Back
[06:32:32] <linuxcnc-build> build #1739 of 4016.deb-wheezy-i386 is complete: Failure [4failed shell_1] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/4016.deb-wheezy-i386/builds/1739 blamelist: Dewey Garrett <dgarrett@panix.com>
[07:54:12] <skunkworks> https://groups.yahoo.com/neo/groups/mach1mach2cnc/conversations/messages/150211
[08:22:47] <jepler> zlog:
[08:25:35] <archivist> classic "Now, done locally by the software making the tool path, such as Linux, you get perfect tracking,"
[08:26:21] <jepler> somebody should figure out how to make m--h use linuxcnc as its motion controller :-P not really, please don't
[08:26:44] <archivist> hehe
[08:57:00] <skunkworks> then we would call it machinekit
[08:57:02] <skunkworks> ;)
[09:01:31] <cradek> I think art mistakenly thinks that linuxcnc only has software step generation?
[09:02:12] <cradek> he can be forgiven for not perfectly understanding our software, but it leads to incorrect statements like how "Linux" cuts threads slower than mach
[09:09:21] <skunkworks> I think he may be using steps in a generic term..
[09:12:16] <pcw_home> Sure nice that LinuxCNCs code for threading and rigid tapping is public, written in a common language,
[09:12:17] <pcw_home> runs on a robust processor with FP, works on all hardware, and any new real motion time extensions
[09:12:19] <pcw_home> have the same advantage
[10:13:39] <cradek> (... and it's been working for ten years come next month)
[10:24:36] <cradek> pcw_home: (... and it's been working for ten years come next month)
[10:49:30] <seb_kuzminsky> o
[10:49:34] <seb_kuzminsky> oops
[10:49:35] <seb_kuzminsky> morning
[10:49:40] * seb_kuzminsky needs another coffee clearly
[10:54:59] <seb_kuzminsky> did you guys see this? zfs will be in ubuntu 16.04:
http://blog.dustinkirkland.com/2016/02/zfs-is-fs-for-containers-in-ubuntu-1604.html
[11:45:49] <cradek> I don't know what a container is, but I do know that zfs on linux still doesn't have delegation
[11:46:32] <cradek> > You'll find zfs.ko automatically built and installed on your Ubuntu systems. No more DKMS-built modules!
[11:46:46] <cradek> and I wonder how they managed that
[11:48:17] <seb_kuzminsky> a container is like a chroot, but with the kernel "namespaces" feature enabled, so you can mock things other then the filesystem
[11:48:37] <cradek> oh good, they've finally implemented jails
[11:48:37] <seb_kuzminsky> it's a lightweight virtualization technique
[11:48:43] <seb_kuzminsky> yeah
[11:49:05] <cradek> maybe soon linux will have everything freebsd has
[11:49:06] <seb_kuzminsky> inside the container you can have a private namespace for network ports and pids and things like that
[11:49:23] <cradek> that's great news
[11:49:28] <seb_kuzminsky> i started taking freebsd more seriously after pkgng
[11:49:53] <cradek> yeah it really helps
[11:50:51] <cradek> maybe zol will get more attention too, if ubuntu is interested
[11:52:59] <cradek> https://github.com/zfsonlinux/zfs/issues/434
[11:53:22] <jepler> canonical has its own containerization system, lxd. I am sure it has all the problems you've come to associate with canonical, namely the CLA and (canonical-originating!) FUD about whether it's actually DFSG-free software
[11:54:10] <cradek> oh they've NIHed DKMS?
[11:54:34] <jepler> lxd is the jail part, not the kernel module building part
[11:54:46] <cradek> oh right
[11:54:55] <jepler> huh actually lxd is using signed-off-by and doesn't seem to have a cla, so that's nice
[11:55:03] <jepler> https://github.com/lxc/lxd/blob/master/CONTRIBUTING.md
[11:56:25] <cradek> looks like this is not about having the root filesystem on zfs
[11:56:31] <jepler> no, I doubt it
[11:56:41] <cradek> but everyone wants that
[12:00:45] <jepler> apparently zfs-dkms and spl-dkms are provided by linux-image, i.e., they have pre-built the zfs.ko, and have combined CDDL and GPL code into a binary that they redistribute.
