sitquietly

Over in our upstream, the Gentoo portage tree, kde-frameworks ebuilds were updated to the 5.14 release. I think that update occurred 18 days ago. We still haven't pulled those updates into our Funtoo portage tree. I don't know if this is a normal delay pending some review by our own team (thank you all!) or if there is some particular problem. Maybe syncing is problematic right now as Gentoo switches to git? How much delay is normal between updates to Gentoo portage and syncing them into Funtoo portage? I appreciate help understanding the process --- the delay is problematic for me right now because I use the Gentoo KDE Overlay. When that overlay moves packages from the overlay into the main Gentoo portage tree they "disappear" from my Funtoo system and suddenly my system is in an inconsistent state and can't be updated.

I sometimes boot into Funtoo ~amd64, and sometimes boot into FreeBSD 10 stable with current (HEAD) ports. They are on separate drives with their own boot managers but share a common /home on a ZFS mirrored pool. I'd like to try running the 32-bit Linux google-chrome in FreeBSD. It appears to be almost possible to do that using the CentOS 6.x system from ports and utilizing the FreeBSD "linuxulator". In other words FreeBSD can run Linux binaries "natively" (but it only emulates the Linux 2.6.x kernel). I can chroot into the FreeBSD linux subsystem at /compat/linux and run Linux gui apps using the FreeBSD X server. e.g. # script to chroot into bash in linux cp /etc/resolv.conf /compat/linux/etc/ mount -t linprocfs linproc /compat/linux/proc mount_nullfs /dev /compat/linux/dev mount_nullfs /sys /compat/linux/sys mount -t fdescfs fdesc /compat/linux/dev/fd export DISPLAY="192.168.1.1:0.0" xhost +localhost DISPLAY=localhost:0 chroot /compat/linux /bin/bash but I run into problems with missing libraries for Google Chrome -- it would be so much more flexible to just put Funtoo in /compat/linux and use portage to install stuff into that chrooted system. Before I spend time installing Funtoo into a FreeBSD chroot (or optionally a jail) I would appreciate hearing from anyone who has tried running a recent Gentoo or Funtoo in a FreeBSD chroot or jail. Any advice? Pitfalls? Is there any reason it can't work? Can pigs fly?

I installed a hardened amd64 profile for three KDE / xmonad / fluxbox desktops. One by one I had to abandon the hardened profile on each machine and switch them to pure64. "Hardened" was easy to install and mostly well supported, but over time I ran into repeated blocks and emerge failures with that profile. Often problems with hardened packages are fixed within a few days but not always. It was sufficiently frustrating to motivate the effort to switch away from it. Some problems were not just that packages could not be emerged but that my own local software and programming projects would fail because they were incompatible with some hardening feature. I gradually turned off PaX features trying to keep my own stuff compileable. I still run hardened on my file server and am glad for the extra security there. The simpler your "desktop", the more likely that you'll be happy with a hardened system, but it just receives less feedback and bug fixing to make everything work all of the time. It would be a great service, maybe a worthwhile project for you, if you would install a hardened system and commit to making bug reports for every problem that arises over time.

Probably reading `man boot.conf` and considering the examples there will resolve the problem for you. You will need to edit /etc/boot.conf to describe your setup for boot-update. Especially notice that the line in the kernel entry that names the kernel, e.g. kernel vmlinuz[-v] must correspond to your own kernel naming convention. If the kernel in /boot/ have names like bzImage-3.9-gentoo then the line must be kernel bzImage[-v] but if they are named like kernel-3.9-gentoo then the line must be kernel kernel[-v]

