#qi-hardware IRC log for Monday, 2013-05-20

wpwrakfirst ups-induced system failure (with this ups) :-(03:13
wpwrak"Welcome to Ubuntu 13.04" ... "New release '13.04' available. Run 'do-release-upgrade' to upgrade to it." hmm ...03:22
wpwrakbut at least they solved the X server regressions. i have three screens again :)03:23
lekernelwho's going to lvee?08:58
viricwpwrak: what was the ups problem?09:04
lekernelthere's also http://conference.vde.com/emlc2013/Pages/EMLC2013.aspx just before09:27
whitequarklekernel: what?10:05
lekernelare you going to LVEE?10:05
whitequarklekernel: hm10:06
whitequarkI have about as much knowledge of litography as it is described on the wikipedia page10:06
whitequarkit's probably not worth it10:07
whitequarkplus, 600 euros... definitely no10:08
lekerneldo you really think I'm going to pay that? :)10:08
lekernelacademic conferences have zero security10:08
lekerneland I have no ethical problem dodging such entrance fees - they are just ridiculous10:09
lekerneljust the typical crap I¬¬¬ & co. are doing10:09
lekernelwas also thinking about http://lvee.org/10:10
whitequarkthat at least seems like a topic I can discuss :)10:11
lekerneloh, you actually learn a lot attending conferences outside your domain :)10:12
whitequarklekernel: I'd still like to have a reason of attending10:12
whitequarkright now, I'm neck deep in programming languages...10:12
lekerneleven if it means asking stupid questions during cocktails (which of course you can indulge in as well without a ticket, especially if the conference costs over 1000E)10:13
wpwrakviric: there was a little power hickup (probably < 1 s). ups tried to fill in. so it switched over, but then the ups-provided power collapsed, resetting my workstation. all the other systems (not connected a ups) didn't even notice the brief interruption.10:14
whitequarkwpwrak: I think that's called "precondition failure"10:14
whitequark$ ups blame # => wpwrak10:14
viricwpwrak: ok, I thought it was more a software issue10:15
whitequarklekernel: though. it occured to me that a fair amount of attendees would probably be clueless (of technical details) business people.10:24
wpwrakonce upon a time, the imposters pretended to be doctors doing live-critical medical procedures, or pilots flying around the world (see Frank Abagnale). nowadays, they sneak into expensive conferences ...10:24
lekernelwhitequark, that actually becomes an advantage when gatecrashing. then just engage the BS-filter.10:25
larscthe filter or the canon?10:27
wpwraka BS generator seems to be most useful in such situations10:29
wpwrakelse, you'll just stand in a corner without anyone to talk to10:29
wpwrakwell, that diminishes the risk of detection10:29
whitequarklekernel: yeah yeah, that's what I'm saying10:32
lekernelthe difficulty is switching quickly between filter and cannon. so you get to suck info about lithography or whatever topic the conference is about, while still going undetected.10:33
lekernelyeah, that thing you use when somone asks you what professor brought you there10:37
LunaVoraxHello everyone14:18
LunaVoraxThe Qi news hasn't been updated for two years, is Qi dying? What's happenning? Isn't there anything interesting for the past two years?14:20
lekernelin milkymist yes14:24
lekernelhttp://www.vjunion.se/2013/05/the-mixxeo-hdmidvi-mixer-project/ for example, based on a new SoC architecture14:25
LunaVoraxSo why the RSS fee of Qi-Hardware.com isn't updated about it, do you need a hand?14:28
lekernelyeah, if you have access to it, please do14:30
LunaVoraxI don't14:30
LunaVoraxAlso, I don't know if Qi-Hardware is reday to take some suggestions about an eventual Ben Nanonote 214:32
lekernelI don't think anyone is working on the NN2, and Wolfgang is MIA14:33
lekernelmissing in action14:33
LunaVoraxoh ok14:33
larscI wouldn't say he's missing, but I think he has a lot of other things on his hands these days14:59
wpwrakLunaVorax: the most important input regarding a Ben2 would be money ;-) i think the technical side is quite understood at the moment, but nobody has the resources to actually implement it.15:55
lekerneluse kickstarter! not a solution? ;)15:56
LunaVoraxlekernel is right imho, have you never thought about using kickstarter wpwrak?15:57
LunaVoraxIt'll be worth a try15:57
