<qi-bot>
[commit] Werner Almesberger: libatrf: new function atrf_interrupt_wait for interrupt-driven wait for interrupt http://qi-hw.com/p/ben-wpan/e6f0a1d
<qi-bot>
[commit] Werner Almesberger: tools/lib/: added interrupt_wait support to atusb and atusb-spi driver http://qi-hw.com/p/ben-wpan/110ecf6
<qi-bot>
[commit] Werner Almesberger: tools/lib/atben.c (atben_interrupt_wait): complain about stray interrupts http://qi-hw.com/p/ben-wpan/9952a20
<qi-bot>
[commit] Werner Almesberger: tools/lib/misctxrx.c (flush_interrupts): don't fall back to reading IRQ_STATUS http://qi-hw.com/p/ben-wpan/76707d6
<wpwrak>
rejon: we should announce our FISL talks on the qi-hw list. do you think the hours are reasonably dependable now ?
<rejon>
i haven't seen mine
<rejon>
no idea
<rejon>
agree on announce
<wpwrak>
rejon: you have two talks, Sat Jul 02 13:00 Qi-Hw. Sat Jul 02 17:00 Open Clip Art
<rejon>
i have one more on the 30th i just found out too, but that is another story, about acawiki.org
<rejon>
ok, sure, sounds like they are set
<wpwrak>
rejon: a bit of an evil schedule
<wpwrak>
rejon: wow, 3 talks ! you'll be hoarse ;-)
<rejon>
yeah, its a bit out of hand
<rejon>
its ok, rather be busy than not
<xiangfu>
kyak: Hi
<xiangfu>
kyak: I committed the 'backfire' branch on both openwrt-package and openwrt-xburst
<xiangfu>
kyak: I think this time we using another method to rename the 'trunk' to 'master'
<xiangfu>
kyak: maybe git push origin trunk:master, then we don't need modify the server config file, (I remember last time we rename the branch by 'git br -m' cause some more works)
<rozzin>
cat urls.audio
<rozzin>
er.
<rozzin>
n/m :)
<rjeffries>
i HOPE YJE nOKIA n9 COMES TO MARKEY IN usa SOON. SWEEET. mEEGO BASED.
<wolfspraul>
ok, well I wish you good luck, we need those cases soon...
<wpwrak>
roh, rjeffries: (at86rf230) a crappy predecessor. good luck with all its race conditions ;-)
<roh>
its loads of details moving so many machines needing a lot of 'support framework' to operate afzer all
<wolfspraul>
I can imagine. talk is cheap, I know how much work you have there to really deliver a full case...
<wolfspraul>
I'll ping you again towards the end of the week or so
<wolfspraul>
as long as box, labels and other stuff are still moving you are not the last one :-)
<roh>
well.. i wish i could just focus on one thing every day.. but currently its a wild mix of everything to keep stuff going. quite annoying and tiring
<wolfspraul>
roh: btw, have you updated your m1 recently? it's getting quite nice to use now :-)
<roh>
but i guess as long as you dont buy over a hundred cases a month from me that will not happen soon ;)
<roh>
wolfspraul: nope.. no time yet.
<wolfspraul>
one by one. But I do need those 80 soonish...
<wolfspraul>
I will keep you posted, as discussed.
<wolfspraul>
on one hand there is no need to stress out while other parts are in the making, on the other hand I want to avoid that the entire project is stuck at the end because we wait for cases.
<wolfspraul>
so I balance between those two...
<roh>
sure.
<wpwrak>
roh: btw, if you want to go the 230 route, there's the ATMEGA128RZ series. avr, not avr32. built-in rf230. all going EOL, of course.
<roh>
wpwrak: ah. nope.. i didnt want to go anywhere. just noticed it in that shop
<roh>
wpwrak: btw.. atxmega vs atmega.. big diffence.
<wpwrak>
roh: ah :) yeah. nice. but already obsolete :) i'm quite glad we could switch to the 231. i had some qualms about the 230.