Your requirements list probably excludes all of the binary distros. I've used Arch/Manjaro extensively and can't agree that they are light in any sense except that the original install is a small base. After that their packages have the same extensive, all-inclusive, dependencies as Fedora. For me it was impossible to accept anything less than Funtoo for my last notebook installation (Toshiba Satellite with amd dual-core, radeon graphics - 4GB ram). I did want to know if everything was going to work so I got the latest Calculate Linux iso, partitioned the drive with a swap, two "root" parttiions and a home partition, installed Calculate KDE, and tested the wireless, trackpad, etc. Calculate could easily be the end of the line; it is built with Gentoo, uses standard emerge and allows you to build packages with your own USE flags if necessary. It is a really nice and very flexible binary distro that allows you to "lighten" it as much as you need. Of course it uses OpenRC. I suspect that it ticks all of your checkboxes. But to get exactly what I want, which is a system that uses openrc and avoids unnecessary or trouble-prone daemons (e.g. avahi, pulseaudio, dbus, kdepim, notifications) I build my own set of binary packages on my workstation and share those via http so that I can install Funtoo on the other partition, drawing packages from my own binary repo. In order to support different USE flags and CPU_FLAGS_X86 and CFLAGS in the repo than I have on my workstation I do the package building in a chroot, rather like is done with Poudriere for FreeBSD.

Now I'm laughing... That was a silly question. :D But the answer is probably just as silly. I just haven't discovered my own silliness yet. Thank you for all of your inputs, it has been a really useful sanity check. I may just start over in a few days with a fresh stage3 and a fresh mind. [i've done a half dozen Funtoo installs in the past year and it has always proven to be a dead-simple and foolproof process with current or stable, hardened or not, pure64 or multiarch. They've all installed sweetly.]

No. :( ping always complains of "No route to host" I even checked the Funtoo script /etc/init.d/netif.tmpl for changes but it is identical between the functioning and non-functioning installations. I haven't found any real difference between the good and bad systems, except that the bad system is a new install from a new stage3. I'm afraid of finding that I've done something really stupid (it would not be the first time), but I'm not seeing it.

Here is the netif.eth0 config file I've been using on my previous install (which is ok) and my new install (which does not work): template="interface" ipaddr="192.168.1.1/24" gateway="192.168.1.254" nameservers="8.8.8.8 208.67.222.222" domain="launchmodem.com" Has it been wrong all along??

Here's the addressing for my existing Funtoo installation (I'm accessing this forum using these settings) __> ip route default via 192.168.1.254 dev eth0 127.0.0.0/8 dev lo scope host 192.168.1.0/24 dev eth0 proto kernel scope link src 192.168.1.1 So you see 192.168.1.254 is the address of my router, and 192.168.1.1 is the address of my workstation. So it still seems that my new installation is correctly setup and should be working. But isn't working. :blink:

ip route actually says what we would expect for the interface: default via 192.168.1.254 dev eth0 127.0.0.0/8 dev lo scope host 192.168.1.0/24 dev eth0 proto kernel scope link src 192.168.1.1 and netstat -r reports (after hanging for about a minute) Kernel IP routing table Destination Gateway Genmask Flags MSS Window irtt Iface default 192.168.1.254 0.0.0.0 UG 0 0 0 eth0 loopback 0.0.0.0 255.0.0.0 U 0 0 0 lo 192.168.1.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0 These look ok. But I can't connect to or ping the other computers on the 192.168.1.0/255 network. Can't get out to the internet of course. This seems very weird to me... The eth0 interface is UP. The routing table is correct (but netstat hangs for a long time). The dns info is correctly in /etc/resolv.conf. And the network is not accessible: # ping 192.168.1.2 # my file server PING 192.168.1.2 (192.168.1.2) 56(84) bytes of data. From 192.168.1.1: icmp_seq=1 Destination Host Unreachable From 192.168.1.1: icmp_seq=2 Destination Host Unreachable From 192.168.1.1: icmp_seq=3 Destination Host Unreachable

I've installed Funtoo Current core-i7 hardened onto an empty partition alongside my existing Funtoo system, and followed the usual easy steps http://www.funtoo.org/Funtoo_Linux_Networking for setting up a static interface on eth0, but there is no route to the local network. Everything boots ok. eth0 is UP with correct address and mask. resolv.conf looks good /etc/conf.d/netif.eth0 is the same as on my old (functioning) install BUT I can't ping other machines on my local network. netstat -r hangs for a long time and then does not show any entry for the local network. i.e. the line 192.168.1.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0 is missing [EDIT: if I wait long enough that line is printed] BTW, the interface comes up functioning if I use dhcp instead of the static addressing (on this network dhcp is reserved for guests and assigns addresses that lack certain priviledges). I don't understand how the system (using the Funtoo networking scripts) can fail to create an entry in the routing table for the eth0 interface. Anyone have an idea what could be wrong? I've already come to the end of my knowledge and don't know how to proceed on this networking problem. I've even tried to copy my old /etc directory onto the new system for testing, editing the fstab of course. The new install still fails to bring up a functioning network (using same kernel config, same configuration files in /etc, same hardware).

