weeezes

  • 3 Posts
  • 32 Comments
Joined 2 years ago
cake
Cake day: September 19th, 2022

help-circle





  • weeezestoSuomiKirjailija Miki Liukkonen on kuollut
    link
    fedilink
    suomi
    arrow-up
    3
    ·
    1 year ago

    – Miki Liukkonen oli rakas ystäväni, sydämellinen kuuntelija ja mielestäni sukupolvensa lahjakkain kirjailija, jonka kanssa minulla oli ilo tehdä työtä viiden romaanin, kolmen runokirjan ja yhden lastenkirjan parodian verran. Ne elävät vielä kauan, mutta en pysty hyväksymään sitä, että niiden tekijä on poissa. Mutta menetys on rakkauden kirjanpitoa, sen mittayksikkö aivan kuten metri koostuu senteistä, ja nyt on tilinteon päivä.

    senteistä, ja nyt on tilinteon päivä

    näinhän sinä varmaan sanoisitkin, kustannustoimittaja. Nyt kääritään hilloa?








  • weeezestoAsklemmy@lemmy.mlHow will lemmy scale?
    link
    fedilink
    English
    arrow-up
    1
    ·
    1 year ago

    The part that worries me about scalability in the long term is the push nature of ActivityPub. My server is already getting several POST requests to /inbox per second already, which makes me wonder how that’s gonna work if big instances have to push content updates to thousands of lemmy instances where most of the data probably isn’t even seen. I was surprised it was a push system and not a pull system, as pull is much easier to scale and cache at the CDN level, and can be fetched on demand for people that only checks lemmy once in a while.

    I think there’s a benefit to the push model, as the instances can prioritize who to push to first if there’s scaling issues, instead of having to throttle GETs, effectively the end result is anyway the same that nothing ends up to other instances in real time (which is fine). I don’t know how lemmy works exactly, but could the push model just be a detail of activitypub https://flak.tedunangst.com/post/what-happens-when-you-honk ?







  • Good points. Also it’s a bit of a stretch to say the different parties are using it for free anyway, because professionally maintaining infrastructure in a way where even a bit of resilience is expected is not cheap. The budget for running these self-hosted platforms can easily be eaten by one person’s salary + infrastructure costs. It in many cases would probably be cheaper to just pay for Slack, but because of some requirements they rather pick one of the open source ones and do what they can within the allocated budgets. I doubt there’s any malicious intent in these situations, or any wish to just be freeloading, but something a bit more complex.

    Ofc would be nice if money was flowing the right direction.