:50]

<gremlin[it]>

nerochiaro, ,ah yes you already answer ;)

[01:50]

gremlin[it]: i'm a step ahead :P

[01:50]

crweb: but that conversion only happens *one time* when converting the svg onto code or

[01:50]

<crweb>

nerochiaro: i still don't understand. i only wine taste and have a beer if a bunch of friends are over

[01:50]

<crweb>

JohnDesigner: correct

[01:51]

crweb: so that is done outside of the OSD (OSD plays no part in this svg)

[01:51]

<crweb>

no, qt loads the file and converts it to the "data that is drawn on screen"

[01:52]

<crweb>

which is a high speed/light weight Pixmap algorithm

[01:52]

crweb: nothing to be understand. you either like it or not. it's just a matter of taste like everything else.

[01:52]

crweb: want to give it a shot with the list cursor as an experiment?

[01:52]

<crweb>

svg might be handled differently though

[01:52]

<crweb>

nerochiaro: taste wise, only semi-sweet apple/peach wines.

[01:52]

<crweb>

and, Captian morgan in coke

[01:52]

<gremlin[it]>

nerochiaro, i'm testing a new PM hardware ... i'm in heavy multitasking :D

[01:53]

* JohnDesigner is covering his ears during crweb's stated drinking tastes

[01:53]

<crweb>

JohnDesigner: samual adams october fest?

[01:53]

crweb: i'm more a wine person than a beer person, and drink very little strong alchool drinks

[01:53]

<crweb>

is the onlly beer i can even swallow

[01:53]

crweb: er, actually scratch that, more beer than wine

[01:53]

sorry

[01:53]

<gremlin[it]>

crweb, peach wine ... bellini !

[01:53]

<crweb>

and it only comes out in october..

[01:54]

gremlin[it]: he should try lambrusco, i guess he would love it

[01:54]

not that anyone needs my approval for their drink of choice.

[01:54]

when it comes to booze I only have one rule, You can't drink all day unless you start in the morning.

[01:54]

crweb: you should try guinness sometimes

[01:54]

<gremlin[it]>

nerochiaro, mhhh don't know ... i'm close to prosezzo zone (valdobiadene) ... really nice ;)

[01:54]

<crweb>

I'm always up to new tastes

[01:55]

<crweb>

nerochiaro: nope. to thick

[01:55]

gremlin[it]: :) not bad, not bad

[01:55]

<gremlin[it]>

uff prosezzo -> prosecco

[01:55]

crweb: black and tan then?

[01:55]

<crweb>

JohnDesigner: only the tan part

[01:55]

crweb: red beers then ? too bitter ?

[01:55]

<crweb>

nerochiaro: yep to bitter

[01:55]

<crweb>

oatmeal beers are ok

[01:56]

oh well, if it wasn't 9AM i would go having a bottle ;)

[01:56]

<crweb>

pale ale's i like

[01:56]

<gremlin[it]>

i'm in the east part of north-east (close to Venice) ... heavy wine production and consume zone :P

[01:56]

nerochiaro: how 'bout an irish car bomb (shot of irish whiskey dropped in a guiness)

[01:56]

<crweb>

i like southern comfort on the rocks

[01:57]

<crweb>

but theres only 1 reason to drink that. and i'm not really into that

[01:57]

JohnDesigner: mixing strong alchool and beer is a recipe for disaster to me (and i don't want to explain what kind of disaster)

[01:57]

crweb: okay, so can I give you an svg of the list cursor to test (soon :))

[01:57]

<crweb>

JohnDesigner: you get us some svg's and we can start testing

[01:57]

<crweb>

sounds good

[01:57]

okay

[01:57]

* nerochiaro vote in favor of testing svg

[01:58]

<crweb>

let me do some reading. maybe there is a note aboug svg performanc

[01:58]

but wish there was a way to convert to internal QT instead of parsing the SVG file everytime

[01:58]

<crweb>

nerochiaro: problem with that is svg is vector, its not really an image

[01:58]

nerochiaro: truth be told, I generally stick to Harp if given the choice

[01:58]

nerochiaro: for the reason you stated

[01:58]

<crweb>

nerochiaro: it has to remain free inorder to scale purely

[01:59]

JohnDesigner: that's a nice one. I often have it.

