**** BEGIN LOGGING AT Tue Feb 26 02:59:57 2019 Feb 26 14:36:18 is there work to bring go 1.11 to openembedded? Feb 26 15:01:54 mort: master has go_1.11: http://cgit.openembedded.org/openembedded-core/tree/meta/recipes-devtools/go/go_1.11.bb?h=master Feb 26 15:06:10 hmmm, that work was done in mid-December 2018 so an update to 1.11.5 seems like a good idea. mort care to send a patch? Feb 26 15:42:53 oe-core is freezing as we speak Feb 26 15:44:58 but spring is on the way! Feb 26 15:46:01 rburton: 1.11.4->1.11.5 so I expect that'll squeak in even post-2.7-M3 Feb 26 15:46:37 hmmm, 1.12 was released yesterday, that might not fly for 2.7 but it's not my thing/call. Feb 26 18:23:34 halstead: I know you are sick, but wanted to inform you that patchwork did not catch some of the patches from oe-devel Feb 26 18:26:13 Thank you khem. Is there a specific patchset to look at first? Feb 26 18:30:44 yes Feb 26 18:31:30 http://lists.openembedded.org/pipermail/openembedded-devel/2019-February/198524.html Feb 26 18:31:44 http://lists.openembedded.org/pipermail/openembedded-devel/2019-February/198525.html Feb 26 20:35:32 got X working on qemuarma15! Feb 26 20:36:24 Finally! Feb 26 20:36:32 hahaha Feb 26 20:37:21 mouse isn't working, but one thing at a time ;-) Feb 26 20:37:26 lol Feb 26 20:37:41 you could ask denix for an X15 and skip qemu :) Feb 26 20:37:58 I got a whole rack of real Arm HW Feb 26 20:38:28 but all these x86 build servers need something to run Feb 26 20:38:39 of course, if we could build on Arm servers... Feb 26 20:39:54 I'm looking forward to do doing my arm and x86 builds on my riscv servers Feb 26 20:40:24 I'm obligated by my employer to say that RSIC-V sucks and no one should ever use it Feb 26 20:40:27 or any combinationof the above Feb 26 20:40:45 I'm obligated not to chose sides Feb 26 20:40:55 and if I misspell it, no one can grep for me saying it in the irc log Feb 26 20:41:23 lol Feb 26 20:58:39 jonmason: i think you'll find both arm and riscv suck Feb 26 21:00:56 rburton: no one wants to pay those hefty Intel prices (or apparently hefty Arm licensing fees) Feb 26 21:00:57 certainly for doing OE builds! Feb 26 21:01:25 Crofton: once we get a 1000 core Arm box, then we will win! Feb 26 21:06:39 I can simplify this.. nobody wants to pay any licensing fees.. Feb 26 21:10:21 jonmason:thats great milestone Feb 26 21:11:32 khem: thanks. now just gotta figure out why usb isn't working and I'll send out patches for both qemuarm(a15) and qemuarm64 Feb 26 21:11:37 jonmason: reading your comment about RISC-V you must be working for arm Feb 26 21:11:54 :D Feb 26 21:12:16 but fact is that the time for the idea has come Feb 26 21:13:12 true, and MIPS just opened up their stuff too (and they actually have some hardened toolchains) Feb 26 21:13:35 yes but mips has missed the boat long time ago Feb 26 21:13:51 they let arm ISA bypass them into mobile market Feb 26 21:14:15 Broadcom still has some MIPS SoCs that they are selling Feb 26 21:14:31 oh many have Feb 26 21:14:33 so some life is there Feb 26 21:14:39 is risc-v any better off from a technical standpoint though? Feb 26 21:14:50 but all the cool kids like RISC-V, so its the real enemy now Feb 26 21:14:55 yes but market share is in rounding errors margins Feb 26 21:15:30 georgem: RISC-V is just an ISA, and its small Feb 26 21:16:24 sifive is doing interesting things and I'll be interested to see that talk at SCaLE by WD Feb 26 21:16:25 Well there are the toolchains and kernel support around the ISA as well. Feb 26 21:16:26 all ISA have 1000s of instructions, and check the compiler output how many of those are really used by 99% of code .. less than 40 Feb 26 21:18:32 think of 1991 when Linux releases linux kernel to community, we had Solaris, windows, HPUX and many more Feb 26 21:20:28 Isn't the RISC-V more akin to BSD? Feb 26 21:20:51 license wise I think so Feb 26 21:20:57 without the copyleft of the GPL, Linux wouldn't have taken off Feb 26 21:22:28 it'll be interesting to see what happens, either way Feb 26 21:30:05 jonmason: On github its interesting to see that apache-2.0, MIT and BSD are most favoured licenses now a days it used to be GPL 20 years ago, I guess communities have changed their minds Feb 26 21:30:15 LLVM is not GPL and its thriving Feb 26 21:35:38 good argument to have over beers :) Feb 26 21:36:22 if you have a referee Feb 26 21:36:48 my observation over 24 years.. GPL is great once there is a 'bad' actor in a community... assuming the community is benevolent and you have contributors and users.. then MIT and BSD work great.. Feb 26 21:37:03 what fray said Feb 26 21:37:12 but when that one guy comes in and tries to 'hide' their work, that is when the community looks for a more GPL like license Feb 26 21:37:20 x11 was great with the MIT license, right? Feb 26 21:37:48 there is always someone that wants to use others work and put nothing in Feb 26 21:37:59 or more to the point when that guy comes in, takes your work and tries to hide it... thats when copyleft is brought in by communities.. Feb 26 21:38:13 'users' are ones who take work and dont contribute back.. but even that is ok.. Feb 26 21:38:18 it's