#qi-hardware IRC log for Thursday, 2010-10-14

wpwrakhmm .. i google for "qi-hardware irc log" and get this: http://en.qi-hardware.com/wiki/IRC_Logs00:30
wpwrakcorrect but oh so misleading ...00:30
qi-bot[commit] Bas Wijnen: use proper place for 2 GB ecc http://qi-hw.com/p/iris/7e8cc1d10:12
qwebirc51684hello?13:25
qwebirc51684is anyone here?13:27
qwebirc51684hello?13:30
qbjectI am.13:32
qbjectqwebirc51684: What's up?13:33
qwebirc51684I just recieved my nanonote yesterday13:34
qwebirc51684trying to explore some possibilities13:35
qbjectThere are many! :) Anything specific?13:36
qwebirc51684one minor one is changing keymap to colemak13:41
qwebirc51684do you have one?13:41
qwebirc51684Hello Kris, this is Jane from Jlime boards13:48
qwebirc51684I got my nanonote yesterday13:48
qbjectI have a nanonote but I haven't tried to change the keymap, I'm afraid.13:59
qwebirc51684does anyone know if the outer layer is the screen or is it under a protector?14:01
qbjectI believe there are some assembly images on qi-hw.com which show the screen as separate from the glass in front of it.14:02
qbjectHere we go: http://en.qi-hardware.com/wiki/File:Cleaning_the_PC_Sheet.jpg14:03
Action: wpwrak wonders if this picture shows cleaning or glueing :)14:15
qwebirc51684nice picture14:20
qwebirc51684is it really glass?14:21
qwebirc51684Ok, have to go now14:23
wpwrakthe sheet is plastic14:28
wpwrak(cleaning) same tool as for cleaning the screen. i'll eat my words then :)14:29
qbjectwpwrak: in another forum, someone was asking if there are any barriers to user-disassembly.14:47
qbjectI haven't torn into mine yet, as I would miss it if I broke it, but I was curious, too.14:48
qbjectI assume the screen sheet is mounted with adhesive, but is there anything else which can't be taken apart and reassembled?14:48
zearhey there qbject, have you tried nethack-newt yet? ;)14:49
qbjectNope. Was busier last night than I'd expected. And we're having a pub night for a departing coworker tonight, so I might not get to it until tomorrow.14:50
zear;P14:50
qbjectBut I'll make sure it's installed today so I can at least try it on the train in the morning.14:50
qbjectI don't know how, but I think I broke tar.14:56
qbjectNo matter how small a file is, tar -x just sits there and does nothing.14:56
zearwell.. i don't know if dingux toolchain has a libtar, but maybe you simply overwritten it with the dingux one?14:56
qbjectI'll check, but I doubt it. I moved specific libs when programs missed them. I didn't replace anything that was already there.14:57
zeari see, then it shouldn't be the case14:58
qbjectMm.14:58
Action: qbject shrugs14:58
wpwrakqbject: (disassembly) yes, several parts:14:59
wpwrakqbject: the hinge14:59
wpwrakqbject: the LCM's pcb also looks dangerous to separate from the lcm (i.e., i haven't tried it)15:00
wpwrakqbject: the white metal part on the top cover will probably bend when removed (and be nearly impossible to straighten afterwards)15:01
qbjectwpwrak: It's bent already without attempting removal. :P15:01
wpwrakqbject: perhaps the little metal grill on top of the speaker as well15:02
wpwrakqbject: not sure about the top cover per se. i think it's glued too.15:03
qbjectAre there any retaining snaps/clips between the glossy and matte plastic?15:03
wpwrakqbject: i thnk they're only between the bottom shell and the frame that goes on and around the keyboard15:04
wpwrakqbject: as far as i recall, nothing on top15:04
qbjectHm.15:04
qbjectI'll probably have to get a second and sacrifice one.15:05
wpwrakbut then you have the hinges. i think they're not removable without causing damage15:05
wpwrakwolfgang will like that :)15:05
qbjectThe stuff you're doing with 3D scans is very (VERY) cool, but as a machinist I'd really love to put together a case design which doesn't require injection molding.15:06
wpwraki'd love to have that, too :-) after all, my 3D scanner can also mill, if given the chance :)15:10
qbjectI was curious about that. What variety of mill is it?15:11
wpwrakthe 3D scans are more thought as a reference. even if someone wanted to make a close replica, there are a few things you probably would want to avoid. i.e., anything that snaps15:11
wpwrakit's a roland "Modela" MDX-15. small desktop mill15:11
qbjectGotcha. I've heard of those.15:11
qbjectI had a MaxNC for a bit, but couldn't do as much as I'd have liked after learning on a big Matsuura. =D15:12
