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 feld (feld@friedcheese.us)

  1. Embed this notice
    feld (feld@friedcheese.us)'s status on Saturday, 21-Jun-2025 01:31:44 JST feld feld
    in reply to
    • Julie Webgirl
    @juliewebgirl I've been looking for 2 days of any evidence this is true and it appears to be a weird internet rumor/meme being spread that has no basis in reality

    by now someone would have been shot for doing it, soooooo
    In conversation about 13 hours ago from friedcheese.us permalink
  2. Embed this notice
    feld (feld@friedcheese.us)'s status on Saturday, 21-Jun-2025 01:14:00 JST feld feld
    in reply to
    • feld
    somehow this morning it got better
    In conversation about 13 hours ago from gnusocial.jp permalink

    Attachments


    1. https://media.friedcheese.us/uploads/6f/f0/23/6ff0234e1c5c0b9b58e7a57d99b87a965f73f6656dfd7b2e1403863445f380b9.png
  3. Embed this notice
    feld (feld@friedcheese.us)'s status on Saturday, 21-Jun-2025 01:13:21 JST feld feld
    in reply to
    • fiat volvntas tva
    • wuhan.bat™
    @jae @scathach other than people being obsessed with the license change, I don't see any point in using Forgejo. They haven't added any meaningful changes and in some cases they were slower than Gitea to fix things. I have next to zero enthusiasm for git being somehow integrated with ActivityPub / the fediverse. It's a pretty terrible idea to be honest
    In conversation about 13 hours ago from friedcheese.us permalink
  4. Embed this notice
    feld (feld@friedcheese.us)'s status on Saturday, 21-Jun-2025 00:59:51 JST feld feld
    in reply to
    • Haelwenn /элвэн/ :triskell:
    @lanodan really are the biggest spammers now
    In conversation about 13 hours ago from gnusocial.jp permalink
  5. Embed this notice
    feld (feld@friedcheese.us)'s status on Saturday, 21-Jun-2025 00:07:06 JST feld feld
    in reply to
    • fiat volvntas tva
    • wuhan.bat™
    @jae @scathach as of the last major forgejo release the database schemas have diverged enough you can't easily switch back to gitea
    In conversation about 14 hours ago from friedcheese.us permalink
  6. Embed this notice
    feld (feld@friedcheese.us)'s status on Friday, 20-Jun-2025 12:35:47 JST feld feld
    in reply to
    • 🌈 ☯️Teresita🐧👭
    @linuxgal I had to use that to play my Nintendo on our wooden TV
    In conversation about a day ago from friedcheese.us permalink
  7. Embed this notice
    feld (feld@friedcheese.us)'s status on Friday, 20-Jun-2025 12:06:19 JST feld feld
    in reply to
    • Fish of Rage
    @sun bookmark this it has the important bits. You should just do it, it can probably be done in 5 mins

    https://calumcrawford.com/posts/gitea-sqlite-to-postgres/
    In conversation about a day ago from friedcheese.us permalink

    Attachments

    1. No result found on File_thumbnail lookup.
      Gitea database migration
      from Calum Crawford
      Moving Gitea from SQLite to Postgresql
  8. Embed this notice
    feld (feld@friedcheese.us)'s status on Friday, 20-Jun-2025 11:56:06 JST feld feld
    • fiat volvntas tva
    • wuhan.bat™
    @scathach @jae I have no clue what the cause was for my specific performance issue, but the changes they made in the next release that I installed after I solved my problem were described in the changelog as:

    Add cache for common package queries (#22491)
    Move issue pin to an standalone table for querying performance (#33452)
    Improve commits list performance to reduce unnecessary database queries (#33528)
    Optimize total count of feed when loading activities in user dashboard. (#33841)
    Optimize heatmap query (#33853)

    Other people had also been complaining about long page load times, sometimes up to almost 15 seconds. Allegedly these changes made it fast for them again.

    Well, for me on Postgres this release made it slightly slower but it's still so fast now I shouldn't complain ...
    In conversation about a day ago from friedcheese.us permalink
  9. Embed this notice
    feld (feld@friedcheese.us)'s status on Friday, 20-Jun-2025 11:46:11 JST feld feld
    Reminder that this exists

    pgsty/pigsty: Free RDS for PostgreSQL —— Pigsty
    https://github.com/pgsty/pigsty
    In conversation about a day ago from friedcheese.us permalink

    Attachments

    1. Domain not in remote thumbnail source whitelist: repository-images.githubusercontent.com
      GitHub - pgsty/pigsty: Free RDS for PostgreSQL —— Pigsty
      Free RDS for PostgreSQL —— Pigsty. Contribute to pgsty/pigsty development by creating an account on GitHub.
  10. Embed this notice
    feld (feld@friedcheese.us)'s status on Friday, 20-Jun-2025 11:37:59 JST feld feld
    in reply to
    • cool_boy_mew
    • Fish of Rage
    @sun @coolboymew more like yokohonkers amirite 😁
    In conversation about a day ago from friedcheese.us permalink
  11. Embed this notice
    feld (feld@friedcheese.us)'s status on Friday, 20-Jun-2025 11:35:19 JST feld feld
    in reply to
    • vermaden
    @vermaden well yes the support is second class but these days most of the new features seem to target Linux first
    In conversation about a day ago from friedcheese.us permalink
  12. Embed this notice
    feld (feld@friedcheese.us)'s status on Friday, 20-Jun-2025 11:30:52 JST feld feld
    in reply to
    • wuhan.bat™
    @jae yeah I prefer valkey. The benchmarks aren't a lie, traditional Redis is single threaded which is obviously problematic if you care about it ever being able to use more than 1 CPU core

