illwieckz has quit [Remote host closed the connection]
illwieckz has joined #neo900
b1101 has joined #neo900
b1101 has quit [Quit: b1101]
b1101 has joined #neo900
b1101 has quit [Changing host]
b1101 has joined #neo900
illwieckz has quit [Quit: Ça va couper chérie…]
illwieckz has joined #neo900
illwieckz has quit [Quit: Ça va couper chérie…]
illwieckz has joined #neo900
MonkeyofDoom has joined #neo900
Recharged has joined #neo900
Recharged has left #neo900 [#neo900]
Pali has joined #neo900
infobot has quit [Read error: Connection reset by peer]
infobot has joined #neo900
paulk-aldrin has joined #neo900
arcean has joined #neo900
nnevala has joined #neo900
Laverne has joined #neo900
SylvieLorxu has joined #neo900
heinrich5991 has joined #neo900
jonwil has quit [Quit: ChatZilla 0.9.91.1 [SeaMonkey 2.31/20141202220728]]
billtux has joined #neo900
<billtux>
hi guys just reading about the project on the site just now it sounds great. Really be interested in purchasing one :) subbed on the news letter to stay updated. Good luck!
<sixwheeledbeast>
Oksana: cheers for the christmas time update.
sparetire_ has joined #neo900
paulk-aldrin has quit [Remote host closed the connection]
<billtux>
by dontaing to the project will you get a device to hack about with? or is that just funding for R&D?
<dos1>
billtux: if you donate at least 100 euros, you can be sure that there will be a device produced for you (it's built to order, not to shelf). Also, this amount you pay now will already count to the final price, so if you donate 100 EUR now, you'll pay 100 EUR less when finally ordering the ready device.
<billtux>
Thanks i just asked on the hackernews thread as well ha. Thats great think im going to donate and secure a device :)
<billtux>
well hopefully if it all goes to plan.
<billtux>
is there any ideas around how much the device will cost once its finalised?
<dos1>
http://neo900.org/faq#cost - "According to current estimations, the cost of motherboard should be somewhere between 500 and 700 EUR. Complete device will cost 100-150 EUR more, depending on prices and availability of N900 spare parts. Those prices are just early estimations, are subject to change and depend heavily on number of orders."
illwieckz has quit [Ping timeout: 240 seconds]
illwieckz has joined #neo900
illwieckz has quit [Read error: No route to host]
illwieckz has joined #neo900
illwieckz has quit [Read error: No route to host]
illwieckz has joined #neo900
Richiverse has joined #neo900
Richiverse has quit [Ping timeout: 246 seconds]
drathir has quit [Ping timeout: 244 seconds]
<DocScrutinizer05>
\o/ Fank V. transferred 500.-
<DocScrutinizer05>
heck, somebody must have organized a paypal flashmob :-D
<DocScrutinizer05>
4 payments
<wpwrak>
must be people who got some idroid for xmess but are longing for something real. so they returned it and got their ticked for a neo900 :)
<kerio>
right
<DocScrutinizer05>
maybe CCC
<DocScrutinizer05>
I already wonder if somebody there would be kind enough to mention Neo900 in a sidenote
<dos1>
we've got quite high traffic from hacker news, as there's xmas update posted quite high in comments
<DocScrutinizer05>
cool
<DocScrutinizer05>
excellent your talk and its topic was in there
<kerio>
pet peeve: the screen turning off should be the last thing that happens during the shutdown process
<dos1>
and it's still on main page
<kerio>
alternatively, turn on a led before the shutdown process starts
<dos1>
kerio: I do that on my N900. I modified shutdown led pattern to light up at its end :)
<DocScrutinizer05>
kerio: quite right, but then, on N900 the LED is never really off, even when complete device is shutdown
<kerio>
there's a shutdown led pattern? :o
<dos1>
yup, that white fade-out
<DocScrutinizer05>
NB LP5523 is powered directly from Vbatt
<DocScrutinizer05>
it needs I2C to "shut it down"
<DocScrutinizer05>
or at least a GPIO for reset, sorry I'm still not mentally performing
<dos1>
anyway, looks like kernel does that on shutdown
<DocScrutinizer05>
leranyway yep, it's a known effect that the device isn't off yet when it pretends to be. It takes another 10 or 20s until you can power it up again
<DocScrutinizer05>
kerio: ^^^
<kerio>
yes
<kerio>
which is fucking awful
<kerio>
dos1: giff instructions
<DocScrutinizer05>
it kinda is, yes
<DocScrutinizer05>
kerio: I guess you simply change shutdown pattern in pattern editor to end in a !=0,0,0 pattern
<dos1>
exactly
<kerio>
wtf is pattern editor
<dos1>
it's somewhere in repos :)
<DocScrutinizer05>
LED pattern editor
<DocScrutinizer05>
you of course can hack away in plain "assembler" in etc/mce/mce.ini
<kerio>
anyway, after an apt-get update, it's here
<DocScrutinizer05>
I honestly "wouldn't know to live" without pattern editor
<DocScrutinizer05>
I rarely ever use it, but when I do I'm extremely happy it's there
<DocScrutinizer05>
though it lacks a few options ;-)
<DocScrutinizer05>
iirc you cannot change the type of pattern, so it shows for example even with screen enabled
<DocScrutinizer05>
you need to hack/edit mce.init for that
<DocScrutinizer05>
also creating new proprietary custom patterns (e.g. for H-E-N) doesn't work in LED Pattern Editor. Maybe editing them works, once they got created
<kerio>
it works for usbmode
<kerio>
anyway, i got a stylish purple :3
<bencoh>
hm, I think I'm still using default patterns
<DocScrutinizer05>
I guess usbmode using the same booston.sh script of mine, so the same LED pattern
<DocScrutinizer05>
oh well, maybe not. I guess pali integrated that stuff into the kernel module for bq24150
<kerio>
dos1: ty for the suggestion
<kerio>
DocScrutinizer05: no no, mceledpattern is a dependency cuz it adds the patterns when you install it
<kerio>
there's a pattern for "boost"
<kerio>
...and that's about it, really
<DocScrutinizer05>
dos1: indeed interesting suggestion. Does it shut down hard at CPU shutdown?
