#qi-hardware IRC log for Sunday, 2010-08-15

bartbesI just read the mailing list and saw jirka talking about this weird build system05:23
bartbesfrom what I've seen so far it's pretty easy actually05:23
qi-bot[commit] Jiri Brozovsky: Compilation problems should be fixed. http://qi-hw.com/p/openwrt-packages/34a4bbc05:48
bartbeswhat are the sections in the menu?06:51
bartbesand the categories?06:51
bartbessmall question07:06
bartbesI'm working on the wordgrinder Makefile07:06
bartbesbut what section and category would that be?07:06
bartbesI went for Utilities/Editors in the end07:15
bartbesin case anyone was ever wondering07:15
qi-bot[commit] Mirko Vogt: add new gmu icon http://qi-hw.com/p/openwrt-xburst/cd8e66307:26
qi-bot[commit] Mirko Vogt: adjust config to 2 new packages in feed packages/ http://qi-hw.com/p/openwrt-xburst/11ee72f07:26
qi-bot[commit] Mirko Vogt: [config] preselect bsd-games (tetris, backgammon, primes, worms) and calcurse http://qi-hw.com/p/openwrt-xburst/ddfccf507:26
qi-bot[commit] Mirko Vogt: [config] adjust config to new feed versions (once again) http://qi-hw.com/p/openwrt-xburst/9c13c4407:26
qi-bot[commit] Mirko Vogt: [config] do not include packages <ntpdate> and <ntpclient> http://qi-hw.com/p/openwrt-xburst/df798b607:26
wolfspraulstress test07:27
wolfspraulmirko: I fixed another but in the commitlog, I think (ahem) it should be stable now...07:27
qi-bot[commit] Mirko Vogt: [gmu] update 0.7.0 -> 0.7.1 http://qi-hw.com/p/openwrt-packages/4f0bbdf07:28
qi-bot[commit] Mirko Vogt: Merge branch 'master' of projects.qi-hardware.com:openwrt-packages http://qi-hw.com/p/openwrt-packages/534010307:28
mirkowolfspraul: hehe :) - ok07:29
mirkowolfspraul: let's see when we do the next merge with owrt-backfire-upstream07:29
mirkowolfspraul: startup time should be improved by the way in the next image07:30
mirkowolfspraul: found some bugs causing heavy cpu load and timeouts which could be avoided07:31
wolfsprauloh nice!07:31
mirkowolfspraul: i suggest i'll upload a "testing"-image tomorrow, announce it on the mailing list and we'll wait 1 week for feedback / input07:31
wolfspraulhopefully a lot of the packages that got contributed in recent weeks are included07:32
mirkowolfspraul: after that 1 week there's 1 week left till my holidays07:32
mirkodo there should be enough time then, to fix possible bugs, build ,upload and announce the new official image07:32
mirkowolfspraul: there are07:32
mirkoi update the config ~twice a week because i include new packages07:33
mirko~twice a day i mean07:33
mirkobsd-games, new gmu version, new gmenu2x version, calcurse, mandoc, new qt4-features, NanoMap, alsa-settings and so on...07:35
wolfspraulsounds good, there are a lot now07:36
jirkabbartbes: probably Utilies07:37
wolfspraulI count 80 subfolders in openwrt-packages now07:37
bartbesyeah, I put it in Utilities/Editor07:37
jirkabbartbes: I see no better  category07:37
bartbesI got it to fail on ncurses07:37
mirkohmm, the keypad layout changed in gmenu2x07:37
bartbesso I'm currently compiling ncurses07:37
bartbesbut as you might know that means I got quite a lot already07:38
bartbesmirko: it did07:38
bartbesdo you need any help with the new keys?07:38
mirkowolfspraul: unfortunately some do not compile07:38
mirkobartbes: i need documentation ;)07:38
bartbes*cough* gforth *cough*07:38
mirkobartbes: ?07:38
bartbesmirko: x = enter07:38
bartbesthat's most important :P07:38
mirkobartbes: please write it down somewhere i can link to in the announce mail07:38
bartbesthe full keymap is in /usr/share/gmenu2x/input.conf iirc07:39
bartbessomething like it though07:39
mirkootherwise there'll be lot's of questions what happened to gmenu2x in the new image07:39
jirkabbartbes: does Wordgrinder request wide character support in ncurses?07:39
bartbesf2 is the old esc07:39
bartbesjirkab: it does..07:39
mirkobartbes: please write it down somewhere public07:39
mirkobartbes: :)07:39
bartbesmirko: give me a wiki page and I'll write it there07:39
bartbeswow, the path was spot-on07:40
Action: bartbes gives his memory a cookie07:40
mirkobartbes: ok, mom07:40
mirkobartbes: uh, there is one already07:40
mirkobartbes: http://en.qi-hardware.com/wiki/Gmenu2x - is that the current layout?07:40
bartbeslet's see07:41
bartbesthe "Plan to"07:41
bartbesthat one is correct07:41
mirkohmm, well - changing sections was "p" and "q" and is now "tab" and "l"?07:41
