Boud on Nostr: npub1ykgfh…sg4jv npub1qmcpm…yv6gx npub1cqakj…yuggm Why ignore the small ...
npub1ykgfhsl4uktyvuwmzyg647kzta8jgvhndj0mvsesa65p4sttdhks3sg4jv (npub1ykg…g4jv) npub1qmcpm5qnud8p4az4enwx0arcnw38jxyq70yzphj5xdn9va6q8s4qryv6gx (npub1qmc…v6gx) npub1cqakjxfn5anx7ayky7x63d44ltutsm42c6gz487jy7zluyt8euaqxyuggm (npub1cqa…uggm)
Why ignore the small instances [1]? The power law slope of the instance size distribution evolved
from -0.55 on 21 Nov 2022 to -0.47 on 15 Oct 2023 [2]. There are many more small instances than big ones, but the system has been evolving to strengthen centralisation (commit comment [2] wrongly interprets the data).
Adding a random sample of small instances would better represent the aim of decentralisation.
[1]
[2] https://codeberg.org/boud/fedistats_naive/commit/ba4061efa431a8b5f6eee002acfac1b4fd28cbee
Why ignore the small instances [1]? The power law slope of the instance size distribution evolved
from -0.55 on 21 Nov 2022 to -0.47 on 15 Oct 2023 [2]. There are many more small instances than big ones, but the system has been evolving to strengthen centralisation (commit comment [2] wrongly interprets the data).
Adding a random sample of small instances would better represent the aim of decentralisation.
[1]
[2] https://codeberg.org/boud/fedistats_naive/commit/ba4061efa431a8b5f6eee002acfac1b4fd28cbee