<Pali>
usbmode.sh is using kernel driver bq2415x_charger
<Pali>
which implements boost mode
<dos1>
yw :) I couldn't live without it, as I was switching batteries daily
<DocScrutinizer05>
Pali: who sends the dbus cmd for LED pattern?
<Pali>
LED Pattern Editor from Extras supports also custom patterns
<Pali>
if you create new pattern in mce.ini (either manually or via mceledpattern cli) it will show in LED pattern editor
<Pali>
DocScrutinizer05: when using BME replacement it is done by hald-addon-bme (replacement)
<DocScrutinizer05>
Pali: yep, that's what I said above I think should work
<dos1>
DocScrutinizer05: yup. at least I never noticed that I still had to wait to boot the device again after led turned off
<Pali>
if you are not using BME replacment, then usbmode.sh will send that dbus cmd
<DocScrutinizer05>
Pali: aaah hald-addon
<DocScrutinizer05>
Pali: ta! :-)
<Pali>
usbmode.sh is intelligent and can detect if it needs to send dbus pattern cmd or not
<kerio>
DocScrutinizer05: why aren't using the bme replacement? :c
<kerio>
*you
<DocScrutinizer05>
kerio: because I usually use stock kernel
<DocScrutinizer05>
shame on me, I know
<kerio>
yeah, shame on you
<kerio>
mh i dunno actually
<kerio>
if it's cuz you're lazy, shame on you
<kerio>
if it's cuz of idealistic reasons, still shame on you but a different kind of shame
<DocScrutinizer05>
nah, I'm a coward who needs a 100% reliable phone
<bencoh>
how is stock kernel more reliable than recent kp ?
<kerio>
i had like 280 days of uptime before i rebooted today
<DocScrutinizer05>
bencoh: I dunno, and that's the point
<kerio>
(did a full backup, about to install the newer cssu)
<DocScrutinizer05>
bencoh: I simply didn't need KP and so I didn't feel like messing with my daily phone
<bencoh>
to be honest I dont remember the first reason I installed kp (but it was *not* OC :)
<DocScrutinizer05>
and one thing's for sure: a BM backup made from a system on KP will wail nuclear on restoring it to stock kernel system
<bencoh>
maybe to try smartreflex and how battery performed with it
<DocScrutinizer05>
s/wail/fail/
<DocScrutinizer05>
smartreflex is one of the things I'm massively averse of "trying"
<bencoh>
are there any nat modules on stock ?
<DocScrutinizer05>
no
<bencoh>
I think that was the second reason I switched
<DocScrutinizer05>
valid and good reason
<DocScrutinizer05>
NB I don't say KP is unstable
<DocScrutinizer05>
I say I don't feel like ganbling with it on my daily phone, unless I see a need for it
<kerio>
a BM backup made from a system on KP with kernel-modules still installed will work flawlessly on stock kernel
<DocScrutinizer05>
prolly KP has less bugs than stock kernel
<kerio>
thumby packages otoh...
<DocScrutinizer05>
ooh right
<Pali>
720p dsp support
<DocScrutinizer05>
err, right?
<DocScrutinizer05>
not sure
<kerio>
the "issue" is if you decide to uninstall kernel-modules to free space
<Pali>
fixed UDF FS
<Pali>
ipv6 support
<kerio>
even then, it's only marginally an issue
<kerio>
you can just drop the modules in the tarball from a pc or whatever
<DocScrutinizer05>
sorry, guys. I didn't want to start a KP discussion
<DocScrutinizer05>
KP is great
<DocScrutinizer05>
all hail to Pali for maintaining it
<bencoh>
:-)
<kerio>
hail
<DocScrutinizer05>
btw he could use a standin for times when he's a week on holiday (and for general bus factor improvement). Nobody around to pick up on this duty? :-)
<kerio>
huh
<kerio>
i've got like 4 packages to upgrade
<kerio>
freemangordon where are you
<DocScrutinizer05>
vacancy: vice KP maintainer
<kerio>
why the fuck was fmtx-middleware updated
<bencoh>
haha
<DocScrutinizer05>
wpwrak: a tad of kernel maintenance assistance/backup on your side?
<DocScrutinizer05>
the git/patch mess could use some sanitation anyway ;-)
<kerio>
ooh, fmg is maintaining it
<kerio>
freemangordon: what's new in fmtx-middleware 0.65?
<kerio>
do i need to h4x0r fmtxd again
<DocScrutinizer05>
~seen freemangordon
<infobot>
freemangordon is currently on #maemo-ssu #maemo #neo900. Has said a total of 24 messages. Is idling for 3h 54m 37s, last said: 'ok'.
<DocScrutinizer05>
ooh
<DocScrutinizer05>
any news from merlin1991 and CSSU?
