July 2012 - openSUSE Factory - openSUSE Mailing Lists

[opensuse-factory] Processing submissions of fixes for 12.2
by Jan Engelhardt 03 Jul '12

03 Jul '12

Submissions to Factory are not forwarded to 12.2 by the factory maintainer even if they are clearly labeled as fixes. Submissions to openSUSE:12.2 are automatically rejected by BS with the message one should create a request to openSUSE:12.2:Update instead. So we end up having a build-up of RPMs in /update/12.2 already. Do these update packages ever migrate to the 12.2-base? Otherwise, it seems like the fixes won't be on the goldmaster when the time comes. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe(a)opensuse.org To contact the owner, e-mail: opensuse-factory+owner(a)opensuse.org

4 4

[opensuse-factory] 12.2 installation report
by Lew Wolfgang 03 Jul '12

03 Jul '12

Hi Folks, I'm not sure this is the right place to report this and apologize in advance if factory is the wrong venue. I needed to purchase a new laptop for myself (not for work) last week and decided to get a fairly nice and current one. Settled on a Samsung 15-inch Series 9 NP900X4C-A01US. The important bits are its Intel HD Graphics 4000 system and an Intel Centrino Advanced-N 6235 for wifi and Bluetooth 4.0. It also has a 128-GB mSATA SSD and 8-GB of ram. It also came with Windows 7 Home 64, which I immediately blew away, of course. I had a bit of time and thought I'd try 12.2 beta (openSUSE-DVD-Build0452-x86_64.iso) on a lark. I figured that 12.2 would have a better chance of driving the 6235 wifi system. Alas, I couldn't get it installed. I tried two times, the first with a custom install using a 100-mb ext-2 boot partition, an 8-GB swap, a 24-GB root partition using Reiser with the free-hog for /export/home also with Reiser. The install proceeded to the first boot when it issued a warning that grub2 couldn't install on an unmountable partition. The screen gave me the option of changing things around and I then selected grub1. Grub seemed to work okay and a green screen with the Gecko and fuzzy white blobs floating around appeared. A white progress bar also appeared. I let it go for a bit but the only things that happened was the fuzzy blobs stopped moving around as much, but the progress bar continued to move right at a very slow pace. I hit escape at a point a couple of minutes into this and saw what appeared to be the kernel log spewing out boot information, but it seemed like it was in a loop. It was moving too fast for me to draw a bead on what it was saying. After what appeared to be a half-dozen loops the green Gecko screen reappeared and froze. Bummer. So I tried another install, this time going completely with the defaults. I didn't customize a thing. This time grub2 didn't complain and showed the new grub2 selection screen. I hit the default selection and it proceed to boot to the same green Gecko screen and froze. ESC did nothing. CTL/ALT/DEL didn't work, a hard power cycle was the only way to break the cycle. Bummer. Apparently grub2 doesn't grok Reiserfs, even with an ext-2 /boot. So I tried a new install with 12.1 and things went well enough. Bluetooth worked (mouse), sound worked, and the function key volume controls work. Initially the screen brightness buttons worked, but after a zypper dup they failed. More importantly, wifi doesn't work. The kernel sees the chipset and tries to bring it up, but fails. The latest Knoppix also fails wifi. Graphics work well enough, but glxgears shows a frame rate of about 58 frames per second. The screen seems okay, however. I'd be happy to file a bug report, but there's not enough for me to hang a bug on, that I can see. Just "install fails, green Gecko gets in way of seeing action". I hate to admit this, but my 8-year old grand daughter was clamoring to use the new laptop, so I tried Ubuntu. It installed easily and everything works as intended. Even the web cam. (didn't try the mic) I've been using SuSE since 5.2 and have made hundreds of installs. I don't intended to change, but that Ubuntu install sure was slick... My grand daughter has been using openSuSE since she was 4-years old, but I'd rather she use Ubuntu than Windows. I've got a desktop at work that needs an install, I'll try 12.2 again and see what happens. Regards, Lew -- To unsubscribe, e-mail: opensuse-factory+unsubscribe(a)opensuse.org To contact the owner, e-mail: opensuse-factory+owner(a)opensuse.org

4 5

[opensuse-factory] Disk not bootable after zypper up with grub install problem
by Jogchum Reitsma 03 Jul '12