    However, pinning the work to one core means you have a better chance of the kernel / NUMA implementation allocating memory on the local memory controller which will keep the memory accesses as low latency as possible

    With Valkey it's possible it needs to read from memory attached to another CPU and that could be significant under heavy load

    Well, guess that just means if you're trying to use all your hardware for a busy Redis workload you need to be smart about the hardware you pick

    Trade offs... but normal multi core CPUs on a single socket motherboard? Should smoke Redis
    In conversation about a day ago from friedcheese.us permalink
  13. Embed this notice
    feld (feld@friedcheese.us)'s status on Friday, 20-Jun-2025 09:16:25 JST feld feld
    in reply to
    • feld
    lol Gitea 1.24.0 "performance improvements, fixed slow queries"

    buddy you made it slower for me
    In conversation about a day ago from friedcheese.us permalink

    Attachments


    1. https://media.friedcheese.us/uploads/36/a2/69/36a26932abe8f69d5c1b5d30c5bdcffea8c46e31c03c0308a6ebc8919910c3c8.png
  14. Embed this notice
    feld (feld@friedcheese.us)'s status on Friday, 20-Jun-2025 09:08:54 JST feld feld
    in reply to
    • Tim W RESISTS
    @tim you goddamn right
    In conversation about a day ago from friedcheese.us permalink

    Attachments


  15. Embed this notice
    feld (feld@friedcheese.us)'s status on Friday, 20-Jun-2025 08:26:26 JST feld feld
    it gets eerily quiet, i look to my left

    my dog is sitting staring at me patiently waiting for my attention. 😍 she's so well behaved
    In conversation about a day ago from friedcheese.us permalink
  16. Embed this notice
    feld (feld@friedcheese.us)'s status on Friday, 20-Jun-2025 08:22:45 JST feld feld
    in reply to
    • feld
    • wuhan.bat™
    @jae I could probably disable the redis caches now and let it do the default in-memory and in-db, but considering that it's only using 11MB I guess I don't care... lol ...
    In conversation about a day ago from gnusocial.jp permalink
  17. Embed this notice
    feld (feld@friedcheese.us)'s status on Friday, 20-Jun-2025 08:18:31 JST feld feld
    in reply to
    • wuhan.bat™
    @jae yeah the performance difference is ridiculous, nobody should be using the sqlite deployment for anything important IMHO
    In conversation about a day ago from friedcheese.us permalink
  18. Embed this notice
    feld (feld@friedcheese.us)'s status on Friday, 20-Jun-2025 08:03:40 JST feld feld
    Sometimes sqlite really is shit

    I have a lot of repos in my private gitea server because I mirror repos I care about accessing locally

    well, it's come to the point where loading my dashboard after login takes somewhere between 4-8 seconds to render the page... the template it says renders in a dozen or so milliseconds

    so I enabled all the redis caching options.

    It got better. 1.5 - 2 seconds now.

    I just migrated the gitea database to my postgresql server

    Now it loads in 25ms

    Like holy shit batman, that's a world of difference
    In conversation about a day ago from friedcheese.us permalink
  19. Embed this notice
    kill -9 (kill_9@ieji.de)'s status on Friday, 20-Jun-2025 07:33:23 JST kill -9 kill -9
    in reply to
    • Delta Chat

    @delta it is because of the double ratcheting of the signal protocol. But a decentralised infrastructure will be much more resilient in the dystopia that is approaching rapidly. Moreover even though I really like Signal I don’t think they can resist the tremendous pressure they’ll be put under. Delta chat due to its back to the basics design will have a much better chance. I hope the full dystopia never arrives but….

    In conversation about a day ago from ieji.de permalink Repeated by feld
  20. Embed this notice
    feld (feld@friedcheese.us)'s status on Friday, 20-Jun-2025 06:36:28 JST feld feld
    Post-Compromise Security and Forward Secrecy don't seem to be helping out the people who have their Signal chats entered into evidence in court though

    RT: https://furry.engineer/users/soatok/statuses/114708631216466170
    In conversation about a day ago from friedcheese.us permalink

    Attachments

    1. No result found on File_thumbnail lookup.
      Soatok Dreamseeker (@soatok@furry.engineer)
      from Soatok Dreamseeker
      And the reason I will not be kind is simple: The people doing this are actively peddling Security Theater when they tell people to use something less-cryptographically-secure than Signal. The sheer arrogance of some of these app developers to peddle bullshit "privacy checklists" while doing things like... *checks notes* removing Post-Compromise Security and Forward Secrecy in their Signal fork? Fucking incredible. Kill it with fire.
  • Before

User actions

    feld

    feld

    Live, Laugh, Like and SubscribeFastmail referral link -- get 10% off your first year https://ref.fm/u12606493

    Tags
    • (None)

    Following 0

      Followers 1

      • GNU Too

      Groups 0

        Statistics

        User ID
        274913
        Member since
        6 Aug 2024
        Notices
        7078
        Daily average
        22

        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.