[01:59]

<crweb>

nerochiaro: it could make for an interesting setup though.. if we go pure svg, we can qpixmapcache only svg's

[02:00]

<crweb>

plus icons that are used repeatedly

[02:00]

crweb: free of what ?

[02:00]

<crweb>

nerochiaro: it cant be bound into a pixel format

[02:00]

nerochiaro: I bet a way does exist to convert svg to QT internal version. Seems so anyway since at their heart all of them are code (but I know nothing really)

[02:00]

<crweb>

pixel formats deteriate when scaled

[02:01]

<crweb>

ooo neat

[02:02]

<crweb>

svg's are a whole different system

[02:02]

crweb: as John smartly said, there has to be a way to convert from SVG to QPicture (which is QT's internal "vector" format, basically a compiled library of QT calls, or something)

[02:03]

i know the opposite is possible

[02:03]

<crweb>

QImage, or QSvgWidget

[02:03]

no, no, i mean QPicture

[02:03]

look it up

[02:03]

<crweb>

you can only render svg's on QImage, or QWidget

[02:04]

what i'm saying is, it would make a lot of sense to be able to convert between SVG and QPicture

[02:04]

<crweb>

and i don't see any way to do that

[02:04]

<crweb>

svg's are rendered completely by the SvgRenderer, which can render on QWidget or QImage.

[02:05]

<crweb>

sweet, there are mobile svg profiles

[02:05]

<crweb>

www.w3.org

[02:06]

<crweb>

Embedded Systems

[02:08]

oh well, you're probably right. i was just saying it would've been nice and better optimized than parsing SVG

[02:09]

crweb: look at the rpath patch reply... it's for you

[02:09]

<crweb>

qtopia-core uses SVG 1.2 Tiny

[02:09]

<crweb>

nerochiaro: yeah, we are diag'ing already

[02:09]

ok, keep me posted

[02:09]

<crweb>

i think we have a version glitch

[02:09]

<crweb>

they must have a ld.so.cache that was generated somehow before the mount fix.

[02:10]

<crweb>

I'm getting ready to flash my osd to just do it first person

[02:10]

ok, update that thread, just for the record

[02:20]

<mp0>

you know that white noisy bar on the top of OSD video captures?

[02:21]

mp0: think i do

[02:21]

mp0: but it doesn't seem to be there always

[02:21]

<mp0>

I overlapped a few frames: paul.bluereboot.net

[02:21]

A question: is the osd bootable from SD or from usb disk?

[02:22]

matters: not out of the box

[02:23]

mp0: artsy :) should that tell us anything ?

[02:23]

sigh, that would have been a perfect way to do cheap testing

[02:23]

<mp0>

I'm curious if it's a signal.

[02:23]

<mp0>

There's some very repetitive elements

[02:23]

matters: NFS+TFTP is the cheap way

[02:24]

mp0: i would guess it's noise. maybe we are setting up the memory wrong and recording one row that is out of the actual video data (just a guess)

[02:25]

nerochiaro: it's a nice way but needs pc turned on and so I can't boot OSD as a little webserver that way (dual booting...)

[02:26]

<mp0>

I'm hoping you captured part of the VBI.

[02:26]

<crweb>

mp0: yes, i know of this

[02:26]

matters: do you need a different kernel to make it act as webserver ? i would guess no, right ?

[02:26]

<crweb>

mp0: i have complained about it before. but new firmware made it vanish. then newfirmware it came back, then new firmware made it vanish

[02:27]

<crweb>

all from the same input source. so could have been the show/transmission..

[02:28]

<mp0>

crweb, complain? I'm not complaining about it... it could be useful.

[02:28]

mp0: it can be that it's recording some extra info that is transmitted inside the TV signal, as you say

[02:28]

<crweb>

mp0: its my job to complain about it ;)

[02:28]

mp0: but to actually use it, you need a way to reproduce it consistently. and if you can do that, please report a bug :)

[02:28]

<crweb>

then they tell me i'm crazy

[02:29]

<mp0>

heh, any idea how the de-interlacing filter would affect the TV data before it gets encoded?

[02:30]

nope

[02:33]

<mp0>

maybe it's macrovision

[02:33]

<crweb>

mp0: i was getting it on direct cable.

[02:33]

<crweb>

mp0: you have to have a macrovision scrambler chip correct?