bartbesyeah, I don't like l, but tab is nice07:42
bartbesmakes sense07:42
mirkothese keys are not even in the same row :P07:42
bartbeshehe, no07:42
bartbesbut I don't know they keymap for VOLUP and VOLDOWN either..07:42
bartbeswell, at least enter, tab and f2 make some sense :P07:43
bartbes(note that I did not change the keymap)07:43
mirkolet's see what the folks say, when they have to switch :)07:43
bartbesmeh, the old keymap wasn't too good either imo07:44
mirkobartbes: i don't they the new one is bad - some things just do not make sense yet to me ;)07:44
bartbesjirkab: the only problem I have left (as far as I can tell) is the lack of ncursesw07:45
jirkabbartbes: is sound very good (though I have not idea how much difficult is compilation of ncursesw )07:47
bartbesI have no idea what the big diff is07:47
bartbeslet me look it up07:47
bartbes(compiling-wise that is)07:47
bartbesit sounds like it's just a compile-time option07:48
qi-bot[commit] Mirko Vogt: [mandoc] (re)add mandoc, as compilation issues were fixed http://qi-hw.com/p/openwrt-xburst/37b39cd07:50
bartbesjirkab: I think you need to add "--enable-widec" to the configure flags07:51
bartbesso I simply added that to the makefile of ncurses07:53
bartbesto see if it is indeed that simple07:53
bartbesbtw, is there a way to turn off the screen (like gmu does)?07:54
larscecho 4 > /sys/class/graphics/fb0/blank07:59
bartbesjirkab: ncursesw doesn't build08:01
bartbeslarsc: thanks I'll test it now08:01
bartbesit immediately turns back on though08:03
bartbesah, because of new output, of course08:04
larscno, because somebody turns it on gain08:04
bartbesmaybe it still registers my enter?08:05
bartbesanyway, when I run it via ssh it works flawlessly08:05
bartbesjirkab: shall I send you the makefile and patch?08:11
jirkabbartbes yes, many thanks (sorry for my delayed response - 'm doing too much things at once :-\ )08:13
jirkabbartbes: ncursesw are needed for more applications so it woudl be usefull to make it work08:14
bartbesit would08:14
bartbesbut it's some weird error in one of the header08:14
bartbeswhich I wasn't able to track down in the very short time I spent trying08:14
bartbesjirkab: http://bartbes.ath.cx/wordgrinder.zip08:15
bartbesand this is the error I get when I compile ncurses(w)08:16
jirkabbartbes: thanks. the error looks strange08:19
jirkabbartbes: looks like it's similar to this Gentoo bug: http://bugs.gentoo.org/show_bug.cgi?id=21464208:34
bartbesjirkab: hmm08:46
jirkabbartbes: looks like GCC can't find definition of wchar_t type08:47
bartbesnot wint_t?08:48
bartbesI think it simply needs an extra include statement08:49
bartbes#include <wchar.h>08:49
jirkabbartbes maybe too, but it stops on the line with wchar_t08:49
bartbesno it doesn't?08:49
jirkabbartbes: where you have wchar.h?08:49
bartbesas far as I can tell it errors here: http://paste.ubuntu.com/478359/08:49
bartbesand there's a wchar.h in staging_dir/include08:50
bartbeshm no08:50
bartbesnot there08:50
jirkabbartbes: see include/curses.h.in in ncurses directory08:54
bartbesso how do we set NEED_WCHAR_H?08:55
bartbesshall I try "--with-build-cppflags=-D_GNU_SOURCE" as mentioned in the gentoo thread?08:57
jirkabbartbes: please try it08:58
Action: bartbes compiles08:59
jirkabbartbes: I will try to understand the  curses.h08:59
bartbesheh, that mess? enjoy!08:59
bartbessomething totally unrelated (while compiling) where can I set the persistent options again?09:02
bartbes(like I did with fstab)09:02
bartbesnvm found it09:03
bartbesjirkab: IT BUILT09:05
jirkabbartbes: nice!09:05
bartbesonly problem is it only built ncursesw09:06
bartbesnow to make it build both09:06
jirkabbartbes: probably the easises solution is to make two packages individual - ncurses and ncursesw09:08
bartbesyeah probably09:08
bartbesI can put those in the same makefile, can't I?09:09
bartbesI can't commit this anyway09:09
bartbesso I might as well copy this package over09:09
bartbesand restore the old one09:09
bartbesmake it ncursesw09:09
jirkabbartbes: yes09:10
bartbeswill do09:10
jirkabbartbes: great09:10
jirkabbartbes: don't forget change ncurses->ncursesw in the Makefile ;-)09:11
bartbesyeah, just did that09:12
bartbesnow.. what is the easiest way to restore the old package?09:12
bartbesthe entire thing was a git repo, right?09:12
jirkabbartbes: wait09:12
jirkabgit checkout FILE_NAME09:12
jirkabbartbes: git checkout FILE_NAME09:12
bartbesso git checkout package/ncurses/Makefile?09:13
