DocScrutinizer05 changed the topic of #qi-hardware to: Copyleft hardware - http://qi-hardware.com | hardware hackers join here to discuss Ben NanoNote, atben / atusb 802.15.4 wireless, and other community driven hw projects | public logging at http://en.qi-hardware.com/irclogs and http://irclog.whitequark.org/qi-hardware
pcercuei has quit [Quit: dodo]
dos1 has quit [Ping timeout: 245 seconds]
panda|x201 has joined #qi-hardware
panda|x201 has quit [Ping timeout: 248 seconds]
arhuaco has quit [Ping timeout: 272 seconds]
xiangfu has quit [Ping timeout: 256 seconds]
viric has quit [Ping timeout: 240 seconds]
viric has joined #qi-hardware
xiangfu has joined #qi-hardware
mirko has quit [*.net *.split]
mirko has joined #qi-hardware
panda|x201 has joined #qi-hardware
lekernel has joined #qi-hardware
kyak has joined #qi-hardware
rz2k has joined #qi-hardware
wolfspraul has joined #qi-hardware
panda|x201 has quit [Ping timeout: 272 seconds]
panda|x201 has joined #qi-hardware
dos1 has joined #qi-hardware
xiangfu has quit [Ping timeout: 272 seconds]
xiangfu has joined #qi-hardware
panda|x201 has quit [Ping timeout: 260 seconds]
<sanderr> That's cool! But how open is open?
<whitequark> I was more thinking along the lines of phoneblocks
<whitequark> which was the same thing but the consensus here was that it's dommed
<whitequark> *doomed
<whitequark> I'm still not convinced of technical feasibility but I wonder what is Motorola thinking
<sanderr> I'm reading the phonebloks website.
<sanderr> It's a reaction to the disposable world, and it emphasizes more on the sharing and app economy.
<sanderr> But I like the idea.
<sanderr> whitequark: I think Motorola wants to move into a niche market.
<whitequark> sanderr: I'm not sure Motorola wants to do anything. it's just their R&D division.
<whitequark> And I wonder how does that R&D division wants to solve stuff like mechanical wear, exceeding cost, slow interfaces, ...
<sanderr> Cheap (and probably not fully standard) copy parts.
pcercuei has joined #qi-hardware
<wpwrak> so they now have invisible pink unicorns at the ivory tower ...
<wpwrak> at least this speaks well of the financial situation of googlorolas blue sky research division. but they could try to be a bit more original. that ideas's been floated before, with very similar cad work.
<wpwrak> or one could say "cgi" or "sfx" instead of "cad". makes it sound a bit less serious :)
<sanderr> I like the way you put it. "googlorolas blue sky research division" :) But it's not very original. It's like Lego building blocks, but with a CAN (automotive) bus.
<wpwrak> "After the trip, we asked ourselves. [...]" hmm yes, that seems to be a not so uncommon meme, too :)
<wpwrak> such concepts have been proposed several times already. what they have in common is that they don't really make sense.
<wpwrak> e.g., if your goal is upgradeability, then you'll find that, in, say, two years, you'll not only want a faster CPU (really ?), but also a new modem that supports a new band that's been added in your area
<wpwrak> and when you touch the modem, you're in the middle of RF compatibility issues. so that's the least modular part of all.
<sanderr> So all in all, the concept is perfect on paper, but useless in the physical world?
<whitequark> yeah
<larsc> I wouldn't call in perfect on paper
<wpwrak> i'd say so, yes. better flexibilize the manufacturing process and move from mass production to build-to-order. then you can have modularity at a higher level.
<wpwrak> heh ;-)
FDCX has quit [Read error: Operation timed out]
<wpwrak> let's say that it looks pretty on paper :)
<whitequark> that's for sure
<DocScrutinizer05> any network cracks around?