wpwrakkickstarter is producing things you already have. you still have to finance the prototype.15:58
wpwraks/ is / is for /15:58
wpwrakyou'd also need someone who manages the finances, etc. and with kickstarter, at least some of that work has to be done in the US.15:59
lekerneland producing a palatable video and content is itself a major effort15:59
wpwrakadd the PR effort to making the prototype.15:59
wpwraki'd say making the prototype would cost something in the order of USD 100k-200k16:00
lekernelit has to go through the curators, then have appeal to the masses who provably use fucked up criteria to judge projects16:00
wpwrakwell, there's that risk as well. but since you're already moving a considerable amount of money, you could probably include a first run as well. keep the BOM cost low, of course. then you may be able to do a run of 2000-3000 units by raising your budget by 50%.16:03
larscyou think there is a market for 3000 units?16:04
wpwrakso that means no expensive experiments and moderate specs. main objective would be to avoid mistakes like not having USB host, putting in a better (wider) screen, getting control over the mechanics. no bleeding edge cpu and such. that only wastes resources.16:04
lekernelwhy not? rpi sold millions (-:C16:05
wpwrakthere's probably a market for  a lot more if you spin it right.16:05
lekernelor ouya, or ...16:05
wpwrakif you have full control over the design, you can also sell it as a building block. you can also offer fabrication services - combine smaller orders for modified designs. and so on. there's a lot of things you can do once you've mastered the ownership barrier.16:06
wpwrakalso, work the DIY market a bit more. sort of like a more effective alternative to arduino + WiFI/BT + smartphone16:07
wpwrakof course, with the cost of the smartphone = 0, that's a bit tricky16:08
lekernelhow many people do arduino + WiFI/BT + smartphone?16:08
lekernelthis sounds like behind headed toward marketing failure16:08
LunaVoraxNot to be mean or anything, but everytime the discussion of gathering funds is brought on the table, there's always a good reason NOT to do anything and the conclusion always end up with something like Qi-Hardware won't produce anything ever or something like that.16:10
wpwrakall those who need a reasonably fancy control interface and don't know how to build one from the ground up16:10
LunaVoraxNot that I don't understand, but it always sounds like Qi is doomed or something16:10
wpwrakLunaVorax: your impression may not be incorrect16:11
LunaVoraxIt's a disaster then imho16:11
lekernelQi lacks some "cool" factor, yes16:11
wpwrakqi-hw ran off wolfgang's money to finance product development and a lot of "free" engineering time by various volunteers16:11
LunaVoraxDefine cool factor lekernel 16:11
lekernelbeing capable of producing mediocre devices like rpi and then selling millions16:12
wpwraknow that wolfgang is done (and probably close to broke) and people don't have much time anymore, things don't look particularly good16:12
LunaVoraxIs copyleft hardware a doomed idea then?16:13
LunaVoraxI mean in a liberal capitalist society16:13
LunaVoraxliberal as in liberalism, not "left"16:14
lekernelcopyleft is a factor orders of magnitude less important than marketing16:14
lekernelas far as money is concerned16:14
larscthe rpi probably would have sold as well as it does even if it had been copyleft hardware16:16
LunaVoraxMoney over freedom&16:16
LunaVoraxlarsc is right I think16:16
LunaVoraxStill it should have stayed as cheap16:17
lekernelthat's a more positive way of sayning it ...16:17
larscbeing copyleft is not really a killerfeature16:17
wpwrakcopyleft is less of a game changes in hardware because of the inherent cost of doing things16:18
larscso your product needs to be good16:18
LunaVoraxWhat makes it good at the end is the software that runs it16:18
lekernelwpwrak, even outside of hardware. software support for rpi or, as I was noticing the hard way yesterday, ftdi-chip is far from stellar16:18
wpwrakin sw, it radically changes the cost structure. in hw, it's at best a major reduction of a cost that's still large.16:18
wpwraklekernel: well yes, that's one of the costs of non-open systems :)16:19
LunaVoraxI always think about the GCW-Zero, which is a pimped-up dingoo, but on the end what makes it great is the GNU/Linux distribution that it runs16:19
