#qi-hardware IRC log for Tuesday, 2011-04-26

wpwrakrejon: ah, here's one upper bound for the resolution: the maximum MMC block size is 2048 bytes. so a line couldn't be larger than this many samples/pixels. of course, the real limit may be much lower ...00:53
rjeffries_MMC association closed up shop and merged into JDEC  http://www.mmca.org/01:36
wpwrakrjeffries_: yup01:43
wpwrakrjeffries_: you can still get the docs, though01:44
wpwrakrjeffries_: they make you "register", but they don't require a valid e-mail address01:44
rjeffries_cool beans02:17
rjeffries_your right angle connector for connecting a scope (or anything else) to UBB looks useful02:18
wpwrakrjeffries_: yeah, i'm already putting it to use :)02:28
qi-bot[commit] hauke: mac80211: b43: Add Firmware http://qi-hw.com/p/openwrt-xburst/21ca38e02:34
qi-bot[commit] hauke: mac80211: b43: Expose debug option http://qi-hw.com/p/openwrt-xburst/3c0289d02:34
qi-bot[commit] hauke: mac80211: b43: Expose PIO mode fallback http://qi-hw.com/p/openwrt-xburst/525ee4502:34
qi-bot[commit] hauke: mac80211: b43: Expose N-PHY support http://qi-hw.com/p/openwrt-xburst/f3a65e702:34
qi-bot[commit] nbd: busybox: fix issues with static routes and 6rd option parsing (thx, Florian Fuessl) http://qi-hw.com/p/openwrt-xburst/9649fa802:34
qi-bot[commit] nbd: busybox: refresh patches http://qi-hw.com/p/openwrt-xburst/3564c6702:34
qi-bot[commit] jow: [package] firewall: prevent excessive uci state data aggregation (#9152) http://qi-hw.com/p/openwrt-xburst/8d0ec8902:34
qi-bot[commit] jow: [package] ppp: prevent duplicate hotplug for non uci managed ppp interfaces (#9275) http://qi-hw.com/p/openwrt-xburst/920b7c502:34
qi-bot[commit] jow: [package] ppp: extend the r26742 change to ip-down too http://qi-hw.com/p/openwrt-xburst/2be8e6002:34
qi-bot[commit] nbd: mac80211: update to wireless-testing 2011-04-19, contains several beacon related fixes for ath9k http://qi-hw.com/p/openwrt-xburst/83ef4a502:34
qi-bot[commit] hauke: mac80211: remove unneeded patch http://qi-hw.com/p/openwrt-xburst/e7a096702:34
qi-bot[commit] acoul: linux/ar71xx: add 2.6.38 preliminary support  http://qi-hw.com/p/openwrt-xburst/dfba2f502:34
qi-bot[commit] acoul: linux/ar71xx: update wget2nand script http://qi-hw.com/p/openwrt-xburst/b7a26a702:34
qi-bot[commit] acoul: linux/adm5120: update wget2nand script http://qi-hw.com/p/openwrt-xburst/15b4a0602:34
qi-bot[commit] nbd: hostapd: update to 2011-04-21 http://qi-hw.com/p/openwrt-xburst/454a4d202:34
qi-bot[commit] nbd: mac80211: do rate control updates when the HT configuration changes on an interface http://qi-hw.com/p/openwrt-xburst/4e1f63302:34
qi-bot[commit] nbd: hostapd: fix a few compile errors and warnings http://qi-hw.com/p/openwrt-xburst/4feb07002:34
qi-bot[commit] nbd: cfg80211: revert upstream regdomain handling breakage http://qi-hw.com/p/openwrt-xburst/9df10a202:34
qi-bot[commit] nbd: ath9k: remove the signal strength fix, it causes a lot of confusion and seems to be just as inaccurate as the original version of the code http://qi-hw.com/p/openwrt-xburst/0318dc002:34
qi-bot[commit] nbd: mac80211: replace the regd revert patch with a proper fix, add some more pending patches http://qi-hw.com/p/openwrt-xburst/542d04302:34
wpwrakaha, we're getting a new kernel :)02:35
kristianpaulhttp://www.marxists.org/archive/lafargue/1883/lazy/index.htm :D02:35
xiangfuupdate the 'trunk' branch first. :)02:36
wpwrakkristianpaul: collecting reading material for the 1st of may ? :)02:37
kristianpaulwpwrak: he, no actually, i get to that link because of http://www.copenhagenfreeuniversity.dk/ <--  we have won!!02:44