jirkabI think so09:13
jirkabI never used this09:13
bartbescheckout is for changing commits09:13
bartbesI think you mean git reset09:13
bartbesor not09:13
bartbesgit should document their shit better09:13
jirkabbartbes: I have this in my notes - so it can be incorrect :-(09:14
bartbesoh it worked09:14
Action: bartbes runs menuconfig to enable ncursesw09:14
Action: bartbes compiles09:15
bartbesjirkab: btw, when I depend on this, the folder is called ncursesw but the package libncursesw, which one do I use?09:17
jirkabbartbes: I think it depends on the name whic his used inside the Makefile09:19
bartbesmakes sense09:19
jirkabbartbes: like here: $(eval $(call BuildPackage,libncursesw))09:19
bartbesthough.. I remember this being different when I added lua.. hmm09:19
bartbesah we'll see09:19
bartbesI guess when you do the wrong thing it tells you it lacks the dependency09:20
jirkabbartbes: well, sometimes09:21
bartbesoh look, no errors!09:21
jirkabbartbes: you are right - at installation it will tell09:22
bartbesnow I will try to build wordgrinder09:22
jirkabbartbes: nice09:23
bartbessome fixes I have yet to do ;)09:25
bartbesbecause apparently it forgot to use the toolchain..09:25
Action: bartbes wonders what pm does with environment variables..09:27
jirkabbartbes: it ignores your environment variables?09:27
bartbesI think... I forgot some settings ;)09:28
bartbeswhat were the openwrt set env variables again?09:31
jirkabdon't know09:33
bartbesright, let's see what it does now09:33
bartbesyeah, I believe it was09:33
bartbesif not then it has no compiler now ;)09:33
jirkabmust exit :-(09:33
bartbesalmost there...09:46
bartbesand it built09:47
bartbesnow to install it on my nn09:47
bartbesright, some bugs in the ncursesw file09:49
bartbes(it installs the some of the same stuff as ncurses, so those are now ripped and instead it depends on ncurses)09:51
wolfspraulbartbes: fascinating read, logging out for the day, but good luck!10:23
bartbesthat was just a rant10:23
bartbessleep well wolfspraul10:23
bartbes(I assume you sleep at the end of the day..)10:24
wolfspraulyes, correct :-)10:24
bartbesokay, I'm not quite sure about the rules here10:34
bartbesit builds, but it still needs some fixes10:35
bartbesso, how do I commit it? do I mark it BROKEN, or is that only for broken builds?10:35
bartbes(as in failing compiles)10:35
mirkobartbes: is there's any case, that it might fail zu compile10:35
mirkomark it as broken please10:35
mirkootherwise full-builds stop10:36
bartbeswell, don't think so10:36
bartbesit only needs some fixes regarding screen size and keymap10:36
mirkothen feel free :)10:36
bartbesthe biggest problem however10:36
bartbesis ncursesw10:37
bartbesI managed to get it to build10:37
bartbeshowever, it is a separate package (due tohow ncurses' is set up you'd need 2 separate builds anyway)10:37
bartbesbut, since they use the same source package10:37
bartbesthey keep overwriting each other10:37
bartbesso every time you compile something that depends on ncursesw it recompiles both ncurses and ncursesw10:37
bartbeswhich.. is not a fun way to spend your time10:38
mirkowhen why not modifying the ncurses-package?10:40
bartbesbecause.. you *need* to build it twice10:41
bartbesyou can't build ncurses and ncursesw at the same time anyway10:41
bartbesand this meant we didn't have to fork ncurses from openwrt10:41
mirkobartbes: yes, so it sounds like it's better having build_dir/target-*/{ncurses,ncursesw}-version directories10:42
bartbesyes, it is10:42
larscwhat you want is BuildVariants10:42
bartbeslarsc: are you going to save me, you holy man?10:42
mirkolarsc: yep10:42
bartbesare you going to elaborate?10:44
mirkolooking for a package which uses that currently...10:44
bartbesno openwrt docs?10:44
larscno docs, as usual ;)10:45
larsctake a look at the uboot-xburst package10:45
bartbesdon't I just need PKG_BUILD_DIR?10:47
mirkobartbes: that way would need 2 Makefiles10:47
bartbesthe rest looks a bit black-magic-y and a bit over-the-top10:47
larscyou add a VARIANT field to each define Package/ncurses{,w}10:47
mirkobartbes: with build-variance you can use just one Makefile10:47
larscand when Build/Compile or Build/Configure you'll have the VARIANT for your package in $BUILD_VARIANT10:48
bartbeswell that isn't entirely possible either10:48
bartbesif you look at the ncurses makefile you see CONFIGURE_ARGS10:48