<roh>
atxmega is that thing between avr32 and avr. like a 'big brother' to the avr.
<wpwrak>
roh: don't know the xmega. fresher patents, i suppose ? ;)
<roh>
wpwrak: but good to know that the 230 and the 231 are that different
<roh>
wpwrak: its quite nice. has some really cool added features.. cleaned up the io pinning a bit.. faster, bigger. still really cheap.
<wpwrak>
roh: 230 has some bugs. like incoming packets overwriting old not yet retrieved packets. so you have to be very very quick. the 231 can avoid that.
<roh>
stuff like msample adc/dac and some irq framework which makes it nice for 'realtime' stuff
<wpwrak>
roh: then the 231 has the higher data rates. and they cleaned up some more things. ah yes, if you care about external RF PA or antenna diversity, then it's also the 231 and not the 230
<wpwrak>
roh: (xmega) hmm. so they dropped the fossils of avr. that would be a good move indeed ;-)
<roh>
its very similar. but different
<wolfspraul>
rejon: I find it so amazing how the networks in the US are racking up prices on wireless data now
<wolfspraul>
all but one have now removed the (never really) 'unlimited' data plans
<wolfspraul>
and the new plans are _expensive_
<rjeffries>
from am arduino board re xmega: "Event System" that enables the peripherals, in conjunction with the DMA controllers, to do a lot of "common" things without having to involve the CPU at all.  Aside from allowing higher performance if the CPU is awake, this also allows the CPU to spend more time in assorted SLEEP modes, consuming less power.
<wolfspraul>
someone wants to reap the rewards of their investment dollars
<wolfspraul>
a 2GB plan 30 USD / month, a 10 GB plan 80 USD / month (just reading about Verizon)
<rjeffries>
wo;s[taul what doe you pay in China for a data plan?
<rejon>
yeah, not sure wolfspraul
<rejon>
seems like they are getting cheaper to me
<wolfspraul>
which prices do you see?
<wolfspraul>
I was just comparing with atben, so atben can transmit 30KiB/sec, that means a little less than 10 hours for 1 GB
<wolfspraul>
if 1 GB at Verizon costs 8 USD now, that means atben can save you 20 USD a day :-)
<wpwrak>
wolfspraul: now we just have to bridge those few extra kilometers ;-)
<roh>
hm. can we just disable that feat of the bot? its 'not helping' having it flooding the channel
<kyak>
no, it's totally helpful
<kyak>
it only floods sometimes
<kyak>
during rebase
<roh>
not when its flooding lots of commits in that style. is really hard to read so everybody will use a browser of git client anyhow. one line with a link to that would be enough.
<roh>
instead of 2 screens full
<kyak>
xiangfu: i will now 'git co master; git pull', right?
<kyak>
roh: it's usually just one line
<xiangfu>
kyak: wait
<wolfspraul>
roh: we limit it to 20 lines now, we can limit it further to 10 or so
<xiangfu>
kyak: it should works fine. but I advice, 'git co -b temp; git br -D master; git co -b master origin/master'
<wolfspraul>
but it gives an impression of a git merge so I don't think it's completely useless
<wolfspraul>
back when we had 500 lines it was not so much fun :-)
<roh>
hrhr.. the bot wasnt kicked due to flooding?
<xiangfu>
kyak: make sure the 'master' is right. then you can remove the 'temp' and 'trunk'
<wolfspraul>
nah, the bot slows down
<wolfspraul>
it's a bot, it knows how to play the game...
<xiangfu>
kyak: 'git co master; git pull' will cause small problems, since there are some file leave by 'trunk' and not delete in you local disk, remove those files should works fine.
<kyak>
xiangfu: following your advice, something is not right...
<kyak>
i was in trunk branch, i saved it with 'git co -b temp'. Then deleted master 'git br -D master'. Then stashed my changed 'git stash'. Then checkout remote master 'git co -b master origin/master'
<kyak>
now i try to restore my stashed changed
<kyak>
*changes
<kyak>
by theere is a conflict
<kyak>
but there shouldn't be any conflict
<kyak>
because i went there right from trunk, which is master now
<xiangfu>
kyak: have you run 'git fetch -a' first?
