Recent searches
Search options
Administered by:
@errhead@podcastindex.social@alex
That is a wonderfully in depth article. That they picked Soapbox for this is really a high quality endorsement.
Though so far it's making me more sympathetic to Gab's claims that federation at scale was out of their technological grasp and wasn't dropped purely for control reasons.
@errhead @alex The author did have an interesting follow-up comment on hn: news.ycombinator.com/item?id=37137110
> We did notice on a less powerful machines the browser getting overwhelmed with the rate of new content (even though we're only streaming 10/s instead of the full 3.5k/s actually happening on the backend). I don't know if the poor performance in this context is due to Soapbox, the browser, or just the hardware.
Have you tested Soapbox with that much volume before? Doubt Spinster or Poast are putting that much load on it.
> We did notice on a less powerful machines the browser getting overwhelmed with the rate of new content (even though we're only streaming 10/s instead of the full 3.5k/s actually happening on the backend). I don't know if the poor performance in this context is due to Soapbox, the browser, or just the hardware.
Have you tested Soapbox with that much volume before? Doubt Spinster or Poast are putting that much load on it.
news.ycombinator.comWe reduced the cost of building Mastodon at Twitter-scale by 100x | Hacker News
Well, Gab was subscribed to large federation relays, meaning it received lots of content no Gab user was interested in.
And they didn't even backport changes from Mastodon that would actually improve performance.
And they didn't even backport changes from Mastodon that would actually improve performance.