[02:34]

mp0: i got it even from a cheap-o dvd player that has no idea what macrovision is, probably

[02:34]

<crweb>

the actual signal is clear. it is only scrambled if you have a chip

[02:34]

<crweb>

the scrambling happens on the input

[02:34]

<mp0>

okay, I'll rule that out

[02:35]

<crweb>

its totally retarded. they should eliminate it

[02:35]

<crweb>

input is only scrambled if your device has a macrovision chip. how f'd up is that

[02:36]

crweb: no less retarded than stuff like DVD CSS and such

[02:36]

crweb: i mean DVD CSS is only holding because the law protect it, otherwise it would be crackable even by my cats

[02:37]

with all due respect to my cats

[02:37]

<mp0>

VCRs are required to "respect" macrovision due to a specific provision in copyright law :-/

[02:37]

mp0: exactly what i'm saying

[02:37]

<crweb>

USB-based video interfaces designed to allow DVD recording on PCs are legally required to detect the presence of Macrovision signals on any analog signals input to them, and if so, inhibit the recording, but few, if any, do so.

[02:38]

crweb: those who don't are technically illegal

[02:38]

<mp0>

I thought it was only required of VCRs?

[02:38]

* mp0 looks up the law

[02:38]

<crweb>

that was from wikipedia

[02:39]

<crweb>

you can build a macrovision decoder fairly easy

[02:39]

<crweb>

by hand

[02:39]

<crweb>

"video stabilizers"

[02:39]

<crweb>

en.wikipedia.org

[02:45]

<mp0>

I don't think wikipedia is correct about USB-based devices... section 1201(k) of the copyright act is specific to devices that record an analog signal on magnetic tape.

[02:46]

<crweb>

mp0: there was a new law in 2007 that required some sort of protection chip

[02:46]

<crweb>

i remember it a big deal cause I bought all my tv cards and equipment before it took effect

[02:46]

<mp0>

ah, the broadcast flag you mean?

[02:46]

<crweb>

oh thats right

[02:46]

<crweb>

nvrmind

[02:46]

<crweb>

couldn't remember what it was about :(

[02:47]

crweb: i know you can bypass macrovision easily, just as you can bypass CSS easily. doesn't mean it's legal to do so, though

[02:47]

<crweb>

nerochiaro: it would be legal to use a signal corrector

[02:47]

<crweb>

since they are just standard devices

[02:48]

well, i'm not a lawyer and i'm not american, but if you use it for the only purpose of bypassing macrovision, then it's illegal

[02:48]

<crweb>

i don't think the law applies to consumer. it applies to vcr as mp0 said

[02:49]

<crweb>

they couldn't limit you from analog copy. so they tried to require copy machines to disable copy

[02:51]

either way, any protection scheme that is trivial and requires a law to protect it, is an aberration

[02:52]

<mp0>

Taking the devil's advocate position, I would argue that the law is the only viable protection... because any technological measure can be broken.

[02:54]

mp0: the key is making it technologically hard to break, so that it is not worth it to do so for most content.

[02:54]

<mp0>

has that ever worked?

[02:55]

<mp0>

At some point, I have to see/view the content. That's a fundamental problem.

[02:57]

mp0: i'm talking about perfect copy of the original content, not passing it via analog and record it again

[02:57]

nerochiaro: Just curious, for the app-startup 'busy' message, why was the existing 'busy banner' not used (the one with the animated neuros rings)?

[02:57]

mp0: i.e. ripping a CD versus copying it from the soundcard output

[02:58]

JohnDesigner: it's heavy as a dead mammoth. i have removed it on other places too and plan to remove it completely from the system actually

[02:58]

JohnDesigner: you may want to design a static replacement for that too, maybe

[02:58]

nerochiaro: while i have never tried to lift a dead mammoth, that sounds pretty heavy ideed.

[02:58]

JohnDesigner: heavy, and with the added floppyness of a dead body

[02:59]

<mp0>

Even if you want a perfect digital copy, we can do things like grab from the LCD control lines. The existance of a decoded being delivered to the adversary is a problem.

[02:59]

<crweb>

JohnDesigner: my explination wasn't good enough for you huh?

[02:59]

nerochiaro: well that is what I am thinking. Design one busy message to be used for all needs.

