IRC log of #maemo-ssu for Sunday, 2012-05-13

* Sc0rpius yawns00:21
*** _rd has quit IRC00:25
*** Sc0rpius has quit IRC00:41
*** mirandir has left #maemo-ssu00:42
*** Sc0rpius has joined #maemo-ssu00:44
*** modeless has joined #maemo-ssu00:58
*** modeless has left #maemo-ssu00:58
*** NIN102 has quit IRC01:12
*** smoku has quit IRC02:22
*** arcean has quit IRC02:29
*** obironbo has left #maemo-ssu03:39
*** wmarone has quit IRC03:48
*** wmarone has joined #maemo-ssu04:13
*** nox- has quit IRC05:07
*** amiconn has quit IRC05:09
*** amiconn_ has joined #maemo-ssu05:09
*** amiconn_ is now known as amiconn05:09
*** nox- has joined #maemo-ssu05:17
*** nox- has quit IRC09:10
*** Pali has joined #maemo-ssu10:46
*** NIN101 has joined #maemo-ssu11:02
*** Free-MG has joined #maemo-ssu11:19
*** Free-MG has quit IRC12:01
*** _rd has joined #maemo-ssu12:27
*** _rd has quit IRC13:02
*** arcean has joined #maemo-ssu13:03
*** _rd has joined #maemo-ssu13:13
*** smoku has joined #maemo-ssu13:21
*** mase76 has joined #maemo-ssu13:31
*** jonwil has joined #maemo-ssu13:50
*** _rd has quit IRC13:51
*** arcean_ has joined #maemo-ssu13:54
*** arcean has quit IRC13:55
*** _rd has joined #maemo-ssu14:26
*** DocScrutinizer has quit IRC14:34
*** DocScrutinizer has joined #maemo-ssu14:34
*** _rd has quit IRC14:42
*** smoku has left #maemo-ssu17:08
*** smoku has joined #maemo-ssu17:09
smokuconfigure: error: Package requirements (hildon-libs >= 0.12.0) were not met17:10
smokuwhat is hildon-libs?17:10
smokuis it some closed package?  as I can find hildon-libs-l10n-public only in repository17:11
merlin1991smoku: which package are you trying to build?17:27
smokumerlin1991: osso-xterm17:28
smokuI already found out - it's an old name of libhildon, which it falls back to when hildon-1.pc is not present17:29
DocScrutinizero/ merlin199117:35
merlin1991o/ DocScrutinizer17:36
* DocScrutinizer just idly wondered if chemist and other tmo mods have a permanent appointment with their psychologist, to not lose their mind about tmo shit17:47
MohammadAGI'm assuming the meeting is still up for tomorrow?17:53
*** Sc0rpius has quit IRC17:54
DocScrutinizersure thing17:54
DocScrutinizerany comments in advance, to think about til tomorow?17:54
MohammadAGSince the speed of light was recently broken, what other physics statements are BS? :P17:55
DocScrutinizer????17:55
*** Sc0rpius has joined #maemo-ssu17:55
DocScrutinizerspeed of light been "broken" since ages17:55
MohammadAGhttp://gizmodo.com/5908206/did-scientists-really-just-break-the-speed-of-light17:56
MohammadAGrebroken then17:56
DocScrutinizerthere's just no way info propagates fatser than light in vacuum, to a place that's not already entangled with your location17:56
DocScrutinizerIOW e.g. tunneling happens with trans-SOL, but obviously for tunneling both ends of the tunnel previously got entangled by any sub-SOL process like installing a physical tube or whatever17:58
DocScrutinizermost spectacular such super-SOL information transfer been demonstrated via entangled photons, over distances like width of a river - when one riverside you do sth to the one photon, on the other riverside the info was instantly available on entangled "sibling"-photon18:01
DocScrutinizerquite obviously prior to that the one photon had to cross the river at light speed18:01
DocScrutinizeras you can entangle them only when they share same location in space-time18:02
DocScrutinizerso you as well could consider this like info is traveling back in time to the moment both photons shared same location, then madatorily takes opposite direction in time moving forward to same moment but on other photon18:03
DocScrutinizerquite intriguing a topic18:04
MohammadAGsee? I'm good at brainstormers :p18:04
DocScrutinizerbut never really broke the fundamental meaning of SOL being the speed limit18:04
DocScrutinizerthere's even cherenkov-radiation happening per definition when particles move at trans-SOL. Obviously it only can happen in any medium but not in vacuum, as SOL in vacuum is speed limit18:06
DocScrutinizerblue glowing of water moderated atomic reactors commonly atrributed to cherenkov radiation (though I seem to recall that's actually incorrect)18:08