03 Jul '12

Hi, Yesterday I did a zypper up on the 12.2 beta 2 (Mantis) located on my test disk. Some 2000 packages were to be upgraded, one of them grub 2. Installation of grub halted at 100%, so I decided to give it a Ctr-C, whereupon zypper complained that grub was unable to read from /dev/fd0. No surprise, since a floppy station is present, but no floppy in it. Install of grub being at 100% I chose Ignore, and the install of the rest went on OK. But now I cannot boot from that disk anymore. It's only a test disk, so no harm done, but is the a chance to make it bootable again whthout a fresh install of 12.2? I should add that I have a working 12.1 disk set on the same machine (which I 'm using right now). Besides that, the install behaviour seems wrong to me. In can't find a bug (against grub) for this one, I take it I should file one? regards, Jogchum -- To unsubscribe, e-mail: opensuse-factory+unsubscribe(a)opensuse.org To contact the owner, e-mail: opensuse-factory+owner(a)opensuse.org

2 2

[opensuse-factory] taglib-sharp
by Nelson Marques 03 Jul '12

03 Jul '12

I'm updating taglib-sharp for 2.0.5.0 which was released ~2 days ago. Forwarding to Factory from Banshee/taglib-sharp. NM -- Nelson Marques // I've stopped trying to understand sandwiches with a third piece of bread in the middle... -- To unsubscribe, e-mail: opensuse-factory+unsubscribe(a)opensuse.org To contact the owner, e-mail: opensuse-factory+owner(a)opensuse.org

1 0

[opensuse-factory] Multiple packages fail in 12.2 while build well for Factory and 12.1
by Ilya Chernykh 03 Jul '12

03 Jul '12

Hi! I have multiple packages failing in 12.2 while build well for Factory and 12.1. I wonder why it is so. Examples include kde3-kdar https://build.opensuse.org/package/live_build_log?arch=i586&package=gtk1-co… gtk1-compat https://build.opensuse.org/package/live_build_log?arch=i586&package=kde3-kd… hal https://build.opensuse.org/package/live_build_log?arch=i586&package=hal&pro… -- To unsubscribe, e-mail: opensuse-factory+unsubscribe(a)opensuse.org To contact the owner, e-mail: opensuse-factory+owner(a)opensuse.org

1 0

[opensuse-factory] KDE's webcam application: kamoso
by Ray Chen 02 Jul '12

02 Jul '12

https://build.opensuse.org/package/show?project=KDE%3AExtra&package=kamoso kamoso works fine in my openSUSE 12.2beta2 I've used this application since 12.1 why not commit to Factory and include this application in KDE Desktop and do some integrations e.g. when edit personal information, you can call kamoso to take a picture as avatar Cheers -- To unsubscribe, e-mail: opensuse-factory+unsubscribe(a)opensuse.org To contact the owner, e-mail: opensuse-factory+owner(a)opensuse.org

4 3

[opensuse-factory] Re: [opensuse] problems building perl 5.14.2 or 5.16.0 using RPMBUILD, build from tar or perlbrew (a cpan util)
by Linda Walsh 02 Jul '12

02 Jul '12

Philipp Thomas wrote: > * Linda Walsh (suse(a)tlinx.org) [20120625 20:28]: > >> COLLECT_GCC_OPTIONS='-v' '--version' '-mtune=generic' '-march=x86-64' >> cc1 -quiet -v -iprefix ../lib64/gcc/x86_64-suse-linux/4.6/ >> help-dummy -quiet -dumpbase help-dummy -mtune=generic -march=x86-64 >> -auxbase help-dummy -version --version -o /tmp/ccggVtI6.s >> gcc: error trying to exec 'cc1': execvp: No such file or directory > > No wonder! Something like COMPILER_PATH=/usr/lib64/gcc/x86_64-suse-linux/4.6/ > is missing in the output so it's no wonder the compiler driver doesn't find > the languge specific compilers which reside in a version specific directory. > Which version of SuSE is this, I assume 12.1? Is this the gcc that comes with > that distribution? > > Why don't you use 'osc build' which will build the package locally but > fetches all that's needed from the obs? > > Philipp > ---- Why doesn't the compiler look where it needs to when it is run? Seems like using a special build system to get around a GCC bug is missing the point. What variables is GCC not setting so it can call it's "cc1" phase"? My PATH is set ok, but we never used to have to set internal paths for compilers for them to work. Seriously -- this is a good reason NOT to build on obs -- because things like this don't get discovered. You didn't see Ritchie using obs when he build "Hello World" -- But at that point I couldn't even write "Hello World " and have it work -- that's pretty bad. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe(a)opensuse.org To contact the owner, e-mail: opensuse-factory+owner(a)opensuse.org

5 5

[opensuse-factory] Fwd: commit xf86-input-synaptics for openSUSE:12.2
by Dominique Leuenberger a.k.a DimStar 02 Jul '12