wolfspraulyes very cool. we have won!02:48
wpwrak"This is why we haven’t published papers or dissertations to wrap up the research projects that we have worked with."  keeping a rather low profile, are we ? :)02:48
xiangfuakiwiguy: hi04:02
akiwiguyHi xiangfu09:47
xiangfuakiwiguy: Hi09:48
xiangfuakiwiguy: I have to reboot my system. back online later. :)09:53
kyako hi David09:56
kyaki'm trying to build the latest MPlayer09:57
kyakwhen ffmpeg checkout, i have this error: fatal: reference is not a tree: 3d462373f8e8fa58b697a9b2d93a5edf7084193709:57
kyakso far, i will download you copy of Mplayer http://mosquito.dyndns.tv/david/nanonote/MPlayer-r33304.tar.gz to proceed further09:57
dvdkkyak: huh?  why did it work for me? hmm.  cannot access my sources from work.09:57
dvdkkyak: going to look into this maybe tomorrow09:58
dvdkbtw what I catually came here to report: :)09:58
kyakdvdk: i read a bit, it seems such things can happend on big commit to git09:58
dvdkFinally somebody is fighting back against the MPEGLA maffia:09:58
dvdkkyak: any ideas how to prevent such a thing in the future?09:59
dvdkkyak: the tgz you downloaded is exactly from my openwrt/dl directory, so 100% matches my config10:00
kyakdvdk: i'm not sure that such things happen very often to worry about it..10:00
dvdkkyak: well from the sample we have, it happens 1 out of 2 times :)10:00
kyaki'm not a git superuser at all, i found this explanation:10:01
kyak"Oops, someone made a super-project commit that refers to an unpublished commit in the submodule sub"10:01
dvdkkyak: already gave up on comprehinding git :)10:01
kyakso we need to pick another commit :)10:01
dvdkkyak: feel free to search one that actually works :)10:01
dvdks/works/still works10:02
dvdkwolfspraul: look who joins the google-webm cross licensing: MIPs Technologies (see link above)10:02
kyakis it a fair assumption that the latest one shoudl work?10:02
dvdkkyak: the one i used, was the latest one from maybe saturday.  so just try.10:03
kyakdvdk: anyway, the build of mplayer from your sources has just finished successfully10:03
dvdki.e. "yes"10:03
dvdkkyak: makefile still missing code to add back proprietary codecs when build_patented set.10:03
dvdkhope you're not disappointed10:03
dvdkalso i removed mpeg1 support, mayb can add that back, since too old for patents.10:04
kyakdvdk: is OK for me as long as your driver works find and ogg is playable10:13
dvdkkyak: ok.  btw without the ffmpeg ogg demuxer, it only uses libtheora (not fftheora).  with libtheora some high-detail scenes jerk at 320x240, 25fps.10:16
dvdkkyak: so you might also want to try webm (seems to work, just the encoder is slower and not as well-tuned as ffmpeg2theora)10:17
dvdkmaybe recompiling libtheora with -O2 might help.  dunno10:17
kyakok.. are you using the same settings for ffmpeg2theora as you described in your e-mail? or something has changed?10:18
dvdkkyak: wrote two emails. first try was 256x144 with upscaling, which shouldn't be a problem, even at 30fps.10:18
dvdkhowever 320x176 may jerk, at least at 30fps.10:18
dvdkkyak: no , nothing has changed.  just that i did some more encodes with diffferent materials for testing over the last days.10:19
dvdkwolfspraul: saw the webm-cross-licensing they just started? http://www.webm-ccl.org/10:21
wolfspraulyes great10:21
wolfspraulit doesn't help much specifically (legally), but it builds up a certain 'counterforce'10:21
wolfspraulalways good10:21
wolfspraulthey got some nice partners too, like LG and Samsung10:22
dvdkwolfspraul: not sure whether they can use it for counter-suing.  nothing about that in there agreement.10:22