<kerio>
T10 was released like yesterday
<kerio>
or some days ago, anyway
<DocScrutinizer05>
\o/
<bencoh>
DocScrutinizer05: wow, so you really cut loose ;p
<DocScrutinizer05>
how many fatalities so far? ;-)
<kerio>
bencoh: he /rageparted #maemo*
<bencoh>
yeah I know ;)
<DocScrutinizer05>
bencoh: I had to detach from maemo at large, or I would have turned into a bad man
<bencoh>
:)
<kerio>
i thought I was the bad man
<kerio>
for the poettering hitman thing
<DocScrutinizer05>
you're a nice troll
<DocScrutinizer05>
;-)
<DocScrutinizer05>
everybody loves you
<bencoh>
I think the anti-poettering thing is one of the things that brings the community together ... isnt it ? :p
<DocScrutinizer05>
except poettering (maybe)
<kerio>
do i yolo upgrade all the 31 packages or do i stay on the -thumb versions for now
<DocScrutinizer05>
I could figure even poettering is secretly happy about your hitman trolling
<kerio>
oh he loves it
<kerio>
he gets to be a martyr for no good reason
<kerio>
i don't want a non-thumby microb-engine :(
<DocScrutinizer05>
he got something to whine about
<DocScrutinizer05>
yeah, martyr, that's it :-P
<DocScrutinizer05>
Wolfgang Spraul once told me "don't worry, this is FOSS. You wouldn't believe what threats I received already. Just ignore it!"
<DocScrutinizer05>
(owtte)
<bencoh>
:]
<Pali>
kerio: in cssu-devel is new version of fmtx-middleware
<DocScrutinizer05>
without MP an abomination like cherry hadn't any chance to survive the bright light of daily usage
<kerio>
?
<kerio>
that's just preinstalled
<kerio>
you can uninstall it without any problem
<kerio>
can't you
<DocScrutinizer05>
yes, like everything
<kerio>
no i mean
<kerio>
it's not even in the metapackage
<DocScrutinizer05>
ooh?
<DocScrutinizer05>
maybe not anymore, for CSSU
<dos1>
I think it is, or was. I remember that disabling it meant tricking it into believing that it already registered, not uninstalling the package
<DocScrutinizer05>
anyway, uninstalling MP is easy, yes. But it also breaks any upgrade since CSSU still ships MP upgrades
<DocScrutinizer05>
dos1: exactly
lexik has joined #neo900
<DocScrutinizer05>
the whole purpose of not-my-nokia package
<DocScrutinizer05>
uninstalling option been rejected iirc, since it would have the opposite effect of sending an SMS *again* on next update re-installing it
<DocScrutinizer05>
CSSU is a layered additional repo on top of stock, it's not simple to uninstall stuff via CSSU that comes per default from underlying stock
<DocScrutinizer05>
you can update stuff - i.e. replace by newer version - but not remove. AIUI
<kerio>
that's when you ship a neutered "cherry" package that does nothing
* DocScrutinizer05
writes "Alban Crequy" on some special list
drathir has quit [Quit: Lost terminal]
<sixwheeledbeast>
yep a cherry update that is an empty deb. The issue is putting neutered dummy package in the correct repo? also it probably will not update with the gpg warning?
<DocScrutinizer05>
ok, no "cherry" in /opt/var/cache/apt/archives/mp-fremantle-community-pr_20.2010.36-2maemo12.5_all.deb (untarred by mc)
<DocScrutinizer05>
sorry, I seem to enter C4 or overtemp shutdown state, mentally. Can't find *any* "depends" in /opt/var/cache/apt/archives/mp-fremantle-generic-pr_20.2010.36-2_all.deb
<DocScrutinizer05>
nah, there it is
<DocScrutinizer05>
seems there *is* a "cherry" somewhere in /opt/var/cache/apt/archives/mp-fremantle-generic-pr_20.2010.36-2_all.deb
<DocScrutinizer05>
mc jumps to the one extralong line " Depends:" that supposedly contains it, but I can't spot it in that wall of tags
<kerio>
grep for "cherry" maybe? i dunno
<DocScrutinizer05>
HAH!! Konsole sh-ctrl-F helped me out :-)
<DocScrutinizer05>
mc is smart enough to simply "open2 it
<DocScrutinizer05>
s/2/"/
<DocScrutinizer05>
so yes, CSSU SHALL include a neutered cherry package
<DocScrutinizer05>
~2119
<infobot>
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119.
<DocScrutinizer05>
dunno why we decided not to do, back when
<DocScrutinizer05>
maybe it was eben me who had a mental hickup, arguing we should not force users to lose MyNokia
<DocScrutinizer05>
#maemo-ssu chanlog might know
<sixwheeledbeast>
ha
<DocScrutinizer05>
but honestly, back on topic, CSSU should get rid of MP concept all together
<sixwheeledbeast>
Well in that case this could be pushed to cssu-devel and then T10.1
<DocScrutinizer05>
yes, that's exactly what cssu-devel is meant for: testing update issues
<kerio>
DocScrutinizer05: HAM won't upgrade anything that's not marked as a user package
<DocScrutinizer05>
I *guess* (no packaging expert here) that all the stuff that comes inside MP is not showing up in HAM as user package since you shall not be able to uninstall any of those pkgs
<kerio>
remove that restriction, and HAM becomes slow as FUCK
<DocScrutinizer05>
kerio: HAM already *is* slow as *f*
<kerio>
no no
<kerio>
actually slow
<kerio>
as in you won't ever see operations complete because you *will* close it in frustration
<DocScrutinizer05>
actaully slow? like "it takes not 15 minutes but 15 hours"?
