**** BEGIN LOGGING AT Thu Jan 18 02:59:57 2007 Jan 18 03:40:53 anyone around? Jan 18 06:18:06 03bzhou * r5224 10optware/trunk/make/miau.mk: miau: 0.6.2 -> 0.6.3 Jan 18 12:55:15 03oleo * r5225 10optware/trunk/ (3 files in 2 dirs): transmisison: CGI excape fixes and watchdog chmod/own Jan 18 14:44:36 03gda * r5226 10optware/trunk/make/ds101g-kernel-modules.mk: ds101g-kernel-modules: small fix Jan 18 14:48:22 03gda * r5227 10optware/trunk/make/ds101g-kernel-modules-fuse.mk: ds101g-kernel-modules-fuse: small fix Jan 18 15:03:07 03operron * r5228 10optware/trunk/make/ds101j-kernel-modules.mk: ds101j-kernel-modules: added tun and ethertap + sub packages now named kernel-module-* instead of kernel-modules-* Jan 18 15:03:45 03operron * r5229 10optware/trunk/sources/ds101j-kernel-modules/armeb-config: ds101j-kernel-modules: added tun and ethertap Jan 18 15:47:13 03oleo * r5230 10optware/trunk/ (make/transmission.mk sources/transmission/transmission.cgi): transmission: \\ fix in sed Jan 18 18:10:52 <`shr00m> guys... anyone know if hostap and prism drivers will compiles on openslug? Jan 18 18:15:49 morning Jan 18 18:16:50 Howdy rwhitby! Jan 18 18:16:59 hey joshin Jan 18 18:17:50 oleo: rebuild stopped on esniper Jan 18 18:17:53 checking for cURL version... 7.16.0 Jan 18 18:17:53 checking for cURL SSL support... configure: error: no SSL Jan 18 18:17:53 cURL is available from http://curl.haxx.se/ Jan 18 18:17:53 make: *** [/home/slug/optware/oleg/builds/esniper/.configured] Error 1 Jan 18 18:18:10 (same for ddwrt feed) Jan 18 18:29:51 oleo: missing openssl-stage in libcurl Jan 18 18:32:25 03rwhitby * r5231 10optware/trunk/make/libcurl.mk: libcurl: missing openssl-stage Jan 18 18:41:58 03bzhou * r5232 10optware/trunk/make/py-simplejson.mk: py-simplejson: 1.4 -> 1.5 Jan 18 19:07:01 03bzhou * r5233 10optware/trunk/ (Makefile make/endian.mk): added and promoted endian, a small endianness report util Jan 18 20:02:57 03rwhitby * r672 10kernel/trunk/patches/2.6.20/defconfig: Added CONFIG_INPUT_UINPUT=m Jan 18 20:08:22 : any status on the next version of Unslung? Jan 18 20:09:07 s/:/mwester:/ Jan 18 20:09:08 fcarolo meant: mwester: any status on the next version of Unslung? Jan 18 20:23:33 mwester: I may have found a glibc bug that explains the problems with mmap() mentioned in the mailing list Jan 18 20:26:13 fcarolo: it's still in the works -- we have plenty of time to track down and sqaush bugs still. Jan 18 20:27:58 fcarolo: i found http://sourceware.org/ml/libc-alpha/2000-01/msg00102.html Jan 18 20:30:08 mwester: for the next unslung release, do you think that it might be good to enhance the meaning of /.ext3flash by mounting /dev, /var and maybe /tmp in memory? Jan 18 20:30:31 see: http://tech.groups.yahoo.com/group/nslu2-asterisk/message/457 Jan 18 20:32:48 eno: it looks similar, but I guess this is the exact one http://sources.redhat.com/bugzilla/show_bug.cgi?id=957 Jan 18 20:34:20 mwester: will the next version use the same glibc (2.2.5)? Jan 18 20:43:18 can anyone give me a clue as to why this might not work? Jan 18 20:43:18 [root@TS-101 etc]# su repvik Jan 18 20:43:19 su: cannot run /bin/bash: Permission denied Jan 18 20:44:16 nm, incorrect permissions on libs Jan 18 20:46:34 osas: I don't think that would be considered an "enhancement" by those running applications low on memory. Jan 18 20:47:02 it would be for those who are running from flash :) Jan 18 20:47:09 maybe a new flag? Jan 18 20:47:59 like /.ext3ramdisk ? Jan 18 20:48:35 Not necessarily for all flash users. It would have to be a new flag. I'm still concerned that it's an attempt to push unslung into an area that is not well-suited