Log from #neuros IRC channel

statfs64("/mnt/removable", 84, {f_type=0x9fa1, f_bsize=16384, f_blocks=1220339, f_bfree=1220223, f_bavail=1220223, f_files=0, f_ffree=0, f_fsid={0, 0}, f_namelen=260}) = 0

[02:05]

mbryner: now that's damn weird.... your df is getting a good answer, then throws it out

[02:05]

hope that makes sense to you derobert :)

[02:05]

sorune: well, sounds like sorune should use statfs on unix

[02:06]

what os?

[02:06]

linux mandrake 10

[02:07]

weird, i use to run that with no problems

[02:07]

:(

[02:07]

sorune: probably some weirdness with automount

[02:09]

figured it out

[02:09]

df is aliased to `df -h -x supermount'

ah, that'd suck

that would explain it :)

[02:10]

mbryner: does /bin/df -a get around the alias?

[02:10]

running # 'df' gives:

[02:10]

none 19525424 1856 19523568 1% /mnt/removable

[02:11]

i didn't set it up that way though!!!!!!!!

[02:11]

it's a mandrake thing.

[02:11]

i wonder why i didn't have that alias...

[02:11]

are you running official

[02:11]

yup

[02:12]

going to take out that alias now.

[02:12]

sorune: use the 4-argument open; I think that will avoid the shell

[02:12]

ah, i was running the preview release (whatever they called it)

[02:13]

derobert: thanks, i'll try that

[02:13]

<DeepB>

what's a good partition table consistency checker?

[02:13]

thank you to both, combined heads do wonders.

[02:14]

sorune: if that doesn't work, there is always the open yourself, then use 3-arg exec

[02:15]

sorune: ok, perlipc page says the 4-argument open doesn't spawn the shell (under "Safe Pipe Opens") so that should work.

[02:18]

here's a link to why they set up the alias: lists.ethernal.org

[02:19]

that's not good...

[02:21]

derobert: do you know where they might have put the alias? it's not in /etc/bashrc, /etc/profile, or user versions of those.

[02:21]

mbryner: grep -r on /etc is useful for situations like that :-(

[02:23]

sorune: there is also a Filesys::Statvfs module which you might find easier.

[02:23]

sorune: no idea why it isn't part of POSIX module

[02:25]

wow, a zombied df due to supermount... that's just bad

[02:27]

ok, i'll search around for something that works better (thanks for the tips)

[02:28]

<DeepB>

so, no idea how can i correct the partition table integrity?

[02:28]

sorune: Filesys::Statvfs will work nicely, except that its yet another module people would need to have installed. Though on debian testing/unstable, its apt-getable.

[02:28]

DeepB: partition table integrity?

[02:29]

<DeepB>

that's weird cfdisk shows no errors, fdisk complains about head limit, parted shows weird non-existing partitions, pqmagic won't run...

[02:29]

deepb: seeing an error from fdisk?

[02:29]

off topic now, but i found a cool trick: '/etc/profile.d/alias.sh' has all aliases systemwide (at least for mandrake)

[02:30]

<DeepB>

the drive works OK, but i just want to resize some, and i can't, cos it is inconsistent

[02:31]

using dos fdisk?

[02:31]

<DeepB>

the linux one

[02:32]

<DeepB>

the 'v' options spits something like this

[02:32]

<DeepB>

Partition 1: head 255 greater than maximum 16

[02:32]

<DeepB>

Partition 3 does not end on cylinder boundary.

[02:32]

back in a few

[02:32]

<DeepB>

probably the BIOS and the kernel don't agree with the head limit

[02:33]

DeepB: hey, could be worse, you could of seen the fun I had with those tools on a >2TiB disk

[02:33]

<DeepB>

i pity you, dude

[02:33]

<DeepB>

that could be insane, especially with valuable data on it

[02:34]

DeepB: thankfully, it was blank, considering what they did....

[02:34]

<DeepB>

i'm just installing so i can wipe it all and start from scratch... all i would lose is time

[02:35]

DeepB: how did it get that messed up?

[02:35]

<DeepB>

wow.. that's a long story

[02:36]

<DeepB>

to make it short...

[02:36]

<DeepB>

this notebook came with a full sized NTFS XP partition, i resized it with qtparted, and make room for a XFS one

[02:37]

<DeepB>

then installed Ubuntu, then wiped the NTFS one

[02:38]

well, honestly, if its already a C.F. and you can just wipe it all and start over, I would...