02 Jul '12

Hi, thanks for patching up Xorg again (and sorry: I was the one removing some of those patches)... BUT are we actually doing a better job this time around and are sending patches upstream? We'll end up with a mess again that nobody wants and can touch, as the patches ARE a pain to maintain... let's not be 'the elite' that has to keep everything for themselves... let's be good downstreams and contribute our patches to upstream, shall we? Yes, I know it's work. Sometimes upstreams don't want our patches. In this case there are two ways: - ignore upstream, keep the patch and maintain it forever - Understand the reasoning of upstream. They sometimes even do have a point :) Dominique

1 0

[opensuse-factory] new & improved 'spwd' function for bash prompt?
by Linda Walsh 02 Jul '12

02 Jul '12

I liked the spwd idea and ended up adjusting it to be a bit more precise. I just did some more testing with it to make sure it handled complicated directories without choking (though if they embed control chars, I don't filter them out... )... But rather than just looking at the number of dirs and displaying the whole path if <=3 elements or ..., like the current one does -- this one uses a progressive shortening, based on the length of the directory names. It aims to keep the total length down to 1/2 the # columns (so 40 on a standard screen)... Despite my many manipulations of the path, It works with dir paths like: dp> ll total 0 drwxrwsr-x 2 6 Jul 1 13:42 ^ difficult dir(to) "parse" ['correctly'], *{maybe}? ~; or #not/ drwxrwsr-x 2 6 Jul 1 14:35 _^ d r(2)s['rt'],*{m}; #r ~ ! \\not?/ The first one is too long to display on an 80 column screen, but it displays (weirdly) when used (it has a line-feed in the dir name): Ishtar:law/dp/^ difficult dir(to) "parse" ['correctly'], *{maybe}? ~; or #not> _ Ishtar:law/dp> cd _\^\ d\ r$2$s\[\'rt\'\]\,\*\{m\}\;\ #r\ ~\ \!\ \\\\not\?/ Ishtar:.../_^ d r(2)s['rt'],*{m}; #r ~ ! \not?> _ The cut-down in paths was put in an array: alias int=declare\ -i _e=echo _pf=printf exp=export ret=return exp __dpf__='local -a PF=( "/$1/$2/$3/../\${$[$#-1]}/\${$#}" "/$1/$2/../\${$[$#-1]}/\${$#}" "/$1/../\${$[$#-1]}/\${$#}" "/$1/../\${$#}" ".../\${$#}" "..." )' It tries to use a bit of the first part and some of the last part... Using more if you have dirs like "a/b/c...y/z". The function is: function spwd () { \ (($#)) || { _e "spwd called with null arg"; return 1; }; \ int w=COLUMNS/2 ;\ ( _pf -v _p "%s" "$1" ; export IFS=/ ;\ set $_p; shift; unset IFS ;\ t="${_p#${HOME%${USER}}}"; int tl=${#t} ;\ if (($#<=6 && tl<w));then ((tl<=2)) && \ { _e -En "${_p}";return 0; } ;\ else \ eval "$__dpf__" ;\ int i pfl=${#PF[*]} ;\ for ((i=0; i<pfl; ++i)); do eval \ "_pf -v _pa %s \"${PF[i]}\"" ;\ _p="$(eval "_pf %s \"$_pa\"")" ;\ ((${#_p}<w)) && break; done ;\ fi ;\ _e -En "${_p#${HOME%${USER}}}" ) } ---- I've been using some form for a few years...had a bunch of conditions in the function at one point...looked bad - so I extracted them into an array and loop through that, instead, and use the first one that passes my length test. The aliases aren't really needed if you want to sub in the text, but if you want to use it as is, you should turn on "shopt -s expand_aliases"... -l -- To unsubscribe, e-mail: opensuse-factory+unsubscribe(a)opensuse.org To contact the owner, e-mail: opensuse-factory+owner(a)opensuse.org

1 0

[opensuse-factory] bug in "src" rpms "arch" field (in 11.4 & 12.1) -- i.e. a process problem.
by Linda Walsh 02 Jul '12

02 Jul '12

I was writing a util to examine poorly named 'rpm' files to see what was in them and what architecture they were for. I was surprised to find out that all of the "src" rpms don't say "arch=src" or "arch=none", but arch=i586 OR arch=x86_64... Are there separate sources for each architecture? If not, could this information be set that "src"'s "arch" field indicates it is "arch" agnostic"? -- so one can look at the arch field and know it's not a binary installable? I'm querying the fields inside the RPM's because historically, the names on the outside of the RPM were often difficult to parse -- especially in relation to the NVR fields. I sorta got caught with with my "pants down", so to speak, relying on querying the information inside the RPM, entirely, thinking that doing so would be the "Authoritative" source for the rpm's contents... -- To unsubscribe, e-mail: opensuse-factory+unsubscribe(a)opensuse.org To contact the owner, e-mail: opensuse-factory+owner(a)opensuse.org

3 3