DocScrutinizersigh, I could've read that URL'd  page prior to elaborating on my physics knowledge. They absically explained it already18:13
*** Pali has quit IRC18:20
*** Pali has joined #maemo-ssu18:22
* DocScrutinizer just had a great idea to follow a MrTesla-alike concept and start propagating a field of entangled photons unidirectional, so in 10 years everybody could communicate with any point in a 10 lightyears diameter from earth with zero delay18:26
DocScrutinizeran obviously useless idea, but anyway, I think I might find sponsors18:27
DocScrutinizer:-D18:28
DocScrutinizeranother funny question: maybe *all* barionic matter is entangled since big bang?18:38
DocScrutinizeranyway, not exactly on topic for tomorrow's meeting18:39
merlin1991hhr18:39
DocScrutinizerMohammadAG: what's your take on the optional packages for CSSU?18:39
merlin1991I'm still trying to find a proper way todo it18:39
merlin19912 packages + transistional package is bad (rootfs space)18:40
merlin19911 package is impossible since ham would put it under updates18:40
merlin19912 packages with conflict to first is something it'd have to check if ham does it properly18:40
DocScrutinizerafter we wanna hear some semi-feasible reasonable suggestions how to proceed and implement that needed functionality, tomorrow18:41
merlin1991yeah I'll have to play a bit with ham tonight :D18:41
DocScrutinizermaybe a true new repo was most straight way to go?18:41
MohammadAGeasy fix18:42
merlin1991that would only allow us an all or nothing apprach for the 1 package solution18:42
MohammadAGmay not be clean, but HAM is far from that18:42
MohammadAGsince CSSU-extras depends on HAM, patch HAM to not show packages from it18:42
MohammadAGor add XB-Maemo-Flags: not-a-freaking-update18:42
merlin1991and HOW do you install them then?18:42
DocScrutinizermerlin1991: I fail to get your "one pkg" stanza18:42
MohammadAGwell, that flag would make them show under install18:43
MohammadAGor download, or whatever the middle button's text is18:43
merlin1991MohammadAG: and what is for updates then?18:43
merlin1991ie you have installed re version X18:43
merlin1991and version Y comes out but is a real update now?18:43
MohammadAGif (not-an-update && currentmaintainer: Nokia) show under install18:43
MohammadAGif (not-an-update && currentmaintainer: else) update18:44
DocScrutinizerwe dn't need updates, do we? if yes then for what?18:44
merlin1991also this way we#d break apt18:44
MohammadAGno18:44
merlin1991DocScrutinizer: fix bugs? :D18:44
MohammadAGapt-get update is advanced, and would upgrade from Nokia to new18:44
merlin1991would you like to install nikocam and stay with its current state? :D18:44
DocScrutinizerno, I would like to pick nikocam from cssu-extras as a simple optional package, like I do with bash from extras18:45
merlin1991yeah but when youve installed it already then you want updates18:45
merlin1991and ofc you want them in your updates window18:45
DocScrutinizersure thing18:45
merlin1991so a flag "this is no freaggin update" only works for the first time18:45
merlin1991so the 1 package way is m00t anyway18:45
DocScrutinizerwhere's the problem with shipping an update in cssu-exrtas repo18:46
MohammadAGDocScrutinizer, the way HAM/apt work is that if we have nicocam in cssu-extras it'll show in updates18:46
merlin1991if we keep the package name same as nokia name18:46
DocScrutinizernot until it's installed?18:46
DocScrutinizeraaah18:46
DocScrutinizerthen that's the problem18:46
DocScrutinizerpackage names conflict, hmhmhmhm18:47
DocScrutinizerI see18:47
DocScrutinizerdang18:47
DocScrutinizerjaffa had some idea how to do all that, no?18:48
merlin1991his idea was 2 distinct packages18:49
* DocScrutinizer re-reads jaffa's mail18:49
merlin1991and one to swap .desktop .dbus .whatever files18:49
merlin1991but that's not feasible either18:49
merlin1991since we have to keep our rootfs footnote as small as possible, and installing each closed blob side by side with the re package is just going to eat up space quite fast18:50
DocScrutinizercould we implement a "replace" menu checkbox in HAM, and all pkgs from cssu-extras have some "#ifdef cssu-replace" in their postinstall script?18:51