<kyak>
nope
<xiangfu>
update your local origin/*
<kyak>
that's a good idea :)
<xiangfu>
it's needs run 'git fetch -a' first sorry
<kyak>
yep, i udnerstand now
<xiangfu>
or you master still the old :)
<kyak>
yeash
<xiangfu>
if it's works fine in your side, let's do this on openwrt-xburst, then send email to list.
<kyak>
ok, everything went fine
<kyak>
i will now remove local trunk and temp
<kyak>
xiangfu: will you remove remotes/origin/trunk as well?
<xiangfu>
kyak: in the server side, it's already removed, in locally, we need run 'git remote prune  origin'
<kyak>
xiangfu: i'll just add one lien to compile-openwrt-xburst.sh (to create failed_packages.txt)
<xiangfu>
kyak: you can compare the compile-openwrt-xburst-trunk.sh and compile-openwrt-xburst.sh  a little. then we just remove the compile-openwrt-xburst-trunk.sh.
<DocScrutinizer>
(<wolfspraul> it's a bot, it knows how to play the game...) LOL
<xiangfu>
kyak: then update 'compile-openwrt.sh'
<DocScrutinizer>
moinmoin everybody
<kyak>
xiangfu: btw, there is a git repo in ~/bin
<kyak>
xiangfu: ok, i think compile-openwrt-xburst.sh should be ready now
<kyak>
xiangfu: should i stop the current build and restart it with compile-openwrt-xburst.sh?
<xiangfu>
yes. let's do it
<xiangfu>
I think we should remote the entire openwrt-xburst.trunk-full_system, useless from now on.
<kyak>
i left the tree latest trunk-full_system, just for us to check if something goes wrong
<kyak>
oh, you mean the build directory itself
<kyak>
i suggest that we wait until the openwrt-xburst.full_system-06212011-1026 build is finished
<xiangfu>
I mean the ~/openwrt-xburst.trunk-full_system
<xiangfu>
the build image is fine
<xiangfu>
kyak: yes. that sounds better.
<wolfspraul>
DocScrutinizer: moin
<kyak>
xiangfu: i removed the link public_html/openwrt-xburst.trunk-full_system to avoid confusion.. Let's keep it locallly for some time :)
<xiangfu>
yes
<kyak>
so now we wait :)
<DocScrutinizer>
unrelated: Nokia announced their new phone N9
<DocScrutinizer>
no hw kbd, capacitive ts... pffff
<DocScrutinizer>
no uSD slot, double-plus-PFFFF
<DocScrutinizer>
now if at least it came with dual-sim
<DocScrutinizer>
no OTG, no hostmode - three extra loud PFFFF LOSER
<Jay7>
why they release it then? :)
<Jay7>
it have no difference with 1xxx series :)
<DocScrutinizer>
indeed
<DocScrutinizer>
they promised a "meego" devie for this year though
<DocScrutinizer>
considering to get a spare spare as long as there are some available
<DocScrutinizer>
especially since Nokia can't even repair them now just 18 months after rollout - they swap for an N8 when you send in a N900 for warranty repair
<kyak>
are they not produced anymore?
<DocScrutinizer>
obviously not
<kyak>
heh, why would they do that to N900?
<kyak>
i heard from many geeks they love it
<DocScrutinizer>
and seems they even missed to keep a reasonable stock for replacements
<DocScrutinizer>
I guess, due to several problems with quality etc, Nokia decided to call it broken by design and to discontinue and rollback
<DocScrutinizer>
they better came up with a N900-i and a improved more open maemo
<DocScrutinizer>
but meh, you know, Elopocalypse
<kyak>
so they gave a toy to a child and then took it away
<DocScrutinizer>
yeah
<kyak>
despite of the fact that the child doesn't really care if the toy is broken or not :)
<DocScrutinizer>
they gave a violin to a 4 year old, and when he grew out of it at age of 9 they refused to give a proper violin for grownups
<kyak>
i think my Nokia N73 is slowly dying and it's time i start thinking about replacement...
