Join us now at the IRC channel.
| MinceR | https://www.quantamagazine.org/quantum-darwinism-an-idea-to-explain-objective-reality-passes-first-tests-20190722/ | Jul 30 00:09 |
|---|---|---|
| -TechrightsBot-tr/#techrights-www.quantamagazine.org | Quanta Magazine | Jul 30 00:09 | |
| <--en__ has quit (Ping timeout: 248 seconds) | Jul 30 00:25 | |
| MinceR | https://img.pr0gramm.com/2019/06/03/8b197e893471e7ac.jpg | Jul 30 00:34 |
| -viera/#techrights-Tux Machines: Linus Torvalds prepares to wave goodbye to Linux floppy drives http://www.tuxmachines.org/node/126408 [https://pleroma.site/objects/ed1b8594-e649-464f-807b-2bc9c2e2731e] | Jul 30 01:07 | |
| -viera/#techrights-Tux Machines: #Android Leftovers http://www.tuxmachines.org/node/126409 [https://pleroma.site/objects/73412c37-d90f-4dec-aff6-bfed6b069952] | Jul 30 01:08 | |
| schestowitz | https://www.reddit.com/r/linuxmasterrace/comments/cjdtrr/microsoft_back_at_it_again_with_trying_to_be_a/ | Jul 30 01:10 |
| schestowitz | "A wolf in a sheep costume. They will never truly change." | Jul 30 01:10 |
| -TechrightsBot-tr/#techrights-www.reddit.com | Microsoft back at it again with trying to be a good company towards us : linuxmasterrace | Jul 30 01:10 | |
| cubexyz | open source windows NT, then we'll talk | Jul 30 01:11 |
| MinceR | stop suing and blackmailing vendors of Linux-based devices | Jul 30 01:15 |
| MinceR | then we'll talk | Jul 30 01:15 |
| -viera/#techrights-Tux Machines: 5 open-source Firefox alternatives for Linux users http://www.tuxmachines.org/node/126410 [https://pleroma.site/objects/f5570e74-398a-4c2d-84e3-32ce2161a0e7] | Jul 30 01:27 | |
| <--gde33 has quit (Ping timeout: 258 seconds) | Jul 30 01:51 | |
| schestowitz | kaniini: pleroma 'hung up' ~3 mins | Jul 30 01:59 |
| -viera/#techrights-Tux Machines: #Kubernetes News/Views http://www.tuxmachines.org/node/126411 [https://pleroma.site/objects/ee14853f-0ee6-435d-a2c9-f10435763a4e] | Jul 30 02:00 | |
| schestowitz | kaniini: aaannnd... it's back :-) | Jul 30 02:00 |
| -viera/#techrights-Tux Machines: today's howtoshttp://www.tuxmachines.org/node/126412 [https://pleroma.site/objects/3ddc1107-8a1e-40d6-95bd-6a30f8b34fb3] | Jul 30 02:02 | |
| <--amarsh04 has quit (Read error: Connection reset by peer) | Jul 30 02:06 | |
| kaniini | schestowitz: yeah i am moving some things around | Jul 30 02:06 |
| -viera/#techrights-Tux Machines: Train Valley and Train Valley 2 released DRM-free on GOG with Linux support http://www.tuxmachines.org/node/126413 [https://pleroma.site/objects/9854d7ed-4460-4fe2-8328-d2b8c85350c9] | Jul 30 02:07 | |
| -->amarsh04 (~amarsh04@118.211.39.107) has joined #techrights | Jul 30 02:12 | |
| -->gde33 (~gde33@84-106-24-118.cable.dynamic.v4.ziggo.nl) has joined #techrights | Jul 30 02:18 | |
| -viera/#techrights-Tux Machines: SUSE displaces Red Hat @ Istanbul Technical University http://www.tuxmachines.org/node/126414 [https://pleroma.site/objects/6c6df38c-892a-4c89-a86a-00fab0a63dbb] | Jul 30 02:19 | |
| <--Firee has quit (Quit: lolmoi) | Jul 30 02:23 | |
| schestowitz | zoobab: are you on the fediverse yet? | Jul 30 02:24 |
| schestowitz | zoobab: I don't want to click anything "twitter" | Jul 30 02:24 |
| schestowitz | would rather follow you on the decentralised FOSS world | Jul 30 02:24 |
| schestowitz | I only access twitter to check epoorg tweets and yours | Jul 30 02:25 |
| schestowitz | I don't expect epoorg to join the fediverse | Jul 30 02:25 |
| kaniini | i have scripts now that can set up a mastodon or pleroma instance automatically ;0 | Jul 30 02:26 |
| kaniini | ;)* | Jul 30 02:27 |
| schestowitz | i hope they gain, I hear from many these days who quit Twitter | Jul 30 02:28 |
| -viera/#techrights-Tux Machines: Python Programming Leftovers http://www.tuxmachines.org/node/126415 [https://pleroma.site/objects/7d4e2d08-76a5-4312-8d74-a8b39e418335] | Jul 30 02:28 | |
| -viera/#techrights-Tux Machines: Open Hardware and Linux Devices http://www.tuxmachines.org/node/126416 [https://pleroma.site/objects/0ccab0ae-361f-4cd6-ba56-77a12f202fd0] | Jul 30 02:32 | |
| schestowitz | kaniini: I am returning to writing (hopefully) about 10 articles a day | Jul 30 02:32 |
| schestowitz | if you have a subject you want covered or think should be covered, let me know... | Jul 30 02:33 |
| schestowitz | can be fediverse stuff too | Jul 30 02:33 |
| -viera/#techrights-Tux Machines: One More Reason Elementary OS Is Ideal For New Linux Users http://www.tuxmachines.org/node/126417 [https://pleroma.site/objects/4fbbcd7f-db07-4c5e-8114-7bf6fc4eac95] | Jul 30 02:39 | |
| kaniini | well i think stuff like the fediverse is important but FOSS in general is also important | Jul 30 02:42 |
| -viera/#techrights-Tux Machines: Linux for Chromebooks 101: Getting started with the command line http://www.tuxmachines.org/node/126418 [https://pleroma.site/objects/cff3ec41-7239-4e25-81ec-aa0defa92e8a] | Jul 30 02:42 | |
| <--kaniini has quit (Ping timeout: 276 seconds) | Jul 30 02:58 | |
| -->kaniini (~kaniini@2602:ffdb:100:1:fc89:b2ff:fee1:323e) has joined #techrights | Jul 30 02:59 | |
| <--schestowitz has quit (Quit: Konversation terminated!) | Jul 30 03:27 | |
| -->schestowitz (~roy@host81-152-239-135.range81-152.btcentralplus.com) has joined #techrights | Jul 30 03:27 | |
| <--schestowitz has quit (Changing host) | Jul 30 03:27 | |
| -->schestowitz (~roy@unaffiliated/schestowitz) has joined #techrights | Jul 30 03:27 | |
| <--kaniini has quit (Ping timeout: 264 seconds) | Jul 30 03:34 | |
| -->kaniini (~kaniini@2602:ffdb:100:1:fc89:b2ff:fee1:323e) has joined #techrights | Jul 30 03:39 | |
| schestowitz | kaniini: eta? | Jul 30 03:45 |
| oiaohm | https://www.cbsnews.com/news/online-reputation-management-cbs-news-investigation-finds-fraudulent-court-orders-used-to-change-google-search/ This is kind of expected. | Jul 30 03:55 |
| -TechrightsBot-tr/#techrights-www.cbsnews.com | Online reputation management: CBS News investigation finds fraudulent court orders used to change Google results exclusive - CBS News | Jul 30 03:55 | |
| schestowitz | is this 'a news'? | Jul 30 04:01 |
| schestowitz | I think it has been well known that this is done | Jul 30 04:01 |
| -viera/#techrights-Tux Machines: A better LibreOffice than LibreOffice and ODF Gains http://www.tuxmachines.org/node/126419 [https://pleroma.site/objects/f973343e-8c39-47e3-846c-b4242a78abb9] | Jul 30 04:10 | |
| -viera/#techrights-Tux Machines: today’s #tuxmachines leftovershttp://www.tuxmachines.org/node/126420 [https://pleroma.site/objects/005f0b84-8895-4db3-b39f-f5c5d962f74b] | Jul 30 04:10 | |
| -viera/#techrights-Tux Machines: Links 30/7/2019: Linux Lite 4.6 RC1, Mint Release Imminent, No More Linux Floppies http://techrights.org/2019/07/29/no-more-linux-floppies/ [https://pleroma.site/objects/f6b789a3-4620-41f5-a2d8-692246f5e15c] | Jul 30 04:10 | |
| -viera/#techrights-Tux Machines: Today in Techrightshttp://www.tuxmachines.org/node/126421 [https://pleroma.site/objects/d06d33fe-cfda-46ea-a583-3c6454354ae1] | Jul 30 04:10 | |
| kaniini | schestowitz: almost done | Jul 30 04:13 |
| <--kaniini has quit (Ping timeout: 276 seconds) | Jul 30 04:17 | |
| -->kaniini (~kaniini@2602:ffdb:100:1:fc89:b2ff:fee1:323e) has joined #techrights | Jul 30 04:34 | |
| schestowitz | kaniini: excellent, I just maintain a queue and resume posting when availability is OK | Jul 30 04:39 |
| schestowitz | MinceR: your mate jonobacon | Jul 30 04:41 |
| schestowitz | the champ of FOSS and "community" | Jul 30 04:41 |
| schestowitz | https://www.jonobacon.com/2019/07/29/conversations-with-bacon-ryan-bethencourt-wild-earth/ | Jul 30 04:41 |
| -TechrightsBot-tr/#techrights-www.jonobacon.com | Ryan Bethencourt on Growing Sustainable Food, Shark Tank, and Wild Earth - Jono Bacon | Jul 30 04:41 | |
| schestowitz | still not finding his place in this world... | Jul 30 04:41 |
| schestowitz | "Most recently though, he formed Wild Earth, a company producing sustainably produced pet food. He took his company onto Shark Tank and raised one of the largest amounts for his season. He works weekly with Mark Cuban." | Jul 30 04:42 |
| schestowitz | He's like a PR agent for hire | Jul 30 04:42 |
| schestowitz | some sort of "I'll get you positive press for some cash" | Jul 30 04:42 |
| kaniini | schestowitz: i have dedicated postgresql node now. aarch64 | Jul 30 04:43 |
| schestowitz | I didn't know it was a psql backend | Jul 30 04:44 |
| schestowitz | I assumed something like mysql | Jul 30 04:44 |
| kaniini | yes well | Jul 30 05:01 |
| kaniini | going from 10 to 11 | Jul 30 05:01 |
| kaniini | is a massive pain | Jul 30 05:01 |
| schestowitz | I only use psql at work | Jul 30 05:04 |
| <--gde33 has quit (Ping timeout: 258 seconds) | Jul 30 05:06 | |
| schestowitz | https://www.suse.com/c/the-digital-transformation-journey-are-we-there-yet/ | Jul 30 05:14 |