wolfsprauljust watch out - this kind of deal only means that those companies cross-license to each other :-)10:22
wolfsprauloh no10:22
wolfspraulthe typical patent troll totally doesn't care about this10:22
wolfspraulneither will mpeg la10:22
wolfspraulbut it's good for the press and general audience10:22
wolfspraul"lots of big companies - must be good"10:22
wolfspraulthat kind of thing10:22
wolfspraulI'm glad they do it, otherwise the other side has their loudspeakers on all the time...10:23
dvdkwolfspraul: even MIPS Inc joined the party.10:23
dvdkwolfspraul: well, they could turn webm-ccl into a pantent-pool for h264 :)  every licensee would have to promise never suing against webm users :)10:24
kyakdvdk: ok, great.. i hope i have some time to do some testing in next few days :)10:33
wpwrakroh: hmm, i have to leave joker.com ... is godaddy.com an acceptable domain registrar ? (just for almesberger.net)11:02
rohwpwrak: eeh.. dunno. i never bought domains that way11:05
rohi always told somebody and ever few years they ask some decent for money11:05
roheh.. the for needs to shl 2 words11:06
wpwrakroh: heh, outsourcing is nice ;-) ah well, i guess i'll just try my luck11:08
rohwpwrak: wait a moment.. asking him11:08
mthwpwrak: they weren't very popular in a recent slashdot discussion: http://yro.slashdot.org/story/11/04/22/0144213/IMSLP-Taken-Down-By-UK-Publishers-Group11:09
rohwpwrak: why do you have to leave them?11:09
rohgodaddy isnt good i hear... troublesome to get domains back when they got them11:09
akiwiguyGodaddy sucks.11:10
rohwpwrak: my expert says: 'stay at joker.com'11:12
wpwrakroh: almesberger.net is about to expire and renewal by credit card requires "verified by visa" or the mastercard equivalent. alternatively, i could wire the money, but for this, i'd have to register as a reseller, which adds even more bureaucracy11:12
rohi see. and you got no cc?11:12
wpwrakroh: i have visa and mastercard, but i don't want those extra verifications. because once you agree to them, a lot of online transactions get more complicated/unreliable, and the contract terms are changed as well11:14
rohwpwrak: i see.11:15
rohakiwiguy: irgh.11:15
rohwpwrak: how much time is left? maybe you can find somebody with a cc in europe to do the payment for you11:15
akiwiguyroh: Whats wrong with that?11:15
rohakiwiguy: dark side ;)11:16
wpwrakroh: about three weeks left11:17
rohwpwrak: i see.. maybe gismo can help. i think he got some cc11:17
rohi dont11:17
wpwrakroh: (payment via proxy) seems rather convoluted :) naw, i'll just take my business somewhere where they're not so difficult11:18
rohwpwrak: its a generic problem.11:19
wpwrakroh: it's the first relevant placed i've seen that doesn't offer sensible payment options. normally, you just pay with your cc with no further ado.11:20
rohwpwrak: cc billing for non-eu/us customers is hell.11:22
rohi can tell you that for sure after living through it with om11:23
rohSEPA rules btw11:23
wpwrakroh: dunno. never had trouble so far. worst case, they ask you to e-mail a photo of your passport.11:25
wpwrakwolfspraul: i see that you're using enom.com. are you happy with them ?11:27
wpwrakenom.com looks reasonably clean. good faq. no cheesy stuff like "domain auctions". well, they have "premium domains", so they're in the squatter biz, too11:41
wpwrakroh: (cc for non-eu/us) most of the weird things are connected to the issuing bank, not the card holder's country of residence. so with my swiss-issued cards, i have no trouble, where the argentina-issued one is basically useless abroad. both have exactly the same owner information.11:47