wpwrakthis one: http://www.rolanddga.com/products/scanners/mdx15/15:12
wpwrakmaxnc already looks quite professional15:13
qbjectSomewhat, but it's all aluminum. I'd rather have one of Sherline's CNC conversions if I'm going to have a dovetail mill.15:14
qbjectAnd compared with this: http://www.fox-prints.com/images/matsuura_small.jpg15:15
wpwrakokay, that's another kind of beast :)15:15
wpwrakmy roland can barely mill aluminium. if i go shallow and slow, it works. but takes forever.15:16
qbjectHave you tried brass? It's a ton more forgiving.15:16
wpwrakwhat's nice is that it's quiet enough that, when in a closed box, it can run 24/7 without keeping the neighbours awake :)15:16
wpwrakno, haven't tried brass yes. i would have thought it would be even harder ?15:17
wpwraks/yes/yet/15:17
qbjectThe matsuura was a great experience. Held almost .01mm all day on the Z axis after it had warmed up.15:17
wpwrak(0.01 mm) sigh :)15:17
qbjectIt is harder and heavier, but it also cuts cleaner and doesn't build up in the flutes. You can use carbide endmills and wind them up to about 4x the spindle speed of HSS.15:18
qbjectI've had endmills that were about to snap off in aluminum keep cutting nicely in brass for several parts.15:19
qbjectMade the brass parts of this after the aluminum ones: http://kahrl.net/portfolio/wand.jpg15:20
qbject(That was on the MaxNC, which did NOT like the aluminum.)15:20
wpwrakthe mdx-15 has a constant 6500 rpm. i wonder where there's a good place to get inexpensive endmills. pmtnow are nice but a bit pricy.15:21
wpwrak(brass) looks pretty clean15:22
wpwrakhow does brass behave without lubrication ?15:23
qbjectYep. We call that "360" brass in the US. Screw machine grade. Leaded, Freecutting. There is nothing more fun to machine.15:23
qbjectThat leaded stuff is basically self-lubricating and we only used coolant for deep-hole drilling. The rest of the time we just put the compressed air to it.15:25
wpwraknice :) i'll have to ask my metal shop about this.15:25
wpwrakalas, i don't have any cooling infrastructure. would probably be too noisy anyway. (apartment building)15:26
wpwrakbut then, i guess just going slow will do the trick, too15:26
qbjectActually, I would suggest keeping a small depth-of-cut and going faster. That'll help avoid chatter, which is the noisy part.15:27
wpwrakah, i see.15:28
wpwrakmy reaction to problems was always shallower and slower. perhaps i just made them worse then.15:29
wpwrakhow's burr with the brass ?15:29
qbjectPresent, but thin, work-hardened, and ready to break off.15:30
qbjectA small fine mill file is usually enough to remove it.15:31
qbjectOr one of those funny hook-shaped deburring tools.15:31
wpwrakokay, not too bad then. for flat things (pcbs and the like), i just use a chisel.15:33
wpwraklately, my mill gets to do little but pcbs and scanning :)15:33
qbjectI know what you mean about trying to go shallower/slower. And shallower is important with a low-powered tool. But chatter is best handled by feeding faster. You want each tooth taking a slightly bigger bite because the chatter is being caused by the end mill vibrating in the cut.15:34
wpwrakokay. i'll try that. now i really need to get new endmills :)15:35
qbjectIt's true.15:35
qbjectIn the US a lot of small operators find that Enco has competitive prices for small quantities. http://www.use-enco.com15:37
qbjectBut you're in South America, right?15:37
wpwrakyup, in argentina15:39
wpwrakso i need one that ships internationally15:39
wpwrakenco apparently do. good. let's compare prices ...15:40
qbjectYou might want to shoot an email to this guy: http://www.calocustomknives.com/docs/contact.htm15:41
qbjectBefore I started thinking about custom portable computers, I wanted to be a custom cutler. Met a lot of extremely helpful guys on BladeForums.com, including him.15:43
qbjectDunno if he posts there much any more, but it's a possible lead.15:44
wpwrak(ruben calco) he's got a nice workshop :)15:45
qbjectKnifemakers are an interesting bunch. In the US they tend to be quite conservative, quite capitalist, but that group is also extremely suspicious of trade secrets. They figure that if you won't tell people how you did it, you probably don't actually know that much.15:47
wpwrakhehe ;-)15:48
wpwrakbtw, nice laptop bag on your site15:49
qbjectThanks! I take it everywhere.15:49