| schestowitz | lol | Jul 30 05:14 |
| -TechrightsBot-tr/#techrights-The Digital Transformation Journey – Are We There Yet? - SUSE Communities | Jul 30 05:14 | |
| -->gde33 (~gde33@84-106-24-118.cable.dynamic.v4.ziggo.nl) has joined #techrights | Jul 30 05:42 | |
| schestowitz | https://nl.linkedin.com/in/greg-kroah-hartman | Jul 30 05:51 |
| -TechrightsBot-tr/#techrights- ( status 999 @ https://nl.linkedin.com/in/greg-kroah-hartman ) | Jul 30 05:51 | |
| schestowitz | " | Jul 30 05:51 |
| schestowitz |  | Jul 30 05:51 |
| schestowitz | Software Engineer | Jul 30 05:51 |
| schestowitz | IBM | Jul 30 05:51 |
| schestowitz | 2001 – 2005 4 years | Jul 30 05:51 |
| schestowitz |  | Jul 30 05:51 |
| schestowitz | Software Engineer | Jul 30 05:51 |
| schestowitz | WireX | Jul 30 05:51 |
| schestowitz | 1999 – 2001 2 years | Jul 30 05:51 |
| schestowitz |  | Jul 30 05:51 |
| schestowitz | Software Engineer | Jul 30 05:51 |
| schestowitz | PSC Inc. | Jul 30 05:51 |
| schestowitz | 1997 – 1999 2 years | Jul 30 05:51 |
| schestowitz |  | Jul 30 05:51 |
| schestowitz | Software Engineer | Jul 30 05:51 |
| schestowitz | MedSelect | Jul 30 05:51 |
| schestowitz | 1995 – 1997 2 years | Jul 30 05:51 |
| schestowitz |  | Jul 30 05:51 |
| schestowitz | Software Engineer | Jul 30 05:51 |
| schestowitz | Datamax | Jul 30 05:51 |
| schestowitz | 1992 – 1994 2 years | Jul 30 05:51 |
| schestowitz | " | Jul 30 05:51 |
| schestowitz | why did I hear somewhere he was once a bouncer/doorman? | Jul 30 05:51 |
| schestowitz | cannot seem to corroborate that | Jul 30 05:51 |
| schestowitz | (not that it matters much, either) | Jul 30 05:51 |
| oiaohm | schestowitz: I know that Greg did that door in the early days of the Linux Conference Australia. But that was not a paying job. So that could have been some of the speak about being a bouncer/doorman. | Jul 30 05:59 |
| schestowitz | thanks, I see... | Jul 30 05:59 |
| schestowitz | Makes sense now | Jul 30 05:59 |
| oiaohm | Lot of people over the years have done the Linux Conference Australia door including Linus. | Jul 30 06:00 |
| schestowitz | how tall is linus? | Jul 30 06:02 |
| schestowitz | I'm guessing about 5 10 | Jul 30 06:02 |
| oiaohm | schestowitz: 1.77 m its online if you google for it. | Jul 30 06:03 |
| oiaohm | Yep roughtly 5.10 He is closer to 6foot at times in real life because he does not mind thick sole shoes. | Jul 30 06:05 |
| schestowitz | so my guess was right :-) | Jul 30 06:07 |
| oiaohm | Being a a Australian Linux Conference and seeing him on different days. He has quite a variation in shoe choice I don't know if that aligns with who is talking with on the day. | Jul 30 06:07 |
| schestowitz | iirc, oiaohm, you're 6 5 | Jul 30 06:07 |
| schestowitz | DaemonFC[m] is also like 6 3 | Jul 30 06:08 |
| schestowitz | i'm 6 0 | Jul 30 06:08 |
| oiaohm | 6 4 without shoes 6 5 with. | Jul 30 06:08 |
| schestowitz | same as greg k-h more or less | Jul 30 06:08 |
| oiaohm | 6 6 if I am shoes on for basketball | Jul 30 06:08 |
| schestowitz | but anyway, he was never a bouncer | Jul 30 06:08 |
| schestowitz | false rumour | Jul 30 06:08 |
| schestowitz | seems to have a solid technical career track record | Jul 30 06:08 |
| oiaohm | doorman at conferences would be true. | Jul 30 06:08 |
| schestowitz | how many? | Jul 30 06:09 |
| oiaohm | I know long time ago he did lca because people were talking about getting signed in by him back in the day and him laughing about it. | Jul 30 06:09 |
| oiaohm | from what I heard it would the Linux kernel plumbers conference as well. But there could be more. | Jul 30 06:10 |
| oiaohm | But that is not a bouncer role. That is check name against list and let in role. | Jul 30 06:11 |
| oiaohm | But its not uncommon for the lead maintainers of Linux to have done this at different times either. | Jul 30 06:11 |
| oiaohm | I would say the bouncer one party comes from description of Gregs role in the staging branch not real job mixed up with the fact he done the door role at different conferences. | Jul 30 06:12 |
| oiaohm | schestowitz: so its a Chinese whispers thing with grains of truth but final meaning false. | Jul 30 06:13 |
| oiaohm | schestowitz: that just based on the information I have. | Jul 30 06:14 |
| kaniini | postgres-xl is annoying | Jul 30 06:24 |
| schestowitz | seems so based on downtime duration | Jul 30 06:28 |
| schestowitz | " | Jul 30 06:34 |
| schestowitz | I've imported that key thanks. An RSA sub key? From 2015? 880A3EC9 Let's see. I've disabled the older key, 74572E8E. | Jul 30 06:34 |
| schestowitz | Thanks for being careful. I imagine I'm on "threat" lists and that the usual spooks and creeps will be watching for a way to screw me. | Jul 30 06:34 |
| schestowitz | " | Jul 30 06:34 |
| schestowitz | we got some potentially big story | Jul 30 06:37 |
| schestowitz | lots of files, too | Jul 30 06:37 |
| <--kushal has quit (Excess Flood) | Jul 30 07:05 | |
| -->kushal (~hopless@fedora/kushal) has joined #techrights | Jul 30 07:05 | |
| ---MinceR gives voice to kushal gde33 schestowitz kaniini | Jul 30 07:10 | |
| ---MinceR gives voice to amarsh04 | Jul 30 07:10 | |
| MinceR | https://full.pr0gramm.com/2019/07/22/00c327751dcf91a3.jpg | Jul 30 07:25 |
| -->Firee (~Firee@unaffiliated/firee) has joined #techrights | Jul 30 08:00 | |
| schestowitz | aha | Jul 30 08:21 |
| schestowitz | kaniini: back | Jul 30 08:21 |
| kaniini | yes, sort of | Jul 30 08:21 |
| schestowitz | ah, can't post | Jul 30 08:22 |
| schestowitz | R-O? | Jul 30 08:22 |
| schestowitz | kaniini: seems like sometimes first attempt it fails | Jul 30 08:23 |
| schestowitz | then it succeeds | Jul 30 08:23 |
| ---MinceR gives voice to Firee | Jul 30 08:23 | |
| kaniini | i'm still working on finishing the migration | Jul 30 08:24 |
| kaniini | postgres "xtreme lattice" is not being very "xtreme" unfortunately | Jul 30 08:24 |
| schestowitz | once it's done you can rest | Jul 30 08:24 |
| kaniini | i'm creating some indexes in the background | Jul 30 08:26 |
| schestowitz | seems to be E-Xtremely fussy if I upload an image | Jul 30 08:27 |
| kaniini | yeah | Jul 30 08:28 |
| kaniini | there's an object permissions index | Jul 30 08:28 |
| kaniini | that has to be regenerated | Jul 30 08:28 |
| kaniini | it will probably take several hours ;) | Jul 30 08:28 |
| kaniini | however, the database cluster and the instances talk to each other over airgapped ipv6 network now | Jul 30 08:29 |
| kaniini | still more to do, but can't progress further until database indexes are done building | Jul 30 08:29 |
| MinceR | https://full.pr0gramm.com/2019/07/22/e66f5c9e8e4f4b5f.jpg | Jul 30 08:45 |
| schestowitz | MinceR: just nature... | Jul 30 08:55 |
| MinceR | indeed | Jul 30 08:55 |
| MinceR | https://img.pr0gramm.com/2019/07/22/dd691f423f5e20dd.png | Jul 30 08:58 |
| -viera/#techrights-Tux Machines: The Linux Foundation: Openwashing as a Service (OaaS) http://techrights.org/2019/07/30/the-linux-foundation-oaas/ [https://pleroma.site/objects/eb499908-3050-4507-b125-e2be63c612a5] | Jul 30 09:08 | |
| kaniini | schestowitz: should be fine now | Jul 30 09:13 |
| *kaniini bed | Jul 30 09:13 | |
| schestowitz | kaniini: https://joindiaspora.com/posts/15560108 | Jul 30 09:14 |
| -TechrightsBot-tr/#techrights-@schestowitz@joindiaspora.com: Overnight I realised that when #pleroma performs maintenance that involves downtime I'm at times hopeless. Dependent on this fantastic network (part of #fediverse sans the awful people). | Jul 30 09:14 | |
| -TechrightsBot-tr/#techrights- Photo by schestowitz@joindiaspora.com: https://joindiaspora.com/uploads/images/thumb_medium_624a34e745324a38e349.jpg | Jul 30 09:14 | |
| -viera/#techrights-Tux Machines: today's howtos http://www.tuxmachines.org/node/126422 [https://pleroma.site/objects/9cadcfd5-6d1b-4af8-8d42-f95d6b5176b5] | Jul 30 09:19 | |
| <--Firee has quit (Quit: lolmoi) | Jul 30 09:43 | |
| schestowitz | > Hey All, | Jul 30 09:44 |
| schestowitz | > | Jul 30 09:44 |
| schestowitz | > As you know we are going through the ISO processes - I have been asked | Jul 30 09:44 |
| schestowitz | > to gather some information from everyone at Sirius to create a list of | Jul 30 09:44 |
| schestowitz | > all assets used by employees of Sirius whether it belong to the company | Jul 30 09:44 |
| schestowitz | > or the employee so if I can have the item name and serial number that | Jul 30 09:44 |
| schestowitz | > would be great. They have also asked which anti virus you all use. | Jul 30 09:44 |
| schestowitz | > | Jul 30 09:44 |
| schestowitz | > Are you all able to send me the required information ASAP please? | Jul 30 09:44 |
| schestowitz | Hi Lou, | Jul 30 09:44 |
| schestowitz | I use three laptops, ASUS one, Acer one, and an HP one. All run GNU/Linux and do not require AV (which would likely create more security issues than it claims to actually tackle). | Jul 30 09:44 |