<DocScrutinizer>
rather refuse to even service the baby violin anymore
<kyak>
yeah, that's a kind of strange logic..
<DocScrutinizer>
""look, there's that nice synthesizer kbd from Redmond, go use that for playing music""
<kyak>
and Nokia surely lost some of their fanboys after coupling with MS
<DocScrutinizer>
quite some, yes
<kyak>
where did they go?
<kyak>
not to Samsung, obvisouly
<DocScrutinizer>
lost even more though of the maemo community when they deliberately put them down the 3rd time
<DocScrutinizer>
most went to andridiot
<DocScrutinizer>
some still cling to "true meego"
<DocScrutinizer>
aka #meego-arm
<kyak>
interesting.. like, running meego on HTC devices?
<DocScrutinizer>
it's all a giant fsckng deja-vu regarding what OM did
<DocScrutinizer>
nah, I guess they'll try to bring true meego handset UX to N9 now
<DocScrutinizer>
they are about (or did already) to rollout meego for N900
<DocScrutinizer>
called meego 1.2 CE
<DocScrutinizer>
for Community Edition
<kyak>
'CE' ffs
<kyak>
not the best choice of abbreviation :)
<DocScrutinizer>
wellknown Carsten Munk of Mer a leading head of that enterprise
<DocScrutinizer>
it's been DE for developer edition formerly ;-P
<DocScrutinizer>
is pondering to watch some threads on talk.maemo.org explode
<DocScrutinizer>
I bet you can't read as fast as they scroll
<DocScrutinizer>
the announcement been like 6h ago
<kyak>
it's interesting how fast it happens. I had no clue about Nokia development of N9/N950 just yesterday
<DocScrutinizer>
it's been in the rumours for like 9 months now
<kyak>
makes me wonder what would happens "tomorrow"
<DocScrutinizer>
nuttin
<kyak>
yeah, probably, i'm just not inside this subject
<DocScrutinizer>
except maybe Elop gets killed by dropping a stinking oistrich egg at him ;-P
<DocScrutinizer>
wonders where to get the rotten ostrich egg from
<kyak>
you should get yourself an alibi in case it happens for real tomorrow :)
<DocScrutinizer>
:-D
<kyak>
remember - public logging!@
<DocScrutinizer>
I'll log tomorrow as well, ( <- checks alibi)
<kyak>
they'd say you had used Eliza bot :)
<DocScrutinizer>
I *am* Eliza bot ;-)
<kyak>
--)
<Jay7>
I'm happy to see that SonyEricsson and Samsung are opening devices slowly
<Jay7>
at least no crypted loaders
<Jay7>
ah, I understand
<Jay7>
DocScrutinizer: they will say now that Meego devices are unpopular and close this line
<kyak>
thanks to crackers, companies now decide that it is better to keep it open from the beginning than to embarass themselves when it is cracked anyway
<DocScrutinizer>
Jay7: indeed the idea to roll out a meego device without hw kbd looks like another perfect step in a plan to kill off the whole "linux-on-a-phone" thing
<DocScrutinizer>
as somebody just mentioned at #maemo, you won't please hardcore maemo/meego-hackers with the N9, and it's highly unlikely users that never were interested in linux at all now will switch to that new "ecosystem" that has no apps, no competitor devices, nuttin
<kyak>
Jay7: yeah, i've read it. Can't help admiring the stupidity of Nokia representative
<lekernel>
Openmoko should just have looked like the iPhone
<lekernel>
so this would excuse price
<lekernel>
and make a very attractive device
<Jay7>
hehe
<Jay7>
time for NanoNote phone editions
<lekernel>
time to get rid of cruft like X11, pulseaudio etc.