<kerio>
i'm not sure how to get a count of the user packages
<DocScrutinizer05>
not all of them need to go "user"
<kerio>
but much less
<kerio>
DocScrutinizer05: yes
<DocScrutinizer05>
only those that formerly were in MP
<kerio>
or you end up doing shit that's even more awkward than metapackage
<kerio>
no no
<kerio>
you don't get it
<kerio>
ham will pretend the non-user packages don't exist
<kerio>
unless it's upgrading a user package that has a dependency on a >= version
<DocScrutinizer05>
*nod*
<DocScrutinizer05>
all fine to me, e.g. for modest
<DocScrutinizer05>
there's absolutely no good reason why modest is in MP
<kerio>
you can't do rolling releases when 90% of your shit isn't round
<DocScrutinizer05>
same applies for ~90% of the packages in there
<kerio>
an even better question is
<kerio>
do we just drop extras(-*) for fptf?
<DocScrutinizer05>
huh?
<kerio>
and just rehost packages in a more sensible repo
<DocScrutinizer05>
no
<kerio>
rootsh is in extras, dude
<kerio>
that shit would be kicked out of the debian repos so fucking hard
<DocScrutinizer05>
lol
<bencoh>
wtf ?
<kerio>
also, 90% of the things in extras-devel are outdated
<DocScrutinizer05>
rootsh is flawed in many ways, but it's a working package
<kerio>
i *really* hope fptf won't have the gainroot bullshit
<kerio>
or rdmode
<kerio>
like, i hope it won't be a thing with the neo900
<DocScrutinizer05>
the real problem is not rootsh pkg but HAM not asking for root password
<bencoh>
me too, but that's a different ting than dropping extras
<bencoh>
extras is outdated 1. because of lack of man power to maintain packages 2. because fremantle core is outdated and some things cannot be upgraded in a peacefull way (but you know that already, better than I do)
<bencoh>
so unless we simply rebase fremantle (the specific bits) on top of debian ...
<bencoh>
s/simply// (that wont be "simple" :D)
<kerio>
i guess you have to find a balance between backward compatibility and having modern shit
<bencoh>
yeah, kinda, but that takes even more human resources
<kerio>
...do you, really?
<DocScrutinizer05>
you seem mistaken on what's fptf. It's not a new repo, it's a set of patches on top of stock fremantle
<bencoh>
(simple example : gstreamer, or PA)
<kerio>
yeah that shit needs to be upgraded
<DocScrutinizer05>
the goal is to keep compatibility to all extras-* packages on Neo900
<DocScrutinizer05>
without need to rebase, recompile, re-$whatever
<DocScrutinizer05>
otherwise we could have moved on to snapdragon CPU already
<bencoh>
(noooo :D)
<DocScrutinizer05>
;-)
<kerio>
DocScrutinizer05: is /sbin/preinit going to be a thing
<kerio>
to keep compatibility with bootmenu
<bencoh>
erm
<bencoh>
eww :)
<DocScrutinizer05>
I'm pretty sure preinit will see some patches but needs to keep "API" compatibility
kerio has left #neo900 ["i'm getting the FUCK away from this shit"]
<DocScrutinizer05>
o.O
<bencoh>
haha
<bencoh>
/invite kerio
<DocScrutinizer05>
I guess he accidentally hit alt-F4 or whatever
kerio has joined #neo900
kerio has left #neo900 ["it was intentional"]
<DocScrutinizer05>
prolly the 7kg Serrano fell on his kbd
<bencoh>
yeah, and his cat typed the F word and everything around it, sure :p
<DocScrutinizer05>
hehehe
* DocScrutinizer05
feels like... pondering what to do about own wetware-CPU thermal management
kerio has joined #neo900
<kerio>
have you guys regained sanity
<DocScrutinizer05>
step 1: check actual recent temperature
<DocScrutinizer05>
kerio: yep, while you were gone, we plotted a plan how to raid your home and take you for hostage
<kerio>
DocScrutinizer05: /sbin/preinit is an implementation detail, not an official API
<kerio>
so packages that rely on that will have to be updated
<DocScrutinizer05>
kerio: look fella, we base on a number of closed blobs in fremantle which we cannot host anywhere. We only can provide an overlay similar to CSSU to add some patches/updates to make fremantle work on Neo900
<bencoh>
kerio: I've been missing with gstreamer, python, vlc, dsp/omx codecs and hls today, I think sanity isnt something I should regain anytime soon :p
<bencoh>
s/missing/messing/
Richiverse has joined #neo900
<kerio>
DocScrutinizer05: no we don't
<kerio>
we provide an updated version of apt
<kerio>
that can do http redirects
<kerio>
no illegal hosting :>
<DocScrutinizer05>
cool
<DocScrutinizer05>
please write a whitepaper, ideally add a PoC
<kerio>
i dunno man shits hard
<kerio>
as in shit's hard
<DocScrutinizer05>
man shits hard, usually. Yes ;-P
<kerio>
DocScrutinizer05: are we actually going to keep NOLO and stuff like that?
<kerio>
cuz otherwise i don't think preinit will work anyway
<Richiverse>
Is it possible to send in my n900 to neo-ize it?
<DocScrutinizer05>
Richiverse: probably not
<DocScrutinizer05>
not right now for sure ;-)
<DocScrutinizer05>
Richiverse: we've still not decided if we can offer such service. Shipping alone is probbaly too much overhead
<DocScrutinizer05>
picking up parcels from abroad at customs alone is way more hassle than the whole hw upgrade
<Richiverse>
K thx
<DocScrutinizer05>
we pay import taxes on your device
<DocScrutinizer05>
and there's the possible warranty conflict when a device isn't as healthy as the sender thought it was (or maybe it actually been at time of sending, but not when it arrived)
<sixwheeledbeast>
I am sure when the Neo is realised a video of how to upgrade will be released
<DocScrutinizer05>
so the general answer is "cannot be done", though we might allow exceptions
<DocScrutinizer05>
sixwheeledbeast: sure
<DocScrutinizer05>
it's not rocket science
<kerio>
more importantly
<kerio>
is it possible to send in the neo900 board when you realize you're not capable of installing it in your n900
<sixwheeledbeast>
It's not hard even for a complete noob. Maybe even Neo can sell/supply PH00 screwdrivers
<kerio>
and get it installed by people with hands
<DocScrutinizer05>
sixwheeledbeast: we will supply all needed tools
<sixwheeledbeast>
I would think if you wanted a Neo9 you expect to be able to make it work.