| schestowitz | Do you need model name or individual (factory) serial numbers of each? | Jul 30 09:44 |
| schestowitz | Kind regards, | Jul 30 09:44 |
| schestowitz | hahaha | Jul 30 09:44 |
| schestowitz | OSI... | Jul 30 09:44 |
| schestowitz | ISO I mean | Jul 30 09:44 |
| schestowitz | they assume people use Windows | Jul 30 09:44 |
| schestowitz | even in technical companies | Jul 30 09:44 |
| schestowitz | >> “Greg Kroah-Hartman previously commented on this site and Linus Torvalds | Jul 30 10:14 |
| schestowitz | >> I’ve only spoken to privately.” | Jul 30 10:14 |
| schestowitz | >> http://techrights.org/2019/07/28/the-linux-foundation-on-microsoft/#comment-325241 | Jul 30 10:14 |
| schestowitz | >> | Jul 30 10:14 |
| -TechrightsBot-tr/#techrights-techrights.org | The Linux Foundation is Totally Not Aware of Efforts to Undermine GNU/Linux | Techrights | Jul 30 10:14 | |
| schestowitz | >> | Jul 30 10:14 |
| schestowitz | >> What you you think? Should I email them? | Jul 30 10:14 |
| schestowitz | > I'm not sure. I agree with your reply that either Greg or Linus would | Jul 30 10:14 |
| schestowitz | > be reluctant to comment on their employer, albeit probably for vastly | Jul 30 10:15 |
| schestowitz | > different reasons. Greg for sure would be unlikely to comment, so I | Jul 30 10:15 |
| schestowitz | > would save him for when you need to write "contacted last week for | Jul 30 10:15 |
| schestowitz | > comment but as of time of publication has not responded" or something | Jul 30 10:15 |
| schestowitz | > like that. Linus, however, should be supported in whatever way | Jul 30 10:15 |
| schestowitz | > possible. While he could take his ball and go home, he has been more | Jul 30 10:15 |
| schestowitz | > interested on the technical aspects, where he excels. Also, changing | Jul 30 10:15 |
| schestowitz | > jobs at his age, with kids in college and probably a hefty mortgage, | Jul 30 10:15 |
| schestowitz | > would be detrimental to both him and the code base. Also, his life | Jul 30 10:15 |
| schestowitz | > might be on the line and more than a few consider that to be a reason he | Jul 30 10:15 |
| schestowitz | > has meticulously stayed out of anything non-technical. That is until | Jul 30 10:15 |
| schestowitz | > "they" got him with the CoC. | Jul 30 10:15 |
| schestowitz | I'll drop a quick line to Linus. I know he'll read it; I doubt he'll reply... | Jul 30 10:15 |
| schestowitz | " | Jul 30 10:28 |
| schestowitz | Hi Linus, | Jul 30 10:28 |
| schestowitz | I have been writing for many years about threats to Linux and more recently I focused on threats to Git (development processes, centralisation, censorship etc.) as well. | Jul 30 10:28 |
| schestowitz | I know you may not feel comfortable replying to me, knowing my criticism of the Novell deal (with Microsoft) and the Foundation alike. My criticism was always sincere and not intended to be provocative, only reactionary. | Jul 30 10:28 |
| schestowitz | This E-mail is not a 'bait' and it is definitely not hostile. I was advised by some readers to ask for your views on the current situation. I want Linux to succeed and not become just a zero-cost commodity for proprietary software giants that could not care any less about Open Source. There are forces in this game that try to change Linux from the inside and, IMHO, not in a good way. They try to make your job harder, knowing you have | Jul 30 10:28 |
| schestowitz | stronger grip on the project as its founder and "community darling". The media paints you rather negatively while they pass rules that disproportionately affect non-corporate participants (as part of their job they cannot say or do certain things that would get them fired, whereas community players/actors such as Con K. cannot be fired but excluded). | Jul 30 10:28 |
| schestowitz | I've spent 15 years studying malicious internal documents; a lot of the strategy put explicitly in those documents echoes or mirrors recent developments. | Jul 30 10:28 |
| schestowitz | Kind regards and with utmost respect, | Jul 30 10:28 |
| schestowitz | Roy | Jul 30 10:28 |
| schestowitz | " | Jul 30 10:28 |
| -->decnet (~decnet@gateway/tor-sasl/decnet) has joined #techrights | Jul 30 11:07 | |
| <--decnet has quit (Client Quit) | Jul 30 11:08 | |
| schestowitz | zoobab: : https://pleroma.site/search?query=zoobab | Jul 30 11:36 |
| -TechrightsBot-tr/#techrights-pleroma.site | Pleroma | Jul 30 11:36 | |
| schestowitz | do you use diaspora maybe? | Jul 30 11:36 |
| schestowitz | I don't want to go to Twitter to see your posts | Jul 30 11:36 |
| -->mmu_man (~revol@vaf26-2-82-244-111-82.fbx.proxad.net) has joined #techrights | Jul 30 12:17 | |
| <--XFaCE has quit (Ping timeout: 272 seconds) | Jul 30 12:19 | |
| -->XFaCE (XFaCE@unaffiliated/xface) has joined #techrights | Jul 30 12:19 | |
| XRevan86 | schestowitz: GNU social targets MariaDB, Pleroma and Mastodon target PostgreSQL. | Jul 30 12:38 |
| schestowitz | I never tried to installl these | Jul 30 12:41 |
| schestowitz | I just know pleroma is all elixir-y | Jul 30 12:41 |
| -viera/#techrights-Tux Machines: Microsoft as Chief Censor of Torvalds' Git and Torvalds' Linux (Foundation): How Did We Get Here? http://techrights.org/2019/07/30/torvalds-robbed/ [https://pleroma.site/objects/4c9e2fa8-dc9b-4a79-b0c6-3a275a7e0464] | Jul 30 12:49 | |
| -viera/#techrights-Tux Machines: Microsoft loves #Microsoft #Linux http://www.tuxmachines.org/node/126423 [https://pleroma.site/objects/b2be7647-2191-4090-96de-b745553a8ed6] | Jul 30 13:28 | |
| -viera/#techrights-Tux Machines: Top 10+ Best Linux Docks To Make Your Desktop Beautiful http://www.tuxmachines.org/node/126424 [https://pleroma.site/objects/f8f7fcd6-b1cf-4212-b9b3-db26682e0a9f] | Jul 30 15:49 | |
| -viera/#techrights-Tux Machines: The 12 Best Gnome Shell Extensions http://www.tuxmachines.org/node/126425 [https://pleroma.site/objects/fa40a068-55ac-4cb0-b3c9-985718c955ec] | Jul 30 15:52 | |
| -viera/#techrights-Tux Machines: An Overview of Debian 10 "Buster" from the GNOME Edition http://www.tuxmachines.org/node/126426 [https://pleroma.site/objects/7b397a64-d04b-4161-87d0-94244ae7722c] | Jul 30 16:49 | |
| -viera/#techrights-Tux Machines: Open-spec board supports blockchain-based IoT with Ethereum http://www.tuxmachines.org/node/126427 [https://pleroma.site/objects/db8911d6-56f2-4b5e-ab45-f588f0063695] | Jul 30 16:52 | |
| -viera/#techrights-Tux Machines: Events: Ubucon Europe 2019:, LibreOffice Asia Conference 2019 and the Future Of HOPE http://www.tuxmachines.org/node/126428 [https://pleroma.site/objects/7a724847-6213-42ef-9c18-3dc4c5d93464] | Jul 30 16:55 | |
| -viera/#techrights-Tux Machines: Security Leftovers: Elections, Microsoft, Docker and Open Vulnerability and Assessment Language (OVAL) http://www.tuxmachines.org/node/126429 [https://pleroma.site/objects/ba4133d6-e680-4b92-af9c-7f2184b2fd67] | Jul 30 17:01 | |
| -viera/#techrights-Tux Machines: Manjaro Snaps Collaboration With Ubuntu and More News About Ubuntu http://www.tuxmachines.org/node/126430 [https://pleroma.site/objects/2a2bee1d-8204-4acc-85e1-f7c91b4233e1] | Jul 30 17:05 | |
| -viera/#techrights-Tux Machines: Programming: LLVM 9.0 RC1, Python, C++ and GCC http://www.tuxmachines.org/node/126431 [https://pleroma.site/objects/0a4886f4-54c4-4d00-b838-175b3448632c] | Jul 30 17:08 | |
| -viera/#techrights-Tux Machines: Programming: LLVM 9.0 RC1, Python, C++ and GCC http://www.tuxmachines.org/node/126431 [https://pleroma.site/objects/9020c602-0abf-404c-854b-f1f31f7f339a] | Jul 30 17:08 | |
| psydroid | https://learnlinuxandlibreoffice.org/1-why-switch-to-linux/1-4-uefi-the-microsoft-kill-switch | Jul 30 17:10 |
| -TechrightsBot-tr/#techrights-learnlinuxandlibreoffice.org | 1.4 UEFI... The Microsoft Kill Switch | Jul 30 17:10 | |
| -viera/#techrights-Tux Machines: Games: Valve, Mono Trap and Bash Shell Games http://www.tuxmachines.org/node/126432 [https://pleroma.site/objects/67bcfcb5-90d7-44dd-b03f-559d940653c8] | Jul 30 17:24 | |
| schestowitz | psydroid: new page? | Jul 30 17:25 |
| schestowitz | also, yours? | Jul 30 17:26 |
| schestowitz | if it's new, we can add to http://techrights.org/wiki/index.php/UEFI | Jul 30 17:26 |
| -TechrightsBot-tr/#techrights-techrights.org | UEFI - Techrights | Jul 30 17:26 | |
| oiaohm | psydroid: that microsoft kill switch misses the fact that coreboot on x86 still has to use the intel/amd provided firmware cores that can start up Baseboard Management Controllers and other horrible. | Jul 30 17:28 |
| oiaohm | Its like the new amd ryzen chips with AMD is going to turn off PCI 4.0 on older boards even in cases where it can work. | Jul 30 17:29 |
| -viera/#techrights-Tux Machines: Graphics: AMD Graphics on Linux, MoltenVK Now Supports More Vulkan Extensions http://www.tuxmachines.org/node/126433 [https://pleroma.site/objects/f74b2c8d-3526-4dc1-a94e-bcd22e390a3a] | Jul 30 17:30 | |