<DocScrutinizer05> I'd need some help on debugging a weird behaviour on 3 of my WLAN devices
<DocScrutinizer05> DNS resolving doesn't work enamore after several hours of device "idling"
<DocScrutinizer05> anymore*
<DocScrutinizer05> my network topology is a bit unusual: I got two NAT/routers between WLAN and the internet
xiangfu has quit [Ping timeout: 248 seconds]
<DocScrutinizer05> 2nd level router (the one giving DHCP leases) is set up to hand out sttaic public DNS server IPs
xiangfu has joined #qi-hardware
<DocScrutinizer05> /tmp/resolv.conf.wlan (dnsmasq specific way to handle DNS) has the correct public DNS server IPs on a device that refuses to ping ct.de
<DocScrutinizer05> ("unkown addr")
<DocScrutinizer05> disconecting/reconnecting WLAN fixes the issue for a day or so
FDCX has joined #qi-hardware
FDCX has quit [Read error: Operation timed out]
<DocScrutinizer05> the PC - connected via cat5 - on same subnet never experiencing any such issue, I can't rule out it's due to PC never "idles"
<sanderr> DocScrutinizer05: Does your wlan go into sleep mode?
<DocScrutinizer05> err sleep mode? it does PSM yes
<DocScrutinizer05> I never shut it down though
<DocScrutinizer05> to me it actually seems like NAT sessions closing eventually
<DocScrutinizer05> once device that is running IRC never goes offline for IRC, and never goes offline for ssh session. It however every now and then goes offline for SIP :-S .oO(???)
<DocScrutinizer05> and moreoften than not I find an occasional web browser I open telling me "server not existing"
<sanderr> Seems that it closes 'stale' network connections after a given time period.
<DocScrutinizer05> when I do a ping ct.de then, it also fails. Then I unregister/reconnect WLAN and everything (browser, SIP, ping) works again
<DocScrutinizer05> yeah, but shouldn't the local dhcpcd or whatever open a new connection whenever needed?
<sanderr> And if you keep pinging it in the background every 60 seconds? (as in, keep the connection open?)
<sanderr> DocScrutinizer05: I guess it should.
<DocScrutinizer05> route looked good
<DocScrutinizer05> /tmp/resolv.conf.wlan looked good
<DocScrutinizer05> ping 8.8.8.8 works
<sanderr> but your network map is: inet --> router1 -(dhcp)-> router2 -(static)-> device
<DocScrutinizer05> ping ct.de fails
<sanderr> Can you still ping router2?
<DocScrutinizer05> I even can ping the intarwebs
<sanderr> You should first find out which connection is dropped, the dhcp-one or the static one.
<DocScrutinizer05> it's strictly DNS that fails
<sanderr> What dns servers are given out?
<DocScrutinizer05> plus of course everything that depends on DNS
<sanderr> you can set up 8.8.8.8 and 8.8.4.4 on the client systems as dns servers.
<DocScrutinizer05> the statically configured ones
<sanderr> It's a workaround, but might work.
<DocScrutinizer05> I have statically configured DNS IP addr on router level2
<DocScrutinizer05> I don't see the difference when I configure local DNS IP
<DocScrutinizer05> the ones listed in /tmp/resolv.conf.wlan were exactly those I configured in rouer-l2
<sanderr> And you could still ping those?
<DocScrutinizer05> so it seems DHCP works, from device POV
<sanderr> So there was a cutoff on port 53 (dns).
<DocScrutinizer05> THAT is a good qustion :-)
<DocScrutinizer05> I only pinged 8.78.8
<DocScrutinizer05> 8.8.8.8
<sanderr> that's the one I always ping too. :)
<DocScrutinizer05> but not my DNS I'm using
<sanderr> try that next time, and try to find out how far you can get locally
<DocScrutinizer05> I wonder if friggin fritzbox (L1-router) is intercepting port53
<sanderr> ewwww... a fritzbox
<DocScrutinizer05> indeed
<DocScrutinizer05> not my decicion, not my domain
<DocScrutinizer05> I can configure it, but that's it
<sanderr> are there by any chance sip-clients behind it?