for. Openslug or openwrt are better if you wish to run in such a manner. Jan 18 20:49:52 yeah ... I can see what you are saying ... Jan 18 20:49:54 I guess I think that if you run all that from ramdisk, and disable swap, then there's really no point in unslinging in the first place -- what you really want is to mount a flash device as "/opt", run from the internal flash, and force ipkg to work on the external /opt. Jan 18 20:51:04 Now that would be an interesting option to add to Unslung, and can probably be done with a simple set of steps in place of the "unsling" process. Jan 18 20:51:05 I see Jan 18 20:51:26 well, ppl were asking and that's why I brought it up Jan 18 20:52:55 I want to make glibc-i18n ipks, is this something that is only need for the diskstation? Jan 18 20:53:00 The need is real - I think the solution might be worth discussing a bit. There are probably some significant disadvantages to the alternative I'm proposing that bear some more discussion. Jan 18 20:53:38 I think that the idea will have traction on asterisk comunity Jan 18 20:53:56 you can plug in a really small flash and run asterisk Jan 18 20:54:09 osas: did you fix asterisk-stage :) Jan 18 20:54:31 you don't need to go through the pain of manualy partitioning small flashes for 6.8 Jan 18 20:54:56 gda_: you are asking about optware or openembedded? Jan 18 20:55:02 optware Jan 18 20:55:16 what is wrong with it? Jan 18 20:55:39 I made a bug report Jan 18 20:56:02 can you point me to it? Jan 18 20:56:09 I didn't saw it :( Jan 18 20:56:09 anyway, it installs into $(STAGING_DIR)/opt/usr Jan 18 20:57:47 no problem, will you fix it? Jan 18 20:57:50 I will check Jan 18 20:58:15 do you have a fix? Jan 18 20:58:32 tell me when it has happend, I have a package that needs to be fixed then Jan 18 20:58:53 If I would have a fix I would have done it myself ;) Jan 18 20:59:14 too much working places currently Jan 18 20:59:39 need i18n support for my diskstation for squirrelmail Jan 18 21:00:05 ok Jan 18 21:01:00 I have made a chan_capi module for asterisk, that currently has a workaround for the asterisk problem Jan 18 21:01:37 this module allows asterisk to use the isdn hardware on my dsl router Jan 18 21:01:40 over tcp Jan 18 21:02:33 ok Jan 18 21:02:34 eno: are you there? Jan 18 21:03:01 it should go in $(STAGING_DIR)/opt Jan 18 21:03:05 I will check Jan 18 21:03:12 osas: yes Jan 18 21:16:52 gda_: just back Jan 18 21:26:38 eno: I want to make glibc-i18n packages Jan 18 21:26:50 shall I make them only for ds101g? Jan 18 21:27:45 anyway wanna help me with some ipkg issues i'm having after a make mpd-ipk command? Jan 18 21:28:13 gda_: what will it contain? any comparable package in debian? Jan 18 21:28:41 don't know enough about debian Jan 18 21:28:42 eno: are you back? Jan 18 21:28:49 eno: i have some stuff to tell you Jan 18 21:29:10 gda_: what will it contain? Jan 18 21:29:15 scant: go ahead Jan 18 21:29:42 eno: esound is currently broken Jan 18 21:29:47 eno: it will contain the files that you can find under /usr/share/locale Jan 18 21:29:56 but changing the aclocal to 1.10 from 1.09 in the mk file worked Jan 18 21:30:07 broken up for languages Jan 18 21:30:32 and a package for common files Jan 18 21:30:37 i'm trying to compile mpd, again, but having some issues Jan 18 21:30:51 i changed the mpd.mk file to point to a newer source Jan 18 21:30:57 I get nearly all files from the glibc in the toolchain Jan 18 21:31:27 gda_: i c, let's take a two step approach Jan 18 21:31:40 i changed these lines: MPD_VERSION=0.12.1 & MPD_SOURCE=mpd-$(MPD_VERSION).tar.gz Jan 18 21:31:54 1. make it a ds101g specific package first Jan 18 