wpwrakLunaVorax: the problem is that most of the "cool" software can also run on a proprietary base16:20
wpwrakLunaVorax: the value of free software is in the longer run. e.g., being able to keep your 3 years old device supplied with new stuff16:20
lekernelmixxeo cannot :) but of course it's a niche market16:21
lekerneldoesn't mean it won't work money-wise, though16:21
wpwrakas long as your planning horizon is < 1 year, it's very hard to convince you of free software. and with < 1 year, you're also not someone who builds things16:21
wpwrak(of free sw in relation to hw, so freedom-friendly hw, but not necessarily open hardware)16:22
LunaVoraxWhat heppens next?16:24
lekernelas far as I'm concerned, make a device that cannot be easily done with traditional proprietary parts, and get the UX/design/product part right this time16:27
lekerneland get out of the hacker/floss-directed stuff, it's a niche market *and* they are cheap bastards, so ...16:35
LunaVoraxcheap bastards?16:36
LunaVoraxAnyway can you generally precise your thoughts lekernel?16:37
lekernelwho's buying a LEON ASIC? hmm? it's a GPL'd chip :)16:37
wpwrakof course, if you require openness to be an enabler, the hacker/floss corner is quite important16:38
lekernelyes, all the "open source" hardware for hackers that sells is merely tiny microcontroller kits16:38
wpwrakyou can try educational, but that's tough market16:38
wpwrakhacker/floss is also important to bring down development cost. if you have to pay for everything yourself, it gets a lot more expensive16:39
LunaVoraxThis all "market" thing seems fake imho. For exemple the Rpi was designed for the "educationnal market" but now everyone is having one16:39
LunaVoraxVersatile products have versatile customers16:39
wpwrakyou still need to reach the people who buy them :-) but yes, sometimes a product meant for a given group can reach a lot more people16:46
LunaVoraxI'm sure that if the Ben2 is well designed and advertised it could reach a lot of people16:47
wpwrakbut you need that group you're reasonably certain of for securing the initial funding. that is, unless you finance yourself exclusively through crowdfunding. but then your focus determines how you present the product.16:47
wpwrakso you can't simply dodge that question16:47
lekernelit's versatile because it's cheap, and it's cheap because broadcom sells huge quantities of the SoC on that breakout board to put them in mobile phones etc.16:47
wpwrakwell, picking an inexpensive chip isn't that hard. even the ben did that ;-)16:48
lekernelhad rpi been designed entirely by the rpi foundation, including the chip, the cost per unit would probably be a lot more even with one million sold16:48
wpwraki don't think it makes sense to design asics for this sort of devices16:49
wpwrakmaybe that was something feasible in the days of the sinclair zx80 or maybe even the commodore amiga, but i doubt you can usefully compete with ingenic, allwinner, or even freescale16:50
lekerneltherefore copyleft asics are for expensive niche markets, and expensive is incompatible with most of the hacker community16:50
LunaVoraxSomehow& I always thought that phone-shape would be really cool for the Ben2 https://globalnewsbulletin.files.wordpress.com/2010/10/nokia_c3_03.jpg16:50
LunaVoraxReplacing the "phone" buttons by other label16:51
wpwraknon-clamshell would make the device a bit bulky. but yes, it would simplify the mechanical design.16:53
wpwrakthe plain shell would be fairly easy to make. any moving parts (keyboard and hinge) are where things get trickier.16:54
LunaVoraxAdding usb-host, tv-out and wifi would make the ultimate pocket UNIX computer for me16:54
wpwrakhaving full design control over the keyboard would of course be very attractive. that's where customization is likely to be needed first.16:54
LunaVoraxMaybe I should just get a Nokia C3 and start hacking it :o)16:55
wpwrakTV ?16:55
LunaVoraxTV/Video out, whatever name you want to call it16:55
wpwrakah, VGA :)16:55
wpwrakwell, the ben sort of has that ;-)16:55
LunaVorax"sort of"16:55
LunaVoraxThe GCW-Zero has everything the Ben2 should have, except it's not shaped like a Ben16:56
wpwrakyou could improve the existing code. you can probably make a good enough frame buffer that runs in the background. so it would be like a "real" display.16:56