DocScrutinizernah, off the point18:52
merlin1991DocScrutinizer: we should at least be apt compatible18:52
merlin1991my best shot at it so far is:18:52
DocScrutinizerthe icon / .desktop itself has to manage conflicts when we install nikocam in parallel to stock18:52
merlin1991have the nokia package from the nokia repo, a cssu package ie camera-ui2 that conflicts with the nokia and the 3rd packages, and have the 3rd package that depends on the nokia package and conflicts with the cssu one (needed when user uninstalls cssu pak and wants nokia back)18:54
merlin1991s/3rd packages/3rd package/18:54
infobotmerlin1991 meant: have the nokia package from the nokia repo, a cssu package ie camera-ui2 that conflicts with the nokia and the 3rd package, and have the 3rd package that depends on the nokia package and conflicts with the cssu one (needed when user uninstalls cssu pak an...18:54
DocScrutinizerhmm, something like that, yeah18:55
merlin1991that way we'd leave the user the choice and take care that there is no side by side rootfs madness18:55
DocScrutinizerand build two pkgs for each project, one always tagged/named "${APPNAME}-cssu" and thus the concurrently usable pkg18:56
merlin1991you lost me there18:57
merlin1991so essentially you want a 4th package aswell that can be install concurrently?18:57
DocScrutinizerand ${APPNAME} and ${APPNAME}-cssu obviously conflict18:57
DocScrutinizeryes18:57
merlin1991why would it conflict then?18:58
merlin1991my suggestion was: 1st ${APPNAME} (original nokia), 2nd ${APPNAME}-cssu (the rewrite) and 3rd ${APPNAME}-nokia the helper to install nokia package back18:59
DocScrutinizerbecause ${APPNAME} from extras-cssu is replacing original nokia pkg, while ${APPNAME}-cssu installs same binary in a way it can coexist with original nokia pkg. evidently they should conflict with each other18:59
merlin1991and then have 1,3 conflict 118:59
merlin1991err 1,3 conflict 219:00
DocScrutinizer:nod:19:00
DocScrutinizermy suggestion just augments this19:00
merlin1991your augmention brings no real value, besides calc19:01
DocScrutinizerok, naming conflicts still a problem19:01
merlin1991all other rewrites are either original or rewrite by design19:01
merlin1991unless you love to have doubled control panel entries, and whatnot :D19:01
merlin1991sure for anything that is a REAL application I can see the point to run them side by side19:02
merlin1991(in this case Calc is the only one so far)19:02
DocScrutinizeruser has three possible options: a) keep Nokia binary, b) replace Nokia binary by cssu rewrite, c) install cssu-rewrite side by side with original Nokia binary19:02
merlin1991though rewrites currently in place: 3 control panel things, not excatly feasible side by side, camera-ui impossible side by side, and calc calc makes sense to have the c option19:03
merlin1991I'd apply c only where it makes sense19:03
merlin1991but not in general19:03
DocScrutinizerwhen one or another option is not available for a particular project, we simply omit that pkg19:03
merlin1991with that added I'm fine with your suggestion :D19:04
DocScrutinizere.g for orientation-lock applet we only got [b)|c)] option. it's an optional pkg that neither can replace any stock Nokia stuff nor live side by side with any19:05
DocScrutinizerso only cssu-orientationlock-applet-cssu pkg should be available19:06
DocScrutinizerI'm unhappy with my own naming sheme though :-D19:06
DocScrutinizerlet's define: $PKG == Nokia stock, ${PKG}-cssuR == replacing, ${PKG}-cssu == optional aka side-by-side (if applicable)19:08
DocScrutinizerquite obviously a ${PKG}-cssuR would install .desktop etc that are same name as ${PKG}19:09
DocScrutinizerto completely replace stuff19:09
DocScrutinizerwhile a ${PKG}-cssu would install .desktop with new name that's not conflicting with original Nokia stuff19:10
DocScrutinizerand yes, in HAM with cssu-extras repo enabled there should also show up ${PKG}-nokia, which allows user to install original binary - basically it's a synonym for uninstalling ${PKG}-cssuR19:13
merlin1991which essentially would give us only 1 ${PKG}-cssu from our current packages19:13
merlin1991yep19:13
