03:03:54 isn't illumos ZFS the master version? 03:04:37 please correct me if i am wrong 03:18:52 not anymore, dangergrrl 03:20:14 do you know why Illumos is no longer part of the collective effort of maintaining a common codebase, richlowe? I do remember people working on Illumos taking part in early OpenZFS meetings, back when the project of merging zfsonlinux & freebsd's implementation were in the early stages 03:20:16 openzfs is explicitly not for linux not that anyone will sue canonical for it but they ship it on all their boot media 03:21:04 I was super excited at the perspective of a cross-os fs codebase and actually thought it could bring more people to illumos distros eventually? 03:21:26 sorry, i'm part of funtoo also and we can't put it on install media 03:29:31 "openzfs is explicitly not for linux" and don't mind the fact that it's developed for linux (and freebsd) now 03:31:55 CDDL was designed to not be GPL compatible 03:32:26 I'm happy to contrinute CDDL code 03:32:42 but it's not GPL compatible 03:34:02 it was a Sun thing, not larry 03:35:59 freebsd uses a license that's compatible 03:37:18 anyway afaik oracle holds the rights and could sue canonical at any time and larry ellison doesn't even breathe air 03:40:35 veg, illumos is Unix 03:40:48 BSD is Unix also 03:40:56 linux is not 03:44:00 licensing is important 03:53:47 Linus says Oracle could sue, Canonical's lawyers disagree 03:55:36 practically speaking, most open-source development is now being carried by OpenZFS that works on GNUL/Linux & FreeBSD, with the macOS being on the roadmap, and I'm afraid to see Illumos fall behind, since I have a particular affection for UNIX history & Solaris in particular 03:56:28 hence my question about the possibility for Illumos to benefit from the work being carried by the OpenZFS project 03:57:12 moving pools between systems would be amazing 03:57:48 and the collective effort of maintaining a common codebase for a diversity of UNIX like systems is a very strong one in my opinion 03:58:42 i never worked for Sun 03:58:53 I worked for IBM and DEC 03:59:07 they both had filesystyems 03:59:18 one of which is ported to linux 03:59:20 that being said, there may be a lof of practical obstacles I'm not aware of, but I was wondering why Illumos was no longer part of the OpenZFS process, since its maintainers seemed to be involved in the beginning, and it did contriute to my excitement about OpenZFS ;) 04:00:16 JFS isn't bad 04:00:33 it's just not good 04:01:21 zfs seems to be the best widely ported filesystem 04:02:32 i can't compare it to the DEC code because not widely ported and HP won't release the rights 04:04:07 zfs seems to be the best filesystem released though 04:06:01 veg, ask jclulow or danmcd or richlowe or one of the other originals 04:06:28 yes, i pinged you, sorry 04:09:00 anyway i think at least two of them are part of joyent and might be able to answer 04:09:44 it's late in the USA and there is a horrible storm affecting the midwest 04:09:55 (effecting?) 04:12:00 it's late in Europe too :) have a good night, safe from storms I hope! 04:16:16 it's early in europe 04:18:04 weg, we have sub arctic temperatures 04:18:22 oh, sorry, veg 04:47:22 veg: I'm not skipping the meetings on purpose, I've just been swamped at work 04:47:54 Porting the code back to illumos represents a lot of work, and it's the file system so changes represent a lot of risk obviously 05:00:06 thanks for the answer, jclulow, I didn't mean to critisize by any means, I figure the work is immense! Congrats for being part of that awesome effort anyhow, I guess I'm just hopeful OpenZFS can contribute to bring Illumos back into the spotlight at some point, it would be well deserved! 10:34:27 jclulow: so far so good with the forced copying of the viona tx buffer, only had 1 vm crash with a small vqsize but not the full hang, just cifs.ko that barfed at something to do with the encryption, so that seems related as i could still ssh to the vm and the nfs mount was still working 10:35:44 nfstat is showing less latancy spikes too for writes, makes sense i guess if those were somehow related to running out/low on tx buffers 12:52:15 veg the thing is, no-one else than illumos [developers] can bring illumos "back into the spotlight". It really does not matter what was once done, it does matter what will be done to get there. 23:18:08 I also think it's best to focus on having software that works for you, rather than focusing on evangelism and popularity 23:24:18 jclulow: the rust paradox? 23:27:15 spotlights are scary, too