[02:38]

<DeepB>

instaled W2k, on a new primary one, and make 2 logical units on a extended partition

[02:39]

damn, how many OS's are you going for?

[02:39]

<DeepB>

so far, qtparted tells me at this point that i have seven (7) partitions with holes in the middle

[02:40]

anyone wanna help me with mounting my neuros on linux, i always get this from dmesg after i plug it in

[02:40]

usb.c: USB device 4 (vend/prod 0x451/0x5416) is not claimed by any active driver.

[02:40]

<DeepB>

just two, but i refuse to make a 60Gb FAT32 partition and i wanted to have some room to comunicate between OSes

[02:40]

noiz77: do you have usb-storage module loaded?

[02:41]

derobert: yeah

[02:41]

shows up in lsmod

[02:42]

i loaded usb-storage, sd_mod, and scsi_mod before i plugged it in

[02:42]

hmm....

[02:45]

anything you can think of?

[02:45]

I seem to remember a long time ago the N2's backpack gave a weird vendor ID....

[02:45]

what kernel version is this?

[02:46]

"For example, the recorder uses the TI DSC25 processor. This is a dual core processor with an Arm7 and a 54xDSP"

[02:46]

DSC25 == DSC255?

[02:46]

or is it something completely different?

[02:46]

(I'm not sure about the weird ID thing... I don't even have a N2, only a N1. No USB2 ports around here...)

[02:46]

<Yono>

hi

[02:46]

2.4.27

[02:48]

* derobert goes to check the forums, sees if he can find it...

[02:49]

* noiz77 is too lazy to search threw tons of topics looking for answers...

[02:49]

* noiz77 *through

[02:50]

<iswm>

noiz77, ehci something or other?

[02:50]

no idea, im pretty new to linux

[02:50]

<iswm>

Get this when I mount mine: usb 1-4: new high speed USB device using ehci_hcd and address 2

[02:51]

hub.c: new USB device 00:1d.7-3, assigned address 4

[02:51]

usb.c: USB device 4 (vend/prod 0x451/0x5416) is not claimed by any active driver.

[02:51]

thats all i get

[02:51]

<DeepB>

any reason why a boot record can't be throwed in into a partition holding an XFS filesystem? is that just stupid or am i missing something?

[02:52]

<iswm>

noiz77, do you have EHCI HCD enabled in your kernel?

[02:52]

DeepB: XFS probably wasn't designed for it, coming from SGI systems

[02:53]

looks like its on

[02:53]

but says its unused

[02:53]

<iswm>

It's in device drivers -> usb support (I use 2.6.10 though)

[02:53]

<iswm>

Hmm

[02:53]

usbcore 52268 1 [usb-ohci usb-storage hid ehci-hcd usb-uhci]

[02:54]

noiz77: Ah, found it in the forums: www.neurosaudio.com

[02:54]

noiz77: read the last few messages. You can work around it in the kernel, but its apparently a bug in certain N2s

[02:54]

k

[02:55]

noiz77: same thing here www.neurosaudio.com

[02:55]

(once again, last few messages)

[02:56]

k

[02:57]

<DeepB>

i love this kind of support messages...

[02:57]

<DeepB>

Fatal: Filesystem would be destroyed by LILO boot sector: /dev/hda3

[02:57]

DeepB: well, good thing it told you.

[02:57]

<DeepB>

thank you LILO, i love you too

[02:57]

lol

[03:00]

hmmm, so it looks like i'd have to recompile the kernel, or send my unit in since the chip inside is messed up...

[03:03]

<DeepB>

ok, advice to all of you out there, use XFS (if you would) for everything but / , i'm paying my n00bness here

[03:03]

noiz77: yep, or get kernel.org or your distro to include that patch in their kernel by default

[03:07]

DeepB: why not just use ext3 for a small /, then have a much bigger XFS partition?

[03:08]

<DeepB>

mbryner: that's why i wanted to resize ;)

[03:09]

:) oh, sorry, missed that

[03:09]

<DeepB>

but to make room for a smaller /boot one, but i can't (read my whinings above)

[03:10]

pardon, wasn't on for a while. that's what i get for coming into the middle of a conversation.