| psydroid | schestowitz, it's not mine, but I came across Secure Boot as I replaced Windows with Linux on my uncle's laptop. As I installed Virtualbox I got a message asking me to put in a password, which I didn't but disabled Secure Boot in UEFI instead | Jul 30 17:31 |
| schestowitz | #WSL waiting for #gnu #linux to become hard if not impossible to boot | Jul 30 17:32 |
| -viera/#techrights-Tux Machines: 5 Free Partition Managers for Linux http://www.tuxmachines.org/node/126434 [https://pleroma.site/objects/18b0acda-de76-4ced-9117-33b6838752f8] | Jul 30 17:37 | |
| psydroid | oiaohm, I am not even in the market for new x86 hardware (although I have 2 laptops with Intel chips from 2016). I still have plenty of other hardware to use and been dogfooding on non-x86 hardware for over a decade. I have an ARM SBC running permanently as my desktop now with fully upstream sources, although I did have to compile a piece of ARM Trusted Firmware for u-boot | Jul 30 17:38 |
| oiaohm | psydroid: really depending on where you some places like in china are having to look to risc-v and the like. | Jul 30 17:40 |
| oiaohm | because of the fact ARM, Intel and AMD are possible to be embargoed by the USA. | Jul 30 17:41 |
| psydroid | oiaohm, I do have Fedora Rawhide for RISC-V running in QEMU, there is still some work to do but it's good to see the progress that has been made so far | Jul 30 17:42 |
| -viera/#techrights-Tux Machines: Servers: Twitter Moves to Kubernetes, Red Hat/IBM News and Tips http://www.tuxmachines.org/node/126435 [https://pleroma.site/objects/42f23cca-4bb4-4889-a39b-9ce0bdbd8637] | Jul 30 17:44 | |
| schestowitz | emulated? | Jul 30 17:44 |
| schestowitz | no actual board of risc-v design? | Jul 30 17:44 |
| oiaohm | schestowitz: lot of risc-v work is emulated at this stage. https://www.eetimes.com/document.asp?doc_id=1334966# but we should see decent stuff start appearing this year. | Jul 30 17:46 |
| psydroid | I went to the SiFive Tech Symposium in Amsterdam at the end of May and they told me affordable boards would be coming later this year | Jul 30 17:48 |
| -viera/#techrights-Tux Machines: today's howtos http://www.tuxmachines.org/node/126436 [https://pleroma.site/objects/7ee27707-3db8-4fa7-ac92-d797af604963] | Jul 30 17:48 | |
| <--TechrightsBot-tr has quit (Ping timeout: 258 seconds) | Jul 30 17:49 | |
| -viera/#techrights-Tux Machines: today's howtos http://www.tuxmachines.org/node/126436 [https://pleroma.site/objects/5770bde0-2aa7-42ef-8736-e5d942a65fbb] | Jul 30 17:49 | |
| -->TechrightsBot-tr (~TR@199.19.78.19) has joined #techrights | Jul 30 17:49 | |
| schestowitz | makes sense | Jul 30 17:49 |
| oiaohm | Decent risc-v boards could appear at the same time most of the debian is risc-v ready. | Jul 30 17:49 |
| schestowitz | oiaohm: I think your url crashed the bot | Jul 30 17:50 |
| schestowitz | did not diagnose, just guessing | Jul 30 17:50 |
| oiaohm | eetimes is not the most bot friendly. | Jul 30 17:50 |
| schestowitz | <li><h5><a href="https://www.phoronix.com/scan.php?page=news_item&px=Linux-KVM-RISC-V-Patches">Linux's KVM Sees Patches For RISC-V Support</a></h5> | Jul 30 17:55 |
| schestowitz | <blockquote><p>In continuation of the article last week how the RISC-V Linux kernel support has been maturing and various missing gaps filled in, another feature just arrived in patch form: support for KVM virtualization. | Jul 30 17:55 |
| schestowitz | Western Digital while associated with hard drives has been working big on RISC-V and already contributed Linux patches in the past. One of their engineers is the one to send out the RISC-V KVM support on Monday. | Jul 30 17:55 |
| schestowitz | </p></blockquote></li> | Jul 30 17:55 |
| schestowitz | (today) | Jul 30 17:55 |
| -viera/#techrights-Tux Machines: today's leftovers http://www.tuxmachines.org/node/126437 [https://pleroma.site/objects/9f34e857-602d-4a17-a4ed-fddc02f0a13c] | Jul 30 17:56 | |
| ---MinceR gives voice to mmu_man TechrightsBot-tr XFaCE | Jul 30 17:59 | |
| -viera/#techrights-Tux Machines: Links 30/7/2019: Manjaro Snap Support, LLVM 9.0 RC1 http://techrights.org/2019/07/30/manjaro-snap-support/ [https://pleroma.site/objects/e0db2e5a-48aa-4bdf-b5f3-9090627c532e] | Jul 30 18:04 | |
| oiaohm | https://buildd.debian.org/stats/graph-ports-week-big.png debian is about 85% of their packages build for risc-v. | Jul 30 18:06 |
| oiaohm | https://buildd.debian.org/stats/graph-week-big.png that is still about 10 percent down on where you want it to be. | Jul 30 18:08 |
| MinceR | i don't want it to be anywhere | Jul 30 18:08 |
| MinceR | debian is just as pointless as every other systemd distro | Jul 30 18:08 |
| oiaohm | MinceR: give me another distribution that you can get stats on how many packages are risc-v ready. | Jul 30 18:10 |
| MinceR | who cares? | Jul 30 18:10 |
| oiaohm | MinceR: I ask because these numbers are useful to me. | Jul 30 18:11 |
| oiaohm | So who cares is basically you don't have another answer. | Jul 30 18:11 |
| <--aindilis has quit (Read error: Connection reset by peer) | Jul 30 18:12 | |
| MinceR | what's the point of using a backdoorless cpu when your OS is full of backdoors? | Jul 30 18:12 |
| oiaohm | Reverse is the problem. If your hardware is flawed you cannot change the software to fix it. | Jul 30 18:13 |
| MinceR | and if you obey ibm/rh, you also cannot change your software to fix it | Jul 30 18:14 |
| oiaohm | The fact debian can build a lot of software for a platform does not mean long term you have to use debian. | Jul 30 18:14 |
| oiaohm | but it will mean others can built software for that platform without complete hell. | Jul 30 18:14 |
| XRevan86 | MinceR: But you're in #devuan | Jul 30 18:15 |
| MinceR | yes | Jul 30 18:15 |
| MinceR | and devuan is not debian | Jul 30 18:15 |
| MinceR | it is not a systemd distribution | Jul 30 18:15 |
| XRevan86 | MinceR: And Devuan is also completely dependent on it. | Jul 30 18:15 |
| MinceR | it's a GNU/Linux distribution, and as such, there is a point to it | Jul 30 18:15 |
| MinceR | yes, unfortunately | Jul 30 18:15 |
| XRevan86 | If Debian builds for RISC-V, it matters for Devuan. | Jul 30 18:15 |
| oiaohm | devuan does not provide stats and is complete dependant on debian for most of their packages. | Jul 30 18:15 |
| MinceR | and as long as it is, it is doomed | Jul 30 18:15 |
| MinceR | it is completely dependent on debian because not enough people are interested in freedom | Jul 30 18:15 |
| oiaohm | No so far devuan init offerences are not that appearling. | Jul 30 18:16 |
| MinceR | to you | Jul 30 18:17 |
| MinceR | which is fine, because you have plenty of distros controlled by rh/ibm to choose from | Jul 30 18:17 |
| MinceR | or you could skip the middle man and buy windows | Jul 30 18:17 |
| oiaohm | MinceR: sysvinit as default with the other options in different graded of broken. | Jul 30 18:18 |
| *MinceR yawns | Jul 30 18:18 | |
| oiaohm | MinceR: sysvinit has it faults. | Jul 30 18:18 |
| XRevan86 | oiaohm: OpenRC works fine on it. | Jul 30 18:18 |
| MinceR | but cancerd is flawless! | Jul 30 18:18 |
| MinceR | right? | Jul 30 18:18 |
| MinceR | it does so much | Jul 30 18:18 |
| oiaohm | No I pushed for openrc in the systemd debate on debian. | Jul 30 18:18 |
| MinceR | it only can't boot reliably or shutdown reliably, but who needs that in an init system? | Jul 30 18:18 |
| oiaohm | So I don't class systemd as flawless. | Jul 30 18:18 |
| oiaohm | I just don't class the options being put up by devuan as good enough yet. | Jul 30 18:19 |
| MinceR | openrc uses the same init binary you bash every day | Jul 30 18:19 |
| *XRevan86 has been booting and shutting down systemd for so many years… | Jul 30 18:19 | |
| MinceR | well, if it isn't good enough yet, it's time to stop using computers. | Jul 30 18:19 |
| *MinceR has attempted booting and shutting down that piece of shit several times... | Jul 30 18:19 | |
| oiaohm | I expect service management that works as well as initing. | Jul 30 18:19 |
| XRevan86 | MinceR: On Ubuntu, right? | Jul 30 18:19 |
| MinceR | right | Jul 30 18:20 |
| MinceR | probably still the most popular vehicle for ibm/rh to deliver malware to PCs | Jul 30 18:20 |
| oiaohm | Not half done service management like sysvinit. | Jul 30 18:20 |
| MinceR | service management does not need to be done in pid1 | Jul 30 18:20 |
| -->aindilis (~aindilis@172-12-3-117.lightspeed.sgnwmi.sbcglobal.net) has joined #techrights | Jul 30 18:20 | |
| MinceR | also, half done service management is still better than service management based mandatorily on a kernel feature that is known not to work | Jul 30 18:21 |
| XRevan86 | MinceR: Whatever starts services should preferably deal with them. | Jul 30 18:21 |