LunaVoraxNot right now, I'm a terrible terrible C coder16:57
LunaVoraxAnd a disastrous UNIX coder16:57
wpwraksounds like a great learning opportunity :)16:58
LunaVoraxHope you're right17:02
paul_boddieIt might be nice if we could still channel some of the news and other goings-on through the qi-hardware site, though. When it is up, it looks almost static now.17:10
LunaVoraxAnd the "planet" link on the wiki doesn't work for me17:12
paul_boddieNo, I think the planet has been switched off.17:13
paul_boddieIt was a bit heavy on the bandwidth, though. One huge, multi-month page with images every time is going to eat up bandwidth.17:14
larscpaul_boddie: so you just volunteered for the editor-in-chief position for QI News? ;)17:17
paul_boddieWell, I do surf the Internet a bit, but lekernel's recent announcement was completely new to me. I guess I don't read enough Swedish VJ news to pick up on the really exciting stuff. ;-)17:18
paul_boddieHave any of you been following the Rhombus Tech project? They actually seem to have working hardware now.17:20
larsclekernel: see, you need to work on your PR strategy17:20
LunaVoraxIt's too bad, the planet link was very interesting17:21
wpwrakyou'd have to hang out on #milkymist to see what's going on in the somewhat parallel universe17:21
LunaVoraxWhat do you mean wpwrak 17:22
LunaVoraxBesides I don't own a Milkymist17:22
wpwrakthat was for paul_boddie. milkymist-related things are mainly discussed there.17:23
paul_boddieRhombus Tech: http://rhombus-tech.net/allwinner_a10/news/ (warning: improperly sized images load slowly in page)17:23
paul_boddiewpwrak: Indeed. I only ever followed the qi-hardware channel.17:23
larscthat's like a rpi but with a a10?17:24
paul_boddieThis is pretty interesting, too: http://rhombus-tech.net/community_ideas/games_console/17:24
LunaVoraxWhat is this paul_boddie17:24
LunaVoraxWhy PCMCIA?17:24
paul_boddieWell, the idea is that they use a standard form factor like PCMCIA for upgradability reasons, so it's not completely like the RPi where people have to build their cases out of Lego and where the hardware is yet another board to be badly supported for the minimum possible time.17:26
paul_boddieIf you want a RPi with A10 or similar, you could just go to Olimex and get something from them given that they've been doing that kind of thing for a while now. Or get the Cubieboard, or whatever.17:26
wpwrakkewl. that pcmcia connector alone probbaly costs more than the whole pcb should :)17:26
paul_boddieI guess they know about the pricing trade-offs. The assertion is that it's still mass-produced and lkcl seems to be pretty price-aware.17:27
paul_boddieWe had a discussion about this a while back (maybe on the list), but people didn't think it would work for NanoNote 2.0. I guess that games console concept would indicate whether it would or not.17:28
wpwrakwell, it's unshielded. so that's half the price. still a rather expensive item to have.17:30
paul_boddieI guess you'd have to read the front page of their site to get the full rationale for using PCMCIA. I think it's good that someone wants to explore this route, even if there's going to be a lot of naysaying about it.17:31
wpwrakwell yes, if you do an exhaustive search on the state space, that guarantees that you'll find all the optimal solutions. so far, so good. but ... :)17:32
paul_boddieIf they manage to revive the KDE tablet idea with that work then just that, all by itself, may have been worth it.17:34
paul_boddieIs the Milkymist IRC channel archived, by the way?17:49
lekernelyes, http://en.qi-hardware.com/mmlogs/17:50
paul_boddieI was looking for a link on the front page or the Milkymist page, but not intensively.17:50
whitequarkalso http://irclog.whitequark.org/milkymist17:52
paul_boddieAs far as what LunaVorax was saying, I think things have changed quite a bit since 2009 or so, and lots of people are trying to make devices/gadgets now.17:58
paul_boddieIt's not exactly all happening in a coordinated way, which I imagine was one of the motivations for Qi-Hw.17:59
paul_boddieAnd so there's a lot of stuff getting funded and made that probably is what people want, but it perhaps doesn't lead to very sustainable further development.18:01