<sixwheeledbeast>
cool free tools
<bencoh>
"free" :)
<kerio>
sixwheeledbeast: i can't hands
<kerio>
the last time i picked up a screwdriver my washing machine broke
<kerio>
and i was opening my laptop
<bencoh>
haha
<DocScrutinizer05>
tentatively the set will consist of: Torx6 driver, Phillips00 driver, a plastic tool to gently unplug the flat ribbon cable, the space&kbd frame plus washers/whatever it comes with, and a uSD card that has some OS and a video and pdf tutorial
<sixwheeledbeast>
lol, seriously it's not going to be difficult. I am sure a phone shop could do it too.
<DocScrutinizer05>
ooh, I forgot: screws
<kerio>
what's the torx one for
<kerio>
:o
<DocScrutinizer05>
the kbd frame screws
<kerio>
also why does it need a usd to boot?
<kerio>
DocScrutinizer05: is there an actual benefit to using torx screws?
<kerio>
or is it only done to prevent people without torx screwdrivers from unscrewing the screws?
<DocScrutinizer05>
it doesn't *need* a uSD, but it's a convenient way to ship our factory/BSP image to users in a bootable form
<kerio>
will the board have a nand memory? i forgot
<DocScrutinizer05>
kerio: generally Torx is the better alternative to Phillips
<kerio>
better for whom? the installers?
<DocScrutinizer05>
technically
<kerio>
or does it actually provide a benefit to the structure of what you're screwing
<sixwheeledbeast>
I think you missed the important bit in your list :D
<DocScrutinizer05>
the question rather is: could we use Torx for the metric screws as well?
<DocScrutinizer05>
sixwheeledbeast: no, I talked about the mech toolkit
<sixwheeledbeast>
I know
<kerio>
silly thing: are torx screws more resistant cuz the head can be smaller or something like that
<DocScrutinizer05>
the NeoN board comes either with the toolkit or embedded into a new case
<DocScrutinizer05>
kerio: they are more sturdy, yes
<kerio>
then why the fuck aren't they used everywhere
<DocScrutinizer05>
but inevitably have less flat heads
<kerio>
also what about pentalobe screws? :>
<sixwheeledbeast>
torx is better it was designed for better torque on the head. No slipping or rounding etc..
<DocScrutinizer05>
exactly
<DocScrutinizer05>
but you need higher head 2depth"
<DocScrutinizer05>
s/2/"/
* DocScrutinizer05
ponders to set up an outo correction on IRC client
<DocScrutinizer05>
auto
<sixwheeledbeast>
:nod: more like a better version of an allen key, it has to have the depth in the head. So not always easy to replace if original has a shallow head.
<sixwheeledbeast>
or countersunk
<kerio>
DocScrutinizer05: my laptop has pentalobe screws on the chassis :(
* DocScrutinizer05
wonders if Konversation knows regex in autoreplace
<DocScrutinizer05>
kerio: OUCH!
<kerio>
MacBook Pro with Retina Display 15" Late 2013 Repairability Score: 1 out of 10 (10 is easiest to repair).
<DocScrutinizer05>
>>The pentalobe screw drive is a five-pointed **tamper-resistant** system primarily used by Apple in its products<<
<sixwheeledbeast>
I gues it's a mac
<sixwheeledbeast>
oh
<DocScrutinizer05>
~wiki pentalobe_screw
<infobot>
At http://en.wikipedia.org/wiki/Pentalobe_screw (URL), Wikipedia explains: "{{Infobox product | title = Pentalobe screw | image = | caption = Pentalobe screw diagram | inventor = | launch year = 2009 | company = Apple | available = In use | current supplier = | last production = | notes = Tamper resistant screw | url = }} The 'pentalobe screw' drive is a five-pointed tamper-resistant system primarily used by Apple in its products.{{cite web |url ...
<kerio>
The lithium-polymer battery is glued rather than screwed into the case, which increases the chances that it'll break during disassembly. The battery also covers the trackpad cable, which tremendously increases the chance that the user will shear the cable in the battery removal process.
<kerio>
The display assembly is completely fused, and there’s no glass protecting it. If anything ever fails inside the display, you will need to replace the entire, extremely expensive assembly.