| oiaohm | service management being done in pid1 in systemd is caused by history kernel mistake. | Jul 30 18:21 |
| MinceR | XRevan86: exactly | Jul 30 18:21 |
| MinceR | which is exactly what runit and daemontool do | Jul 30 18:21 |
| MinceR | neither of them in pid1 | Jul 30 18:21 |
| oiaohm | yes the idea that cgroups had to be controlled by PID1 appeared in section of cgroup v1 code. | Jul 30 18:21 |
| MinceR | no, service management being done in pid1 in systemd is caused by the same thing that causes everything else being done in pid1 in systemd | Jul 30 18:22 |
| oiaohm | cgroup v1 need to die with major case of fire. | Jul 30 18:22 |
| MinceR | poettering's extreme incompetence | Jul 30 18:22 |
| oiaohm | Early mailing list debates on it you missed. | Jul 30 18:22 |
| *MinceR yawns | Jul 30 18:22 | |
| oiaohm | everything was shoved into PID1 because of that early namespace/cgroup screw up on the idea of security grounds. | Jul 30 18:22 |
| oiaohm | Reason why I backed openrc developed a little latter. | Jul 30 18:22 |
| MinceR | doesn't matter what their excuse for it is, it's bad engineering | Jul 30 18:23 |
| oiaohm | They avoided that mega stupidity. | Jul 30 18:23 |
| MinceR | remember cohesion? | Jul 30 18:23 |
| oiaohm | Kind of bad engineering if you want to support cgroup/namespaces around services on the oldest kernels possible you have to have that crap in PID1. | Jul 30 18:23 |
| oiaohm | The engineering mistake starts in the Linux kernel and spreads to systemd. | Jul 30 18:24 |
| oiaohm | In this case. | Jul 30 18:24 |
| oiaohm | MinceR: basically systemd is not the only source of really bad engineering. | Jul 30 18:25 |
| MinceR | there's an obvious way this could have been avoided | Jul 30 18:25 |
| MinceR | first finish cgroups and release it | Jul 30 18:25 |
| MinceR | then develop your service manager built on top of it | Jul 30 18:25 |
| MinceR | then use it in your allegedly stable distro releases | Jul 30 18:25 |
| MinceR | and never force it on people | Jul 30 18:25 |
| oiaohm | cgroup v1 was so called finished when systemd started. | Jul 30 18:26 |
| MinceR | this is the diametrical opposite of what your favorite cult did | Jul 30 18:26 |
| oiaohm | But had never been properly tested. | Jul 30 18:26 |
| MinceR | i'm not interested in "so called finished" | Jul 30 18:26 |
| oiaohm | systemd properly tested and and found it was totally broken. | Jul 30 18:26 |
| MinceR | just as in "so called stable", as in what rh/ibm did always and what debian started doing with release 8 | Jul 30 18:26 |
| MinceR | systemd "tested" in the production environments of other people | Jul 30 18:26 |
| oiaohm | sysvinit for Linux was developed the same way. | Jul 30 18:27 |
| oiaohm | History of lack of proper enginering goes back the day one of Linux kernel. | Jul 30 18:27 |
| MinceR | what was used before it and what distros/releases were affected? | Jul 30 18:27 |
| oiaohm | sysvinit was used before systemd in most distributions with distribution package makers doing custom workaround individually to lots of sysvinit faults. | Jul 30 18:29 |
| MinceR | that does not answer the question | Jul 30 18:29 |
| oiaohm | So most distributions were effected by broken init system and long list of out standing bugs before systemd started. | Jul 30 18:29 |
| MinceR | i see, so it wasn't actually done before | Jul 30 18:30 |
| MinceR | and you believe that if it was, that would be an excuse to perpetuate this bad practice forever | Jul 30 18:30 |
| oiaohm | cgroup manager would repleace pid1 on older distributions. | Jul 30 18:30 |
| MinceR | the "but someone else already did the same thing" school of software "engineering" | Jul 30 18:30 |
| oiaohm | So leading to sysvinit even more messed up. | Jul 30 18:30 |
| MinceR | at least cgmanager didn't try to do everything | Jul 30 18:30 |
| oiaohm | MinceR: really it was still running extra crap as PID1. | Jul 30 18:31 |
| MinceR | and the solution is obvious: run even more extra crap in pid1 | Jul 30 18:31 |
| MinceR | fuck up even harder. | Jul 30 18:31 |
| oiaohm | In fact cgmanger run more stuff as PID1 than systemd at it worse. | Jul 30 18:31 |
| oiaohm | So stuff was mega screwed before system. | Jul 30 18:32 |
| MinceR | interesting | Jul 30 18:32 |
| oiaohm | system/systemd | Jul 30 18:32 |
| oiaohm | systemd improved slightly over cgmanger pid1 hack. | Jul 30 18:32 |
| MinceR | so cgmanager replaced init, rc, udev, [x]inetd, crond, atd, syslogd, firewall, bootloader, dns cache, webserver, container provisioner and ntpd | Jul 30 18:32 |
| MinceR | it's nice to learn something new every day | Jul 30 18:33 |
| oiaohm | You were talking about PID1. | Jul 30 18:33 |
| oiaohm | systemd is a collection of applications. | Jul 30 18:33 |
| MinceR | riiiiight | Jul 30 18:33 |
| MinceR | for example, systemctl is an application | Jul 30 18:33 |
| MinceR | totally independent, works juuuust fine on its own | Jul 30 18:34 |
| oiaohm | container provisioner and init and logging... as PID1 this was both cgmanager PID1 hack and systemd PID1. | Jul 30 18:34 |
| MinceR | just drop a fuckton of binaries in your build process and your design being a worthless pile of crap just doesn't matter anymore! | Jul 30 18:34 |
| oiaohm | and that was caused by a stupid security thing done in Linux kernel that only PID1 could modify cgroup configuration. | Jul 30 18:34 |
| MinceR | but if you believe in your design so much, why force it on people? | Jul 30 18:35 |
| MinceR | let them choose based on merit | Jul 30 18:35 |
| MinceR | then maybe fix the fucking kernel before you build on it? | Jul 30 18:35 |
| oiaohm | Sorry that is not how the Linux kernel development model works. | Jul 30 18:35 |
| oiaohm | I wish it did. | Jul 30 18:35 |
| oiaohm | systemd was able to demo cgroup v1 was busted so able to get it rebuilt correctly as cgroupv2. | Jul 30 18:36 |
| oiaohm | Without being able to demo how the Linux kernel is broken the Linux kernel developers don't fix areas. | Jul 30 18:36 |
| MinceR | then the Linux kernel development model should be restricted to the kernel | Jul 30 18:36 |
| MinceR | instead of infecting userspace | Jul 30 18:36 |
| oiaohm | Linux kernel interfaces to userspace need userspace programs to demo their flaws. | Jul 30 18:37 |
| oiaohm | We have a chicken/egg problem here. | Jul 30 18:37 |
| MinceR | no we do not | Jul 30 18:38 |
| MinceR | also, the Linux kernel development model should not infect distributions either | Jul 30 18:38 |
| oiaohm | systemd was the one that demoed flaws in cgroup v1 leading to cgroupv2. This should have created the foundations for something as feature powerful in init /service management as systemd. | Jul 30 18:38 |
| oiaohm | without is horrible flaws. | Jul 30 18:38 |
| MinceR | fine, demo flaws | Jul 30 18:38 |
| MinceR | in your test systems | Jul 30 18:38 |
| MinceR | outside alleged "stable" releases of distributions | Jul 30 18:39 |
| MinceR | especially outside alleged "stable" releases of distributions made by other people | Jul 30 18:39 |
| oiaohm | That is also required to get alterations in upstream Linux kernel demoing usage cases. | Jul 30 18:39 |
| oiaohm | Linux kernel development model truly suxs. | Jul 30 18:39 |
| oiaohm | Its like the pidfd thing it taken 15+ years to in fact demo the problem of a race condition to with killing processes to get a proper fix done. | Jul 30 18:41 |
| MinceR | demos don't belong on production systems | Jul 30 18:41 |
| MinceR | if you can't understand this, engineering is not for you. | Jul 30 18:41 |
| oiaohm | To have standing in the Linux kernel world yes it does. | Jul 30 18:41 |
| MinceR | then the Linux kernel doesn't belong on production systems | Jul 30 18:41 |
| oiaohm | I know this is not good engineering practice. | Jul 30 18:42 |
| oiaohm | The Linux kernel itself development model is not. | Jul 30 18:42 |
| schestowitz | roy@ted:~$ uptime | Jul 30 18:42 |
| schestowitz | 18:42:18 up 293 days, 22:14, 5 users, load average: 0.27, 0.43, 0.36 | Jul 30 18:42 |
| schestowitz | roy@ted:~$ ps aux | grep systemd | Jul 30 18:42 |
| schestowitz | root 1 0.0 0.1 25216 3124 ? Ss 2018 6:35 /lib/systemd/systemd --system --deserialize 21 | Jul 30 18:42 |
| MinceR | ew | Jul 30 18:43 |
| schestowitz | roy@ted:~$ ps aux | grep systemd |wc | Jul 30 18:43 |
| schestowitz | 9 111 912 | Jul 30 18:43 |
| schestowitz | 9 processed | Jul 30 18:43 |
| schestowitz | processes inc. the grep | Jul 30 18:43 |
| schestowitz | wth? | Jul 30 18:43 |
| oiaohm | MinceR: BSD world has always said the Linux kernel development is crap. They had ways of properly killing processes in 1993. Linux kernel is only getting that this year. | Jul 30 18:45 |
| schestowitz | https://joindiaspora.com/posts/15564445 | Jul 30 18:45 |
