<qi-bot> [commit] Werner Almesberger: Put scans on index page in top-down order and added placeholders for still http://qi-hw.com/p/ben-scans/3b14f67
<wpwrak> kristianpaul: looks like a start :) how thick are your walls ? the whole case seems very tall.
<DocScrutinizer> server move accomplished?
<kristianpaul> wpwrak: wall thickness is 1mm i hope yeah is kind of tall but is just 13mm :)
<kristianpaul> may be i can draw a side of the nanonote and you'll see how it looks next to him
<kristianpaul> s/him/it
<wpwrak> 13 mm seems HUGE. with 4 mm walls, you shouldn't need more than about 4.5 mm: 1 mm top wall, 1.2 mm clearance above PCB, 0.8 mm PCB, 0.5 clearance below PCB, 1 mm bottom wall. maybe even less clearance.
<kristianpaul> ahh was my fault i tought that 1 in your case.fig was 1cm ! (what was i thinking??)
<kristianpaul> i'll fix to night and make test print :)
<wpwrak> also, i think that huge it would be even above the keyboard level.
<kristianpaul> yes :)
<wpwrak> oh :) naw, it's the micro-SD card :)
<kristianpaul> how is that? wpan pan share ports with a sd card also?
<wpwrak> the one that goes into the display won't need any extra case :) but that one's for the future. right now, everything interesting can be already tested with the one that goes into uSD, plus it doesn't need any freakish assembly :)
<kristianpaul> lol
<kristianpaul> sure :)
<wpwrak> the atusd variant connects via the uSD signals. it goes directly into the uSD slot: http://downloads.qi-hardware.com/people/werner/wpan/atusd/atusd-20100908-inserted.jpg
<wpwrak> you can't share the port with a card, though. there can be only one ;-)
<kristianpaul> sure sure i misunderstand ( naw, it's the micro-SD card :))
<wpwrak> qi-bot has an identity crisis
<kristianpaul> lol
<kristianpaul> qi-bot6: ping
<rafa> tuxbrain_away: hola
<rafa> estas ?
<rafa> tuxbrain_away: queda por subir los siguientes free software games: tecnoballz, powermanga y opentyrian
<kristianpaul> what's wrong with qi-bot?
<wpwrak> my guess would be that it's not registering with nickserv, so everytime it tries to change its nick, it gets changed back
<wolfspra1l> wpwrak: yes correct
<wolfspra1l> I'm on it, need to improve the tcl scripts a bit
<wpwrak> kewl
<wpwrak> btw, how did the 1000 units run go ? everything smooth ?
<wolfspra1l> I have a question about the usd breakout board
<wolfspra1l> sure, all fine with the run
<wolfspra1l> I'll send a mail once I have more data
<wolfspra1l> for the pulyfuse, you said we should pick a current rating after we know how much we can draw
<wolfspra1l> so how much is that?
<wpwrak> (run) great ! in a previous life, i wouldn't have dared to use the word "sure" in the same sentence as a successful run :)
<wolfspra1l> your url pointed to a 200mA fuse
<wolfspra1l> can we make an educated guess what would be a meaningful value in our case?
<wolfspra1l> and also not limit the breakout board maybe for other devices unnecessarily
<wpwrak> the limit is a good question :) it would be the maximum current the power supply can deliver, minus the typical peak consumption of the regular ben
<wolfspra1l> would a capacitor between VDD and GND make any sense?///
<wpwrak> i was hoping carlos and/or adam would have a good estimate of the current situation.
<wolfspra1l> ah good idea I can ask Adam, although he will probably refuse to give me a number until he did all calculations :-) he knows I will just go ahead 'produce something'...
<wpwrak> i woldn't add any extra capacitors without need, considering that what's already there is a bit troublesome
<wpwrak> ;-)))
<wpwrak> i don't know to what extent the ben's circuit was engineered by sharism and to what extent it was just copied from the original design. so this can be anywhere between adam just polling the numbers out from design notes to him actually having to do the whole set of calculations and measurements.
<wpwrak> well, or someone else :)
<wolfspra1l> maybe no fuse then for now
<wolfspra1l> I will ask Adam in parallel
<wolfspra1l> but adding a nice polyfuse and not really knowing which current rating it should have really looks stupid.