[02:59]

crweb: you told me to ask nero

[02:59]

<crweb>

lol

[02:59]

<crweb>

looks like you just copy and pasted the question ;)

[02:59]

mp0: i am sure in the end one can even connect something to the retina to capture the video. the point is making it hard enough that it's too expensive to do for most people

[03:00]

* JohnDesigner only does as he is told, regardless of the consequences.

[03:00]

<crweb>

consider youself cut off from this knowledge base

[03:00]

ack

[03:00]

JohnDesigner: if you want to keep the not-so-funny quips, then they have to be two, but they can be based on the same graphical design, sure

[03:00]

<mp0>

night guys, I've got work today.

[03:01]

* JohnDesigner restates his stance that crweb *told* me to ask nero.

[03:01]

mp0: same here. see you next time

[03:01]

<crweb>

JohnDesigner: minor details

[03:01]

night mp0

[03:02]

nerochiaro: those are funny messages I think. I'd like to keep them and fold in the capability to display a couple of text lines into the new busy message

[03:03]

with a different animation. Those rings did *not* work out as planned.

[03:04]

JohnDesigner: anything without an animation, is *a lot* better in a "waiting" message box

[03:04]

since an animation is just stealing time from the thing you are waiting for

[03:04]

thus making you wait longer

[03:04]

but i understand its use

[03:04]

so make it very small and with very few frames, if you must

[03:05]

<crweb>

nerochiaro: i don't understand. from flash, ld.so.cache doesn't generate with /media/ext/* libs

[03:05]

right, but some type of animation, however economical, is worth the cost to indicate that the unit has not frozen.

[03:05]

ya, i get that. blinking dot in a corner :P

[03:06]

nerochiaro: roger that

[03:06]

crweb: i don't either

[03:06]

crweb: why are they any different ?

[03:06]

crweb: maybe not mounted yet when you regen

[03:06]

?

[03:06]

<crweb>

it has to be mounted by then..

[03:06]

<crweb>

otherwise nothing will work..

[03:08]

<crweb>

hda: hda1 hda2

[03:08]

<crweb>

yaffs: dev is 3145730 name is "hda2"

[03:08]

<crweb>

yaffs: dev is 3145730 name is "hda2"

[03:08]

<crweb>

mount: mounting /dev/hda2 on /mnt/tmpfs/mount_CF-card2 failed

[03:08]

<crweb>

/etc/rcS.d/S00_mountall: line 161: /dev/hda2: Permission denied

[03:09]

crweb: what is that ?

[03:09]

<crweb>

the only error during bootup

[03:09]

crweb: if the card has 2 partitions, i guess ?

[03:09]

<crweb>

i guess it does..

[03:09]

<crweb>

i didn't make them

[03:10]

<crweb>

JoeBorn: We need to talk.

[03:10]

* crweb makes growl face at JoeBorn

[03:11]

<gremlin[it]>

hi JoeBorn !

[03:11]

* nerochiaro smell crweb's lust for money

[03:11]

<crweb>

heh, no

[03:11]

just joking, dude

[03:11]

<crweb>

we've been left out of a major loop.

[03:14]

<crweb>

i dunno what hda2 is

[03:14]

<crweb>

the upk created it

[03:15]

<crweb>

ooo

[03:15]

crweb: i think it's all in the docs

[03:15]

<crweb>

my flash failed.

[03:15]

<crweb>

/media/ext/.osd-extended/programs doesn't exist

[03:18]

are flashing with a CF card in ?

[03:18]

<crweb>

i'm being told now the osd won't blank the CF card

[03:18]

are you upgrading from TORFU ?

[03:18]

<crweb>

was

[03:19]

what FS did you have on the card ?

[03:19]

<crweb>

fast32

[03:20]

<crweb>

i used it for newpackage/r3.upk

[03:20]

<crweb>

just like any user could

[03:20]

no, i mean on the CF card

[03:20]

<crweb>

yeah me too

[03:21]

<crweb>

its the same card i use for flashing. fat32 with newpackage/r3.upk

[03:21]

you should not do that

[03:21]

<crweb>

why not? any user can do that

[03:22]

i don't think you're supposed to update FROM the same CF card that you will use to upgrade TO

[03:22]

<crweb>

and probably will, since they do have a CF card in hand..