| -TechrightsBot-tr/#techrights-@schestowitz@joindiaspora.com: I'm no #systemd expert. but can the so-called 'experts' (Red Hat/IBM advocates, who use this lock-in to sell support contracts) explain why a so-called 'init system' needs to run 8 processes on my machine which is limited to just 2 GB of RAM? | Jul 30 18:45 | |
| MinceR | is it getting that? | Jul 30 18:46 |
| MinceR | last time i checked they were complaining that cgroups couldn't even do this properly | Jul 30 18:46 |
| oiaohm | MinceR: cgroup is for tracking did not fix the problem where signal was going to wrong process. pidfd is what is required to fix up the Linux kernel signal system properly so it works. | Jul 30 18:47 |
| oiaohm | MinceR: nothing like attempting to use tracking and resource limitation ie cgroups/namespaces for correct delivery of signals and wondering why you are end up in rabbit hole all the way down. | Jul 30 18:48 |
| oiaohm | Round peg square hole. What cgroup/namespaces are meant todo they do well. | Jul 30 18:49 |
| MinceR | is that why cancerd SIGKILLs the shell on shutdown? | Jul 30 18:50 |
| oiaohm | Some of it attempting to work around the broken pid handling results to some stupid solutions. Some of it is simplly like applications not in fact shutdown when users told them to. | Jul 30 18:51 |
| oiaohm | systemd is basically the problem we had to have to bring the cgroup/namespace and pid handling with singals and other things to the floor and finally fixed in the Linux kernel. | Jul 30 18:52 |
| MinceR | no, they didn't even tell the shell to shutdown | Jul 30 18:53 |
| MinceR | they just killed it | Jul 30 18:53 |
| MinceR | and wrote a comment that this is the best cgroups can do | Jul 30 18:53 |
| MinceR | (even though it's obvious that it could have been done better, albeit with a minor race condition) | Jul 30 18:53 |
| oiaohm | Because it was you are aware that due to the PID faults you cannot in fact wait on a process to end correctly. | Jul 30 18:54 |
| MinceR | you can send SIGTERM to all the processes you know about, though | Jul 30 18:54 |
| oiaohm | MinceR: Completing the pidfd API << on lwn. | Jul 30 18:54 |
| MinceR | if they keep spawning new ones, too bad | Jul 30 18:54 |
| oiaohm | with pidfd a way better shutdown should be doable. | Jul 30 18:55 |
| oiaohm | The stupid part was you could send sigterm to processes and have a error message back of please do not term as well. | Jul 30 18:56 |
| oiaohm | MinceR: you are complete under estermate how badly broken the pid system has been. | Jul 30 18:57 |
| oiaohm | Between signals going to the processes. Not being able to track what processes has correctly sent you a signal. | Jul 30 18:57 |
| oiaohm | and then processes able to change their location in the process tree(this is what cgroups/namespaces deal with). | Jul 30 18:57 |
| oiaohm | pidfd deals with the first 2. | Jul 30 18:58 |
| oiaohm | 1 signals going to wrong process and 2 not knowing what process a signal has come from absolutely << These are problems pidfd fixes. Attempting to use cgroup/namespaces to fix both just leads into a rabbit hole. | Jul 30 18:59 |
| MinceR | i doubt the shell cares who sent SIGTERM to it | Jul 30 19:00 |
| oiaohm | MinceR: did you forgot a process getting a SIGTERM can send back a message please don't kill/shutdown yet. Problem is you really need to know what service that is. systemd with cgroups worked out bugger it kill it is really about the best that can be done with the level breakage. | Jul 30 19:00 |
| oiaohm | I do hope systemd does pick up pidfd usage. | Jul 30 19:01 |
| XRevan86 | schestowitz: systemd processes aren't heavy. On 2G that's nothing. | Jul 30 19:02 |
| MinceR | even if you ignore that, you still get a better system than what systemd did and probably still does | Jul 30 19:02 |
| oiaohm | with pidfd done and cgroup v2 it should be possible to make a really good and light replacement to systemd. | Jul 30 19:04 |
| oiaohm | That in fact works in all cases. | Jul 30 19:04 |
| XRevan86 | Except for journald I guess. | Jul 30 19:04 |
| oiaohm | MinceR: sending sigterm and having items send back messages could leave you with a growing stack of messages to process by bad applications. This is why you need to know who the sigterm comes from. This was a historic way to stall up a sysvinit shutdown as well. | Jul 30 19:06 |
| MinceR | considering how atrocious and bloated cancerd is, it's trivial to make a really good and light replacement to it | Jul 30 19:07 |
| oiaohm | MinceR: the signal mess really did need fixing and really need that 20 years ago. | Jul 30 19:07 |
| MinceR | oiaohm: it won't be stalled if you don't care about processes asking for more time | Jul 30 19:07 |
| XRevan86 | 7.4 MiB systemd | Jul 30 19:08 |
| XRevan86 | 5.3 MiB systemd-resolved | Jul 30 19:08 |
| XRevan86 | 1.46 MiB systemd-networkd | Jul 30 19:08 |
| XRevan86 | 1.38 MiB systemd-logind | Jul 30 19:08 |
| XRevan86 | 1.15 MiB systemd-timesyncd | Jul 30 19:08 |
| XRevan86 | 1.04 MiB systemd-udevd | Jul 30 19:08 |
| XRevan86 | 26.8 MiB systemd-journald | Jul 30 19:08 |
| XRevan86 | Here's what I see on a systemd-powered system. | Jul 30 19:08 |
| oiaohm | MinceR: no you can stall it by stacking the signal handing. This is another flaw that has been recently fixed. | Jul 30 19:08 |
| oiaohm | MinceR: recently I mean Last Linux kernel release. | Jul 30 19:09 |
| oiaohm | MinceR: the level of broken at Linux kernel level has been insane. | Jul 30 19:09 |
| XRevan86 | resolved is probably caching a lot. | Jul 30 19:09 |
| MinceR | oiaohm: you could at least try | Jul 30 19:10 |
| MinceR | it will probably be more reliable than what cancerd does | Jul 30 19:10 |
| MinceR | and then you'll have a system you can use while you develop a replacement if you want | Jul 30 19:10 |
| MinceR | as for cancerd being light, https://blind.guru/daemon-cpu.html | Jul 30 19:10 |
| -TechrightsBot-tr/#techrights-blind.guru | I am sorry, but this looks insane | Jul 30 19:10 | |
| oiaohm | MinceR: there was embedded compare between systemd and normal sysvinit equals result is that systemd is lighter when you are using feature equal. | Jul 30 19:12 |
| MinceR | yeah, i'm sure there was | Jul 30 19:13 |
| oiaohm | But systemd is still heavier than it really need to be. | Jul 30 19:13 |
| MinceR | and it was probably done by poettering | Jul 30 19:13 |
| oiaohm | If it was not working around so many kernel level defects. | Jul 30 19:13 |
| XRevan86 | 73 days of uptime, the systemd process used up 5:25 of plain CPU time | Jul 30 19:13 |
| oiaohm | Or have left over code layout required to work around old defects that the Linux kernel use to have. | Jul 30 19:13 |
| oiaohm | I do hope openrc is able to make a solid lightwieght version with the features of systemd without the overheads. | Jul 30 19:14 |
| XRevan86 | 7:34 on logind, 1:04 on resolved, and uh… 3871:23 on journald %). | Jul 30 19:14 |
| MinceR | but but but systemd can't work without journald :> | Jul 30 19:15 |
| oiaohm | That a joke. | Jul 30 19:15 |
| MinceR | (they didn't say that it can't work with it either...) | Jul 30 19:15 |
| oiaohm | Thinking that embedded systemd is commonly built without journald | Jul 30 19:15 |
| *XRevan86 is considering disabling journald. | Jul 30 19:15 | |
| MinceR | perhaps you should force your beloved lennart to confront his lies. | Jul 30 19:15 |
| oiaohm | and it works fine just systemctl loses a few features. | Jul 30 19:16 |
| *MinceR disabled journald | Jul 30 19:16 | |
| MinceR | along with the rest of cancerd. | Jul 30 19:16 |
| XRevan86 | 1649:47 rsyslogd – not that this CPU time is surprising for a logger. | Jul 30 19:16 |
| MinceR | and the rest of that shitty distro. | Jul 30 19:16 |
| XRevan86 | but I don't like how much RAM it eats | Jul 30 19:17 |
| XRevan86 | on this VPS with 73 days of uptime (and some verbose logging too) journald ate 89.2 MiB of RAM | Jul 30 19:18 |
| oiaohm | MinceR: https://elinux.org/images/6/69/Demystifying_Systemd.pdf this is what happens when embedded developer puts systemd on a force diet. | Jul 30 19:21 |
| oiaohm | You can get systemd with unit processing down under 1meg. | Jul 30 19:21 |
| XRevan86 | ^ all those numbers are from Debian, of course | Jul 30 19:22 |
| -viera/#techrights-Tux Machines: KDE Plasma 5.16.4 Desktop Environment Released with 18 Changes, Update Now http://www.tuxmachines.org/node/126438 [https://pleroma.site/objects/a19732a3-56d6-4429-b76e-16b369c56d57] | Jul 30 19:25 | |
| oiaohm | The size of systemd most distributions go for is the common everything incluiding the kitchen sink size. | Jul 30 19:26 |
| -viera/#techrights-Tux Machines: Canonical Releases New Linux Kernel Live Patch for Ubuntu 18.04 and 16.04 LTS http://www.tuxmachines.org/node/126439 [https://pleroma.site/objects/0474595d-bc0f-44ba-b768-137ea3479048] | Jul 30 19:27 | |
| oiaohm | Of course systemd source tree does not mandate that. | Jul 30 19:27 |