bartbesthat is what I'm modifying10:48
larscifdef BUILD_VARIANT10:49
bartbesit's that and the install10:49
bartbessee, you guys just throw me some makefile keywords10:50
bartbeshow am I supposed to make sense out of that?!10:50
bartbesso, I modify: define Package/libncurses/install10:50
larscby thinking10:51
bartbesto: define Package/libncurses{,w}/install10:51
bartbesI need to make 2?10:51
larscone for each package10:51
bartbesright, did that10:52
bartbesnow what do I do with Build/InstallDev?10:52
larscyou need to check $BUILD_VARIANT there10:53
bartbesmakefile ifs?10:53
bartbesifeq ($(BUILD_VARIANT),ncurses), that?10:54
bartbesright, did that for Build/InstallDev10:55
bartbesnow I guess the same with the CONFIGURE)ARGS10:55
bartbesso, now what do I do with the eval lines?10:57
bartbesjust leave in both?10:57
larscyes, one for each package10:58
bartbesright, now what else?10:59
bartbesthe PKG_BUILD_DIR?10:59
bartbesor just leave that alone?10:59
larscshould work fine with the default11:00
bartbesso now how do I build them?11:01
bartbesmake package/ncurses/compile should build them both?11:03
larscif both are selected yes11:03
larscbut I just found out, that you need to change PKG_BUILD_DIR to something like PKG_BUILD_DIR:=$(BUILD_DIR)/$(PKG_NAME)-$(BUILD_VARIANT)/$(PKG_NAME)-$(PKG_VERSION)11:04
bartbesI'll compile both of them now11:07
bartbesthen a program depending on them11:07
bartbesto see if they need to recompile11:07
bartbesbut what if it works?11:07
bartbesdo I need to add the ncurses to qi-packages?11:07
bartbesthat seems weird11:08
bartbeswell, it failed11:16
bartbescan i do make package/ncurses/compile?11:17
bartbesbecause obviously the build variant is empty..11:17
bartbes(as the build dir shows)11:17
larscplease show your current Makefile11:25
bartbesgive me a sec while I install xclip11:28
bartbesnow of course I made some horrible mistake..11:29
larscbartbes: you need to add a VARIANT field to each package11:35
bartbeswhat should the contents be?11:35
larscthe contens is what ends up in BUILD_VARIANT, when the package is build11:36
larscso in your case you want ncurses and ncursesw11:37
Action: bartbes fixes it all11:37
qmasterrrI managed to crosscompile mpg123 but if i try to play a mp3 file on my BNN the cpu cant keep up, i thought it would be a fast way of getting mp3 support for gmu11:40
wejpqmasterrr, you have to use the fixed-point-math decoder11:41
wejpwith fpu emulation it is way too slow11:41
qmasterrrwejp, thanks :)11:41
qmasterrrwejp: then mpg123 is not an option and i need to check if i can run GMU with some other mp3 decoder (like mad)?11:48
wejpno, of course mpg123 is an option, jst compile it with the fixed-point decoder instead of the floating point decoder11:49
wejplibmad is slower than mpg123 and has a horrible api, so it is no use11:50
qmasterrri'm new to the openwrt build system and Makefile style so its a bit hard to figure out11:51
wejpyou could also just use libmpg123 from the dingoo package of gmu. works just fine on the ben as well11:53
bartbeslarsc: well, it tells me I have an endef too little11:56
bartbesyet grep + wc tell me the number is the same11:57
qmasterrrthat would be the easy way, i was planning to create a (openwrt) Makefile for libmpg12311:59
larscbartbes: well, i can't help you without seeing the Makfiles content11:59
bartbesheh, nice syntax highlighting12:01
bartbesit calls just about everything illegal ;)12:01
larschm, indeed. looks about right12:03
wejpqmasterrr, ah, i see. there are examples on how to create openwrt packages for programs using a configure script. you could check out those or just have a look at one of the packages already in openwrt. many of them use a configure script. it isn't that difficult12:03
qmasterrryes, i have created a makefile its a little bit of a mess but it works :)12:04
bartbeswejp: it always assumes you use the autotools12:05
qmasterrrnow i have upgradet to gcc version 4.4.4 and im starting to getting errors on Makefiles i haven't touched12:05
wejpbartbes, even if it recreates the configure script each time, it still needs to call it at one point ;)12:06
bartbeslarsc: line 115, it says...12:06
bartbeswejp: no, I mean it always does that12:06
bartbesby default12:06
bartbesyou only need to specify it separately if you have additional arguments12:06
bartbesanyway, dinner12:06
wejpthat's the point, it needs to be called with additional arguments12:07