21:32:00 but now i'm having an ipkg issue, it's not causing any major issues, but i'm having some file name conflicts Jan 18 21:32:19 2. test on other platforms, and enable other platforms Jan 18 21:32:28 just like what we did for perl Jan 18 21:32:46 can you help me troubleshoot? the old version was 0.11 the new version is .12, the last line of the make ipk is: Packaged contents of /mpd-0.12.1-ipk into mpd_0.11.5-1_armeb.ipk Jan 18 21:32:56 why is there a file name conflict? Jan 18 21:32:59 okay Jan 18 21:34:01 but don't stay tuned for it, am already tired Jan 18 21:34:16 eno: r u there? Jan 18 21:34:18 gda_: sure Jan 18 21:34:32 scant: just one sec, let me look at the mpd.mk Jan 18 21:35:20 do you understand my issue? it's compiling the newer version after i updated the mpd.mk file, but the ipkg file that make ipk outputts has the wrong version as a postfix in the filename Jan 18 21:35:44 scant: yes i understand the issue Jan 18 21:36:09 did you "make mpd-dirclean" ? Jan 18 21:36:16 let me try that Jan 18 21:36:30 then do i try make mpd mpd-ipk? Jan 18 21:36:38 yes, try that again Jan 18 21:36:45 remember, i'm doing this natively Jan 18 21:37:05 it's compiling, i'll tell you what the version is in the mpd ipkg filename Jan 18 21:38:01 if you have a more powerful x86 pc, i'd highly recommend you either setup a cross environment, or at least use distcc Jan 18 21:38:14 i have no clue how to setup a cross devel env Jan 18 21:38:23 http://www.nslu2-linux.org/wiki/Optware/Distcc Jan 18 21:39:05 when you reported esound broken, what was not working? Jan 18 21:39:22 it's using aclocal 1.09 instead of 1.10 Jan 18 21:39:23 build problem or runtime problem? Jan 18 21:39:25 and won't compile Jan 18 21:40:36 i told you that for native build, autoreconf is not necessary, so you can comment out the entire two lines Jan 18 21:40:57 oh... Jan 18 21:41:02 i changed it to 1.10 and it worked Jan 18 21:42:05 great, i can patch esound.mk to behave differently depending on native or cross Jan 18 21:42:38 i didn't try a native compile by commenting out the aclocal-1.09 line though Jan 18 21:42:38 but keep in mind most developers have cross environment already setup Jan 18 21:42:55 it's cool, i just wanted to tell you my progress Jan 18 21:43:11 i'm more concerned about unslung not having a libao package Jan 18 21:43:14 or an alsa package Jan 18 21:43:18 or a mpd package Jan 18 21:43:19 and making it work on both cross and native is not of the highest priority Jan 18 21:44:09 you're very welcome to contribute these Jan 18 21:44:23 i'm such a newb though Jan 18 21:44:43 all linux devel i do on the unslung cause it's how the tutorials were setup Jan 18 21:44:52 i don't think there is a tutorial on how to cross compile Jan 18 21:44:55 once you're ready (hope it's ready for cross), add it to CROSS_PACKAGES_READY_FOR_TESTING Jan 18 21:45:27 eno: so the make mpd-dirclean didn't fix the filename issue Jan 18 21:45:38 what should i try next? Jan 18 21:46:37 ah, make/mpd.mk is using sources/mpd/control Jan 18 21:46:52 mpd.mk needs an upgrade Jan 18 21:46:54 do i change the version in the control file? Jan 18 21:47:07 for the short term, yes Jan 18 21:47:21 how hard is it to change the mpd.mk file? Jan 18 21:48:00 but control file change will not be accepted in the feed Jan 18 21:48:12 very easy, look at make/template.mk Jan 18 21:48:24 eno: i'm not trying to put my changes back into the feed Jan 18 21:48:43 and see the new way of generating control file Jan 18 21:48:47 i'm surprised tho, that no unslug developers use mpd Jan 18 21:49:01 mpd on a slug is like so awesome Jan 18 21:49:32 see line 202 of make/template.mk