wpwrakroh: ironically, the cards from .ch are even less trouble inside argentina (here, they very often require shops to call in to verify a transaction. but they never do this with my foreign cards)11:48
wolfspraulwpwrak: enom is fine, but I have no moral standards, I think that industry stinks11:56
wolfspraulI find them expensive11:57
wolfsprauland yes they do lots of cheesy things, like everybody else11:57
wolfspraulI was seriously thinking sometimes about advertising IP addresses instead ;-)11:57
wolfspraulbut maybe that's a bit too radical...11:57
wpwrakwolfspraul: (ip addr) ;-))) may get tricky with e-mail11:59
kristianpaulhey no,  advertising IP addresses is not a bad idea if you consider that from ipv6 you may get some sense ip number to remenber easilly ;)13:50
kristianpaulhmm email13:51
kristianpaulhash based email ;)13:52
kristianpauli like that indea from the i2p network13:52
kristianpaulwhat gell this does?   :(){ :|: & };:16:00
larsckristianpaul: forkbomb16:13
larscdeclares a function namend ":" which calles itself twice16:14
kristianpauli tought the : was a null character16:16
kristianpaulwhat the ; for?16:16
kristianpaulah, just to close the brackets..16:17
kristianpaulhe, anywya16:17
xiangfukristianpaul: larsc make my system, as you see, I have to reboot.16:18
xiangfumake my system freeze16:18
kristianpaulah wikipedia have an article about it.nice16:21
larscyou can do all sorts of smiley-bombs :-() {:-| :-} ;-|16:24
wpwrakkristianpaul: how sweet, a fork-bomb. didn't think you could redefine ":"16:24
kristianpaulme either16:27
wpwraklarsc: that one may actually be recoverable, since it's linear16:27
Action: kristianpaul what other special chars can be redefined..16:29
wpwrak= too ;-)16:30
larscis : a special char in shell?16:32
kristianpauldunno, is it?16:33
kristianpauli just tought it was16:33
wpwraklarsc: not in a function name, only in a variable name16:33
larschave upyou ever used it when writing shell scripts?16:33
kristianpaulnot me..16:34
wpwrakhere's a nice one (safe to use):  2+2() { echo 5; }16:34
wpwraklarsc: yes, the single : yes. but never thought of redefining it.16:34
kristianpaulhmm zsh may protect me :)16:35
wpwraklarsc: makes sense, though. after all, you can also have things line  git-commit  but not  foo-bar=216:35
larscah... : is a nop16:39
larscbut well, you can alias other commands as well16:41
wpwrakgrmbl. need to remount / ...16:41
kristianpaulnop, yeah, at first it tought it was so inoffensive16:42
wpwrakfile system still broken16:46
rjeffrieswpwrak I have used godaddy to register domains, never had an issue20:57
mthwhitequark: it seems the autotools support in jzboot in the qi-hardware repository is incomplete22:15
mthautomake refuses to process Makefile.am since there is no conifgure.ac22:15
mthor configure.in, I always forget what is the old name and what is the new one22:17
xMff.in is old22:22
xMff.ac is new22:22
wpwrakrejon: if looking for consortia, maybe also consider CELF23:30
rejondo you know anyone at CELF?23:36
rejonany intros appreciated23:36
Jay7rejon: Tim Bird23:37
rejonplease email me intro23:37
Jay7rejon: what do you need there?23:38
Jay7and for your info, CELF is merged with LF23:39
wpwrakJay7: oh, i didn't know of this merger23:41
Jay7Richard Purdie may be right person to contact too23:41
Jay7he is LF Fellow now23:41
rejonok, great23:42
rejonlets get some more people involved in what we doing is the goal23:42
Jay7rejon: I'll email you their contacts23:43
Jay7rejon: sent23:46
Jay7rejon: /whois RP or RP__23:48
Jay7this is Richard23:48
Action: Jay7 -> sleep23:49
--- Wed Apr 27 201100:00

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