GNU social JP
  • FAQ
  • Login
GNU social JPは日本のGNU socialサーバーです。
Usage/ToS/admin/test/Pleroma FE

  • Public

    • Public
    • Network
    • Groups
    • Featured
    • Popular
    • People

Notices by Gantua (gantua@pleroma.marchera-pas.fr)

  1. Embed this notice
    Gantua (gantua@pleroma.marchera-pas.fr)'s status on Sunday, 26-Apr-2026 06:02:48 JST Gantua Gantua
    in reply to
    • Ryan Castellucci (they/them) :nonbinary_flag:
    @ryanc

    * Rogue DNS search/domain issued by ISP
    * Too many resolvers configured
    * DNS Response not routed correctly (ICMP redirects)

    I'll stop there.
    In conversation about 2 months ago from pleroma.marchera-pas.fr permalink
  2. Embed this notice
    Gantua (gantua@pleroma.marchera-pas.fr)'s status on Sunday, 26-Apr-2026 05:38:06 JST Gantua Gantua
    in reply to
    • Ryan Castellucci (they/them) :nonbinary_flag:
    @ryanc The one that says "probably lennart's fault somehow anyway" ?

    * problems with TCP fallback (firewall, or bug in dnsmasq)
    * Some random program (chrome?) is hammering DNS resolver
    * authoritative DNS resolver used as recursive
    * outdated DNSSEC root of trust or algorithm support
    In conversation about 2 months ago from pleroma.marchera-pas.fr permalink
  3. Embed this notice
    Gantua (gantua@pleroma.marchera-pas.fr)'s status on Sunday, 26-Apr-2026 05:18:43 JST Gantua Gantua
    in reply to
    • Ryan Castellucci (they/them) :nonbinary_flag:
    @ryanc That's a useful hint!

    * NetworkManager
    * netplan.io
    * Bad interaction with policy routing, done by wireguard or someone else.
    * Network card corrupts byte 45 on half the packets because systemd enables weird ethtool settings
    * firewalld or similar installed without notice during an upgrade
    * two DHCP clients fighting each other
    * /etc/resolv.conf is missing or configured to use localhost and two local DNS resolvers are fighting each other
    In conversation about 2 months ago from pleroma.marchera-pas.fr permalink

    Attachments

    1. No result found on File_thumbnail lookup.
      Netplan | Canonical Netplan
      Backend-agnostic network configuration in YAML.
  4. Embed this notice
    Gantua (gantua@pleroma.marchera-pas.fr)'s status on Sunday, 26-Apr-2026 05:03:58 JST Gantua Gantua
    in reply to
    • Ryan Castellucci (they/them) :nonbinary_flag:
    @ryanc

    * unwanted RA (RAscal) on the network announcing RDNSS or rogue DHCPv6.
    * DNS filtering by evil ISP
    * systemd
    In conversation about 2 months ago from pleroma.marchera-pas.fr permalink
  5. Embed this notice
    Gantua (gantua@pleroma.marchera-pas.fr)'s status on Friday, 17-Apr-2026 19:39:25 JST Gantua Gantua
    in reply to
    • Rich Felker
    • kimapr
    @dalias @kimapr This is C++. If your code is exception-safe, it is likely cancellation-safe for free.
    In conversation about 2 months ago from pleroma.marchera-pas.fr permalink
  6. Embed this notice
    Gantua (gantua@pleroma.marchera-pas.fr)'s status on Friday, 03-Apr-2026 04:28:30 JST Gantua Gantua
    in reply to
    • Rich Felker
    @dalias heat pumps can move heat from a hot source to a cold source when external power is applied to them, because you ignore the heat produced by the external power generator.

    If you stop ignoring the heat produced by the external power generator and include it into the hot source, the resulting system cannot cool itself.
    In conversation about 2 months ago from pleroma.marchera-pas.fr permalink
  7. Embed this notice
    Gantua (gantua@pleroma.marchera-pas.fr)'s status on Friday, 03-Apr-2026 03:32:34 JST Gantua Gantua
    in reply to
    @dalias This sounds like a violation of the second principle of thermodynamics.

    Your body+compressor would be a thermodynamic system that can cool itself using only its internal energy without any external power being applied to it.

    In practice, cooling without access to a cold source is very inefficient.
    In conversation about 2 months ago from pleroma.marchera-pas.fr permalink
  8. Embed this notice
    Gantua (gantua@pleroma.marchera-pas.fr)'s status on Thursday, 19-Mar-2026 03:22:00 JST Gantua Gantua
    in reply to
    • Rich Felker
    • William D. Jones
    • pinskia
    • disregard Joe Groff
    • nilix
    @dalias @nilix @joe @r @pinskia @cr1901 @astraleureka @mirabilos There are also laptops and even desktops motherboard with fullmac devices.