wejpbut it looks like qmasterrr got it working already12:07
qmasterrrkind of :D12:08
qmasterrrquick question what gcc version do you use?12:09
wejp4.3.3 i think12:09
wejpthe one for openwrt that is12:09
wejpmy native gcc is 4.5.012:09
qmasterrrhum, if it complains about an makefile since i have updatet to make 3.82 should get wondering?12:13
qmasterrrstupid me makefile syntax error is not gcc its Make :)12:13
qmasterrropenwrt-xburst/build_dir/target-mipsel_uClibc-0.9.32/busybox-1.15.3 \ Makefile:431: *** mixed implicit and normal rules.  Stop.12:17
qmasterrrit seems to me that the new Make version doesn't like "target %target:"12:18
lekernelseems to be good shit12:51
lekernelaccording to the paper it's better than quartus12:52
lekernel"The BDS-pga program is for academic purposes use only" "This research has been funded by Altera Corporation."12:52
lekernelwhoa there's even a verilog compiler to the BLIW language they use12:57
lekerneli need to try it12:57
lekernelso far it's the most serious open source (but not free it seems) synthesis tool i've seen12:58
lekerneldamn, it compiled... this is so rare for academic tools13:00
bartbeslarsc: apparently endef can't be indented...13:06
bartbeslarsc: thx for the help, it all works now13:15
bartbesnow, how should I commit this?13:15
lekernel...but it segfaults. back to normal :)13:16
larscbartbes: send a patch13:19
bartbesehm, I can commit13:19
bartbesI just mean, how should I commit the ncurses stuff?13:19
larscnot to the openwrt repo13:19
bartbesbut then13:19
bartbeswordgrinder is going to have to wait for ages probably13:19
bartbesbecause it depends on ncursesw13:20
bartbesright, so how should I create the patch?13:20
larscsvn diff13:21
bartbesso I need to clone their svn repo, copy my new one over and diff?13:21
larscor just send me the makefile13:22
bartbesheh, that sounds easier ;)13:22
bartbeshow should I send it to you?13:23
larsclars at metafoo.de13:23
bartbesbtw, should I just commit wordgrinder and set it as broken in the mean time?13:25
bartbesmaybe it would be easier for you guys if there was a simple guide which stated what the rules are for the repo (what is to be committed, what not, what is to be marked broken, etc)13:27
larscif in doubt: commit. Its usally easy to revert it later it shouldn't have been commited13:31
bartbestime to commit porn then ;)13:32
bartbesgood thing it's hard to erase history13:32
bartbessafe storage for ever!13:32
larscwith git it is13:34
bartbesbtw, I just add a depend BROKEN right?13:34
larscdoes ncursesw really depend on ncurses?13:35
bartbesit depends on some of the stuff in the ncurses package13:35
bartbesif I include them in both you can't install them side-by-side13:35
bartbesthe terminfo stuff13:36
Action: bartbes cleans up his wordgrinder patch (newlines etc)13:36
larscmaybe we should move the terminfo stuff into it's own package13:38
larscdebian for example has a ncurses-base package13:38
bartbesthat might not be a bad idea13:39
bartbescan you build it separately?13:39
qi-bot[commit] bartbes: Wordgrinder: Builds, marked broken until upstream accepts ncurses patch http://qi-hw.com/p/openwrt-packages/847933213:43
bartbesthere we go13:43
larscwell, you could, but you don't have to13:43
bartbes*if* they accept it of course13:43
larsci can commit to the openwrt repo, so thats not a problem13:44
bartbeswell in that case13:45
bartbesbtw, shall I create the separate termlib package?13:45
larscwell, if you want to, sure13:46
bartbesI wonder if --with-termlib *only* builds it13:46
bartbes"generate separate terminfo library"13:46
bartbessimple solution13:47
bartbesnormal compile13:47
bartbesas far as I can tell it builds both13:48
bartbeswhich.. sucks13:48
bartbeswell, then I'm too lazy13:50
bartbesspent enough time on it already13:50
larschm, ncursesw and ncruses ship header files with the same name but different content13:51
bartbesthey do?13:51
bartbeswell, yes, I guess it makes sense that there's diff content13:52
bartbesdebian's solution is making ncursesw install its headers in /usr/include/ncursesw13:54
bartbeswe can do the same thing I guess13:54
bartbesthat's why I'm glad I'm only lead dev of 1 big project, I'd hate it to have to make these decisions ;)13:54
bartbeslarsc: so, do I need to do anything?14:12
bartbesnot to sound pushy, but when can I expect the 'patch' to be into the openwrt tree?14:22
larscit does not seem to work as expected14:32