| XRevan86 | But even with a kitchen sink Storage=none could deal with journald. | Jul 30 19:29 |
| -viera/#techrights-Tux Machines: Blender 2.80 http://www.tuxmachines.org/node/126440 [https://pleroma.site/objects/caea34da-7bee-44c9-a13d-2820f335b5ab] | Jul 30 19:33 | |
| -viera/#techrights-Tux Machines: Programming: Python, Bash and HTML http://www.tuxmachines.org/node/126441 [https://pleroma.site/objects/2253c2f7-69c1-4578-8de9-a51c701fe629] | Jul 30 19:43 | |
| -viera/#techrights-Tux Machines: #Android Leftovers http://www.tuxmachines.org/node/126442 [https://pleroma.site/objects/1307b451-ec94-4266-8d2c-b1cf889fd038] | Jul 30 19:52 | |
| -viera/#techrights-Tux Machines: Collabora Brings VR Support to Linux Desktop Environments, Sponsored by Valve http://www.tuxmachines.org/node/126443 [https://pleroma.site/objects/6696f1db-59de-495c-8d02-7e074ca75080] | Jul 30 19:55 | |
| -viera/#techrights-Tux Machines: today's howtos http://www.tuxmachines.org/node/126444 [https://pleroma.site/objects/db438138-9c0e-4f76-b6ab-b0281d6bff88] | Jul 30 20:03 | |
| MinceR | oiaohm: or i can get it down to 0 bits and be better off | Jul 30 20:09 |
| MinceR | especially considering that even its unit dependency handling is broken | Jul 30 20:09 |
| schestowitz | XRevan86: still heavy | Jul 30 20:13 |
| XRevan86 | schestowitz: I suppose, but still not by much. On 2G or even 1G this wouldn't make a dent. | Jul 30 20:17 |
| -viera/#techrights-Tux Machines: Security Leftovers http://www.tuxmachines.org/node/126445 [https://pleroma.site/objects/6cc12416-122f-4f21-a9b4-ead43e60474d] | Jul 30 20:17 | |
| -->abeNd-org (~kklenke@rrcs-192-154-182-170.sw.biz.rr.com) has joined #techrights | Jul 30 20:33 | |
| schestowitz | you'd surprised how tight I am on this machine | Jul 30 20:36 |
| MinceR | that's what she said. | Jul 30 20:36 |
| schestowitz | every 50mb help | Jul 30 20:36 |
| ---MinceR gives voice to abeNd-org aindilis | Jul 30 20:36 | |
| schestowitz | MinceR: don't get me started :-) | Jul 30 20:36 |
| MinceR | that is also what she said :> | Jul 30 20:37 |
| schestowitz | don't get me finished | Jul 30 20:38 |
| MinceR | :> | Jul 30 20:38 |
| schestowitz | if you build it, they will come | Jul 30 20:38 |
| schestowitz | as per the joke you posted this week | Jul 30 20:38 |
| schestowitz | chinese machine | Jul 30 20:39 |
| XRevan86 | schestowitz: By the same measurement, OpenRC eats less than a MiB, and rsyslog – 2.1 MiB | Jul 30 20:39 |
| XRevan86 | schestowitz: So if you need an extra 15 MiB, then systemd is indeed too much. | Jul 30 20:40 |
| XRevan86 | schestowitz: And I'd also ask kaniini around about Alpine ;). | Jul 30 20:40 |
| XRevan86 | (unless you need something that doesn't work with musl) | Jul 30 20:41 |
| MinceR | with bedrock, you could presumably use a musl distro for whatever works with musl and a glibc distro for whatever doesn't | Jul 30 20:42 |
| XRevan86 | MinceR: That's more overhead, not less. | Jul 30 20:42 |
| MinceR | maybe you don't need the glibc stuff all the time | Jul 30 20:44 |
| MinceR | maybe you want to optimize for CPU over RAM :> | Jul 30 20:44 |
| XRevan86 | MinceR: I'm not convinced that glibc is less efficient on CPU… | Jul 30 20:44 |
| MinceR | well, maybe having less code to run helps | Jul 30 20:45 |
| MinceR | maybe not | Jul 30 20:45 |
| MinceR | ¯\_(ツ)_/¯ | Jul 30 20:45 |
| XRevan86 | MinceR: glibc has some crazy code, but that's partly because of crazy optimisations. | Jul 30 20:45 |
| XRevan86 | musl's code is simple and straightforward, but that's also because they avoid low-level optimisations. | Jul 30 20:47 |
| XRevan86 | or just obscure hacks that can help run shit faster | Jul 30 20:47 |
| <--abeNd-org has quit (Quit: Leaving.) | Jul 30 20:49 | |
| XRevan86 | Not judging, just saying that the RAM advantage is a lot more obvious than the CPU one. And if the RAM advantage (or storage space) is removed by containering, then musl probably doesn't have much else to offer. | Jul 30 20:49 |
| XRevan86 | schestowitz: Also, you'd probably save a lot more RAM by removing Apache httpd %). | Jul 30 20:52 |
| -->abeNd-org (~kklenke@rrcs-192-154-182-170.sw.biz.rr.com) has joined #techrights | Jul 30 20:52 | |
| MinceR | yeah, especially if you have cache | Jul 30 20:53 |
| ---MinceR gives voice to abeNd-org | Jul 30 20:53 | |
| MinceR | i suppose if you have cache not all cores can use, you might have cache savings in such a hybrid system | Jul 30 20:53 |
| MinceR | if musl-based processes run on a different set of cores from the glibc-based ones | Jul 30 20:53 |
| kaniini | Alpine is boring | Jul 30 20:54 |
| kaniini | abyss is the new hotness | Jul 30 20:54 |
| MinceR | the abyss also gazes into you | Jul 30 20:54 |
| XRevan86 | kaniini: I saw you talk about it. But elaborate :P | Jul 30 20:55 |
| XRevan86 | MinceR: But then you also have multiple sets of basic processes. | Jul 30 20:55 |
| MinceR | do i need them? | Jul 30 20:56 |
| XRevan86 | MinceR: If we're talking about proper containers, then yes | Jul 30 20:56 |
| MinceR | i wasn't talking about containers at all | Jul 30 20:57 |
| XRevan86 | And Bedrock is apparently virtualisation | Jul 30 20:57 |
| MinceR | not really | Jul 30 20:57 |
| MinceR | maybe filesystem virtualization | Jul 30 20:57 |
| MinceR | but /home and stuff like that is common | Jul 30 20:58 |
| XRevan86 | > chroot | Jul 30 20:58 |
| XRevan86 | Ah, no, not virtualisation. | Jul 30 20:58 |
| MinceR | it really exists so you can run binaries from multiple distros together | Jul 30 20:58 |
| XRevan86 | Does it run init? | Jul 30 21:00 |
| XRevan86 | There are problems either way %) | Jul 30 21:00 |
| MinceR | yes, one init | Jul 30 21:00 |
| MinceR | and it lets you pick one | Jul 30 21:00 |
| MinceR | you can reboot to switch inits | Jul 30 21:00 |
| XRevan86 | I of course mean guest system inits. | Jul 30 21:00 |
| XRevan86 | with their guest /etc/init.d/ or /{etc,usr/lib}/systemd/system/ | Jul 30 21:01 |
| MinceR | like i said, it's not virtualization in that sense | Jul 30 21:01 |
| XRevan86 | But then its point dimishes, especially on servers. | Jul 30 21:02 |
| XRevan86 | where services and their dependencies matter a lot | Jul 30 21:03 |
| MinceR | doesn't prevent you from using virtualization | Jul 30 21:03 |
| MinceR | things like docker and cgroups being broken will do that instead :> | Jul 30 21:03 |
| MinceR | unless you run qemu | Jul 30 21:03 |
| MinceR | or something like that | Jul 30 21:04 |
| XRevan86 | MinceR: But then it still means that Bedrock is not useful. | Jul 30 21:04 |
| scientes | qemu itsself is broken | Jul 30 21:04 |
| scientes | don | Jul 30 21:04 |
| MinceR | XRevan86: to you | Jul 30 21:04 |
| scientes | you will quickly run into bugs in qemu | Jul 30 21:04 |
| XRevan86 | MinceR: And to the proposed scenario of running two distros, musl and glibc. | Jul 30 21:04 |
| MinceR | more like running a mixture of two distros | Jul 30 21:05 |
| XRevan86 | but without their initscripts? | Jul 30 21:05 |
| MinceR | probably only with the initscripts of the one that provides your init | Jul 30 21:06 |
| XRevan86 | just like two chroots slapped together | Jul 30 21:06 |
| MinceR | kinda | Jul 30 21:06 |
| scientes | XRevan86, does Debian multi-arch solve your problems | Jul 30 21:07 |
| XRevan86 | scientes: What problems? :) | Jul 30 21:07 |
| scientes | on Debian you can have multiple libcs installed at the same time | Jul 30 21:07 |
| scientes | with the same library compiled multiple times | Jul 30 21:07 |
| MinceR | on debian you can have many more things broken at the same time | Jul 30 21:08 |
| MinceR | like init and apt and gnome | Jul 30 21:08 |
| XRevan86 | scientes: Hm, indeed. But these aren't "my problems". | Jul 30 21:08 |
| XRevan86 | scientes: It's just where the conversation led us %) | Jul 30 21:08 |
| scientes | multi-arch is great | Jul 30 21:09 |
| XRevan86 | scientes: systemd eats more than 10M of RAM, schestowitz doesn't like that, I suggested that if 15M of RAM is important then Alpine should be considered. MinceR suggested combining glibc and musl distros with Bedrock Linux. | Jul 30 21:09 |
| MinceR | not for schestowitz's problem | Jul 30 21:10 |
| scientes | you could also build systemd against musl | Jul 30 21:10 |
| XRevan86 | scientes: I'm trying to prove that this is a bad idea that doesn't solve it anyway. | Jul 30 21:10 |
| MinceR | afaik systemd requires glibc | Jul 30 21:10 |
| XRevan86 | Anyway, Alpine is a lightweight distro in the most literal sense of the word. | Jul 30 21:11 |
| XRevan86 | That's why I suggested it. | Jul 30 21:11 |
| scientes | XRevan86, I don't think installing it on my laptop will make it weigh any less ;) | Jul 30 21:11 |
| XRevan86 | when it comes to RAM of course :) | Jul 30 21:12 |
| XRevan86 | not THAT literal | Jul 30 21:12 |