<wpwrak> the latter is quite a bit of work, if you remember the endless stories about characterizing the current consumption of the gta02. that's why i was hoping some design data was already around.
<wpwrak> ;-)
<wolfspra1l> guessing the value also doesn't make sense, either too low or too high would devalue the fuse quite a bit
<wpwrak> i think the power supply should survive shorting just fine. the problem is that it will take the system down.
<wpwrak> there could be secondary effects when overloading SD power, e.g., exceeding USB current, though.
<wpwrak> some surge protection may be nice to have, too. after all, these signals so get exposed this way, and they run straight to the CPU.
<wpwrak> of course, surge protection ought to be inside the Ben, not on the peripheral. but it would still be better than nothing.
<wolfspra1l> if I have no number for the current rating, I can't add a fuse
<wolfspra1l> but you are right, I will ask Adam
<wpwrak> yup, that's the best way to start. if he doesn't know, we can still do it the hard way.
<kristianpaul> hmm i need a 50mA fuse
<wolfspra1l> kristianpaul: on the breakout cable?
<kristianpaul> wolfspra1l: not for the EVB just for initial tests
<wolfspra1l> I hope qi-bot0 is a safe temporary name for the night :-) I'll go about fixing the actual script tomorrow...
<bartbes> kristianpaul: you were the gforth expert, right?
<bartbes> (of course I may be entirely wrong, but you should at least have mentioned it once :P)
<kristianpaul> bartbes: no
<bartbes> hehe, see?
<kristianpaul> actually i was asking about gforth but no more the other day
<kristianpaul> yes i mentioned it
<kristianpaul> looks interesting
<bartbes> well, it is not something worrying anyway
<bartbes> I just figured it was a fun fact
<kristianpaul> :p
<bartbes> apparently gforth thinks it's 16th of january 1970
<kristianpaul> yeah
<kristianpaul> interesting now?
<kristianpaul> now/no
<kristianpaul> no/inst
<bartbes> very
<bartbes> now
<bartbes> I need to find me something to program in forth that is more useful that rot13 :P
<kristianpaul> if foth support SDL or someway of usehte the display will be cool i think
<wpwrak> if the forth fanatics are progressing like this, wolfgang can make the ya with 32 kB of RAM ;-)
<kristianpaul> lol
<kristianpaul> :)
<kristianpaul> my milkymist avnet board is 2 kB
<kristianpaul> so..
<kristianpaul> :)
<wpwrak> alright. that's the kind of platform where forth would even make sense nowadays :)
<kristianpaul> wpwrak: ahh the case need to be C like due the wall could not be so thin to fit the small space between the PCB and the ben
<kristianpaul> C not the language
<wpwrak> of course, at the same time, we have opkg that needs swap space before it will even run. that seems to be a rather worthwhile target for a bit of optimization ;-)
<wpwrak> (c-shape) yup :) don't design what you can't build ;-))
<kristianpaul> we also have the linux monolitic kernel, but wait there is irisi, why not look at it !
<kristianpaul> iris*
<kristianpaul> or minix ;)
<kristianpaul> at the end the nanontoe should behave like other home appliance so..
<kristianpaul> bartbes: how good is going forth self-compilation in the nanonote?
<bartbes> hmm?
<bartbes> self-compilation?
<kristianpaul> yup
<bartbes> lol, I was asking what you meant
<kristianpaul> is in wikipedia article about forth
<rafa> wpwrak: I will kill opkg
<wpwrak> rafa: that's a worthy challenge ! :) dragons are so old-fashioned. you, valiant knight, are to slay an opkg :)
<wpwrak> kristianpaul: locking the case/board in place will also be a bit tricky. a very simple case would just be some sort of sleeve. but it would either hold the board by friction or the board would have to be glued inside
<kristianpaul> wpwrak: locking could be achieved by holding on of the edges of the PCB i think
<wpwrak> kristianpaul: so you're saying a single part, with friction lock ?
<wpwrak> kristianpaul: or two parts ?
<kristianpaul> wpwrak: two parts
<kristianpaul> yeah friction lock is the right word
<wpwrak> much easier :)
<rafa> wpwrak: I need to know what opkg does :)
<rafa> you need to know well your enemy first :D
<wpwrak> n = count_packages(); for (i = 0; i != 1 << n; i++) malloc(random());   ?