I definitely don't believe in magic, so it was embarrasing to report that the problem resolved with no apparent reason. Sometimes that suggests a race condition -- when I reported the problem I was "-j4" and I intended to check it with "-j1" but then it just compiled to completion with "-j4". Yes I'm using XFS for / but I compile in tmpfs. My report above shows PORTAGE_TMPDIR="/tmp", which is tmpfs, and notice that I don't use -pipe in my CFLAGS. (/home is on ZFS raidz.)

Two days later: eix-sync && emerge -auvDN --quiet-build @world and everything updated. And I have no idea what the problem was. But it resolved itself.

Yeah, it's Evil. It sometimes turns a person Evil. Be careful, don't read that last link, it could happen to you!

I enjoy your stories. Thanks. I don't like emacs either :D and really like vim's modality. It makes a lot of sense so I use a vi mode inside emacs and am mostly happy with org / latex and still my beloved editing modes and ex commands. yeah, thank the holy ones for Funtoo as a haven of sanity. I once did extensive work on '386 and '486 control systems that could handle hundreds of tasks with guarenteed latency of under a millisecond. Now our thousand times faster systems are ten times slower. :< AND they crash...

IMO the problem is systemic in the linux ecosystem. We have 10000 under qualified hackers working on a ball of strings adding more and more brightly colored strings (lots of focus on playing around with the GUI interface and little work on the hard stuff underneath); it's unraveling and now we have an under qualified and badly motivated cabal from Red Hat trying to change the core of the ball, the part that WAS working well. I'm in danger of ranting... After Funtoo it's FreeBSD for me. For now I am very happy with Funtoo without systemd, without pulseaudio, without avahi, without udisks and upower, without gstreamer, without gnome, without consolekit and policykit, running a solid zfs raidz storage system and a full complement of development tools (haskell, smlnj, ocaml, racket, etc.), a great publishing system (org-mode on emacs with latex output), a snappy fluxbox and rox desktop and mostly kde apps.

Oh yes the vapigen problem. I ran into just this identical error while building packages for Archlinux..... Damned if I can remember how I fixed it!@?#!! [rummaging through notes...] Well according to my notes this error is reported to occur in several distros including Archlinux, Ubuntu, and Gentoo. It was discussed in Gentoo forums, see here. It has been reported as an error to the Gnome developers and seems to recur because they keep changing this stuff. If I can think of a workaround for you I'll come back... Good luck with vala. It looks like a very nice modern (functional) language. I'll try installing gedit/vala later in the day and see if I learn anything about your problem...

Those kernel-headers are the reference headers for glibc. They are needed by the latest glibc and will not interfere with your chosen kernel. You'll be fine! You're right -- A funtoo installation should not be difficult. Sounds like you're in the middle of a long learning experience. In grad school (3 times over 3 decades) I always found if I didn't feel a very uncomfortable anxiety and a sense of chaos and confusion I wasn't learning anything. You must be learning! Yeah, _I_ know how to fix all of your problems. But I'm not sitting at your computer and you wouldn't like it if I was. I'll toss out some general advice if you need it. Maybe read some basic docs over again (I don't know your level of expertise so excuse me if you are already more expert than me). I'll go look for your vapigen thread... Best regards.

You might want to try Cross Linux From Scratch. It actually is very well designed and documented, easy to follow and shows how to really build a system from scratch. The team behind it seem to be top shelf engineers. Then there's Cross Linux From Scratch Embedded, using the musl c-library in place of the behemoth glibc. Also a good project and very educational. (yeah, I'm laughing). I've settled in with Funtoo too. Also gave Exherbo a skip. Never could find a reason to dabble there.