Jan 18 21:50:01 i'll check it out after i compile Jan 18 21:50:39 it is making $(_IPK_DIR)/CONTROL/control according to the rule from line 167 - 180 of make/template.mk Jan 18 21:51:22 the benefit is that once you change _VERSION and _IPK_VERSION Jan 18 21:51:27 eno: can you tell me, after i change the control file, do i do a make mpd-dirclean and then make mpd mpd-ipk? Jan 18 21:51:48 yes. the generated CONTROL file will automatically get the change Jan 18 21:52:19 eno: i only changed the version number and the source entries in the control file, now i'm recompiling... Jan 18 21:52:57 also, all optware platforms will get the change because Architecture is not hardcoded in control file Jan 18 21:54:11 i'm confused... Jan 18 21:55:50 scant: maybe i jumped ahead Jan 18 21:55:55 eno: i think you said there were two ways to fix this ipkg file name issue, change the control file in sources/mpd or change the mpd.mk file, is this correct? Jan 18 21:56:22 yes, so you're trying the first way Jan 18 21:56:43 yes, i'm trying the first way, which i said won't work if i want to upload the changes back into the feed, is this correct? Jan 18 21:56:49 and i was telling you about how the second way is much more superior Jan 18 21:57:11 eno: after the first way works, i will try the second way... Jan 18 21:57:17 yes Jan 18 21:58:46 libao is the first name in optware/Makefile PACKAGES_THAT_NEED_TO_BE_FIXED Jan 18 21:58:59 gda_: ping Jan 18 21:59:05 with comment "has runtime trouble" Jan 18 21:59:34 eno: is it like this like that needs to be changed? mpd.mk:178 install -d $(MPD_IPK_DIR)/CONTROL Jan 18 22:00:31 scant: change 178 to $(MAKE) $(MPD__IPK_DIR)/CONTROL/control Jan 18 22:01:23 eno: is that the only line that needs to be changed? Jan 18 22:02:03 and provide the rules to generate $(MPD__IPK_DIR)/CONTROL/control (see template.mk 167-180) Jan 18 22:02:27 * rwhitby praises eno on the amount of dedication he has to this project. Jan 18 22:02:44 yes, eno rocks Jan 18 22:02:47 he's helped me out a lot Jan 18 22:03:06 i had esound aclocal issues compiling natively and now this filename issue Jan 18 22:03:31 and provide all the MPD_ variables used in the $(MPD__IPK_DIR)/CONTROL/control rules (see template.mk 29-45) Jan 18 22:04:00 i'm trying to get all the help we can get :-) Jan 18 22:04:11 eno, is this correct? mpd.mk:178 install -d $(MAKE) $(MPD__IPK_DIR)/CONTROL/control Jan 18 22:04:25 no Jan 18 22:04:47 mpd.mk:178$(MAKE) $(MPD__IPK_DIR)/CONTROL/control Jan 18 22:05:31 ? Jan 18 22:06:02 the tab character, which probably appears as 8 spaces Jan 18 22:06:19 so don't actually put in on the line, right? Jan 18 22:06:27 right Jan 18 22:06:37 hehe cool Jan 18 22:06:52 (but you must put a tab character, not 8 spaces) Jan 18 22:07:19 ok Jan 18 22:07:26 does anyone in here use mpd? Jan 18 22:07:30 .mk files are really sensitive about tab vs spaces Jan 18 22:07:56 mpd is like a gift to music Jan 18 22:08:20 especially if someone has speakers in every room in their house Jan 18 22:08:27 excellent Jan 18 22:09:10 eno: i see the variables in the template.mk file lines 29-40 Jan 18 22:10:00 basically you want to copy the lines from sources/mpd/control to the corresponding variable Jan 18 22:10:40 yes, but how do i do that in mpd.mk Jan 18 22:11:03 the mpd control file has this line in it now, Version: 0.12.1-1 Jan 18 22:11:18 and i see the template.mk also has _VERSION Jan 18 22:11:36 i figure i have to specify the version number through the variable in the mpd.mk file, but how? Jan 18 22:11:56 you want to write in mpd.mk MPD__VERSION=0.12.1 Jan 18 22:12:17 sorry, MPD_VERSION=0.12.1 Jan 18 22:12:48 so, after $(MAKE) $(MPD__IPK_DIR)/CONTROL/control, make a new line and put MPD_VERSION=0.12.1 in it? Jan 18 22:13:27 better follow make/template.mk and put MPD_VERSION=0.12.1 almost at the top Jan 18 22:13:51 it's already there Jan 18 22:14:08 that was one of the lines i changed a long time ago Jan 18 22:14:17 and have MPD_DESCRIPTION=The Music Player Deamon Jan 18 22:14:27 and so on and so forth Jan 18 22:14:40 these lines were in mpd.mk when i was having the filename issue Jan 18 22:14:40 MPD_SITE=http://www.musicpd.org/uploads/files Jan 18 22:14:40 MPD_VERSION=0.12.1 Jan 18 22:14:40 MPD_SOURCE=mpd-$(MPD_VERSION).tar.gz Jan 18 22:14:40 MPD_DIR=mpd-$(MPD_VERSION) Jan 18 22:14:40 MPD_UNZIP=zcat Jan 18 22:14:56 but you can see the MPD_DESCRIPTION was missing Jan 18 22:15:06 mpd.mk was too old Jan 18 22:15:26 eno: after i make these changes Jan 18 22:15:40 can i erase the mpd control file in sources/mpd? Jan 18 22:15:45 that's what i meant by upgradeing mpd.mk, so that it can be as close to template.mk as possible Jan 18 22:16:31 yes, packages generating control file no longer need sources//control Jan 18 22:16:43 hmm Jan 18 22:17:17 we also have the added benefit that generated control file is no longer Architecture specific Jan 18 22:17:32 it can be built for other optware targets as well Jan 18 22:17:57 cool Jan 18 22:18:01 i'm updateing the variables Jan 18 22:18:07 i'll show them to you when i'm done Jan 18 22:18:25 great Jan 18 22:19:26 this is what i did: Jan 18 22:19:30 MPD_SITE=http://www.musicpd.org/uploads/files Jan 18 22:19:30 MPD_VERSION=0.12.1 Jan 18 22:19:30 MPD_SOURCE=mpd-$(MPD_VERSION).tar.gz Jan 18 22:19:30 MPD_DIR=mpd-$(MPD_VERSION) Jan 18 22:19:30 MPD_UNZIP=zcat Jan 18 22:19:31 MPD_MAINTAINER=NSLU2 Linux Jan 18 22:19:33 MPD_DESCRIPTION=Music Player Daemon (MPD) allows remote access for playing music (MP3, Ogg Vorbis, FLAC, AAC, Mod, and wave files) and managing playlists. Jan 18 22:19:36 MPD_SECTION= Jan 18 22:19:38 MPD_PRIORITY=optional Jan 18 22:19:40 MPD_DEPENDS=libao, libmad, libid3tag, libogg, libvorbis Jan 18 22:19:42 MPD_SUGGESTS= Jan 18 22:19:44 MPD_CONFLICTS= Jan 18 22:19:58 MPD_SECTION should not be empty Jan 18 22:20:09 it probably should be audio Jan 18 22:20:31 ok Jan 18 22:20:52 I think MPD_DEPENDS also needs flac, libvorbisidec Jan 18 22:21:23 have you added lines 167-180 of make/template.mk to make/mpd.mk ? Jan 18 22:22:19 not yet Jan 18 22:22:27 i'm trying to get those variables right first Jan 18 22:22:36 please do, and replace with MPD Jan 18 22:22:50 replace with mpd Jan 18 22:22:53 ya, i'm not that much of a newb Jan 18 22:22:54 =) Jan 18 22:23:08 i'm not not a linux or a make file guy Jan 18 22:23:38 so eno, in mpd.mk file there is this section that starts with: # Jan 18 22:23:38 # This builds the IPK file. Jan 18 22:23:38 # Jan 18 22:23:57 # starts comment Jan 18 22:24:02 is that where i put template.mk lines 167-180? Jan 18 22:24:43 according to make/template.mk, you should add 167-180 before # This builds the IPK file Jan 18 22:24:56 yeah, i just noticed that Jan 18 22:25:01 although the postition is not that significant Jan 18 22:25:09 in this case Jan 18 22:25:40 if you start from scratch, you would copy make/template.mk to make/mpd.mk Jan 18 22:25:51 this is what i have now: Jan 18 22:25:53 it just happened that we have an old mpd.mk here Jan 18 22:25:55 $(MPD_IPK_DIR)/CONTROL/control: Jan 18 22:25:55 @install -d $(@D) Jan 18 22:25:55 @rm -f $@ Jan 18 22:25:55 @echo "Package: MPD" >>$@ Jan 18 22:25:55 @echo "Architecture: $(TARGET_ARCH)" >>$@ Jan 18 22:25:56 @echo "Priority: $(MPD_PRIORITY)" >>$@ Jan 18 22:25:58 @echo "Section: $(MPD_SECTION)" >>$@ Jan 18 22:26:00 @echo "Version: $(MPD_VERSION)-$(_IPK_VERSION)" >>$@ Jan 18 22:26:02 @echo "Maintainer: $(MPD_MAINTAINER)" >>$@ Jan 18 22:26:04 @echo "Source: $(MPD_SITE)/$(_SOURCE)" >>$@ Jan 18 22:26:06 @echo "Description: $(MPD_DESCRIPTION)" >>$@ Jan 18 22:26:08 @echo "Depends: $(MPD_DEPENDS)" >>$@ Jan 18 22:26:10 @echo "Suggests: $(MPD_SUGGESTS)" >>$@ Jan 18 22:26:12 @echo "Conflicts: $(MPD_CONFLICTS)" >>$@ Jan 18 22:26:22 there're a couple of not replaced Jan 18 22:26:32 do a global search and replace Jan 18 22:26:36 fixed Jan 18 22:27:02 yeah, that will get you going, you can even delete sources/mpd/control now Jan 18 22:27:23 what about these lines: Jan 18 22:27:24 install -m 644 $(MPD_SOURCE_DIR)/control $(MPD_IPK_DIR)/CONTROL/control Jan 18 22:27:24 install -m 755 $(MPD_SOURCE_DIR)/postinst $(MPD_IPK_DIR)/CONTROL/postinst Jan 18 22:27:24 install -m 755 $(MPD_SOURCE_DIR)/prerm $(MPD_IPK_DIR)/CONTROL/prerm Jan 18 22:27:24 echo $(MPD_CONFFILES) | sed -e 's/ /\n/g' > $(MPD_IPK_DIR)/CONTROL/conffiles Jan 18 22:27:44 do those stay? Jan 18 22:28:29 remove the first line Jan 18 22:28:39 the other lines stay Jan 18 22:28:40 this one: install -m 644 $(MPD_SOURCE_DIR)/control $(MPD_IPK_DIR)/CONTROL/control? Jan 18 22:28:49 yes Jan 18 22:29:09 ok, i'm gonna delete the control file and remake Jan 18 22:29:50 i hope those tabs don't mess things up Jan 18 22:30:24 so no one here uses mpd Jan 18 22:30:34 eno: what do you use your slug for? Jan 18 22:31:47 i'm more interested in the area of web servers and different computer languages Jan 18 22:32:14 do you run servers for outside your network or inside? Jan 18 22:32:27 just inside Jan 18 22:32:55 do you do it for fun or for practice? Jan 18 22:33:05 mostly for fun Jan 18 22:33:47 how does your slug help you learn different lanugages? like just by making things compile? Jan 18 22:34:43 it's one way of evaluating technologies, to see if it can scale down or not Jan 18 22:34:59 i only know c++, java, and lisp Jan 18 22:35:03 what do you know? Jan 18 22:35:12 http://www.nslu2-linux.org/wiki/Profiles/BrianZhou Jan 18 22:35:28 you know lisp, very impressive Jan 18 22:36:16 unfortunately ecl (lisp) is broken on slug right now Jan 18 22:36:31 i see you are working on that Jan 18 22:37:46 eno: make[1]: *** No rule to make target `/CONTROL/control'. Stop. Jan 18 22:39:27 scant: this should be an exercise Jan 18 22:39:43 i'm exercising Jan 18 22:39:46 ;P Jan 18 22:40:45 can you do a "grep /CONTROL/control make/mpd.mk" ? Jan 18 22:42:43 eno: yeah, it appears twice Jan 18 22:43:21 paste the output here Jan 18 22:43:38 $(MPD_IPK_DIR)/CONTROL/control: Jan 18 22:43:38 $(MAKE) $(MPD__IPK_DIR)/CONTROL/control Jan 18 22:43:38 # install -m 644 $(MPD_SOURCE_DIR)/control $(MPD_IPK_DIR)/CONTROL/control Jan 18 22:43:57 MPD__IPK_DIR Jan 18 22:44:09 should be MPD_IPK_DIR Jan 18 22:44:38 * rwhitby notes that eno's commit rate is now at 9.77 hours, edging up on my 7.91 hours on average between commits. Jan 18 22:45:28 http://cia.navi.cx/stats/project/nslu2-linux Jan 18 22:45:30 bringing optware to slugosbe helps my commit rate Jan 18 22:46:07 129 nslu2-linux developers. Jan 18 22:46:45 hi rwhitby Jan 18 22:46:56 Hmm Jan 18 22:47:03 hi NAiL Jan 18 22:47:05 I'm down to 16.41 hours Jan 18 22:47:07 hi scant Jan 18 22:47:41 what's a commit rate? how often you feed stuff back? Jan 18 22:48:32 how often, on average, you commit a change to the nslu2-linux repositories. Jan 18 22:49:22 rhitby: are you #1 on commits? Jan 18 22:49:44 If I combine my usernames, I'm at 12.9 hours. Still a lot less than it used to be :-P Jan 18 22:50:06 I was close to having a commit every second