bartbescare to elaborate?14:33
larsconly the dev files for ncuresesw get installed if both are selected14:33
bartbescan I help?14:34
larscdon't think so14:34
larsci guess InstallDev doesn't handle VARIANTs correctly yet14:35
bartbesI'm pretty sure you need to if that14:35
bartbesevening Textmode14:36
Action: Textmode huggles bartbes 14:36
bartbeswordgrinder is pretty nice btw14:43
bartbesbut it really can't handle terminals smaller than 80x24 I guess14:43
bartbesits menus are simply hardcoded14:43
Chubsis it ok to buy a GTX 480 for an old Q9400 2,67 ghz Quadcore, or would it just be slowed down?15:44
bartbesthat seems.. slightly.. off-topic15:44
Chubscant find a right tech help chat15:47
jirkabbartbes: WordGrinder fits easily at 53x20 console15:55
jirkabbartbes: I think this is one of the resolutions of future release15:56
bartbeswell in that case15:57
bartbesyou read the logs, didn't you?15:58
jirkabbartbes: I just looked at http://en.qi-hardware.com/irclogs/latest.log.html15:58
bartbesthat's what I said15:58
bartbesright, I think that the keymap is the worst problem15:59
bartbesand you can change it easily15:59
bartbeshowever, it stores it in the document15:59
bartbesnot in a conf file15:59
jirkabbartbes: that's not very good15:59
bartbesit is not16:00
bartbesbtw, I did some nice tricks16:00
bartbesfor example it used to use luac to compile the lua into bytecode16:00
bartbesbut since that is non-portable and there's no mipsel luac to run16:00
bartbesI simply made it use cat16:00
bartbesand now it runs nicely ;)16:01
bartbesI've committed it already btw16:01
bartbes( http://qi-hw.com/p/openwrt-packages/8479332 )16:01
jirkabmust see and try :-) congratulations!16:01
larschu? I though luac was portable16:02
bartbesthe bytecode isn't16:02
bartbesit's platform-dependent version-dependent16:02
bartbesand much more-dependent16:02
bartbesthough man pages seem to disagree with me16:03
bartbes"The binary files created by luac are portable to all architectures with the same word size. This means that binary files created on a 32-bit platform (such as Intel) can be read without change in another 32-bit platform (such as Sparc), even if the byte order (``endianess'') is different. On the other hand, binary files created on a 16-bit platform cannot be read in a 32-bit platform. "16:03
bartbesHowever, I know from personal experience it is not16:03
bartbes(it errored before I fixed it of course, but I have previous experience)16:04
jirkabnothing is perfect ;-)16:04
jirkabbut it's good to know that it doesn't work16:05
bartbesjirkab: if you want to compile it I can send you the makefile of ncurses16:05
larscbut shouldn't luac simply preprocess the lua file just as a lua interpreter does when it loads the same lua script?16:06
bartbesnot.. entirely16:06
bartbeswell, a bit yes16:06
bartbesbut it compiles16:06
bartbesand since we're using a modified version of lua16:06
bartbes(lnum most notably)16:06
jirkabbartbes: yes, thanks. I want to try it ;-)16:07
bartbesthe bytecode might not be interpreted the same way16:07
bartbes(in this case it already failed in the header, as it always should, I guess)16:07
bartbesjirkab: just tell me where and how to send it16:07
larscah ok, luac is run on the host building the package16:07
bartbeslarsc: yes16:07
bartbeslarsc: btw, have you got the installdev sections to work yet, or shall I take a look at it?16:09
larscbartbes: not yet16:12
bartbeslarsc: does only ncursesw install headers in your attempt?16:21
larscwell, both install them, but wenn ncursesw headers are installed ncurses headers are removed again16:23
bartbesI see16:23
bartbesthat explains the weird behavior16:24
bartbescan't we just merge them both into one InstallDev?16:25
larsci don't think so16:25
bartbesI tried, it's compiling now16:26
bartbesanother idea16:26
larscwe'll need to add support for this to the build system16:30
bartbesthere is no support for multiple InstallDevs yet?16:31
bartbesI even tried puttin an if in the define16:31
bartbesthat failed too16:31
bartbes--: eval: line 0: syntax error near unexpected token `libncurses,libncursesw'16:31
bartbeswe *can* add multiple hooks16:33
bartbes$(foreach hook,$(Hooks/InstallDev/Post),\16:33
bartbesdoesn't that mean it loops over them?16:33
larscyes, but that wont help us16:33
larscthe problem is that files installed by DevInstall are removed the next time DevInstall for the same Makefile is run16:34