Today's emerge of the glibc update from 2.18-r1 to 2.19 failed to complete the compile phase (near the end). I'm on a profile of x86-64bit, current, desktop, hardened. All packages were up to date as of yesterday. I have two questions: (1) should I perhaps update binutils manually first with the old glibc and then try building a clean toolchain with the new glibc? (2) should I abandon the hardened profile, if that is part of the problem, and if so how to switch "hardened" off? The glibc error is here: make[3]: Entering directory '/tmp/portage/sys-libs/glibc-2.19/work/glibc-2.19/elf' make subdir=csu -C ../csu ..=../ objdir=/tmp/portage/sys-libs/glibc-2.19/work/build-x86-x86_64-pc-linux-gnu-nptl -f Makefile -f ../elf/rtld-Rules rtld-al l rtld-modules='rtld-check_fds.os rtld-errno.os rtld-divdi3.os rtld-init-arch.os' make subdir=dirent -C ../dirent ..=../ objdir=/tmp/portage/sys-libs/glibc-2.19/work/build-x86-x86_64-pc-linux-gnu-nptl -f Makefile -f ../elf/rtld-Rules r tld-all rtld-modules='rtld-closedir.os rtld-readdir.os rtld-rewinddir.os rtld-getdents.os rtld-fdopendir.os' make subdir=gmon -C ../gmon ..=../ objdir=/tmp/portage/sys-libs/glibc-2.19/work/build-x86-x86_64-pc-linux-gnu-nptl -f Makefile -f ../elf/rtld-Rules rtld- all rtld-modules='rtld-profil.os rtld-prof-freq.os' make subdir=io -C ../io ..=../ objdir=/tmp/portage/sys-libs/glibc-2.19/work/build-x86-x86_64-pc-linux-gnu-nptl -f Makefile -f ../elf/rtld-Rules rtld-all rtld-modules='rtld-xstat64.os rtld-fxstat64.os rtld-lxstat64.os rtld-open.os rtld-read.os rtld-write.os rtld-lseek.os rtld-access.os rtld-fcntl.os rtld-c lose.os' make[4]: Entering directory '/tmp/portage/sys-libs/glibc-2.19/work/glibc-2.19/csu' ../Makerules:252: *** Cannot allocate memory. Stop. make[4]: Leaving directory '/tmp/portage/sys-libs/glibc-2.19/work/glibc-2.19/csu' ../o-iterator.mk:9: recipe for target '/tmp/portage/sys-libs/glibc-2.19/work/build-x86-x86_64-pc-linux-gnu-nptl/csu/rtld-check_fds.os' failed make[3]: *** [/tmp/portage/sys-libs/glibc-2.19/work/build-x86-x86_64-pc-linux-gnu-nptl/csu/rtld-check_fds.os] Error 2 make[3]: *** Waiting for unfinished jobs.... make[4]: Entering directory '/tmp/portage/sys-libs/glibc-2.19/work/glibc-2.19/io' make[4]: Entering directory '/tmp/portage/sys-libs/glibc-2.19/work/glibc-2.19/dirent' make[4]: Entering directory '/tmp/portage/sys-libs/glibc-2.19/work/glibc-2.19/gmon' Notice the message "Cannot allocate memory" -- there is 16 Gbytes of ram here and plenty of free mem during this compile. My best wild guess is that my hardened binutils/gcc are throwing the error. Here's emerge --info glibc: __> emerge --info glibc Portage 2.3.6-r9 (funtoo/1.0/linux-gnu/arch/x86-64bit, gcc-4.8.2, glibc-2.18-r1, 3.16.5-gentoo x86_64) ================================================================= System Settings ================================================================= System uname: Linux-3.16.5-gentoo-x86_64-Intel-R-_Core-TM-_i5-2405S_CPU_@_2.50GHz-with-gentoo-2.2.0 KiB Mem: 16317652 total, 2232580 free KiB Swap: 13096956 total, 13096956 free Timestamp of tree: Unknown ld GNU ld (Gentoo 2.23.2 p1.0) 2.23.2 app-shells/bash: 4.3_p30 dev-lang/perl: 5.20.1-r2 dev-lang/python: 2.7.8-r1000, 3.3.5-r1000 dev-util/cmake: 3.0.2 sys-apps/baselayout: 2.2.0-r6 sys-apps/openrc: 0.12.3-r3 sys-apps/sandbox: 2.6-r2 sys-devel/autoconf: 2.13, 2.69 sys-devel/automake: 1.11.6, 1.13.4 sys-devel/binutils: 2.23.2 sys-devel/gcc: 4.8.2-r3 sys-devel/gcc-config: 1.8 sys-devel/libtool: 2.4.3-r2 sys-devel/make: 4.1-r1 sys-kernel/linux-headers: 3.17-r1 (virtual/os-headers) sys-libs/glibc: 2.18-r1 Repositories: gentoo stargate Installed sets: @kernel ACCEPT_KEYWORDS="amd64 ~amd64" ACCEPT_LICENSE="* -@EULA" CBUILD="x86_64-pc-linux-gnu" CFLAGS="-march=native -O2" CHOST="x86_64-pc-linux-gnu" CONFIG_PROTECT="/etc /usr/share/config /usr/share/gnupg/qualified.txt" CONFIG_PROTECT_MASK="/etc/ca-certificates.conf /etc/env.d /etc/fonts/fonts.conf /etc/gconf /etc/gentoo-release /etc/revdep-rebuild /etc/sandbox.d /etc/terminfo /etc/texmf/language.dat.d /etc/texmf/language.def.d /etc/texmf/updmap.d /etc/texmf/web2c" CXXFLAGS="-march=native -O2" DISTDIR="/usr/portage/distfiles" FEATURES="assume-digests binpkg-logs config-protect-if-modified distlocks