[03:12]

<DeepB>

seems like the whole universe (but my obstinacy) want to make me start this whole, boring set-up process from scratch...

[03:12]

<DeepB>

i still kindly refuse, though

[03:13]

DeepB: well, can you put the boot sector on on the disk, instead of a partition? e.g., /dev/hda instead of /dev/hda3 ?

[03:14]

<DeepB>

yep, that's where i ended, but i loved my former XOSL :(

[03:15]

<DeepB>

universe, send me a fucking virus or something, so i have a real reason, damn it

[03:15]

<DeepB>

:)

[03:15]

* derobert coughs on DeepB

[03:16]

* DeepB 's putah fails to catch the viruses as they decay on a several thousand miles trip

[03:17]

damn

[03:17]

DeepB: where do you live?

[03:17]

<DeepB>

but it was a nice try, i have to say

[03:17]

* [Chameleon] is back (gone 17:22:09)

[03:17]

<DeepB>

i'm Spanish

[03:17]

<Yono>

finally cham

[03:17]

<Yono>

I was typing up an email :-P

[03:17]

DeepB: damn, yeah, that'd be hard to get them there intact

[03:17]

<[Chameleon]>

ah

[03:18]

<Yono>

yeah, yeah

[03:18]

<[Chameleon]>

<Yono> what are the chances of porting uclinux or an equivlent to a device without an ARM core?

[03:18]

<Yono>

lets take this to pm

[03:18]

<[Chameleon]>

uClinux is for CPUs that do not have a MMU

[03:18]

<[Chameleon]>

no

[03:18]

<[Chameleon]>

keep it in this channel so that others who may have input or a desire to learn can do so.

[03:19]

yeah, I'd like to hear too :-)

[03:19]

It was my suggestion after all...

[03:19]

<[Chameleon]>

:)

[03:19]

<Yono>

fine, if I get in trouble with joe I blame it on americantechpush :-P

[03:19]

<[Chameleon]>

LOL

[03:19]

ok

[03:19]

<[Chameleon]>

just don't divulge privy info

[03:20]

<Yono>

lets see

[03:20]

<Yono>

the pocketplayer isn't going to have an ARM core

[03:20]

<[Chameleon]>

OK... Just a DSP like the current Neuros?

[03:20]

<Yono>

that is probably general knowledge that could be found if you googled anyway

[03:20]

<Yono>

ummm kind of

[03:20]

I think Joe has mentioned the chipset on the forums before...

[03:20]

I'd have to search for it

[03:21]

<Yono>

TMS320DA255, in the C55 family

[03:21]

<[Chameleon]>

yeah, just a DSP

[03:21]

<[Chameleon]>

figures

[03:21]

<[Chameleon]>

easier to use the same base firmware for all devices that way

[03:21]

<Yono>

this is a company that has linux running without an arm: linuxdevices.com

[03:22]

<Yono>

this is basically what the pocketplayer is an oem of: 64.233.167.104

[03:22]

<Yono>

but the device that had linux working was a floating point C6x, more advanced

[03:23]

<[Chameleon]>

yup

[03:24]

but the FP shouldn't be a problem, because (AFAIK) FP isn't used inside the kernel

[03:24]

<Yono>

this all would be easier if NA had chosen to oem a more linux-friendly device, but oh well

[03:24]

<Yono>

N3 should have ARM, so thats whats really important

[03:24]

<[Chameleon]>

yeah

[03:25]

yes. OTOH, someone mentioned here that the neruos mini was in doubt here...

[03:25]

<[Chameleon]>

I think that there will likely be a code branch out for the MPEG-4 Recorder & N3 devices.

[03:25]

<[Chameleon]>

and future devices on similar platforms.

[03:26]

does the TMS320DA255 have a GCC port?

[03:27]

<[Chameleon]>

do we look like google?

[03:28]

well, google isn't helping me much :-(

[03:28]

<Yono>

derobert: its C55x

[03:29]

<DeepB>

Yono: i've read a while ago about a company selling a custom gcc toolchain for C5x targets

[03:29]

<Yono>

really?

[03:30]

<DeepB>

and this guy already ported uCLinux to those DSC25 targets

[03:30]

<DeepB>

www.scguild.com

[03:30]

gcc.gnu.org gives the c4x family...