hour ;) Jan 18 22:50:15 rwhitby's commit rate is very hard to beat Jan 18 22:51:16 yup Jan 18 22:51:20 oh well Jan 18 22:51:21 we have had 9510 commits in total. I have 2611 of those in 2.36 years, eno has 1644 in 1.83 years. Jan 18 22:51:22 * NAiL goes to bed Jan 18 22:51:41 I've got a total of 1038 commits over 1.5x years Jan 18 22:51:41 night NAiL Jan 18 22:51:46 1.54, I think. Jan 18 22:51:47 g'nite NAiL Jan 18 22:52:34 ah, 1038+89 :P Jan 18 22:52:56 bye NAiL Jan 18 22:52:58 nite al Jan 18 22:54:22 argh Jan 18 22:54:24 another error Jan 18 22:56:02 eno: r u still there? Jan 18 22:56:39 yes Jan 18 22:57:13 Package: MPD -> Error: Package name contains illegal characters, (other than [a-z0-9.+-]) Jan 18 22:57:18 change it to Package: mpd? Jan 18 22:58:41 yes, the line should be @echo "Package: mpd" >>$@ Jan 18 22:58:55 why must it be lowercase? that is weird Jan 18 22:59:09 what's the resoning behind package names being lowercase? Jan 18 22:59:14 style only? Jan 18 22:59:19 style Jan 18 22:59:35 ok, i changed it, recompiling... Jan 18 23:00:08 in fact, you can just rm builds/mpd_*.ipk Jan 18 23:00:27 too late Jan 18 23:00:29 hehe Jan 18 23:00:30 and then "make mpd-ipk" Jan 18 23:00:44 thanks for the tip though Jan 18 23:01:33 it's also a good idea to add the mpd-check target according to make/template.mk Jan 18 23:01:46 to make/mpd.mk Jan 18 23:02:01 woah Jan 18 23:03:22 MPD-check: $(MPD_IPK) Jan 18 23:03:22 perl scripts/optware-check-package.pl --target=$(OPTWARE_TARGET) $(MPD_IPK) Jan 18 23:03:26 is that it? Jan 18 23:03:54 yup Jan 18 23:04:24 hm, what does that do? Jan 18 23:04:27 sanity check? Jan 18 23:04:34 try it Jan 18 23:04:57 i guess it's checking syntax or something Jan 18 23:05:07 actually, you should use mpd-check and not MPD-check Jan 18 23:05:59 hmmm Jan 18 23:06:34 i think it all worked Jan 18 23:06:44 how about mpd-check? Jan 18 23:07:21 i haven't done that yet Jan 18 23:08:09 if there's error from "make mpd-check" then it's not ready Jan 18 23:08:09 it works, i have integer decoding of ogg with mpd now Jan 18 23:09:37 send your make/mpd.mk to the mailing list and request svn commit access Jan 18 23:10:08 eno: i have one last problem... but i think i wore you out Jan 18 23:10:43 rwhitby: i can't pm cause i'm not registered or something Jan 18 23:11:00 you should definitely setup a cross build environment if you want to do any more change Jan 18 23:11:10 eno: let's have a dialoge out mpd though Jan 18 23:11:33 here is the issue: i wanted mpd on my slug, so i tried to see if there were packages avail Jan 18 23:12:01 k Jan 18 23:12:23 but the alsa, libao, and mpd packages aren't available Jan 18 23:12:45 you can help us change that ;-) Jan 18 23:13:23 rwhitby: what is this e-mail address used for? Jan 18 23:13:44 scant: so I can get your details to get you svn write access Jan 18 23:13:54 eno: so this mpd.mk file was only one part of the issue, namely, updating it to a new version Jan 18 23:14:19 scant: we use openssl client certificates for access control. Jan 18 23:15:13 if you do runtime test of the alsa/libao/mpd chain, i can take care (or help you cross test) the native -> cross part Jan 18 23:15:56 the mpd package still isn't ready tho Jan 18 23:16:00 i mean it is, but it isn't Jan 18 23:16:16 whoever made the original package i think messed up on some additional parts Jan 18 23:17:25 rwhitby: i sent, might bounce because i changed the header of the e-mail, i hope it says the correct e-mail address Jan 18 23:17:49 someone added make/mpd.mk 2005-02-24 and it was never made READY_FOR_TESTING Jan 18 23:18:35 eno: what does that mean exactly? that someone put it in the