bartbesthat makes absolute sense though16:34
bartbesso your build variants aren't the working solution after all (for now) :P16:34
larscwell, we need to teach the DevInstall stuff about variants16:35
larscor introduce a per paket DevInstall16:35
bartbeswell, what is the easiest, and what is the most useful?16:36
larscthe later is more usefull16:38
larscbut the other is easier (just did it)16:38
bartbeseasy way is best :P16:38
bartbesso, did it work?16:46
bartbesso I even made sure the openwrt build system 'improves'16:47
bartbeswhat an amazingly productive day ;)16:47
bartbesin retrospect wordg... something was easy to port compared to the trouble I went through to get ncursesw16:48
rafalarsc: could you tell me how to create the /dev/rtc or /dev/misc/rtc device for nn? I built 2.6.34 with RTC support and I see the rtc driver init at boot time.. but hwclock -w complains17:10
larscrafa: hm, you need CONFIG_RTC_DEV or something like that17:11
rafait (hwclock) tries to use /dev/misc/rtc.. but no sure which proper mknod arguments to use17:11
rafalarsc: yes, kernel has rtc support I guess17:11
rafathe problem is the lack of /dev/*rtc* device files17:12
rafaI think.17:12
rafalarsc: I set rtc driver in menuconfig. and kernel messages at boot time says:17:14
rafa$ dmesg | grep -i rtc17:14
rafajz4740-rtc jz4740-rtc: rtc core: registered jz4740-rtc as rtc017:14
rafajz4740-rtc jz4740-rtc: setting system clock to 1970-01-04 21:06:54 UTC (335214)17:14
larscyes. there is a seperate option for the /dev/rtc devices17:14
rafalarsc: it was there. I mknod /dev/rtc0 and then I did a symbolic link rtc -> rtc0 and it works now. I do not have udev right now17:22
rafathat is why I needed to create those17:22
bartbesjirkab: for when you're reading this (you will, I know that), I found *a* 'keymap' in src/c/arch/unix/cursesw/dpy.c17:34
Textmodebartbes: ?18:26
bartbesTextmode: ?18:26
Textmodebartbes: !18:26
bartbesTextmode: so what do you want to know?18:30
Textmodebartbes: I was wondering why a "keymap" was notable.18:32
bartbesbecause it isn't available in a config18:39
bartbesyet is set on a per-file basis18:39
rafalarsc: wolfspraul: any idea how to add swap using nand?.. if that is possible.. a swapfile on ubifs would fail it seems21:44
wolfspraulI think swap is equally bad on NAND and SD, it's only a temporary solution (the user turns it on, does something, turns it off).21:45
wolfspraulbut that means it cannot be the 'default' for an end user21:45
wolfspraulfrom those 2, SD is probably still better at least the microSD is easily replaceable :-)21:46
rafawolfspraul: why do you think that? we have used swap partitions on sd with jlime for years21:46
wolfspraulrafa: ok then you know more than me that's good! Most people try to reduce NAND writes, so generally don't swap on nand.21:47
rafawolfspraul: with just 16 mb of machines, so kernel used that often. For nn we have better environment with 32MB21:47
rafaof Ram.21:47
wolfspraulanyway, you want swap on nand, so why not! :-) But I cannot answer your question, unfortunately.21:47
kristianpaulgood point avoid nand writes21:48
wolfspraulubifs may not even implement the special locked-only code paths that are normally needed to support swap files (don't know Linux internals about this, just guessing)21:48
kristianpaulrafa: you said once linux require swap why is not the same with openwrt?21:49
wolfspraulso maybe you need a new partition as a block device, then put swap on it?21:49
rafawolfspraul: that I know is that swap is a nice thing for the system almost always, for programs doing mallocs at least (many).21:49
rafakristianpaul: I do not know much current openwrt. For the openwrt image on nn it looks like just a sdl application on fb, which is not eating much ram. But I can give you an example for openwrt where an application will not work because the kernl will kill it without swap.21:51
rafakristianpaul: just do a proper malloc still if your program will not use that memory.21:51
rafawolfspraul: I was asking to test it on nand. But I do not know if that works. I read that flash filesystem are not useful for swapfiles, and when I tried swapon it fails.21:53
kristianpaulis not posible format NAND as ext?21:53
kristianpaulare you using ubifs isnt?21:54
kristianpaulany way i need sleep (just arrived from trip), g8 4 all !21:56
rafakristianpaul: perhaps you find this cool to read: http://sourcefrog.net/weblog/software/linux-kernel/swap.html  (still if it is obsolete, nice to read)21:57