[03:32]

<Yono>

DeepB: would u mind pointing those out to chreekat if you see him on here?

[03:32]

<[Chameleon]>

that Nutt guy could do a lot of good for Neuros Audio

[03:32]

<[Chameleon]>

We should forward his resume to Joe.

[03:32]

<DeepB>

Yono: ok

[03:33]

yeah, totally

[03:34]

<DeepB>

Yono: i still can't find where i read about that C5x toolchain, though

[03:35]

<DeepB>

maybe i pointed it out on the forums when i first read it

[03:35]

<DeepB>

lemme check

[03:37]

<Yono>

ok

[03:37]

<DeepB>

yep, i did

[03:37]

<DeepB>

www.neurosaudio.com

[03:38]

<Yono>

it Dr Hayes hired or working privately?

[03:39]

<DeepB>

that's the link: pedigree.phenominet.com

[03:40]

<Yono>

it was done for both C5470 and 71

[03:40]

<DeepB>

Yono: dunno

[03:41]

Hmmm, that link says there are Linux command-line tools from T1 for the C54x

[03:41]

<DeepB>

derobert: yep, that was what turned me on the most

[03:41]

but that probably doesn't mean GCC.

[03:42]

<Yono>

DeepB: the C5471 has ARM7

[03:42]

probably means "give TI $$$$$$"

[03:46]

<DeepB>

Yono: i've told all this info before, but as it seems Joe is now seriously into OS, and you seem to be pretty close to him, be sure to remind him about it

[03:47]

<DeepB>

i need some rest

[03:47]

<DeepB>

have a nice one, gnite all

[03:48]

<Yono>

later DeepB, I sent him the resume

[04:21]

sorune: you still around?

[04:22]

yep

[04:24]

i'm probably one of the few people who use the sorune gui in linux. :) anyway, why does the sync go only ~500k/s through the gui (USB2).

[04:25]

when i transfer manually and then just rebuild the database, i get ~5 mb/s

[04:26]

when i take a look at CPU usage w/ top while it's running, it reports sorune taking more than 50-70% of CPU, and system slows to a crawl.

[04:26]

not sure, try turning off the xfer validation

[04:26]

hmmm

[04:27]

ok i found that. i'll try it.

[04:30]

that fixed it. now getting ~7mb/s. i didn't know xfer valid was there. must have got turned on when i updated. thanks.

[04:30]

cool

[04:30]

too bad that doesn't work for you though

[04:33]

on my mandrake system, it would take about 1-2 secs to do the verification per song

[04:33]

but that's with an N1

[04:34]

i wouldn't think an N2 would take more than a second

[04:34]

may have something to do with your supermount again???

[04:39]

g'nite all

[04:44]

in case anyone wants to see, here are some more pics that Yono took of the recorder

[04:44]

www.sorune.com

[05:44]

<[Chameleon]>

noiz77: I haven't received mine yet.

[05:53]

we should all pitch in and buy yono a camera...

[06:02]

I'll take some nice pics once I get mine... Well, at least nicer than Yono's pics

[07:03]

<[Chameleon]>

LOL, me too

[07:03]

<[Chameleon]>

I bought a new 5 MP camera for Christmas

[07:03]

<[Chameleon]>

takes nice pix

[12:32]

<erwan_taf>

[Chameleon]: :)

[12:59]

erwan_taf: did your friend and joe have a chance to talk yet ?

[12:59]

<erwan_taf>

seems they've been, by mail

[12:59]

oh, that's good news. anything else you know about that ?

[12:59]

<erwan_taf>

nope

[13:00]

<erwan_taf>

sorry

[13:00]

np. keep up posted, in case you learn anything else

[13:00]

*us

[13:00]

<erwan_taf>

of course

[13:15]

* nerochiaro is at the office, with nothing to do, no one else around and a terrible amount of sleep. should he give up to the temptation and take a nap ?

[13:16]

<erwan_taf>

:)

[13:16]

shakespearian dilemma, "to sleep or not to sleep ?" ;)

[15:18]

<erwan_taf>

nerochiaro: be carefull you are sleeping on your keyboard then you're joining/quitting all the time :)

[15:22]

erwan_taf: no, i had to reset the pc, and when i come back i ghosted myself by mistake :P

[15:23]

in the end i chose not to sleep

[16:04]

cool - Linux on DSP core :)

[16:05]

wonder if my company would like to build a DSP board :)

[16:13]

JoshMalone: where did you see that ?