<rafa> malloc(random()) ; exit(1);
<rafa> wpwrak: another possibility could be void main(void) { printf("error"); exit(random()); }
<kristianpaul> rafa: you are evil ! :)
<wpwrak> rafa: perhaps you meant   return ((int (*)(void))({ long foo = random(); &foo; }))();   ?
<rafa> kristianpaul: and you tell me that I am evil?? ! :)
<rafa> wpwrak: ;-))
<kristianpaul> runs away
<kristianpaul> hehe look this http://boredzo.org/helloworld/
<wpwrak> kristianpaul: sigh. when will people learn proper use of operator precedence and omit redundant parentheses ... (zoo == 0) ? foo : bar; /* blargh */
<bartbes> so gtk+ should work without problems?
<bartbes> right?
<rafa> bartbes: on fb?
<bartbes> hides from the cops
<bartbes> rafa: why?
<rafa> mmh ¡?
<rafa> I just tried to understand where gtk+ should work without problems
<rafa> surely you are talking about openwrt
<bartbes> oh on framebuffer
<bartbes> hahaha
<rafa> but I wanted to be sure
<bartbes> yeah, I meant on the nn
<bartbes> default openwrt
<rafa> what did you understand?
<bartbes> facebook
<rafa> ??????????'
<rafa> :P
<bartbes> so, it should just work, right?
<rafa> no idea.. I read that qi had a gtk+ lib which works on fb, so maybe yes.
<bartbes> I thought stardict was just gtk+
<bartbes> but yeah, I wanted to know sure before actually trying
<bartbes> it would be a waste of effort otherwise
<rafa> I also guess that nn does not like gtk2 much..
<rafa> stardict : maybe it is sdl?
<rafa> no idea
<wpwrak> rafa: why should it not like gtk2 ?
<rafa> wpwrak: i have tested only.. and it is so slow.. IIRC, gtk2 libs do a lot of floating point calculations.. but I could be wrong
<wpwrak> oh, fp would be bad
<bartbes> :(
<rafa> hehehe.. I was talking for a while with a supertux developer.. and I modified a bit the game :)
<rafa> now.. enjoying it on nn ;)
<kristianpaul> rafa: yes?
<kristianpaul> ohh
<kristianpaul> rafa: i tried some day in jlime but was so slowww
<rafa> kristianpaul: I did a new build.. no the OE package
<rafa> kristianpaul: the speed is great
<rafa> with sound as well
<kristianpaul> good :)
<kristianpaul> what is wrong with OE packages?
<kristianpaul> package**
<rafa> nothin
<rafa> nothing
<rafa> supertux is not a game for 320x240
<rafa> the OE package is okey
<rafa> and it builds the proper current supertux
<kristianpaul> i see
<rafa> that is why I was talking with a supertux developer
<rafa> so he guided me to use other sources
<rafa> and to modify a bit those
<kristianpaul> btw , have you tried frozzen bubble in jlime?
<kristianpaul> i not sure if is on the repo
<rafa> no yet..
<rafa> just the games I gave tuxbrain
<rafa> can you play frozzen bubble using 320x240? (on pc)
<kristianpaul> hmm
<kristianpaul> i dont have 320x240 on my PC
<kristianpaul> but is giess if i have will be too small indeed
<kristianpaul> i just plaied  once ina  nokia/s60 phone was great :)
<kristianpaul> 240 x 320 pixels, 2.2 inches
<kristianpaul> nokia 5700 :)
<kristianpaul> wich is not same as  320x240 ;)
<kristianpaul> what a problem
<rafa> I mean.. if you can run the game using some command line argument or config to test using 320x240
<rafa> that is the best before to try a game on nn
<kristianpaul> i see
<rafa> first to try to run the game on pc using a 320x240 window
<rafa> if the game does not have some argument to set that size
<rafa> then that is not a good idea to try on nn..
<kristianpaul> oh it uses SDL :)
<kristianpaul> but i dont see where set screen resolution afaik :/
<kristianpaul> the s60 version does but guess is too s60 specific..
<qwebirc54914> has anyone figured out how to get a working gui on debian with the nanonote yet?
<qwebirc6704> I as well as someone else wanted to know if anyone can run the gui with the debian on the nanonote