ebuild-locks fixlafiles merge-sync news parallel-fetch preserve-libs protect-owned sandbox sfperms strict unknown-features-warn unmerge-logs unmerge-orphans userfetch xattr" FFLAGS="" GENTOO_MIRRORS="http://distfiles.gentoo.org" LANG="en_US.UTF-8" LDFLAGS="-Wl,-O1 -Wl,--sort-common -Wl,--as-needed" MAKEOPTS="-j4" PKGDIR="/usr/portage/packages" PORTAGE_CONFIGROOT="/" PORTAGE_TMPDIR="/tmp" PORTDIR="/usr/portage" PORTDIR_OVERLAY="/usr/local/portage" SYNC="git://github.com/funtoo/ports-2012.git" SYNC_USER="root" USE="X aac acl alsa amd64 apng avx bash-completion bs2b bzip2 cairo cdda cddb cdio cdr cracklib crypt cups cxx dbus djvu dot dri dvd dvdr dvdread encode ffmpeg fftw flac fluidsynth fontconfig frei0r gdbm gif gpm graphviz gtk guile hardened harfbuzz iconv icu id3tag ipv6 jpeg kipi ladspa lame latex lcdfilter libass mad matroska mmx modplug modules mp3 mp4 mpeg mudflap multilib ncurses nptl nsplugin ocaml ocamlopt ogg opengl openmp opus pam pango pax_kernel pcre pdf pic png postscript pppd python qt3support readline resolvconf speex spell sqlite sse sse2 sse4_1 sse4_2 ssl ssse3 svg t1lib taglib tcpd theora tiff truetype twolame udev unicode urandom vaapi vim-syntax vorbis wavpack win32codecs x264 x265 xattr xcb xft xml xmp xtpax xv xvid zlib" ABI_X86="64" ALSA_CARDS="ali5451 als4000 atiixp atiixp-modem bt87x ca0106 cmipci emu10k1x ens1370 ens1371 es1938 es1968 fm801 hda-intel ice1724 intel8x0 intel8x0m maestro3 trident usb-audio via82xx via82xx-modem ymfpci" ALSA_PCM_PLUGINS="adpcm alaw asym copy dmix dshare dsnoop empty extplug file hooks iec958 ioplug ladspa lfloat linear meter mmap_emul mulaw multi null plug rate route share shm softvol" APACHE2_MODULES="actions alias auth_basic authn_alias authn_anon authn_dbm authn_default authn_file authz_dbm authz_default authz_groupfile authz_host authz_owner authz_user autoindex cache cgi cgid dav dav_fs dav_lock deflate dir disk_cache env expires ext_filter file_cache filter headers include info log_config logio mem_cache mime mime_magic negotiation rewrite setenvif speling status unique_id userdir usertrack vhost_alias authn_core authz_core socache_shmcb unixd" CALLIGRA_FEATURES="braindump flow krita stage words" CAMERAS="ptp2" COLLECTD_PLUGINS="df interface irq load memory rrdtool swap syslog" ELIBC="glibc" GPSD_PROTOCOLS="ashtech aivdm earthmate evermore fv18 garmin garmintxt gpsclock itrax mtk3301 nmea ntrip navcom oceanserver oldstyle oncore rtcm104v2 rtcm104v3 sirf superstar2 timing tsip tripmate tnt ublox ubx" INPUT_DEVICES="evdev keyboard mouse wacom" KERNEL="linux" LCD_DEVICES="bayrad cfontz cfontz633 glk hd44780 lb216 lcdm001 mtxorb ncurses text" LIBREOFFICE_EXTENSIONS="presenter-console presenter-minimizer" OFFICE_IMPLEMENTATION="libreoffice" PHP_TARGETS="php5-5" PYTHON_ABIS="2.7 3.3" PYTHON_SINGLE_TARGET="python2_7" PYTHON_TARGETS="python2_7 python3_3" QEMU_SOFTMMU_TARGETS="i386 x86_64" QEMU_USER_TARGETS="i386 x86_64" RUBY_TARGETS="ruby21" USERLAND="GNU" VIDEO_CARDS="intel fbdev vesa" XTABLES_ADDONS="quota2 psd pknock lscan length2 ipv4options ipset ipp2p iface geoip fuzzy condition tee tarpit sysrq steal rawnat logmark ipmark dhcpmac delude chaos account" Unset: CPPFLAGS, CTARGET, EMERGE_DEFAULT_OPTS, INSTALL_MASK, LC_ALL, LINGUAS, PORTAGE_BUNZIP2_COMMAND, PORTAGE_COMPRESS, PORTAGE_COMPRESS_FLAGS, SYNC_UMASK ================================================================= Package Settings ================================================================= sys-libs/glibc-2.18-r1 was built with the following: USE="hardened (multilib) -debug -gd -nscd -profile (-selinux) -suid -systemtap -vanilla" CFLAGS="-march=native -O2 -fno-strict-aliasing" CXXFLAGS="-march=native -O2 -fno-strict-aliasing" Advice? [EDIT] I have had Funtoo pure64 - current - desktop - NOT hardened - running on another partition. Glibc and friends updated just fine there of course. If no one has other advice for me I'll switch back to that installation, add zfs to it and get it working with my zfs raidz pool. Am (Was) I the only one running a hardened profile on Funtoo?? This is my second show-stopper with "hardened": I couldn't run zfs on a hardened kernel and had to switch to gentoo-sources, and now can't emerge glibc-2.19 on a hardened toolchain.