[16:16]

linuxdevices.com

[16:16]

reading thru channel archives

[16:19]

I'm getting ready to order parts for my media player deck; should get paid next week and finally have the $$$

[16:22]

that dsp mentioned there is probably a lot more powerful than the 54x, i would guess

[16:24]

and as usual "MediaBios is built around TI's hard real-time kernel DSPBIOS"

[16:24]

yeah - not suitable for smaller DSPs Id imagine, but still cool

[16:24]

(and cheaper than a $30-40 X-scale CPU)

that's sure

heh - welcome back

THis company sells VIA EPIA boards with linuxbios bre-installed :)

[16:54]

<DeepB>

??AA strikes again

<DeepB>

LOL - that's funny

so that guy pocketed the "legal defense" money people donated?

[16:59]

<DeepB>

could be, he raised some real funds, enough to spend some nice holidays on Brazil

[17:01]

yeah - even minus the undisclosed "settlement" amount

[17:03]

<DeepB>

april's fool is not close enough anyway

[17:03]

nope *sigh*

[17:03]

<DeepB>

all major torrent sites are now officially fucked up

[17:04]

All of the US is officially fucked up

[17:04]

<DeepB>

??AA lawyers do they work pretty well

[17:04]

we're getting closer every day to being run by greedy corporations than by "We, the people"

[17:05]

<DeepB>

closer? is that a joke?

[17:05]

closer to 100%

[17:05]

(it's prolly 60 now, I'd estimate - we can still choose to not buy things :)

[17:06]

<DeepB>

hahaha

[17:06]

when we start having mandatory consumption rates is when I break out the automatic weapons and the molitov

[17:08]

<DeepB>

i love the american way of mass comsumption/destruction ;)

[17:08]

<DeepB>

"i don't need it, but it's cheap, let's buy it"

[17:09]

<DeepB>

"I don't like it, and it's weak, let's bomb it"

[17:10]

:(

[17:13]

<DeepB>

anyway, i should not pay attention to Alex Jones', Moore's and the likes speechs... they're turning me into a whole of a paranoid

[17:14]

<DeepB>

i promise

[17:14]

I've given up on Moore

[17:14]

he's just as bad as the people he satires

[17:15]

<DeepB>

really? i kinda suspected it

[17:17]

I think so

[17:17]

he's not exactly forthcoming and honest in his documentaries

[17:17]

although I think he's more right than wrong, he still bends the truth

[17:18]

and that's not cool, in my book, for a filmmaker who wants people to believe him

[17:19]

dam guys, as much as i would like to pitch in into this interesting and inflammatory discussion, i can't. in 40 minutes i need to go sign the contract to get my new job :)

[17:19]

woohoo - congrats man

[17:19]

what's the gig?

[17:19]

<DeepB>

congrats nero, hope that will be for good

[17:20]

<ivel|Work>

congrats

[17:20]

<ivel|Work>

JoshMalone: i need to talk to you about this media box you're building...

[17:20]

it's a job as .NET programmer, not web (i needed a change from that stuff, after years doing that)

[17:20]

ivel: sure man - gimme a sec though - need more drinkage

[17:21]

nerochiaro: sounds all right, though

[17:21]

<ivel|Work>

take your time...i'm at work

[17:21]

hell - any programming job U can get these days is good :)

[17:21]

indeed

[17:21]

ivel: me, too - bbiab

[17:21]

<ivel|Work>

errr...any IT job too

[17:21]

gotta run for a sec - brb all

[17:21]

especially if it's better paid and has better perspectives of growth than the current one ;)

[17:21]

<DeepB>

IT jobs are easy to get, properly paid ones are not

[17:22]

<DeepB>

at least here in third world

[17:22]

i'll be out there in a few seconds. catch you all tonight as i come back from there. wish me good luck

[17:22]

* nerochiaro is out

[17:22]

* DeepB wishes the best to nerochiaro

[17:27]

ivel: so - wassup?

[17:28]

<LoRez>

DeepB: you got all the car mode issues worked out?

[17:30]

<DeepB>

LoRez: so far i'm having only one side effect, the unit does not turn itself off after unplug, it it was charging on stand-by, when car mode is set to off

[17:30]

<ivel|Work>