    All fullmac devices run a firmware that contains the equivalent of mac80211 (MLME), so if you can rewrite their firmware, you can probably turn them to softmac devices. But that's a big if, since vendors lock everything down.

    fullmac devices still have some advantages. They offload work off the CPU, and the CPU can even go to sleep while still connected to an AP, so the system uses less power. The offload is also useful for modern gigafast devices (think 802.11ax/be/bn) where the CPU can be a limiting factor. In fact, many modern devices are a blend of softmac and fullmac.
    In conversation about 3 months ago from gnusocial.jp permalink
  9. Embed this notice
    Gantua (gantua@pleroma.marchera-pas.fr)'s status on Thursday, 19-Mar-2026 02:51:27 JST Gantua Gantua
    in reply to
    • Rich Felker
    • William D. Jones
    • pinskia
    • disregard Joe Groff
    • nilix
    @dalias @nilix @joe @r @pinskia @cr1901 @astraleureka @mirabilos You must be thinking of cfg80211, which is the common API for all modern wireless drivers. mac80211 is just a layer between cfg80211 and softmac drivers. fullmac drivers uses cfg80211 directly.

    Most (if not all) wireless chips found in smartphones are fullmac. The raspberry pi also uses a fullmac chipset.
    In conversation about 3 months ago from gnusocial.jp permalink
  10. Embed this notice
    Gantua (gantua@pleroma.marchera-pas.fr)'s status on Thursday, 19-Mar-2026 02:38:46 JST Gantua Gantua
    in reply to
    • Rich Felker
    • William D. Jones
    • pinskia
    • disregard Joe Groff
    • nilix
    @dalias @astraleureka @r @mirabilos @cr1901 @pinskia @joe @nilix No that's not what the doc says. it says "All mac80211 drivers". mac80211 is only for softmac devices.

    fullmac devices (which are very common) typically do not support 4-addr mode. In fact, with a rapid grep, I could only find one fullmac driver (qtnfmac) that supports 4-addr, and only in STA mode.
    In conversation about 3 months ago from gnusocial.jp permalink
  11. Embed this notice
    Gantua (gantua@pleroma.marchera-pas.fr)'s status on Tuesday, 04-Nov-2025 09:38:52 JST Gantua Gantua
    in reply to
    • ✧✦Catherine✦✧
    • Henri Sivonen
    @whitequark @hsivonen Some of these systems are quite hard to replace by efficient alternatives. I retired a 15 years old i386 fanless industrial pc with 2 eth, many serial and USB1 ports, screen, two SATA drives and a PCI port. It consumed ~15W.

    To get the same DMIPS with less watt these days, you would probably need an ARM SBC, except most of those are missing the connectivity of a PC, have an out-of-tree kernel with no security updates and you are expected to design the case and the thermal management yourself, otherwise it will crash every week.

    In the end, i replaced it with a modern PC with more DMIPS for the same wattage, hoping it will last ten years. Hardly a win for e-waste.
    In conversation about 7 months ago from pleroma.marchera-pas.fr permalink
  12. Embed this notice
    Gantua (gantua@pleroma.marchera-pas.fr)'s status on Thursday, 03-Jul-2025 02:11:07 JST Gantua Gantua
    in reply to
    • Rich Felker
    • Jan Schaumann
    @dalias @jschauma Host appeared in HTTP/1.1 (as of RFC 2068) even if many programs (e.g. curl) includes it in HTTP/1.0 requests too.
    In conversation about a year ago from pleroma.marchera-pas.fr permalink
  13. Embed this notice
    Gantua (gantua@pleroma.marchera-pas.fr)'s status on Sunday, 30-Mar-2025 01:08:23 JST Gantua Gantua
    in reply to
    • Rich Felker
    • Jess👾
    @dalias @JessTheUnstill We already have something like that in France: the government gives your employer the percentage of income that must be paid. It's still displayed on the paystub.

    In practice, it means your employer can infer personal details from that percentage alone, and track it over time. If it's high, you're probably single. If it's medium, you're probably in a small family. If it's low, you're probably in a big family. If it's higher than high, you're earning more money via another gig...

    There is an option to give your employer a "standard" percentage and pay/refund the rest, but that percentage is between medium and high, so penalizes everyone but singles, so almost no one bothers with it.
    In conversation about a year ago from pleroma.marchera-pas.fr permalink

User actions

    Gantua

    Gantua

    Encore un n-ième développeur. À toujours faim.

    Tags
    • (None)

    Following 0

      Followers 0

        Groups 0

          Statistics

          User ID
          309059
          Member since
          24 Dec 2024
          Notices
          13
          Daily average
          0

          Feeds

          • Atom
          • Help
          • About
          • FAQ
          • TOS
          • Privacy
          • Source
          • Version
          • Contact

          GNU social JP is a social network, courtesy of GNU social JP管理人. It runs on GNU social, version 2.0.2-dev, available under the GNU Affero General Public License.

          Creative Commons Attribution 3.0 All GNU social JP content and data are available under the Creative Commons Attribution 3.0 license.