<kerio>
pretty good huh
<DocScrutinizer05>
well, you bought Aplle product. You asked for it
<sixwheeledbeast>
well it is an apple product
<bencoh>
oooh right, the pentalobe stuff
<DocScrutinizer05>
hehe
<kerio>
the pentalobe shit is obviously just for evil
<DocScrutinizer05>
Apple products are made for being shiny and convenient to use as long as they last, and to earn Apple more money by selling a new product to their faithful users pretty soonish
<bencoh>
meh :)
<kerio>
but i think that they actually need stuff like glued batteries and glued screens to achieve maximum thinness
<DocScrutinizer05>
that too
<bencoh>
DocScrutinizer05: it's like maemo, you really need a good reason to stick with it and all its blobs ;p
<kerio>
rofl
<bencoh>
(right, we're not friday)
<DocScrutinizer05>
>>This was the only internal usage of pentalobe screws; all following MacBook Pros use the "Tri-Wing" security bit to attach the battery to the internal frame.<< \o/
<DocScrutinizer05>
((friday)) right :-) But almost
<DocScrutinizer05>
duh! infobot does't know s***t about ~friday
<DocScrutinizer05>
o.O
<DocScrutinizer05>
~listkeys friday
<infobot>
Factoid search of 'friday' by key (13): fridayitis ;; 02friday ;; just started late friday. uclinux' makefile ;; jargon virtual friday2 ;; black friday ;; friday #DEL# ;; weathercode friday ;; not for me. our fridays ;; emulate friday ;; friday night ;; friday's ;; jargon virtual friday ;; cmd: friday (.*?).
<DocScrutinizer05>
friday 13 ?? ;-P
<DocScrutinizer05>
~jargon virtual friday2
<infobot>
[jargon virtual friday2] Friday (and Thursday is a virtual Saturday, as is Friday). There are also `virtual Mondays' that are actually Tuesdays, after the three-day weekends associated with many national holidays in the U.S.
<DocScrutinizer05>
o.O
<DocScrutinizer05>
HAH!
<DocScrutinizer05>
~listkeys friday
<infobot>
Factoid search of 'friday' by key (13): jargon virtual friday ;; friday night ;; cmd: friday (.*?) ;; 02friday ;; emulate friday ;; fridayitis ;; jargon virtual friday2 ;; weathercode friday ;; not for me. our fridays ;; friday ;; black friday ;; friday's ;; just started late friday. uclinux' makefile.
<DocScrutinizer05>
~friday
<infobot>
[friday] !!!!!!!! FRIDAY !!!!!!!
<DocScrutinizer05>
~factinfo friday
<infobot>
friday -- last modified at Fri Mar 9 15:55:07 2012 by tccastaldo; it has been requested 64 times, last by DocScrutinizer05, 9s ago.
Richiverse has quit [Ping timeout: 246 seconds]
<sixwheeledbeast>
but it's
<sixwheeledbeast>
~sunday
<infobot>
Sunday is the day all trolls swarm to #debian, avoid at all cost to remain sane.
<DocScrutinizer05>
~factinfo sunday
<infobot>
sunday -- last modified at Fri Apr 14 19:08:59 2006 by streuner!n=streuner@p54A5DB34.dip.t-dialin.net; it has been requested 13 times, last by sixwheeledbeast, 11m 23s ago.
<DocScrutinizer05>
:-D
<DocScrutinizer05>
I just sent a flaw report with fix request to Tim :-)
<DocScrutinizer05>
a static factoid should always override a cmd:
<DocScrutinizer05>
not the other way around
<DocScrutinizer05>
anyway I wonder where's the wiki entry for "friday" which explains that on friday everybody on irc turns into a troll
questionfunding has joined #neo900
<sixwheeledbeast>
~black friday
<infobot>
somebody said black friday was evil
<DocScrutinizer05>
LOL this electronic fever thermometer has THIRTY years warranty. Qute brave, particularly for device with LCD
<questionfunding>
Hello there, is it still time to secure a device through funding?
<DocScrutinizer05>
sure! :-)
<kerio>
"can i give you money"
<DocScrutinizer05>
please note that we are about to start a second complementary campaign to raise funds for already buying hard-to-source components like the 1GByte RAM chips, which is not covered by the 100EUR we ask for R&D
<DocScrutinizer05>
IOW please already consider to do a substantially higher down payment than 100 Euro, to allow for sourcing of components for "your" device
<DocScrutinizer05>
questionfunding: also see our latest newsletter
<DocScrutinizer05>
dos1: frontpage news section maybe needs an update, it still says "Don't miss this year's OpenPhoenux Hard & Software Workshop in Munich, Germany on November 29-30."
<dos1>
right
<DocScrutinizer05>
:-)
<questionfunding>
I read the xmas update, and i came here to have a little bit more information :)
<questionfunding>
i can't really do a "substantially higher" payment right now, so i just wanted to know what would happen in the future.
<questionfunding>
is the idea to keep the donations going for the different parts/hard-to-find-components? if yes, what kind of amount can be expected through this year?
<DocScrutinizer05>
questionfunding: no problem when you can't spare more than 100EUR right now. You're in good company so far, lots of users gave 100EUR only, so far. We just need to raise more funds pretty soon, to get components in time for starting production in a maybe 6 or 8 months
<DocScrutinizer05>
last calculations by Werner (wpwrak) resulted in a total BOM cost ~350EUR, however that's including some non-risky parts and misses to count in another few. Nevertheless a first ballpark figure
<questionfunding>
DocScrutinizer05: sounds good to me. do the donations "cumulate" for each device (and for the overall R&D) so that, once everything is funded and done, the devices are shipped to the donators? kind of a dumb question but eh!
<DocScrutinizer05>
e.g. when we look at modem, this alone is a 100EUR and currently I'd not consider it a risk part that needs early (pre production start) sourcing. You howver never can tell what's a risk part until you sourced it or found it starts to get hard to source it
<DocScrutinizer05>
(cumulate) yep, sure
<DocScrutinizer05>
questionfunding: anyway all "donations" to Neo900 are actually credits we owe you payback the very moment you order a device
<DocScrutinizer05>
pretty simple deal
<questionfunding>
DocScrutinizer05: perfect, thank you very much!