paul_boddieIn other words, you get yet another Arduino/RPi clone being made, with yet more hardware differences between the last variant and the new one, and everybody has to learn (or find out) the differences, upgrade their stuff, and so on.18:03
LunaVoraxNow maybe we should do something so that nice discussion we had won't fall into the trashbin of history18:13
larsc'we should do' is a bit of a euphemism for 'yea, ok, it will probably never happen'18:19
wpwraki think that's a rather accurate translation :)18:24
hellekin(the French translation)18:29
larscthat doesn't sound french18:29
Fallenoucontraction of "il n'y a qu'à, il faut qu'on"18:30
hellekinthank you Fallenou :)18:31
hellekinI was struggling with my French layout ;o)18:31
Fallenou-> "you only have to, we need to"18:31
Fallenouhehe :)18:31
whitequarkhuh huh, openwrt got a great lot better since 201018:44
whitequarktheir new web interface is amazing18:44
whitequarkyou can even configure wlans on the integrated switch18:44
larscluci has been around for quite some time18:46
larscIt was even there in 201018:46
pcercueiwhitequark: can you transfer files to/from the device on their web interface?18:51
whitequarklarsc: in 2010 it was almost unusable18:52
whitequarkpcercuei: I don't think I ever wanted that18:52
whitequarkdoesn't seem so18:53
whitequarkbut they finally have SCP18:53
LunaVoraxOk that was a seriously offending translation18:54
LunaVoraxI'm genuinely concerned even if I don't have time to chat on #qi-hardware everyday18:54
LunaVoraxWhy so much cynism?18:54
LunaVoraxOn the other hand if there's no hope, explain to me, and that no joke, why #qi-hardware is still active18:55
wpwrakthink of it as the neighbourhood pub where you go for a beer after the day's work is done18:57
wpwrakonly that no beer is being served on #qi-hw, and you can tune in anytime you like18:58
whitequark(except that he didn't stomp on anyone, he fell from the sky. flakey translation.)19:02
wpwrakrussians are weird :)19:04
whitequarkit's very absurd and funny. and I like her voice :)19:04
larscLunaVorax: It's not cynism, if you want to get something done do it yourself and don't wait for others19:27
paul_boddieWell, I see the point being made, which is that people are fairly downbeat about stuff.19:59
paul_boddieThere's not too much general enthusiasm any more, at least not compared to other groups of people. That's the impression one might get.20:02
wpwrakpaul_boddie: i'd say that everybody involved so far has done what they can do. and it wasn't enough for the project to become sustainable. so choices are to burry the remains, perhaps keep the good bits while that's possible (e.g., downloads and the repositories are rather handy), or bring in people who can breathe new life into this project.20:37
paul_boddieFor a strict definition of the project as being concerned with the Ben, I guess I have to agree: the engine has gone cold, as it were.20:39
paul_boddieI guess the lack of any particular focus at the moment, and the lack of energy in finding a new focus, means that there is going to be a sense of hibernation about the whole thing.20:41
wpwrakregarding the ya / next ben, i have a pretty clear idea of what to do, minus some things that may count as original research, e.g., the keyboard. all that needs is financing. that's something i'm not good at organizing. so there's my stopping block.20:41
viricIn general, I never understood how can people group to do something together20:41
viricin my experience, either I do it, or noone will. :)20:41
paul_boddieI don't see why various groups don't work together: I see them struggling, and it would be so much easier for them to team up. The only problem is the existing investment everyone has.20:42
wpwrakthe ben's engine has been cooling for a long while. i tried to put a bit of spark into it at the beginning of this year, with the various ubb projects, but the response was a bit underwhelming. paul_boddie, the medal of the only one caring to even react to all that pretty much goes to you :)20:42
wpwrakviric: a lot of people like to follow strong leaders. so that's the image you have to project.20:43
paul_boddiewpwrak: Well, I did like your different ideas at making open hardware projects interesting and accessible to people.20:43
viricI either fail at giving that image, or I didn't find such people :)20:43
