#soylent | Logs for 2026-09-07
« return
[01:00:12] -!- bender has quit [Remote host closed the connection]
[01:00:21] -!- bender [bender!bot@Soylent/Bot/Bender] has joined #soylent
[01:01:21] -!- Loggie [Loggie!Loggie@Soylent/BotArmy] has joined #soylent
[08:36:02] <chromas> ari furiously masturbating to peter thiel again. by "right wing", he probably means the tiny ring his fist makes when he strokes his submit button
[12:28:58] <Ingar> thiel me baby, thiel me hard
[16:06:34] <fab23> https://daringfireball.net
[16:06:57] <fab23> "Dickover of the Week: Slashdot Put One in Their RSS Feed"
[16:08:09] <chromas> What's wrong with his rss reader that it even supports such a thing?
[16:10:21] <fab23> I guess John "Markdown" Gruber does not have /. in his feed reader, he has just seen it reported from another blogger, but John did also invent the Dickover therm.
[16:10:55] <fab23> the really big question is, why are people still check the /. site :)
[16:16:23] <Ingar> becaue it's green!
[16:16:31] <Ingar> brushing up on my IPv6
[16:16:45] <Ingar> found out that IPv6 actually exposes my entire network to my ISP
[16:17:04] <fab23> is it bridged?
[16:17:30] <Ingar> not yet, it will be
[16:18:02] <Ingar> prefix substitution still exposes all hosts. subnet delegation still exposes all hosts
[16:18:04] <fab23> ugly setup, did you not got a /64 routed to you?
[16:18:21] <Ingar> fab23: from what I've read I should get a /57
[16:18:48] <Ingar> right now I just get adress from my ISP's router
[16:19:37] <Ingar> *addresses
[16:19:43] <fab23> ah, that would be a couple /64 then, as it is recommended to use a /64 for each LAN segment
[16:20:43] <Ingar> frebsd might be able to do actual IPv6 source nat
[16:20:47] <fab23> but still ode, my "business" ISP did assign me a /48.
[16:21:18] <fab23> Ingar: yes should be able, as OPNsense also is capable
[16:22:32] <fab23> Ingar: NPTv6 is the key word, see also https://weberblog.net
[16:23:49] <fab23> there are people who will hate you if you mention IPv6 an NAT, but sometimes it is the only useful thing to do
[16:24:24] <Ingar> wwtf, I click that link nad I get a weird "Choose a firefox profile" window
[16:24:30] <Ingar> and not the one you get with firefox -p
[16:25:01] <fab23> Ingar: ah Firfox did implement some "new" Profiles a while ago.
[16:25:39] <fab23> if you already had old profiles, this option will not be available, as far as I see it on my computers
[16:25:57] <Ingar> tell that to my firefox
[16:26:10] <fab23> :)
[16:26:29] <Ingar> ofc, it's snot that mozilla is competent or anything
[16:27:08] <Ingar> also, afaik NPTv6 only maps prefixes
[16:28:34] <fab23> the idea is that you use internally a private IPv6 range, and that does translate only with the prefix you get assigned from your ISP. Should then also work if you may have 2 IPv6 uplinks from 2 different ISPs
[16:29:02] <fab23> so the internal network has static IP addresses, else things may get messy
[16:29:19] <Ingar> yes, but it doesn't solve the original issue where my ISP knows what hosts I have
[16:30:36] <Ingar> most likely the easiest solution for now though
[16:30:56] <fab23> the ISP does only see traffic from random IPv6 addresses, most clients anyway do use random IPv6 addresses in their range
[16:32:41] <fab23> BSDs do have a sysctl net.inet6.ip6.use_tempaddr and similar
[16:33:15] <Ingar> I can't test anyway atm
[16:33:32] <Ingar> for that I have to install the router and put my modem in bridged
[16:33:51] <fab23> I see
[16:41:34] <Ingar> maybe the easiest solution for now is to setup both an internal and external range
[16:41:53] <Ingar> then I can do a DNS zone for the internal range
[19:54:11] <AlwaysNever> the easiest (and only sane) solution is to skip IPv6 altogether, it's a clusterfuck, and you need IPv4 anyway
[19:55:36] <AlwaysNever> IPv6 is the systemd of the networking world, it does many more things appart from widening the IP space, and those additional things nobody asked for them