qbjectzear: Hrm. Tried installing nethack-newt and got :Packages ... found, but incompatible with the architectures configured.15:59
methril_someone read this? http://www.fsf.org/news/endorsement-criteria16:07
wpwrak(enco) nice prices indeed. at least for standard sizes. for small stuff, they either don't have it or it's more expensive than at pmtnow :-(16:12
wpwrakhmm, how does a larger number of flutes change the mill's behaviour ? i.e., it it safe to go from 2 to 4 ? (so far, i've always used 2, but i see that for one of the sizes i want, 4 seems to be more common)16:22
qbjectConsider what I said earlier about the depth-of-cut per tooth and chatter. Twice as many teeth/flutes means that you can (or must) feed twice as fast at the same spindle speed.16:26
qbjectHowever, doing so does increase the load on the mill's servo or steppers16:29
wpwrakhmm, i was afraid of that conclusion16:30
wpwrakalso, my spindle motor is quite weak somethine like 10 W if i recall correctly16:30
qbjectMy personal opinion is the 2 flutes are great for prototypes and 1-offs. Adding flutes is more beneficial when you start getting into production efficiency.16:31
qbjectIt's good to prove that the program won't break cutters before you switch to more-expensive multi-flute endmills16:36
wpwrakironically, the 4-flute is cheaper than the 2-flute in this case16:37
qbjectHuh. Interesting.16:37
wpwrakstandard end mill, 1/32" size, 1/8" shank, 3/32" length (to have some tolerance when cutting through standard pcbs)16:40
wpwrakhmm, they call the #325-2462 a "stub" yet the length is supposed to be 3/32". how does that compute ?16:41
qbjectA depth-of-cut 3x the diameter still isn't that much...16:45
wpwrakisn't the relation quadratic ? pressure = x*diameter^2  (for constant depth/diameter ratio, same feed rate, etc.)16:55
wpwrakwhile break-off-pressure may only increase linearly. actually, i'm not sure about this.16:56
qbjectHm. Now you're getting into things I've forgotten.17:02
qbjectwpwrak: Some day we'll run EMC with a NanoNote and use it to make cases for other NanoNotes!17:13
wpwraknow there's an idea ! :)17:13
wpwrakare you planning to design a NN case ?17:14
qbjectIncremetally, yes.17:15
qbjectI'd like to start by just replacing the glossy bits with canvas garolite17:15
qbjectAnd go from there.17:17
qbjectIt's too bad there aren't more AVT2s.17:18
wpwrakgarolite ... hmm, isn't it quite brittle ?17:20
qbjectThat depends. They use the canvas and linen composites for things like pulleys on sailboat masts.17:21
wpwrakah, canvas reinforced ... "good mechanical and impact strength"17:21
qbjectBingo.17:21
qbjectAnd a design proved out in CE or LE garolite can then be transferred directly to Grade 10/FR4, which is nearly bulletproof, or Grade 30, which actually IS bulletproof.17:22
wpwraktaking your ben to bad neighbourhoods ? ;-)17:23
qbjectMy pocket seems to be a bad neighborhood.17:23
qbjectTake from that what you will. :P17:23
wpwrakso your ben is already in need of a new case ? :)17:24
qbjectLike I said, the metal trim is already bent.17:25
qbjectThere's some flex that I don't care for when it's closed.17:25
qbjectEtc.17:25
qbjectMy plan is:17:25
qbject1. augment the existing case17:26
qbject2. use lessons learned from that to design a new case with a massive hinge17:26
qbject3. design a new case with full-sized keyboard.17:27
wpwrakah, radical changes17:27
kristianpaulhello, (changes) radical indeed17:28
qbjectI know it's not free, but I was thinking that QTouch applied to a nice heavy piece of G-10/FR4 would make a nice, tough, easily replicated keyboard.17:28
urandom__full-sized keyboard?17:28
kristianpaullets wait touch screen get cheaper17:29
wpwraki'm thinking of making a replacement keyboard layout. same size, though17:29
kristianpaulfull screen no more :)17:29
kristianpauleasy case too17:29
urandom__nah is dont want any touch screen for nanonote17:29
wpwraklike this: http://downloads.qi-hardware.com/people/werner/tmp/layout.pdf17:30
wpwrakget rid of arrow, function keys, etc.17:30
wpwrakah, and CAPS LOCK, king of the useless :)17:31
urandom__how is that an improvemet to the current? (exept having no caps lock)17:31
qbjectkristianpaul: I agree with urandom__. I'm disappointed with the touchscreen performance on my Android phone with a lot more cycles/second than my nn. I'm very happy to keep it gui-less. I just want to make it last forever, and be comfortable to use if it will last forever.17:31