| MinceR | what if you put a bladder of helium in your laptop? | Jul 30 21:12 |
| XRevan86 | It's just that "lightweight" is abused so much these days. | Jul 30 21:12 |
| XRevan86 | for software | Jul 30 21:12 |
| scientes | MinceR, or use ARM's new helium instruction set? | Jul 30 21:13 |
| MinceR | lol | Jul 30 21:13 |
| MinceR | XRevan86: especially by people like oiaohm | Jul 30 21:13 |
| XRevan86 | kaniini: So what's that abyss thing? How many Magog are we talking here about? | Jul 30 21:13 |
| -viera/#techrights-Tux Machines: Even Microsoft Boosters 'Love' Linux, Albeit Only When It's Actually Microsoft's http://techrights.org/2019/07/30/microsoft-sandbox-for-linux/ [https://pleroma.site/objects/7134070c-7560-47ab-acf8-737794e83347] | Jul 30 21:25 | |
| MinceR | https://img.pr0gramm.com/2019/07/22/eef3facc6b847e97.jpg | Jul 30 21:33 |
| -viera/#techrights-Tux Machines: Ubuntu in the Largest Surveillance Host and 'Store' Experience http://www.tuxmachines.org/node/126446 [https://pleroma.site/objects/c292659d-1c2f-400c-b435-5faa0e390151] | Jul 30 21:45 | |
| -viera/#techrights-Tux Machines: Robot kit builds on Jetson Nano http://www.tuxmachines.org/node/126447 [https://pleroma.site/objects/61e3dd78-4ee1-4559-bce7-fd9ba307264c] | Jul 30 21:53 | |
| -viera/#techrights-Tux Machines: 4 Discord alternatives for Linux gamers http://www.tuxmachines.org/node/126448 [https://pleroma.site/objects/adaf3dda-be3a-41e5-8ce7-547d6f0c9b8f] | Jul 30 21:56 | |
| -viera/#techrights-Tux Machines: A month later this balloon is starting to become deflated (the 15th birthday of @tuxmachines site) http://schestowitz.com/royrianne/gallery/index.php/15-Year-Anniversary-Party-for-Tux-Machines/Preparation-for-Party/20190628_170454 [https://pleroma.site/objects/91ba3a10-ae0c-4bb0-acff-0e96fcafab1c] | Jul 30 22:03 | |
| MinceR | https://vid.pr0gramm.com/2019/06/03/9aac8efedc3fd0ac.mp4 | Jul 30 22:19 |
| oiaohm | MinceR: https://lists.freedesktop.org/archives/systemd-devel/2014-September/023177.html I like how you keep on getting things wrong. Its been possible to build systemd without glibc for quite some time. | Jul 30 22:29 |
| -TechrightsBot-tr/#techrights-lists.freedesktop.org | [systemd-devel] [RFC 00/25] Compile against the musl libc | Jul 30 22:29 | |
| oiaohm | MinceR: requires glibc is bullcrap as normal. | Jul 30 22:29 |
| oiaohm | XRevan86: sysvinit with the bash/dash memory eat from the runscripts created envornmental crap in fact works out in a lot of cases worse in memory usage. | Jul 30 22:30 |
| XRevan86 | oiaohm: It runs and then it doesn't. | Jul 30 22:31 |
| XRevan86 | oiaohm: initscripts are one-shot things | Jul 30 22:32 |
| MinceR | this is obviously beyond oiaohm's understanding | Jul 30 22:32 |
| XRevan86 | Which is mostly shitty, because process tracking through pidfiles is crude, but in this case it helps | Jul 30 22:32 |
| MinceR | it probably wasn't on the brochure :> | Jul 30 22:32 |
| MinceR | it's crude, but at least it works | Jul 30 22:32 |
| MinceR | unlike cancerd | Jul 30 22:32 |
| MinceR | and it's portable | Jul 30 22:33 |
| MinceR | unlike cancerd | Jul 30 22:33 |
| XRevan86 | MinceR: systemd knows when a process crashed. | Jul 30 22:33 |
| MinceR | no, systemd _believes_ it knows when a process crashed. | Jul 30 22:33 |
| XRevan86 | MinceR: No… it's pretty good at this | Jul 30 22:34 |
| oiaohm | MinceR: No shut up . | Jul 30 22:34 |
| XRevan86 | Back to the usual friendliness… | Jul 30 22:34 |
| oiaohm | XRevan86: read what I said carefully. envornmental. | Jul 30 22:34 |
| XRevan86 | oiaohm: What do you mean by that? | Jul 30 22:35 |
| XRevan86 | oiaohm: envvars? /run/ pollution? What? | Jul 30 22:35 |
| oiaohm | XRevan86: Environment Vars a lot of them are populated in your old initscripts. Those end up copied into the environment vars of all you down stream running parts. | Jul 30 22:36 |
| XRevan86 | oiaohm: Ah, I see. | Jul 30 22:36 |
| oiaohm | Its one of the sneak evil where in hell did my memory go problems. | Jul 30 22:36 |
| oiaohm | Its why when people started totally benching systemd vs sysvinit systems things did not quite work out to be what was expected. | Jul 30 22:37 |
| XRevan86 | oiaohm: old initscripts… I can only tell about Debian initscripts | Jul 30 22:37 |
| MinceR | that's some entirely new brand of bullshit | Jul 30 22:37 |
| MinceR | blaming environment variables on sysvinit | Jul 30 22:38 |
| XRevan86 | oiaohm: And it's not like "export" is by default | Jul 30 22:38 |
| MinceR | in some last ditch attempt to prop up cancerd | Jul 30 22:38 |
| MinceR | you people are desperate | Jul 30 22:38 |
| MinceR | i just don't see why | Jul 30 22:38 |
| MinceR | you already have debian and ubuntu in your pockets | Jul 30 22:38 |
| oiaohm | People were expecting systemd to be memory heavier but sysvinit in most case cases turned out to have a high total memory usage. | Jul 30 22:38 |
| MinceR | TIL bash is part of sysvinit | Jul 30 22:38 |
| XRevan86 | oiaohm: There has to be *intent* when adding a new variable. Just a stray variable in sh won't cut it | Jul 30 22:39 |
| MinceR | or rather, the environment variable array | Jul 30 22:39 |
| MinceR | XRevan86: they don't care, they try out benchmarks until they find a way to "prove" their massive hairball is something it isn't | Jul 30 22:39 |
| oiaohm | XRevan86: some of it how bash/dash stores functions. | Jul 30 22:39 |
| oiaohm | XRevan86: there are exports in the way init scripts are done up. | Jul 30 22:40 |
| MinceR | hopefully at least the NSA pays them for this massive generation of entropy | Jul 30 22:40 |
| MinceR | if you hate shell scripts so much, go run winblows | Jul 30 22:40 |
| XRevan86 | oiaohm: "store functions" – one-shot, remember? | Jul 30 22:40 |
| MinceR | it already has all the broken bloated fascist bullshit you want | Jul 30 22:40 |
| XRevan86 | # grep -r export /etc/init.d/ | wc -l | Jul 30 22:41 |
| XRevan86 | 2 | Jul 30 22:41 |
| oiaohm | XRevan86: /lib/lsb/init-functions this here populates a stack of functions. It run in most init scripts to populate functions in you envornmetal memory. | Jul 30 22:42 |
| oiaohm | XRevan86: and that is copied to when its starts services. | Jul 30 22:42 |
| XRevan86 | oiaohm: Okay… I can test that | Jul 30 22:42 |
| oiaohm | XRevan86: its not as one shot and gone as you would hope for. | Jul 30 22:42 |
| MinceR | a shell you can tell to stop exporting functions would take a lot less effort to develop than cancerd | Jul 30 22:43 |
| MinceR | or than finding a benchmark that makes cancerd look good | Jul 30 22:43 |
| XRevan86 | a moment, please | Jul 30 22:43 |
| XRevan86 | oiaohm: I don't see it. | Jul 30 22:44 |
| XRevan86 | No extra variables. | Jul 30 22:45 |
| oiaohm | XRevan86: you did check with the set command because the env command will not display functions. | Jul 30 22:51 |
| oiaohm | XRevan86: fun of dash and bash a blackhole where it can put some in the copied between programs environment memory that gets copied between programs and then decides not to show it unless you do the right commands. | Jul 30 22:53 |
| XRevan86 | oiaohm: Wait, sh populates functions? | Jul 30 22:53 |
| XRevan86 | I thought you meant that it wraps them in envvars and sends somewhere | Jul 30 22:54 |
| XRevan86 | Okay, I'll check that… | Jul 30 22:54 |
| MinceR | https://twitter.com/stfnnoo/status/752616647886798848 | Jul 30 22:55 |
| -TechrightsBot-tr/#techrights-@stfnnoo: @systemdsucks starts my daemons w/o getting IPv6 addr first. Bind() fails,daemons stay dead on reboot until manually started.A FUCKIN JOKE. | Jul 30 22:55 | |
| XRevan86 | oiaohm: Checked with set, still nothing. | Jul 30 22:58 |
| XRevan86 | I checked on Devuan Ascii with plain sysvinit. | Jul 30 22:59 |
| oiaohm | XRevan86: hmm maybe that leak finally got fixed. | Jul 30 22:59 |
| oiaohm | I remember from a compare back in 2014 | Jul 30 22:59 |
| MinceR | https://i.redd.it/qqdla830gsex.jpg | Jul 30 23:00 |
| XRevan86 | oiaohm: If it's fixed, then it's not scary anymore | Jul 30 23:01 |
| XRevan86 | I can bootstrap Squeeze and then though :P | Jul 30 23:01 |
| XRevan86 | * and check that though | Jul 30 23:01 |
| XRevan86 | or better Wheezy | Jul 30 23:02 |
| XRevan86 | oiaohm: Still nothing. | Jul 30 23:21 |
| oiaohm | XRevan86: I would guess fix would be backported. I had missed the fix. I did find some internal notes where particular init scripts that depend on the functions set in bash/dash replicating by envornmental vars failed when they no longer did. | Jul 30 23:28 |
| oiaohm | still leaves sysvinit with the PID reuse issues and other races like that that need fixing. | Jul 30 23:30 |
| XRevan86 | oiaohm: I did not plug wheezy-updates in there. | Jul 30 23:39 |
| <--abeNd-org has quit (Quit: Leaving.) | Jul 30 23:48 | |
Generated by irclog2html.py 2.6 by Marius Gedminas - find it at mg.pov.lt!