wpwrakviric: i think the blame's on you :) the planet is full of followers ...20:44
viricdamn it, I'm not a leader.20:44
paul_boddieSay what you like about lkcl, for example, but the guy is stubborn. That means that his PCMCIA initiative will either succeed or fail in style, but he seems to be relentless.20:44
wpwrakviric: of course, an effective leader also has to be slightly psychotic. maybe you're just not demented enough :)20:45
viricwpwrak: I've to acknowledge that I'm quite alone in *using* the software I developed. For me it's a must-have, and very rarely other people like it :)20:45
viricwpwrak: ah yes, that's a very plausible explanation ;)20:45
wpwrakpaul_boddie: i think the question is whether he'll run out of resources before a decision point will be reached20:45
wpwrakviric: lack of mass appeal could create some difficulties :)20:46
paul_boddiewpwrak: The guy is stubborn, as I said - it's almost personal for him now, I think. He also might have made the right connections to secure sufficient resources.20:46
viricwpwrak: right. I'm not trying to lead a project about a banana keyboard20:46
paul_boddieWhat he's doing isn't really unusual: lots of people seem to be making their own boards these days. But if he can pull in enough domain experts to get that KDE tablet out there, it will demonstrate a capability at a different level.20:47
wpwrakpaul_boddie: well, it depends on what he's selling. if he's selling those modules, i'm sceptical. if he's selling the design behind them, i think he's in a much better position.20:48
wpwrakselling = offering and having a buyer20:49
paul_boddiewpwrak: It's both. The controversial thing about it is that he wants people to make their own, which is good, but his organisation supposedly reserves the right to the concept via a couple of patents, which is not good.20:49
wpwrakhah ;-)20:49
wpwrakwell, investors like patents.20:50
wpwrakbut they're also good at cutting off dead branches20:50
wpwrakso ...20:50
paul_boddieI'd be supportive (not just reporting what he's been up to) if he made the licensing status clear and reassured everyone that they aren't going to get sued if they adopt his proposed standard.20:51
paul_boddieBut up to this point such reassurances have been intangible and vague, when they actually need to be concrete and specific as we all know.20:52
wpwrakin fact, i wasn't aware of those issues. my scepticism comes from the potential merits of the project even before considering those evil bits20:52
wpwrakso he's basically trying to trap people. oh well, the industry is full of that20:53
paul_boddieI think a modular standard makes some kind of sense. We've had it in various forms over the years, and you could probably get a lot of mileage out of it for lots of devices before it might become too inflexible.20:54
paul_boddieI think that there's even the idea that it should be a family of form factors, not just PCMCIA. That is just a pragmatic choice of the first generation of gadgetry, I believe.20:55
paul_boddieBut anyway, with regard to things like leadership and financing, I don't think we lack people doing open hardware who can do these things. For example, the Golden Delicious people doing the GTA04 clearly have the ability to finance something which can be regarded as comparable to the Ben.20:56
paul_boddieAnd similarly, you have the OpenPandora people. And the GCW-Zero people. And the Rhombus Tech people, although arguably in a more opaque fashion.20:56
wpwrakyes, they have a working business that can finance their adventures20:57
wpwraki don't know about how sustainable the others are20:57
paul_boddie(I was doing some UBB stuff yesterday, but ran out of steam. Despite having a strategy, I think my electronics skills are lacking and/or my target was a bit too ambitious, and I started to wonder what I was really going to achieve, anyway.)21:00
paul_boddieBeyond a certain point, there's only so much someone can realistically do for themself, though, if they are really interested in some project succeeding.21:47
paul_boddiePeople can only bring what they know to the table, and perhaps commit to learning other things, but that doesn't get a project with lots of different aspects going strong, let alone finished.21:48
--- Tue May 21 201300:00

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