[03:22]

i mean, that CF is supposed to be stuck and glued into the CF slot

[03:23]

we should give them a tube of superglue

[03:23]

<crweb>

ahh my falt

[03:23]

<crweb>

hidded swap partition...

[03:23]

<crweb>

err hidden

[03:23]

and ?

[03:23]

<crweb>

they say there must only be 1 partition on the card

[03:24]

i am thinking this whole business is madness. CF should not be used for upgrade to AZ.

[03:24]

it should be the target of the AZ upgrade

[03:24]

<crweb>

why not?

[03:24]

<crweb>

the upk is copied from the card before the upgrade

[03:25]

<crweb>

now granted.. if the upk won't fit in ram. that could end up being an issue

[03:25]

<crweb>

but we'd have to have 64mb upk

[03:25]

fair enough, but then when upgrading from TORFU to AZ, the CF card should be formatted completely

[03:25]

<crweb>

i agree there

[03:25]

<crweb>

the upgrade should fdisk and format after the upk is copied

[03:25]

yes

[03:25]

<crweb>

not just format primary partition

[03:26]

<crweb>

which it didn't even do cause i had 2 partitions

[03:26]

<crweb>

this needs to be fixed. not all users will use the small card provided

[03:26]

indeed. how do you do that complete format programmatically, not via fdisk

[03:26]

?

[03:27]

<crweb>

carefully...

[03:27]

<crweb>

heh.. aka.. no idea

[03:27]

<crweb>

i'd blank the table. then create with gpartd

[03:27]

crweb: memzero the block device is careful enough ? :P

[03:27]

<crweb>

then mkfs.ext2

[03:28]

<crweb>

should be. just memzero the partition table

[03:28]

<crweb>

not the whoel device

[03:28]

crweb: right

[03:28]

<crweb>

its like the first 512b?

[03:28]

<crweb>

then.. use a small subset of gpartd to create

[03:29]

<crweb>

you only need 1 creation. no need for solaris support, etc

[03:29]

parted itself

[03:30]

<crweb>

if going to use parted, it can simply just overwrite the table with new one

[03:32]

right

[03:33]

i'm going to send out an email about this, see if terry know how to do it by himself, and what Gao thinks. then we'll see

[03:33]

<crweb>

hunt is also doing so

[03:33]

hmm, let me speak with him

[03:34]

<crweb>

many 3rd party users will want to create swap file, etc so can do more on larger cards

[03:34]

they can do it after the upgrade

[03:35]

<crweb>

i agree, but if can't have more than 1 partition... whats that about?

[03:35]

<crweb>

whats it matter if there is a hda2?

[03:35]

crweb: Hunt is opening a chat, we'll talk there

[03:35]

with him too

[03:35]

otherwise it's wasted breath

[03:35]

so to speak

[03:35]

<crweb>

ok

[03:38]

<crweb>

neat. i have chinese fonts installed

[04:00]

<gremlin[it]>

mhhh a question (i don't remember) .. the Ethernet of OSD is auto sensing cross or patch cables ?

[04:00]

<crweb>

gremlin[it]: yes

[04:01]

<gremlin[it]>

so a patch cable is ok also to do a point-to-point right ? (double check)

[04:02]

yes, it's auto-cross

[04:04]

<gremlin[it]>

ok ... :)

[04:06]

nerochiaro: crweb : do you guys have a Recorder 2 in your possesion?

[04:07]

no

[04:07]

<crweb>

nope

[04:07]

* crweb isn't that cool

[04:07]

from time to time I am asked by the bus folk about such and such "R2" functions

[04:07]

bus folk ?

[04:07]

<crweb>

you talk to bums?

[04:08]

<crweb>

what are they using r2's for on buses?

[04:08]

such as I guess the R2 made thumbs from videos in some manner

[04:08]

<crweb>

i'd think they'd rather homes..

[04:08]

business folk

[04:08]

athough you had some good guesses.

[04:08]

<crweb>

that last statement may still apply

[04:09]

<crweb>

it would be possible to grab thumbnails of video. we'd need arm side codecs though

[04:09]

<crweb>

it wouldn't be to slow, since you can just grab a frame..

[04:10]

I'm sure there is some sort of technical gotcha (to use nero's great description) about why the R2 or other non OSS device could do X.

[04:10]

JohnDesigner: they have hardware support for such things

[04:11]