feed for public use w/o testing it? Jan 18 23:19:01 no, adding make/mpd.mk != putting it in the feed Jan 18 23:19:13 oh Jan 18 23:19:17 for new packages, that's not unusual. if it builds and doesn't segfault when run, that's good enough to commit the .mk file. Jan 18 23:19:26 eno: can i describe the last issue with mpd that i have noticed? Jan 18 23:19:37 (and often good enough to push to the feed so users can give feedback on it too) Jan 18 23:20:27 scant: look at optware/Makefile and see how things are organized Jan 18 23:20:49 go ahead with the last issue Jan 18 23:22:04 eno: first off i see libao needs to be fixed in the Makefile Jan 18 23:23:02 i make mpd as root, is that ok/secure? Jan 18 23:23:21 that's not necessary and not recommended Jan 18 23:23:33 Optware must build as a non-root user. Jan 18 23:23:42 should i try making under non-root? Jan 18 23:24:32 always. Jan 18 23:25:15 rwhitby: did you get my e-mail? Jan 18 23:26:41 scant: yes, and replied. Jan 18 23:27:02 bbiab Jan 18 23:27:16 eno: did you disappear? Jan 18 23:31:48 i'm here Jan 18 23:32:19 eno: i don't understand unslug Jan 18 23:32:22 you need to be more specific regarding what needs to be fixed in make/libao.mk Jan 18 23:32:56 before i do anything else, i need to compile as non-root Jan 18 23:33:05 but i don't understand users with unslug Jan 18 23:33:20 is it the case that in order to add a new user i have to do it via a shell? Jan 18 23:33:21 that is not true Jan 18 23:34:14 you can do a lot with binary packages: http://ipkg.nslu2-linux.org/feeds/optware/nslu2/cross/unstable Jan 18 23:34:35 wait, i'm confused, are you talking to me? Jan 18 23:34:37 you can add new user with the web interface Jan 18 23:35:00 yeah, it's not working Jan 18 23:35:26 * kop_ just disassembled his DNS-323 to take some hi-res pictures of its board Jan 18 23:40:26 is rwhitby back? Jan 18 23:40:58 scant: adding a new user basically involves: 1) mkdir -p /home/ Jan 18 23:41:33 2) editing /etc/passwd to copy a line Jan 18 23:42:06 3) customize the new line in /etc/passwd Jan 18 23:42:28 such as user id, user name, prefered shell Jan 18 23:42:48 4) chown -R scant /home/scant Jan 18 23:43:01 eno: OR i can do it via web interface? Jan 18 23:43:10 5) http://www.nslu2-linux.org/wiki/HowTo/ChangePasswordsFromTheCommandLine Jan 18 23:43:29 yes, for new ppl, web interface is a good option Jan 18 23:44:19 well... i create a new user via Administration->Users and create a private folder share, but i can't login as that new user Jan 18 23:44:56 try step 5 of the above Jan 18 23:46:47 try "su - scant" Jan 19 00:18:36 enu: u there? Jan 19 00:26:17 eno: pm me when u get back or something Jan 19 00:29:36 eno: getmail->py-setuptools configure is failing Jan 19 00:31:04 hmm - might be a host staging collision Jan 19 00:33:08 eno: please ignore - looks to be transient Jan 19 00:36:23 If one is trying to create a package for SlugOs, and it requires the kernel source to build, can it be done on a cross-build machine or is it best to do natively on the slug? Jan 19 00:43:55 it should be able to be done on a cross-build machine Jan 19 00:44:11 it's almost never best to compile on the slug. Jan 19 00:46:15 kapookee: for a SlugOS package, you're using openembedded, which has specific support for kernel source compiles in a cross environment. Jan 19 00:58:35 scant: cert sent Jan 19 01:02:43 rwhitby: okay, that'll get me to research further. Jan 19 01:08:39 rwhitby: can u help me with adding a new non-root user? Jan 19 01:10:26 i added it via the web interface but it's doesn't seem like it's an interactive account **** ENDING LOGGING AT Fri Jan 19 02:59:58 2007