merlin1991also since the user can install the -cssuR packages from ham (they are visible after all) they need a way to install the nokia package again (which is invisible ofc)19:13
DocScrutinizeryep19:14
DocScrutinizernota bene for PKGs not existing in stock (like orientation-lock applet), the "-cssu" extension is somewhat optional, since there are no ambiguities19:16
merlin1991exactly :D19:17
*** Pali has quit IRC19:23
DocScrutinizeroh, one thing I'd like to get defined in our HowTo about all that: it's generally mandatory for a R-pkg to deliver a non-R aka -cssu pkg as well, unless this obviously doesn't make any sense (you can't have a -cssu pkg for HD, you however *can* have a -cssu pkg for camera)19:30
DocScrutinizerexcuse my poor wording19:31
merlin1991err?19:31
DocScrutinizeroh, one thing I'd like to get defined in our HowTo about all that: it's generally mandatory for a project delivering a R-pkg to provide a non-R aka -cssu pkg as well, unless this obviously doesn't make any sense (you can't have a -cssu pkg for HD, you however *can* have a -cssu pkg for camera)19:31
merlin1991you can only have a -cssu pkg for camera with quite some work19:31
merlin1991it has to be the only one by design19:32
DocScrutinizeroh? how so?19:32
merlin1991becaus it's coded like that?19:32
merlin1991also hd is a bad example, since anything open-source isn't part of this great scheme19:33
DocScrutinizerI always thought fcam and blessN900 were quite able to delive "-cssu" type pkgs19:33
merlin1991they are not the stock camera ;)19:33
DocScrutinizerright, and that's exactly the rationale behind this19:34
merlin1991well still the nikocam needs some adjustments to work as a -cssu pkg19:34
merlin1991so unless nicolai does them there isn't going to be a side by side package for the camera-ui19:34
merlin1991imo having a side by side option is just icing on the cake19:35
merlin1991first thing the user needs is to have a choice at all19:35
DocScrutinizerI don't like users of stock maemo left behind just because app devels are too lazy to consaider make their stuff work on non-cssu systems as an alternative concurrent app19:35
merlin1991stock maemo users are left behind per definiton19:35
merlin1991since this is all cssu specific19:36
DocScrutinizeryes, but we're not hostile against stock, we actively want to encourrage devels to build their apps for all maemo users19:37
DocScrutinizerand we want to keep freedo of choice where ever possible19:37
merlin1991yeah but cssu is never going to host "apps"19:37
DocScrutinizerchoice to install -cssu pkg rather than -cssuR19:37
merlin1991this is only about hosting rewrites of stuck stuff within cssu19:38
merlin1991s/stuck/stock/19:38
infobotmerlin1991 meant: this is only about hosting rewrites of stock stuff within cssu19:38
DocScrutinizermerlin1991: it's exactly not19:38
merlin1991the meeting is19:38
DocScrutinizerit's also about stuff like orientation-lock applet19:38
merlin1991the whole idea why I started this is19:38
merlin1991and PER DEFINTION OF CSSU WE ARE NOT GOING TO ADD ANYTHING THAT CAN LIVE WITHIN EXTRAS THUS THERE IS NO NEED TO CONSIDER cssuR PACKAGES FOR YOUR AVERAGE APP19:39
merlin1991ffs19:39
merlin1991sry :D19:39
DocScrutinizeryes, all on same page here19:39
DocScrutinizerjust for app rewrites like OMP and camera, I think we owe cssu users to require devels to also provide a -cssu pkg, so users can compare side by side19:41
DocScrutinizerif it can't be done for a particular project, then ok, so it's like that19:42
DocScrutinizerbut if it can get done easily, we shouldn't let devels get away with laziness19:43
merlin1991yeah sure, but most devels already have such things in place anyway19:43
merlin1991(see omp in extras)19:43
merlin1991since essentially anything -cssu could live in extras unless it depends on somehting cssu specific19:43
DocScrutinizerthe more easily we can define this as a usually mandatory best practice19:43
DocScrutinizerof course when there's an extras pkg already, we don't need a -cssu pkg19:44
DocScrutinizerat least if the pkg from extras doesn't fail on CSSU ;-D19:45
DocScrutinizerand only if the pkg from extras is actually not replacing stock app in stock fremantle19:46