<DocScrutinizer05> I can't even install tellnet/ssh on it
<DocScrutinizer05> sure, mine
<DocScrutinizer05> 3 or 4 ;-)
<DocScrutinizer05> (SIP clients)
<sanderr> because what I know, fritzboxes prioritize their own (on the hardware) sip to be "more equal" than others (connecting through it)
<DocScrutinizer05> and SIP pretty much acting up since that friggin fritzbox came into my path
<sanderr> I know, they're a pain
<DocScrutinizer05> 3210, no SIP but seems some ALG
<DocScrutinizer05> I can't pinhole 5060 -> "internally used, sorry"
<DocScrutinizer05> >:-(
<DocScrutinizer05> but that's kinda plausible. What however is NOT is the DNS fsckup
<DocScrutinizer05> it's goddamn annoying
<sanderr> sorry that I can't help any further, but you need to investigate a bit more
<DocScrutinizer05> and strange
<DocScrutinizer05> thanks for the (basically obvious, but i missed it) hint to ping my actual DSN servers
<sanderr> connect your level2 router with utp to the fritzbox?
<DocScrutinizer05> DNS*
<DocScrutinizer05> yep
FDCX has joined #qi-hardware
<sanderr> well, fingers crossed!
<DocScrutinizer05> (inernet)<-->FritzBox<-->LINKSYS<---(WLAN)
<DocScrutinizer05> before there was no fritzbox in this picture, and everything fine
<DocScrutinizer05> err what?
<DocScrutinizer05> I miss the context there
<DocScrutinizer05> seems like answers to an unknown OP
<DocScrutinizer05> ohnoes, phonebloks again
<DocScrutinizer05> bhone-blokes
<sanderr> DocScrutinizer05: That's on Phonebloks and Motorola/Google going at it again.
<DocScrutinizer05> fools
<sanderr> We discussed that before the Fritzbox debacle. :s
<DocScrutinizer05> seems an idea needs a certain minimum of *idiocy* to really kick of a hype
<DocScrutinizer05> off, even
<DocScrutinizer05> well, when they increase the size of phoens err phablets further at same rate, we soon will be able to use PCIe cards
<DocScrutinizer05> THEN the whole modularity idea starts to become feasible, thoufg still nonsensical
<DocScrutinizer05> after all, look what PCIe card you got in your current PC (if you're still using a PC and not a laptop already)
<DocScrutinizer05> and particularly which cards are non-existent: decent coproc cards. RAM cards. Even decent GFX cards (need PCIe-x)
<DocScrutinizer05> but hey, maybe they can use a backplane (thry call it 'base'?) full of USB3 hubs ;-P
<whitequark> all your base are belong to us
<DocScrutinizer05> speedevil had a nice characterization for the phonebloks idea
<cde> whitequark: this joke expired 5 years ago
<DocScrutinizer05> [2013-10-25 16:03:44] <SpeedEvil> My response in another place - http://permalink.gmane.org/gmane.comp.handhelds.openmoko.community/63678 This is basically at the level of a 5-year-old trying to design a car.
<DocScrutinizer05> <SpeedEvil> 'Ok - it needs wheels and doors and a ball-pit'. Without the knowledge of what the transmission or suspension is.'
<DocScrutinizer05> >>Can it be done - sure!
<DocScrutinizer05> Will it be twice the price, twice the weight, a quarter of the reliability - very likely.<<
<DocScrutinizer05> I disagree. Very likely it won't fly at twice the size and price. I estimate at least 8 times the volume and 20 times the price
<DocScrutinizer05> just like those jolla-TheOtherHalf freaks suggesting to implement dual-SIM by simply placing a complete 2nd phone into TOH
<DocScrutinizer05> even if it could be done, you run into a hell of a RF-design nightmare, since two aerials won't work nice in such close neighbourhood to each other
arhuaco has joined #qi-hardware
<wpwrak> btw, for those not reading the list, here's a bit of y-box fun: schematics: http://downloads.qi-hardware.com/people/werner/anelok/tmp/ybox-20131025.pdf
<wpwrak> (y-box is the critter that combines a powerless usb host with usb power from some pc/wallwart/etc. and lets some usb device (keyboard or similar) plug into all this.
<DocScrutinizer05> those eagle schematics always are terribly ugly
<DocScrutinizer05> or is this KiCAD even?
<wpwrak> of course it's kicad
<wpwrak> and what's ugly about it ?