<DocScrutinizer05>
yw :-)
<DocScrutinizer05>
sorry for all the weird confusing wording around "donations" and whatnot, it's mainly due to our awkward tax laws here in Germany
<DocScrutinizer05>
we're not going to evade taxes due, but we also don't want to pay taxes we shouldn't have to pay
<DocScrutinizer05>
thus we had to come up with a rather complex concept for the "crowdfunding with payback"
<questionfunding>
you seem pretty available to explain that in simple ways to people like me so that's fine by me :)
<kerio>
DocScrutinizer05: the "playback" znc module is amazing
<DocScrutinizer05>
it got much simpler since the moment a dedicated company Neo900 UG is running the whole business now. It doesn't interfere with other business done by same entity and thus we can simplify things now and make the dinations actual credits - sth we couldn't do with GDC running the project
<DocScrutinizer05>
s/dination/donation/
* DocScrutinizer05
afk
<DocScrutinizer05>
o/
<DocScrutinizer05>
questionfunding: please don't miss to spread the word :-)
<questionfunding>
no problemo im on it ;)
<questionfunding>
thanks again
<DocScrutinizer05>
yw
questionfunding has quit [Quit: Page closed]
paulk-collins has joined #neo900
<DocScrutinizer05>
+1 bank account
<kerio>
DocScrutinizer05: btw why do you have two accounts
<DocScrutinizer05>
hmm?
<DocScrutinizer05>
bank and PP?
ReqGame has joined #neo900
<kerio>
no, irc
<kerio>
05 and 51
<bencoh>
because he's got two eyes
<bencoh>
it's stereoscopic irc
<bencoh>
3D text and all
* ShadowJK
guesses -51 is on RX51
<kerio>
yeah but
<kerio>
why
<DocScrutinizer05>
it's same account but two clients
<DocScrutinizer05>
the desktop client is not linked to bouncer
<DocScrutinizer05>
so I got a fallback when ZNC fails for whatever reason
<kerio>
but y
<kerio>
oh my god you're using konversation
<bencoh>
:))
<kerio>
u pleb
<kerio>
where's your irssi
<kerio>
where's your quassel
<kerio>
where's your weechat
<DocScrutinizer05>
i'm using konversation, xchat and dinnowhat else. all clients except Konversation are connected via ZNC
<kerio>
konversation is about as good as mirc
<DocScrutinizer05>
wfm
<DocScrutinizer05>
mostly
rblanco has joined #neo900
rblanco has left #neo900 [#neo900]
<sixwheeledbeast>
mh does it matter.
ReqGame has quit [Quit: WeeChat 1.0.1]
<sixwheeledbeast>
hmm
* sixwheeledbeast
spotted the prepand timestamp option in ZNC :)
<DocScrutinizer05>
:-D
<kerio>
you should be using a client that speaks servtime
<DocScrutinizer05>
not only that, it even allows free format definition
<kerio>
;-;
<DocScrutinizer05>
kerio: huh?
<DocScrutinizer05>
what's a "client that speaks servtime" ?
<kerio>
basically, if your client asks for it, znc tags every message with the time
<DocScrutinizer05>
uhuh
<kerio>
and it's smart enough to not send the prepended/appended timestamp in that case
<DocScrutinizer05>
well, my client prolly doesn't, I'm happy with the inline tags for buffer playback, and local timestamps for the rest
<kerio>
textual 5 does all sorts of shit
<kerio>
it even supports the playback module
<DocScrutinizer05>
I'm a fan of the simpler solution if there's a choice between simple and sophisticated
<kerio>
to request buffers from the last known timestamp onwards
<kerio>
which is actually something you pondered about in the past, isn't it
<DocScrutinizer05>
yes, I did
<DocScrutinizer05>
should be a pretty simple straight forward syntax/semantics extension to the playback command
<kerio>
kinda, yeah
<kerio>
/msg *playback PLAY <channel(s), can be *> [<timestamp>]
<kerio>
there's the occasional repeated message once in a while
<kerio>
cuz it's not a perfect solution
<kerio>
(the perfect solution would be to assign an ID to every message)
<sixwheeledbeast>
I also have no issue with time of playback with time of msg after it.
Pali has quit [Remote host closed the connection]
<DocScrutinizer05>
the perfect solution would tag each message line with a client-specific ACK on successful TCP transfer
<kerio>
no that's not a good solution
<kerio>
cuz guess what
<kerio>
tcp does that
<DocScrutinizer05>
and playback only those msgs that are not tagged yet for the particular client
<wpwrak>
i guess what you really want is a message sequence number, then refresh the display according to message arrival. e.g., if you get 1-5 then 7-10 and later 6, then you'd first see the exchange without #6, but it would be inserted later. if hole are found, the client could also indicate how many messages are missing.
<DocScrutinizer05>
kerio: sorry you lost me
<kerio>
wpwrak: tcp != udp
<kerio>
it's a stream
<kerio>
DocScrutinizer05: you still have the problem of timeouts
<DocScrutinizer05>
err what?
<kerio>
and then all the other stuff that can go wrong that's not related to transmission
<wpwrak>
kerio: tcp is just a transport. you don't have to care about its detailed semantics :)
<kerio>
wpwrak: i care about the fact that you can't get 1-5 and then 7-10 without 6
<DocScrutinizer05>
tcp is a "secured" transport
<kerio>
you either get everything or it fails
<DocScrutinizer05>
it actually doesn't help for packet swaps afaik
<kerio>
wtf are you talking about
<DocScrutinizer05>
but other than that it should guarantee that the data arrived
<wpwrak>
kerio: yes, but then you have loss of connection and reconnect. your linear model of tcp only works if your connection is eternal
<kerio>
90% of the internet would stop working if data wasn't reordered transparently
<kerio>
no protocol is smart enough for that
<kerio>
except the ones that use UDP, for obvious reasons :>
<wpwrak>
well, tcp itself is, too :)
* DocScrutinizer05
thinks he lost context completely, waves and heads out
billtux has quit [Ping timeout: 244 seconds]
<Oksana>
((better specs)) Putting "faster processor, more memory" into Neo900 would also require larger battery. And then what would we have? Either a giant device, or sensors pushed out by boring CPU-RAM-battery triad. ((five layers of abstraction)) I strongly dislike that, yes. ((Snow)) Wish there was some here, too...