the leeches that are the problem.. Feb 26 21:38:25 embrace/extend/hide Feb 26 21:38:26 preach it brother! Feb 26 21:38:47 lots of legal cases where people are using creative commons pics and putting the in ads and newspapers Feb 26 21:38:54 some communities have embraced the leech.. others have said 'no'.. Feb 26 21:39:12 copyleft style is for those that decide 'no'.. Feb 26 21:39:20 (but int he end it is up to the community to decide) Feb 26 21:39:56 I have usb working now (stupidly no kernel support, seems to be working when added) Feb 26 21:40:21 i'm skeptical of the value of gpl more than i used to be. if "complying" is t hrowing up a random tarball on their web server that no one will ever touch, I'm not sure I care, it's not really participating in the community, the value of that isn't that great really.. Feb 26 21:40:23 my personal preference is GPLv2 for things I think others will find useful and maybe even re-use... while BSD-like is great for sample code someone might want to include in their own work, but doesn't really have much of a value on it's own Feb 26 21:41:04 kergoth, I look at RPi.. if it wasn't for the GPLv2, their kernel patches would get hidden.. since they sure as hell have no desire to work with the actual kernel.. but by posting their 'tarball' (equivalent) it means others can work on and contribute them Feb 26 21:41:07 it's nice for stuff shipping on devices, for the owner of said devices, though Feb 26 21:41:10 kergoth: but then the community can decide to take that tarball. while not optimal, its something Feb 26 21:41:31 fray: exactly Feb 26 21:41:44 exaactly.. I think the forced tarball is a happy medium.. IF the broader community cares, they can take it themselves.. if they don't then the cost as low on the 'leech' side Feb 26 21:41:45 the pi people are very disappointing Feb 26 21:41:48 not saying there's no value, just skeptical as to how much. the kernel is an obvious important case Feb 26 21:42:00 how they happened to patch busybox or wget, yeah, couldn't care less Feb 26 21:42:16 and I don't like the GPLv3 where certain things are required that IMHO shouldn't be.. but that is more philisophical IMHO.... Feb 26 21:42:17 I 100% agree with the origin of BSD. If tax paers pay for it, it should be free for eoneone and everything in every way Feb 26 21:42:32 I don't care so much about 'freedom to modify', as I do 'freedom to get the source and use it elsewhere' Feb 26 21:43:07 depends on the particular perspective, i think. as an open source developer more than latter, as a user of a device, the former takes on more importance Feb 26 21:43:15 I'd love for a device I own to be modifyable by be (though it is a violation of the DMCA) Feb 26 21:43:43 I do think there is a market for purely open devices Feb 26 21:43:45 jonmason: that's less of an issue than it used to be, at least for particular classes of devices. right to repair legislation has made strides over the past year Feb 26 21:44:06 https://ifixit.org/blog/11951/1201-copyright-final-rule/ Feb 26 21:44:33 still a ways to go, of course, but it's a good start Feb 26 21:44:44 there is still the ability, which corp funding, to make it illegal again. Feb 26 21:44:54 ok, I'm getting tinfoil hat on Feb 26 21:45:33 but I consider the GPLv2 a 'right to re-use'... and GPLv3 as 'right to modify'... BSD is 'I'm willing to share'... Feb 26 21:45:46 I'd love if everything was 'right to modify', but I'm not hung up on that Feb 26 21:45:53 all this hippy talk Feb 26 21:46:12 Crofton, that is BSD... :) Feb 26 21:47:06 I use MIT or GPLv2, depending on the project. Though I can see where Apache would be helpful Feb 26 21:47:06 I suppose Feb 26 21:47:33 My time, GPL. If you want me use to use a permissive license, you pay. Through the nose Feb 26 21:47:40 every license has a 'point'.. the key is who and what does it represent.. Feb 26 21:48:15 for the original authors of the code, I may complain about their choice, but in the end I do respect their choice.. Feb 26 21:48:28 just don't do it in bad faith cause you made a decision and then decided you don't like it years alter Feb 26 21:49:43 anyone else going to the LF Leadership thing? Feb 26 21:54:41 Crofton: seems like a no Feb 26 21:54:51 yeah Feb 26 21:55:07 I need more professional polos Feb 26 21:55:25 OE golf hats Feb 26 21:55:32 lol Feb 26 21:55:33 no Feb 26 21:55:34 no Feb 26 21:55:35 no Feb 26 21:55:48 OE ties, I know a company that makes custom ties Feb 26 21:55:50 do you need leadership training trips to Hawaii? Feb 26 21:56:05 I know a guy on the big island Feb 26 21:56:20 leadership training in Bangkok Feb 26 21:56:39 but seriosuly, I will be at leadership thing and if you ahve ideas for me to share with thgouht leaders, let me know Feb 26 21:56:58 get the various LF projects to actually consider working -together- on things Feb 26 21:57:05 already there Feb 26 21:57:27 getting all LF things to use OE as their build system Feb 26 21:57:41 dont be so obvious Feb 26 22:05:28 and give Crofton $$$$ Feb 26 22:05:44 or is that too obvious? Feb 26 22:10:24 hah Feb 26 22:10:32 not me directly of course **** ENDING LOGGING AT Wed Feb 27 02:59:56 2019