rafakristianpaul: yes, ubifs.21:58
rafacya man21:58
wolfspraulyou should definitely test it, would be a nice stress test too :-)21:59
rafawolfspraul: yes, perhaps I need that (a new partition as a block device).. I would need to learn how to do that first ;) (if that is possible for current nand and system)21:59
wolfspraulbut I don't know how to quickly make it work on nand, sorry...21:59
rafawolfspraul: well, If I learn and test then I will let you to know the tests22:00
wolfspraulI don't know how the kernel gets information about the partition starts, maybe over the command line? from u-boot? where does u-boot get it from? statically compiled in somewhere?22:00
wolfspraulmaybe it's statically compiled into the kernel22:00
rafawolfspraul: I will check22:21
xiangfurafa:  the partition is statically compiled into kernel.22:32
lunavoraxMorning everyone22:41
lunavoraxI wonder if there's anyone still awake here22:41
xiangfulunavorax: yes22:42
lunavoraxhey :)22:43
lunavoraxi'm looking for information about compiling a simple hello world for the ben22:43
lunavoraxI'm a real noob so I don't really know where to search22:44
xiangfulunavorax: http://en.qi-hardware.com/wiki/Building_Software_Image22:48
Textmode...compile a boilerplate "hello world" with the xburst compiler?22:48
xiangfulunavorax: you need get the cross compiler first.22:49
lunavoraxThanks xiangfu but what is that22:49
lunavoraxDo I have to compile on the ben or can I do it on my x86_64 computer ?22:49
xiangfulunavorax: do it on your x86_64 computer.22:50
Textmodelunavorax: its a cross-compiler chain22:50
Textmodexiangfu: stop being better at this than me! ;_;22:50
lunavoraxAs I do use Ubuntu I suppose I have to do the whole aptitude install thing as shown on the wiki right ?22:52
lunavoraxOr is there some packages i don't need22:52
lunavoraxOh btw22:55
lunavoraxIs there a way to change the size of the letters ?22:55
lunavoraxWhen I hit Ctrl + Alt + F122:56
lunavoraxThe letters are way to big and I can't find a way to switch to a way smaller police22:56
TextmodeI think it selected during kernel configuration.22:56
lunavoraxOh so it cannot be changed then22:56
xiangfulunavorax: just follow the wiki.  and here is what is cross compiler : http://en.wikipedia.org/wiki/Cross_compiler22:56
lunavoraxOk sorry xiangfu maybe i ask too much questions :/22:57
xiangfulunavorax: it's can not changed in the last release. because in last release 2010-06-15, there is no font package, no kbd utilities.22:58
xiangfulunavorax: no. I like question. that mean someone start working with NanoNote. or learning NanoNote. :)22:58
xiangfulunavorax: we will add more fonts in next release.22:59
lunavoraxWell I bought the nanonote in order to program on it yeah... even if I'm a very beginner :/22:59
xiangfuTextmode: yes. by default it's selected during kernel configuration.22:59
lunavoraxThat's great to hear xiangfu I was tired of doing ssh only to have correct text display22:59
lunavorax(also i'm waiting for screen flickering fix, you know)22:59
lunavoraxDingux have a 4x3 font while displaying text at boot, that's much better23:00
xiangfulunavorax: 4x3 is too small I think. we have test 4x8. 6x10, 6x11, never test 4x3.23:01
xiangfulunavorax: you can found the screenshot here: http://en.qi-hardware.com/wiki/Ben_NanoNote_fonts#Screenshots23:02
lunavoraxI said 4x3 but I don't really know the size. However i'm sure that was the smallest possible23:02
lunavorax4x8 looks ok for me23:04
lunavoraxAlso text fits correctly23:04
lunavoraxOh I almost forgot23:04
lunavoraxI tried an experiment w/ the nanonote but apparently it's wasn't designed for that23:04
lunavoraxI bought a mini-b male usb to usb A female adaptor23:05
lunavoraxAnd plugged in a usb flash key but apparently it doesn't work. I bought this adapter in order to use it w/ a wifi usb key. But if it doesn't work.. <_<23:06
xiangfulunavorax: the usb port in nanonote is Device. not HOST.23:06
Textmodepersonally, I wish it was USB2Go, or somethign of that ilk. but meh.23:07
lunavoraxok so i was wrong23:07
Textmodelunavorax: even if it did work, you'd need drivers.23:07
Textmodelunavorax: there are SDIO cards that work, though they are a pretty penny from what I hear.23:08
lunavoraxToo expensive for me i think23:08
--- Mon Aug 16 201000:00

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