<Oksana>
Good morning, and Happy New Year! :-)
<DocScrutinizer05>
morning Oksana
<DocScrutinizer05>
5 layers??
<Oksana>
Just reading the backscroll on "inefficient programming". I wasn't here for whole weekend.
<DocScrutinizer05>
aah, thanks for more context. I forgot
<DocScrutinizer05>
obviously TCP guarantees transmission of data - resp tells sender about any failure that happened
<Oksana>
One missing feature in Google Translate: show original text side-by-side with translation. To make learning new language easier. Or, is there an alternative to Google Translate?
<DocScrutinizer05>
and it actually even has a sequence counter
<kerio>
ofc
<kerio>
how do you think it worked? :o
<DocScrutinizer05>
but only in one established connection
<DocScrutinizer05>
kerio: the question is: what did *you* think how it worked, that made you state it's not suited for tagging IRC messages as "sent"
<DocScrutinizer05>
in my book a TCP packet sent from IRC server to client is OK and obsolete as soon as client machine's TCP stack ACKs the packet
<kerio>
i'm not even sure that kind of stuff is available without underlying stuff in the protcool
<kerio>
*protocol
<DocScrutinizer05>
I'm absolutely sure it is
<kerio>
not in unix
<DocScrutinizer05>
hah!
<kerio>
hell you don't even get to see how your sequence of bytes terminated by CRLF is actually fragmented in packets
<DocScrutinizer05>
you say in unix a TCP/IP socket write may return OK even when it fails?
<DocScrutinizer05>
I have a hard time to believe the system will not assert an error to calling layer when a lower layer runs into a timeout
<kerio>
sure, a timeout
<kerio>
2 minutes later
<kerio>
or 30 minutes later
<DocScrutinizer05>
that depends
<DocScrutinizer05>
when you allow 30 min timeout for your TCP ACKs, you got a serious configuration problem on your system
<DocScrutinizer05>
you might have 30 min timeout on open idle TCP connections
<DocScrutinizer05>
though even that timeout usually is lower than 2 minutes
<kerio>
No indication of failure to deliver is implicit in a send().
<DocScrutinizer05>
hmm
<DocScrutinizer05>
that kinda invalidates the whole purpose of TCP
<kerio>
do you seriously expect every packet to require a roundtrip
<DocScrutinizer05>
anyway, I'm no expert on it, but it makes me wonder
<kerio>
how do you get a throughput higher than like
<kerio>
8192 bytes/second
<DocScrutinizer05>
please read how TCP works, receiver ACKs a sequence number and implicitly acks all previous packages by doing though
<DocScrutinizer05>
I.E. there may be multiple packets on the fly
<kerio>
yes
<kerio>
and the stack deals with that
<DocScrutinizer05>
doesn't mean sender stops sending
<kerio>
not the application
<DocScrutinizer05>
if you had a ACK timeout of 30min, the whole TCP ACK would be completely useless
<DocScrutinizer05>
sorry, afk
<DocScrutinizer05>
sick
<Oksana>
I am curious about other halves. But, I strongly doubt that: a). Jolla Phone has modem "isolated-separated" from main-processor; b). it would make any sense to make N[eo]900 OH-compatible (Jolla Phone is much thinner?). Anyway, I hope that hacker-bus is more functional-versatile than OH; OH just makes tinkering easier, more fashionable, for mainstream users.
<Oksana>
Get well soon. :-)
lexik has quit [Ping timeout: 265 seconds]
<DocScrutinizer05>
((<kerio> No indication of failure to deliver is implicit in a send(). <kerio> do you seriously expect every packet to require a roundtrip)) no, what I expect is a error message on connection teardown (due to final timeout/retransmission expiry) that tells higher layer about the number of successfully transferred bytes
<kerio>
so you have to keep track of all the shit you sent? ain't nobody got time fo dat
<DocScrutinizer05>
I thought that was exactly what we were talking about: keeping track of sent IRC msgs
<DocScrutinizer05>
anyway, o/
<kerio>
\o
paulk-collins has quit [Quit: Quitte]
<freemangordon>
kerio: hmm? what about fmtx-middleware?
<nadrimajstor>
I see that many projects opted for SoC with PowerVR GPU. Is there any alternative that does not involve SGX?
<nadrimajstor>
When I say "any alternative" I mean any suitable for neo900 and unique constraints of its production.
b1101 has joined #neo900
b1101 has quit [Client Quit]
<nadrimajstor>
Or to rephrase my question: Where can read about reasoning involved in choosing BB-xM?
<wpwrak>
nadrimajstor: the main constraint is compatibility with maemo. this doesn't leave much of a choice regarding the cpu. bb-xm and such are just consequences of this.
<wpwrak>
(compatibility with maemo) or rather what it has evolved into since. but as i understand it, that has still many dependencies on the 3730.
louisdk has joined #neo900
<nadrimajstor>
wpwrak: I'm concerned about the state of "having a fully open stack for SGX". Is it true that the situation is so toxic that no one would go there?