[2022-08-06T00:18:11Z] wael[m]: to avoid a circular dependency

[2022-08-06T01:05:42Z] <wael[m]> true

[2022-08-06T05:09:38Z] we had mailing list. it was dropped iirc, i don't remember why

[2022-08-06T05:10:16Z] dilyn can say more

[2022-08-06T05:27:26Z] <testuser[m]12> Hi

[2022-08-06T05:29:15Z] hi

[2022-08-06T09:22:37Z] Hi

[2022-08-06T09:23:20Z] ioraff: i removed openssl support from openssh and seems it works fine without it

[2022-08-06T09:38:21Z] the only left packages which have hardcoded dep on openssl api are rust, nodejs, python

[2022-08-06T09:38:35Z] not that many, but if considering how bloated they are...

[2022-08-06T09:39:07Z] everything else has an option to enable different crypto backend or can be patched to drop openssl

[2022-08-06T09:42:53Z] btw why rust needs openssl? for cargo to fetch shit?

[2022-08-06T09:47:52Z] also python can be compiled without openssl, but then it won't have tls and https capability. we need to investigate this i think

[2022-08-06T09:48:28Z] but i may be wrong cuz i don't remember details

[2022-08-06T10:53:50Z] illiliti: I've been away too long so I just saw the talk about doing a kiss BSD

[2022-08-06T10:54:08Z] I didn't exactly fail doing it

[2022-08-06T10:54:32Z] I was working with the OpenBSD kernel and I did get it to work

[2022-08-06T10:54:50Z] It's just that the kernel didn't support my hardware and I lost interest and dropped it

[2022-08-06T10:57:07Z] I wasn't super familiar with BSD at the time, and to be honest I came to the conclusion that Linux is much better for desktop

[2022-08-06T10:58:00Z] ah

[2022-08-06T10:58:04Z] cem: link? I've been curious about doing this myself

[2022-08-06T10:58:06Z] you didn't share your progress, that's why i thought you failed :)

[2022-08-06T10:59:55Z] Yeah I didn't spend much time on it, and I lost all of my stuff after a disk failure because I am a psycopath who doesn't commit shit until I'm happy with it

[2022-08-06T11:01:56Z] <testuser[m]12> Bruh

[2022-08-06T11:01:58Z] I don't remember the status of my last commit and I don't care enough to find out

[2022-08-06T11:02:06Z] https://github.com/carbsbsd/repository

[2022-08-06T11:02:41Z] That was freebsd so probably not

[2022-08-06T11:02:54Z] I switched to OpenBSD, don't know why

[2022-08-06T11:03:07Z] s/know/remember/

[2022-08-06T11:04:01Z] illiliti: Any news on the linker script for hare ?

[2022-08-06T11:04:35Z] Ogromny: was that devault's language?

[2022-08-06T11:04:50Z] what kind of news do you need?

[2022-08-06T11:05:15Z] cem: yep

[2022-08-06T11:07:34Z] if you mean submitting patch to upstream, i haven't had a time to work on it and that fix that phoboes suggested is rather hack than actual solution

[2022-08-06T11:08:37Z] illiliti: my bad I thought it wouldn't compile because of the linker script, I misread the github issue

[2022-08-06T11:09:08Z] it won't work with mold, otherwise it should be ok

[2022-08-06T11:09:45Z] mold: https://github.com/rui314/mold

[2022-08-06T11:12:19Z] mold looks really cool, do you think replacing lld or gold with mold will work out of the box ?

[2022-08-06T11:15:26Z] mold isn't suitable to bootstrap stuff, so you won't be able to fully replace ldd/bfd/gold

[2022-08-06T11:15:45Z] some things will work, some don't

[2022-08-06T11:16:16Z] as you see mold doesn't support linker scripts which breaks packages which use them

[2022-08-06T11:17:04Z] Yep unfortunately

[2022-08-06T11:17:42Z] mold is still useful to build heavy software, such as chromium or firefox, cuz it is 100x faster than everything else

[2022-08-06T11:20:05Z] i haven't tried to build them with mold though. my weak computer will blow up if i do

[2022-08-06T11:23:25Z] <testuser[m]12> It's good for debugging too cuz 5s for each link is better than 2 mins

[2022-08-06T11:29:13Z] cem: i see that carbs's main website has ipv6, but not git subdomain. is there any reason why it doesn't support ipv6?

[2022-08-06T11:30:25Z] Huh, should've forgot to add the record

[2022-08-06T11:35:58Z] ok

[2022-08-06T11:41:17Z] Updated the records now

[2022-08-06T11:44:22Z] connection refused

[2022-08-06T11:44:44Z] https://ready.chair6.net/?url=https://git.carbslinux.org

[2022-08-06T12:24:38Z] It should work now?

[2022-08-06T12:29:58Z] <testuser[m]12> Any idea why dhcpcd would say eth0: no IPv6 Routers available ?