<Jay7>
qt4 :)
<lekernel>
and make an outstanding GUI
<kyak>
asking the question like "What do you think would be the price for N9?" to the audience of students and then not being able to defend his grounds
<lekernel>
qt4 is still nowhere near the IOS GUI
<lekernel>
and it's very slow and bloated
<Jay7>
efl then :)
<lekernel>
enlightenment is on the right path, except that they have trouble getting things done
<lekernel>
yes
<lekernel>
:)
<Jay7>
well, toolkit is choosen :)
<Jay7>
people are talking about some developer device with HW keyboard..
<kyak>
last time i heard from bartbes, enlightenment coudln't work without X
<lekernel>
"The EFL begins with Evas, our canvas library. Because Evas is built on several different selectable engines (Linux FrameBuffer, DirectFB, X11, OpenGL/OpenGL-ES, QTopia, etc) the platform is extremely portable, which translates thru to all of the libraries built on top of it."
<DocScrutinizer>
Jay7: N950 won't go retail
<DocScrutinizer>
developer device only
<kyak>
ok, sounds promising.. Perhaps need to have a look
<Jay7>
then it will not work from battery
<Jay7>
I'm sure :)
<DocScrutinizer>
even an extremely rare one: only 250 available to community
<lekernel>
it should have some hw acceleration too... I don't know the status of it
<Jay7>
remembers qt greenphone
<DocScrutinizer>
yeah
<Jay7>
Elopocalypse, as you said..
<lekernel>
things like software scrolling are slow, even on modern embedded devices
<DocScrutinizer>
openmoko's "predecessor"
<bartbes>
kyak: EFL works, e17 doesn't
<kyak>
bartbes: but what can you do with EFL, without e17?
<jivs>
Greetings
<kyak>
also, how would it correlate with their "the platform is extremely portable"?
<kyak>
are these just words, again?
<jivs>
i'm stuck on "kernel panic- not syncing: VFS Unable to mount root fs on unknown-block(0,0) " msg
<bartbes>
kyak: develop applications ;)
<bartbes>
EFL is like gtk
<jivs>
how can i put it to usbboot mode again.
<bartbes>
and e17 is gnome
<jivs>
i tried to reflash it with my own images.
<kyak>
bartbes: ok, good point. I don't expect gnome to run in DirectFB :)
<kyak>
jivs: try powering it on while holding the 'u' key
<kyak>
and watch for dmesg
<kyak>
then you can use reflash_ben.sh to reflash
<jivs>
tried so.. doesn't go to boot mode. when i connect to usb cable. it boots up and the same msg again
<kyak>
bartbes: are you aware of any efl-based applications that can run in DirectFB (i.e. on Ben)?
<jivs>
dmesg: i am not sure about it.
<kyak>
jivs: hold the 'u' while pluggin in the USB cable (i.e. while powering on)
<jivs>
kyak, thank you
<jivs>
looks im in boot mode now. i will try to reflash it now.
<bartbes>
kyak: ehm, basically everything that's not the wm, supposedly :P
<bartbes>
anyway, edje is supposed to be a super-easy way to create applications
<bartbes>
scripted and all
<kyak>
bartbes: it's hard to believe that it is just e17 that doesn't work
<kyak>
i assume there are a lot of apps dependent on e17
<bartbes>
well, as mentioned, efl's graphics thingy has a directfb backend
<bartbes>
so only things that need direct X access (the wm) or opengl won't work
<bartbes>
or well, that's supposedly the truth
<kyak>
okay then... maybe we can try finding something usefull for Ben
<kyak>
and fast? :)
<bartbes>
well, as I said edje applications should work
<bartbes>
so perhaps if you manage to work out how that works you can write an SDL-less gmenu alternative
<bartbes>
it's been a while since I played with the EFL though
<wpwrak>
dvdk: bah, all fancy stuff. that's what you deserve for using bashisms ;-)
<dvdk>
:)
<dvdk>
but without bashisms, sh is hardly turing complete :)
<kristianpaul>
hum..
<wpwrak>
dvdk: perhaps the correct approach would the be to implement a turing machine in /bin/sh, then program that one. this should eliminate your worries, no ? :)