<DocScrutinizer05> dunno, the ground symbols, maybe the font, the colors, and particularly any wire ending in nirvana, with only a (electrically) irrelevant writing indicating what might be the deal with it
<DocScrutinizer05> a wire from a grounf symnol to a box labeled [POWERD]
<wpwrak> the latter is rather common if drawing the lines would be messy
<DocScrutinizer05> not in any professional schematics I ever seen
<wpwrak> well, POWERED is one of the less pleasant kicadisms, i agree. don't like it either, but it keeps the (otherwise useful) ERC complaints away.
<DocScrutinizer05> also not in DIN schamics symbol definitions afaik
<DocScrutinizer05> s/shamic/schematic
<wpwrak> the POWERED ? i'd be surprised if it existed in any standard
<DocScrutinizer05> no, wires ending in nirvana without *any* termination
<DocScrutinizer05> a decent CAD wouldn't allow such stuff fist instance
<wpwrak> naw, a real CAD checks what things you are trying to connect
<DocScrutinizer05> you either use a bus symbol or an offsheet symbol to terminate the wire
<DocScrutinizer05> offsheet symbols look like / similar to your [POWERED], and have to come in pairs
<DocScrutinizer05> as they are virtual invisible wires
<wpwrak> anyway, that's not the issue here. one thing that gives me a bit of a doubt is whether the micro-USB connectors should place the plug "up" or whether it doesn't really matter.
<DocScrutinizer05> hm?
<wpwrak> right now, i'm using non-reversed receptacles, which are intended for bottom-side mounting. since i use them on the top, the plug thus faces "downward"
<DocScrutinizer05> ohwell
<DocScrutinizer05> I dunno if anybody actually cares
<wpwrak> the plug's overmold has a top and a bottom side. the top side may get a USB logo or at least "A" and "B". the bottom side may be blank or have a vendor logo.
<wpwrak> yeah, that's my question :)
<DocScrutinizer05> well, *I* know that
<DocScrutinizer05> I dount anybody else does
<DocScrutinizer05> and even I don't 'use' that knowledge
<DocScrutinizer05> usually users do the try-and-error method on mating plugs
<wpwrak> yeah, the procedure is pretty much: take plug, try to fumble it into the receptacle, if it fits, done, if it doesn't, reverse and repeat, if it still doesn't fit, have a closer look
<wpwrak> indeed :)
<wpwrak> it gets a bit nastier with AB receptacles because they accept the plug in either orientation and it's only the plastic and not the metal preventing it from entering the wrong way
<DocScrutinizer05> I actually look at the *plug* of my usb charger before plugging it into the N900. But not at the symbols in plastic, but at the actual metal sleeve
<wpwrak> that's not a problem in this circuit but the password safe itself has an AB receptacle
<DocScrutinizer05> o/
<wpwrak> (to make things more interesting, the china-made micro-B cable i tried to attach has plastic from the overmold spilled under one of the lock "tongues", so even if inserted correctly, it was almost impossible to push it in, let alone get it out again)
<DocScrutinizer05> need to hurry
<wpwrak> (short-term solution: cut off that "tongue". now it moves properly. long-term solution: accept the lesson that they still don't know how to make decent connectors in china)
arhuaco has quit [Ping timeout: 256 seconds]
* whitequark switches to the channel and reads "cut off that tongue"
<whitequark> should I read backlog?..
<wpwrak> it's a piece of metal in a connector :)
<wpwrak> no preventive whistleblower treatment :)
dos1 has left #qi-hardware [#qi-hardware]
arhuaco has joined #qi-hardware
dos1 has joined #qi-hardware
wolfspraul has quit [Ping timeout: 240 seconds]
wolfspraul has joined #qi-hardware
rz2k has quit []
rz2k has joined #qi-hardware
rz2k has quit [Read error: Connection reset by peer]
rz2k has joined #qi-hardware
rz2k has quit [Read error: Connection reset by peer]
Codora has quit [Quit: om nom nom delicious bitcoins...]
jekhor has joined #qi-hardware
lekernel has quit [Ping timeout: 246 seconds]
lekernel has joined #qi-hardware
mth_ has joined #qi-hardware
mth has quit [*.net *.split]
jekhor has quit [Ping timeout: 252 seconds]
lekernel has quit [Quit: Leaving]
wolfspraul has quit [Quit: leaving]
arhuaco has quit [Ping timeout: 248 seconds]