[2022-08-06T12:30:03Z] <testuser[m]12> Ipv6 is enabled on vps config and kernel

[2022-08-06T12:54:23Z] cem: it is. thanks

[2022-08-06T12:54:42Z] illiliti: sure, thanks for letting me know

[2022-08-06T12:55:48Z] testuser[m]12: maybe ipv6 is disabled in /proc or dhcpcd config

[2022-08-06T13:20:16Z] <testuser[m]12> illiliti: how can u disable ipv6 in /proc

[2022-08-06T13:20:24Z] <testuser[m]12> dhcpcd config is default one

[2022-08-06T13:21:07Z] i mean /proc/sys/net/ipv6/conf/all/disable_ipv6

[2022-08-06T13:23:25Z] <testuser[m]12> Nope

[2022-08-06T13:23:26Z] <testuser[m]12> It's set to 0

[2022-08-06T13:27:59Z] hmm. maybe ipv6 routes don't get advertised for some reason. in this case, you have to set them statically

[2022-08-06T13:28:27Z] better off asking vps provider

[2022-08-06T13:30:16Z] <testuser[m]12> I'll try making static

[2022-08-06T13:40:24Z] <testuser[m]12> illiliti: how do i even do that

[2022-08-06T13:40:58Z] * testuser[m]12 uploaded an image: (89KiB) < https://libera.ems.host/_matrix/media/r0/download/matrix.pratham.cc/iwsCtXhLaUjBReykXyaJjeTV/Screenshot_20220806-191048_Fennec.png >

[2022-08-06T13:47:16Z] <testuser[m]12> https://termbin.com/ava7

[2022-08-06T13:50:44Z] search it. i don't remember the exact command

[2022-08-06T13:51:18Z] afaik you can do that via dhcpcd config

[2022-08-06T13:52:03Z] <testuser[m]12> Seems to be only for ipv4

[2022-08-06T13:52:08Z] <testuser[m]12> I hate networking

[2022-08-06T13:53:08Z] man dhcpcd

[2022-08-06T14:01:01Z] >building ipv6 into your kernel in the first place

[2022-08-06T14:01:04Z] psh, casuals.

[2022-08-06T14:04:48Z] <testuser[m]12> Bruh F it I'll just wait until ipv4 dies

[2022-08-06T15:22:13Z] <testuser[m]12> Anyone using kiss with a ryzen U series craptop?

[2022-08-06T16:18:29Z] <sad_plan> testuser[m]12 i have one. 3700u iirc.

[2022-08-06T16:19:43Z] <sad_plan> Why you ask?

[2022-08-06T16:19:49Z] <testuser[m]12> sad_plan: how long to build gcc in tmpfs

[2022-08-06T16:19:59Z] <testuser[m]12> I'll have to get one next year

[2022-08-06T16:20:13Z] there's no point asking that question, testuser[m]12

[2022-08-06T16:20:26Z] you should estimate using SBUs instead

[2022-08-06T16:20:34Z] <sad_plan> Uh.. 30 mins perhaps? I dont recall, but firefox usually goes aroubd ish 1hr, same with rust

[2022-08-06T18:39:10Z] illiliti: what about gnupg1, mold, mutt, and wpa_supplicant?

[2022-08-06T18:58:47Z] i don't use any of them, but i believe it should be possible to patch them

[2022-08-06T19:00:00Z] does gnupg1 need openssl? i'm not sure it does

[2022-08-06T19:02:43Z] mutt should work with libressl

[2022-08-06T19:06:11Z] you're right, it doesn't

[2022-08-06T19:06:26Z] earlier, did you mean that you just disabled the openssl configure option for openssh?

[2022-08-06T19:08:28Z] I believe mold only needs sha256 routines

[2022-08-06T19:09:26Z] someone made an issue about support for libsodium, but the developer is only sticking with libcrypto because it's the fastest.

[2022-08-06T19:25:59Z] <testuser[m]12> https://github.com/mesonbuild/wrapdb/tree/master/subprojects/packagefiles/openssl

[2022-08-06T19:37:45Z] <testuser[m]12> Can build without perl

[2022-08-06T19:41:54Z] neat

[2022-08-06T19:42:52Z] <testuser[m]12> https://github.com/nodejs/node/tree/main/deps/openssl

[2022-08-06T19:42:54Z] <testuser[m]12> And without meson

[2022-08-06T19:46:14Z] <testuser[m]12> But it generates in some gypi format, can be patched to generate makefile like meson did

[2022-08-06T19:54:26Z] <testuser[m]12> There's a build_all_generated target in openssl makefile aswell which can be used

[2022-08-06T19:59:53Z] <testuser[m]12> Yea works fine

Proxy Information
Original URL
gemini://gemini.ctrl-c.club/~phoebos/logs/kisslinux-2022-08-06.txt
Status Code
Success (20)
Meta
text/plain; charset=utf-8
Capsule Response Time
427.684634 milliseconds
Gemini-to-HTML Time
2.108028 milliseconds

This content has been proxied by September (3851b).