DocScrutinizeras otherwise it's a non-cssu version of the R-pkg19:47
merlin1991omp is the perfect exmaple19:48
merlin1991in extras and does not replace19:48
DocScrutinizeryes19:50
DocScrutinizerall fine with that19:50
DocScrutinizerdoesn't need an explicit -cssu pkg, as there's the extras pkg doing exactly what otherwise we'd require at _least_ for -cssu pkg19:50
DocScrutinizerI'd like to see similar situation for nikocam19:52
DocScrutinizer(actually I don't know if it's maybe already like that)19:53
DocScrutinizeranyway... o/19:57
* DocScrutinizer afk19:57
*** mase_76 has joined #maemo-ssu19:58
*** mase76 has quit IRC20:01
*** mase76 has joined #maemo-ssu20:11
* Jaffa sees his name and goes to read scrollback20:21
Jaffamerlin1991: DocScrutinizer: The "bridge" package could uninstall the original package since CSSU doesn't do the whole complete meta-provides thing20:22
DocScrutinizerJaffa: yup, that's how I understood merlin1991's concept20:23
*** nox- has joined #maemo-ssu20:31
*** mase76 has quit IRC20:36
*** jonwil has quit IRC20:38
*** mase76 has joined #maemo-ssu20:51
*** mase76 has quit IRC20:58
*** mase76 has joined #maemo-ssu20:59
*** arcean_ has quit IRC21:11
*** arcean has joined #maemo-ssu21:15
*** BCMM has joined #maemo-ssu21:23
*** _rd has joined #maemo-ssu21:28
*** _rd has quit IRC21:50
*** _rd has joined #maemo-ssu21:50
*** fw190 has joined #maemo-ssu21:56
fw190helo21:56
fw190in one of the conversations someone wrote that APMD sucks21:59
fw190but what to do if one has a Chinese battery21:59
fw190and stock indicator sucks like hell22:00
fw190after hlaf battery i goes to 022:00
fw190and in AMPD the nokia BME alternative shows more accurate22:00
fw190does AMPD mess the system or something like that?22:01
Sicelobq27200 ftw22:02
fw190I do not understand the ftw thin. IRC shortcuts arent easy for me :)22:04
Sicelo~ftw22:04
infobotFor The Win!22:04
Sicelohttp://enivax.net/jk/n900/bq27200.sh22:05
StyXmanSicelo: any plans to make a widget out of it?22:07
*** fw190 has quit IRC22:08
Siceloi tried in DCE .. but for my purposes, just running it in a terminal is good enough22:08
*** _rd has quit IRC22:08
StyXmanwell, it's a .sh script, it should be doable in any other language :)22:09
Siceloi mostly use it while calibrating battery22:12
*** _rd has joined #maemo-ssu22:27
*** mase76 has quit IRC22:32
*** mase76 has joined #maemo-ssu22:32
DocScrutinizerI think pali and freemangordon (iirc) were about to implement a complete set of replacements for bme, that APM shite thing (I.E. a better bq27200.ko with sys/*/* nodes), hald-addon-bme, and battery applet22:36
DocScrutinizerfw190 gone, meh22:36
DocScrutinizeranyway, for now the approach is so utterly obvious: replace each read() from /sys/*/*/bq27200/* by a system("i2cget bla blub") call, in this APMD shite22:38
DocScrutinizeror even implement a lib-fake-bq27k.so that you LD_PRELOAD for AMPD and similar apps using /sys/*/*/bq27200/* nodes, and this lib transforms each read to any of those nodes to an equivalent direct I2C read22:41
DocScrutinizerAIUI there's a lib doing the exact opposite, for bme in mer/meego-ce22:41
DocScrutinizerthere's even a way to patch 27200.ko in a way it wouldn't lock the I2C bus chip addr permanently, but rather basically would do all the init stuff now done when modprobing the module whenever somebody does a read() on any of the sysnodes, then reads out the value from chip via I2C, and immediately releases I2C again, this not permanently blocking bme and i2ctools like in bq27200.sh22:45
DocScrutinizernobody tackled that, as it would need serious rewrite of the bq27200.ko module. Obviously something nobody is eager to do22:46
*** mase76 has quit IRC23:12
*** mase76 has joined #maemo-ssu23:14
*** bsdmaniak has quit IRC23:15
*** mase76 has quit IRC23:20
*** Sc0rpius has quit IRC23:23
*** BCMM has quit IRC23:32
*** NIN101 has quit IRC23:34
*** Kaptenen_ has joined #maemo-ssu23:36
*** Kaptenen_ has quit IRC23:41
*** gri_ is now known as gri23:45
*** _rd has quit IRC23:49
*** _rd has joined #maemo-ssu23:49
*** Kaptenen_ has joined #maemo-ssu23:52
*** Kaptenen_ has quit IRC23:52

Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!