wpwrakurandom__: wide space, wide enter, no arrow17:32
urandom__so better suited for non-asian hands, i see17:32
wpwrakurandom__: also, simpler shape, with the fn row integrated into the 5x10 matrix. also reduces the number of keys and thus may free some i/os17:32
wpwrakurandom__: (less asian) yeah ;-)17:33
qbjectI had a thought about sticking two nn mobos in one case, one set up as a server, the other as a terminal and text-renderer. I wondered if I could drive a larger screen with that arrangement.17:33
urandom__we just need some kind of pointing device/trackpoint for the nanonote, much better than any touch screen17:33
wpwraki think a touch screen would be nice. you just can't treat it like a mouse17:34
wpwrakqbject: (ben-for-display) i'd try just driving the big screen with a single ben. maybe memory bandwidth is sufficient. well, depends on how big you go.17:35
wpwrakqbject: alas, display-heavy applications tend to do little else, so there may not be such a big speedup when having two bens17:36
wpwrakurandom__: (pointint device) trackpoint ?17:36
urandom__wpwrak pointing stick like some good notebooks have it http://upload.wikimedia.org/wikipedia/commons/thumb/0/0a/Mouse_pointing_stick.jpeg/800px-Mouse_pointing_stick.jpeg17:38
wpwrakwhat's nice about a touch screen is that it doesn't eat extra space and that you can shape the gui elements underneath. what's bad is that it's low resolution and that it gets the screen dirty.17:38
wpwrakyeah, that's the track point. i love those.17:38
qbjectwpwrak: I wasn't thinking about display-heavy apps as much as making an extremely pleasant text interface with nice font rendering.17:39
qbjectA handcrafted open machine designed for long-term relationships with the written word.17:45
wpwrakheh :)18:00
viricHallo18:03
Action: qbject greets viric18:04
viricI managed to get some system as a 'distributed feeds reader'18:04
viricand I can read offline all the rss feeds in the nanonote18:04
viricThis changed the nanonote to something much more useful for me :)18:04
wpwrakqbject: implementing your suggestion of going fast. now my mill is ripping through fr4 at 0.5 mm/s instead of 0.1 mm/s :)18:11
wpwrakmaybe i should even go faster. i've seen it jam into the piece at 10 mm/s by accident and nothing untoward happened. at some point, the adhesive tape holding the small pcb may come off, though.18:12
qbjectIndeed. Tape shear strength is a consideration. :) Not sure how things change with a glass laminate like FR4, but a lot of exotic metals have to be cut aggressively, even when they're really hard to work (like titanium and high-alloy tool steel.) You have to keep cutting ahead of the work-hardening or it eats your tool.18:15
qbjectBut glass is very wear-resistant, and it takes it out on the cutter. So one would think you get more work out of each tooth if it has to cut through the workpiece fewer times.18:17
qbjectHow's that tape holding up? (I really hate it when workpieces come loose. It's almost as scary as crashing the head into the table.)18:18
qbjectwpwrak: (which I did on that big red matsuura once. :O )18:19
wpwrakwith the small mills i use, a workpiece coming loose tends to have the same result as the head hitting something that won't yield ...18:21
wpwrak(big mill crash) did the folks around you applaud ? ;-)18:22
wpwraki went to 1 mm/s now. burr has increased but tolerable. the tape is doing fine so far18:23
qbjectHehe. They did cheer, actually.18:25
qbjectIt was a lot funnier than the time that someone left a half-meter of ~1cm rod sticking out of the chuck of a lathe and spun it up to 6000rpm. It went off-center and bent almost instantly, becoming the world's larges and most dangerous eggbeater.18:26
wpwrakworkpiece shifted a little. but at a place where it doesn't matter.18:28
Action: qbject nods18:28
qbjectAre you shaping the FR4 or cutting away copper cladding?18:28
wpwrak(eggbeater) nice ! a good thing to know if someone you don't like is near the machine. oh so tragic accident :)18:28
wpwrakshaping the pcb. i haven't had much success with cutting just the copper so far.18:29
qbjectAnd what are you doing to keep from filling your apartment with silicosis-inducing dust?18:29
wpwraklemme see if i can find the picture that shows how i prevent this ...18:30
wpwrakah no, you can's really see the box ... http://en.qi-hardware.com/wiki/File:Roland_MDX-15_scanning_nanonote.jpg18:34