http://packages.ubuntu.com/xenial/zfs-dkms
[12:01:00] <jepler> Packages providing zfs-dkms
[12:01:00] <jepler> linux-image-4.4.0-6-generic
[12:01:00] <jepler> Linux kernel image for version 4.4.0 on 64 bit x86 SMP
[12:01:19] <jepler> http://changelogs.ubuntu.com/changelogs/pool/main/l/linux/linux_4.4.0-6.21/changelog
[12:01:25] <jepler> * Miscellaneous Ubuntu changes
[12:01:25] <jepler> - [Packaging] git-ubuntu-log -- fix empty section formatting
[12:01:26] <jepler> - SAUCE: (noup) Update spl to 0.6.5.4-0ubuntu2, zfs to 0.6.5.4-0ubuntu2
[12:01:34] <jepler> their lawyers sure are giving them interesting advice
[12:05:10] <cradek> that's hard to understand
[12:05:44] <jepler> https://insights.ubuntu.com/2016/02/18/zfs-licensing-and-linux/
[12:07:15] <jepler> I notice they make claims about zfs.ko that I don't think can be made about spl.ko.
[12:07:59] <Roguish> hey, i just loaded the wheezy iso with linuxcnc into my older desktop with absolutely not issues. GOOD JOB on the iso package.
[12:08:10] <cradek> and "Equivalent exception ... nvidia" is not true, because that's dkms
[12:08:30] <cradek> Roguish: thanks! I spent a lot of time on it.
[12:09:51] <jepler> zfsonlinux's spl (solaris porting layer) is GPLv2+ without any extra linking exceptions.
[12:10:14] <jepler> https://github.com/zfsonlinux/spl/blob/master/module/spl/spl-atomic.c#L11
[12:15:10] <jepler> http://www.ubuntu.com/legal/ubuntu-advantage/assurance hah this is weak sauce
[12:15:39] <jepler> if you paid ubuntu under $15k last year, their idemnification is capped at that payment amounty
[12:15:56] <jepler> and if you went above, it's only double what you spent and still capped at 2 million
[12:16:45] <cradek> or they might just refund your support fees and call it good
[12:16:48] <jepler> and of course if you pay $0 for your Ubuntu and Oracle hauls you into court, Ubuntu is not standing behind you
[12:17:09] <cradek> oh wait, they only refund 12 months of support fees
[12:17:30] <cradek> > The customer may not modify the software provided by Canonical.
[12:17:41] <jepler> and in the unlikely event that they decide to defend you, you have to keep paying Canonical the whole time the court case drags out for years, at the same level
[12:17:49] <cradek> aaaaaand they also violate the gpl
[12:18:24] <cradek> I guess they mean you become ineligible for this bogus protection if you modify the software
[12:18:53] <cradek> that might be ok...?
[12:18:53] <jepler> they're only punishing you for using your rights under the GPL, not forbidding it
[12:18:56] <jepler> :)
[12:19:20] <cradek> yeah I misread "may not" by not considering how it's under the heading "Eligibility"
[12:20:52] <jepler> Red Hat can also just give you your money back and leave you holding the bag..
https://www.redhat.com/en/about/open-source-assurance-agreement
[12:25:57] <jepler> sys/list.h:#include <linux/list.h>
[12:26:17] <jepler> spl's headers include linux kernel headers that contain nontrivial inline code covered under GPL
[12:26:34] <jepler> I just can't believe their lawyers signed off on this
[12:27:07] <cradek> maybe they just don't care.
[12:27:49] <jepler> "every Android handset infringes GPL and nobody gets sued, so let's do it too"
[13:53:13] <KGB-linuxcnc> 03Dewey Garrett 052.7 6a5a64f 06linuxcnc 10docs/src/gui/ngcgui.txt 10lib/python/pyngcgui.py pyngcgui.py remove --help mention of --height * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=6a5a64f
[16:35:04] <skunkworks> Yay :)
https://groups.yahoo.com/neo/groups/mach1mach2cnc/conversations/messages/150234
[19:20:12] <Tom_itx> https://blog.freenode.net/2016/02/recent-events-and-future-changes/
[19:20:17] <Tom_itx> probably worth posting here
[21:06:19] <skunkworks> I think I have stopped trying to understand gene
[22:00:27] <jepler> http://openschemes.com/2010/11/11/1800w-induction-cooktop-teardown/
[22:01:43] <jepler> from the followup article: All our components already contain this built-in indicator called “smoke” that can easily be used to detect if something was driven just a bit too hard. Heh.