[ Index ][ Official Thread ][ Bottom ]

(@caskd@social.redxen.eu) 17/10/20(Sat)16:10:44 id:1602965444 No. 1


if your software has:
- 30k files
- takes 2 hours to compile
- uses more than 8GB of memory when compiling
- does 30 things

that's not software, it's cancer

(@thatbrickster@shitposter.club) 17/10/20(Sat)16:13:23 id:1602965603 No. 2

'world application' should be renamed 'hell on earth software'

(@benis@cawfee.club) 17/10/20(Sat)16:16:08 id:1602965768 No. 3 >>4

I doubt even chromium takes 2hrs to compile, what on Earth are you doing?

(@caskd@social.redxen.eu) 17/10/20(Sat)16:16:27 id:1602965787 No. 4 >>5 >>7

i have to limit the jobs so it does

(@caskd@social.redxen.eu) 17/10/20(Sat)16:16:43 id:1602965803 No. 5

i compile it with 8 cores

(@blackernel@social.nixnet.services) 17/10/20(Sat)16:17:31 id:1602965851 No. 6

@caskd *that's not software, it's Windows


(@benis@cawfee.club) 17/10/20(Sat)16:19:28 id:1602965968 No. 7 >>8


(@caskd@social.redxen.eu) 17/10/20(Sat)16:20:00 id:1602966000 No. 8 >>9

building on >8 cores takes over 12GB

(@benis@cawfee.club) 17/10/20(Sat)16:51:55 id:1602967915 No. 9 >>10 >>18

did you buy a processor with more than 8 cores and paired it with less than 32GB of RAM?

(@caskd@social.redxen.eu) 17/10/20(Sat)16:52:35 id:1602967955 No. 10

Ryzen 7 3700X with 16GB

(@sacha@ligthert.net) 17/10/20(Sat)18:11:16 id:1602972676 No. 11 >>26

@caskd Read: the average webbrowser.

(@qorg11@social.snopyta.org) 18/10/20(Sun)03:44:38 id:1603007078 No. 12 >>13

@caskd chromium source xode is 30GBs, something like that

(@qorg11@social.snopyta.org) 18/10/20(Sun)03:44:45 id:1603007085 No. 13 >>14


@caskd eithet that or 8gb

(@caskd@social.redxen.eu) 18/10/20(Sun)03:56:47 id:1603007807 No. 14 >>15


@qorg11 compressed it’s only 800M

-rw-r--r-- 1 caskd caskd 795M Oct 17 19:01 /home/caskd/Packages/browsers/chromium/src/chromium-85.0.4183.121.tar.xz

(@qorg11@social.snopyta.org) 18/10/20(Sun)03:57:34 id:1603007854 No. 15 >>16


@caskd mind unpacking it? xz does magic

(@caskd@social.redxen.eu) 18/10/20(Sun)03:59:56 id:1603007996 No. 16 >>17


(@qorg11@social.snopyta.org) 18/10/20(Sun)04:00:30 id:1603008030 No. 17



(@caskd@social.redxen.eu) 18/10/20(Sun)04:03:31 id:1603008211 No. 18 >>19

other than building stuff i always hover between 1.5GB and 6GB usage so it was pretty useless for me

(@benis@cawfee.club) 18/10/20(Sun)04:11:53 id:1603008713 No. 19 >>20

>other than at lunch and dinner I never eat

(@caskd@social.redxen.eu) 18/10/20(Sun)04:12:24 id:1603008744 No. 20 >>21

i barely even compile stuff this heavy so like who cares

(@benis@cawfee.club) 18/10/20(Sun)04:14:13 id:1603008853 No. 21 >>22 >>23

as a rule you should count 2GB per thread/4GB per core next time you build a machine

(@caskd@social.redxen.eu) 18/10/20(Sun)04:15:42 id:1603008942 No. 22

instead of getting more ram i'd rather use lighter stuff
i have legit had no problems with the memory size with 1-2GB/thread other than this specific case

(@gsthnz@fosstodon.org) 18/10/20(Sun)08:37:10 id:1603024630 No. 23 >>24


@benis What kind of rule is that? Why do you think RAM usage scales with CPU thread count?

(@benis@cawfee.club) 18/10/20(Sun)08:40:08 id:1603024808 No. 24 >>25

empirical evidence derived from the abysmal performance of a TR 3990X paired with 64GB of RAM, I think Level1Techs' Wendell found that out

(@gsthnz@fosstodon.org) 18/10/20(Sun)08:49:57 id:1603025397 No. 25


@benis Hmm, interesting, do you have a link for this?

Also, It should be fine to have <16GB of RAM and be able to compile a entire Linux system from scratch. I don't think we should try to cope with shit software that cannot be compiled without ridiculous specs.

(@coldacid@noagendasocial.com) 18/10/20(Sun)16:20:59 id:1603052459 No. 26


@sacha which is cancer by definition

[ Top ] [ Index ] [ Refresh ]