wpwraki made a box for the mill. so most of the dust just stays in there (and noise as well)18:35
qbjectGotcha. Certainly better than nothing. :)18:44
wpwrakyup. besides, it's not as if i'd be cutting pcbs all the time18:48
qbjectNo indeed. You need time to test/break them in between!18:49
wpwrakyup. paticularly when it's break before test :)18:49
kristianpaullooks this http://www.fsf.org/news/endorsement-criteria21:55
kristianpaul"they say the hardware must run free software on every layer that is user upgradeable" tricky ;)21:56
wolfspraulyes21:57
wolfspraulif apple would remove iOS upgrades the iPhone would qualify as well21:57
wolfspraulto be honest, it's idiotic :-)21:57
kristianpaul:p21:58
wolfspraulwe need to chart our own course into new territory, the fsf won't help. well, she helps us keep some confused people away from our project which is good so we waste less time in useless discussions.21:58
wolfspraulkristianpaul: I'm serious. If a 'hardware maker' makes some functionality non-upgradeable, the fsf will endorse the product!21:59
wolfspraulthe fsf actually promotes the locking down of proprietary software21:59
kristianpauli know :(21:59
wolfspraulbut it must be locked down so that the fsf can endorse it21:59
kristianpaulwhut?? : The following arrangements can all be discussed:22:01
kristianpaulUse of the FSF name in your press releases and advertising, including quotes from FSF executives22:01
kristianpaulWTF is that thing?22:01
Action: kristianpaul sigh22:02
kristianpaulyou are right wolfspraul :)22:02
wolfspraulthat's normal. in a press release you want quotes.22:02
kristianpaulok..22:02
wolfspraullet's say you have a product, and on the launch date you want to have a quote from rms in your press release22:02
wolfspraulso basically they say when your product is endorsed, we will help you with a quote22:03
kristianpaulhmm22:03
wolfspraulthat's just a technicality of press releases, but journalists will want to see it...22:04
kristianpaulsounds fair, but i dont this FSF is the best entity to qoute for hardware related things22:04
wolfspraulhe22:04
wolfspraulcorrect!22:04
kristianpaulis confusing22:04
wolfspraul:-)22:04
kristianpaul:D22:04
wolfspraulunfortunately in hardware, the fsf is a bit confused (or rather 'defocused') to distinguish between open and free22:04
wolfspraulalthough the difference between open and free is their speciality in software22:05
kristianpaulindeed22:05
wolfspraulso in software they promote free software, in hardware they promote open hardware (i.e. proprietary hardware with open APIs)22:05
kristianpaulI wonder wht thet came up with this endorsement criteria..22:05
kristianpauls/wht/why22:06
wolfspraulthey want to focus on free software and the gpl22:06
wolfspraulso they need a clear line where hardware starts22:06
wolfsprauleven if that line is an illusion, they will just draw it and start evangelizing22:06
kristianpaulfair point for they22:06
kristianpaul(just draw it and start evangelizing) i hate that22:06
wolfspraulI think after a strong copyleft hardware movement and maybe organization is established, let's say in 10 years, the fsf would also be able to give up this nonsense and endorse real copyleft hardware, and only that22:07
wolfspraulbut for now they endorse open hardware, fine...22:07
wolfspraullike I said, the tendency to lock down more and more software will help them, in a weird way, to find more products they can endorse :-)22:07
wolfspraulthe hardware makers just need to make it impossible to upgrade proprietary software22:08
wolfspraulthen the fsf can call it 'hardware'22:08
kristianpaullockware22:08
wolfspraulhe :-)22:08
wolfspraulYEAH!22:08
wolfspraulgood word22:08
wolfspraulfsf endorses lockware22:08
kristianpaul:p22:08
wolfspraulnew: proprietary software is locked down, so you can feel good using it22:09
wolfspraulI think we should not annoy the fsf with this, it hurts our common cause.22:09
kristianpaulsure :)22:09
wolfspraulthe fsf is doing very important work in software and the gpl22:09
wolfspraulwe are trying in hardware, let's see how far we get...22:09
kristianpaul:)22:10
wolfspraulkristianpaul: 'lockware' is really great, I'll start to use it... thanks!22:25
wpwrakwolfspraul: we can help each other out with quotes ;-)22:27
kristianpaul:-)22:34
kristianpaulgn922:34
kristianpaul822:34
wolfsprauln822:35
--- Fri Oct 15 201000:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!