I enjoyed your story about the thrift shop -- it'll happen to me some day. For now I'm happy with my "old" Sandy Bridge (Gigabyte board with Intel i5-2405S, it caps the power draw, great performance, built-in HD3000 graphics serves me very well with smooth 1080p playback etc.). 16 Gb ram (in four slots) allows me to dedicate 8 GB to /tmp for fast emerging of packages. While you have the chance consider building a super quiet system. I find that it really adds to the enjoyment. Getting rid of the power supply fan really silenced the system for me. A Seasonic 400 W fanless unit (e.g. from Newegg) is adequate for me to power my board, dual Intel Pro pci ethernet, ssd, and 3 1TB hard drives, and the power supply is running near its most efficient loading, and dead silent. I need only three 140mm fans in my case, all running at about 1000 rpm (silent). I love silence! Good luck on your new Funtoo system.

I've taken refuge in Funtoo too. My workstation, netbook and home server now run Funtoo because months of testing proved it to be more adaptable and reliable than Debian or Arch. systemd was part of the problem in the land of popular binary distros -- their forums show thousands of little problems cropping up but enthusiasm causes a blind eye. I wouldn't say that systemd is "not ready", nor that "it's a matter of choice." I don't see the problem being systemd itself but the binding of systemd to higher levels of software. It is bad practice. No "application" should care what init system was used to boot the computer. But the systemd cabal IS pushing forward on a plan to create an interlocked monolith. Here's an experiment for the curious to explore the implications of the interlocking of layers: create a self-hosting source-based linux system that builds itself from sources starting with the absolute minimum of executables (which will be bootstrapped by cross-compiling -- maybe use aboriginal linux). The goal would be to create a bootable system with less than 20 packages in the base and which is capable of building itself up into a complete system. This is a proof that your system is in fact a system (a single well-defined thing) and not just a pile of packages that happen to actually run at the moment. The base will need a compiler (gcc or clang), c library (glibc or musl), a linker and associated tools (binutils or toybox), an init system (toybox, busybox, bsd rc, sysvinit) and a package manager (something simple like butch from sabotage linux). This can be done. Now try to do it with systemd as the init system -- you just failed! The package dependencies just blew up into a huge set and you are entangled in a mess of circular dependencies. Thanks to Dantrell for proving the point that the lock-in is not actually a technical requirement (I run Gnome 3.12 with OpenRC -- no systemd required). Thanks to Robbins for keeping Funtoo rationally designed. Thanks to all you devs who know how to do good system design and keep Funtoo fun :D

Today I installed 3 new 1TB drives with the intention of configuring them as a zfs raidz pool. The zpool create command executed ok: # zpool create -o ashift=12 -f tank raidz ata-xxxx ata-yyyy ata-zzzz where the xxxx, yyyy, zzzz were the disk/by-id names. I now have a /tank filesystem with 1,8 TB available. But when I try to create a dataset the operation fails, leaving some processes busy on /tank but never completing: # zfs create tank/home # Fail I've repeated this process carefully, after re-initializing the zfs config. The error message this second time "Killed". I tried to list the contents of /tank to see if home had been created but the `ls` command locks up; the filesystem there is bad. Indeed the zfs create operation failed. I'm running funtoo current pure64 hardened with a desktop profile on the gentoo-hardened-3.15.10 kernel. Can anyone confirm that it is possible to run zfs on top of hardened kernel, hardened profile? Any ideas on how to proceed?? Of course I thought this installation was going to be very easy and straightforward, nothing can go wronikj@**!?... [EDIT] I installed a non-hardened kernel, gentoo-sources-3.16.5, and rebuilt zfs-kmod-0.6.3 for it. I can create datasets on the pool with the non-hardened kernel. On the same pool when I try to create a new dataset using the hardened kernel the operation fails with "Killed". I also find that I can not write to the existing datasets from within the hardened kernel. This appears to be the death knell for the hardened kernel on this system unless someone can help me find a fix.

Oh yes. Do it! Funtoo seems to be the only place where this kind of fundamental re-evaluation of the "distro packaging" problem is taking place. I've been seeing the problem from my own viewpoint, building a source-based distro that uses Arch Linux style pkgbuilds AND Gentoo style use-flags (which I call "features"). I have about 1500 packages which I upgrade no more than monthly. Almost any change, upgrading a library, changing a feature (e.g. semantic-desktop -> -semantic-desktop) requires a huge rebuild and I see that it may be impossible to get a well-defined and reliable "operating system" while simultaneously cantering to the inner child's gleeful delight in NEW and DIFFERENT stuff every day. Maybe a step toward creating a Funtoo for large scale deployments would be to define one or two standard use-flag sets (e.g. kde or server) against which some volunteers test the portage tree and do a "light freeze" periodically and when it looks good enough define that set of ebuilds as a "tested standard release". Corporate/personal users could draw from the git branch "Release" to get something that at least is free of emerge blocks. They could even apply use-flags other than the standard, but the Release would be known to work well with the Standard Features set (with no user-imposed package masks). I was taught that engineering is all about working toward a goal within limitations. It's dealing with the limitations that makes an engineer. However you do it I commend the idea of accepting limitations in order to produce a good product. I really appreciate Funtoo; I run Debian Testing, my own Kit Linux, and Funtoo. Of those Funtoo is the best operating system. (btw Debian is second...)