HN user

esdf

71 karma
Posts0
Comments39
View on HN
No posts found.

Hosting large datasets can be expensive but the hosting for the danbooru datasets was not. It's "only" a few terabytes in size. A previous release was 3.4TB, so the latest is probably some hundreds of GB, to a TB~, in size larger. The download was hosted on a hetzner IP, which is a provider known for cheap servers. You can pay them $50/m for a server with "unmetered" 1gigabit up/down network + 16TB of disks. $600 a year would not be difficult.

Probably about the protocol being more complex with less implementations or the forced encryption which means acquiring certificates for TLS. Spoofed IPs for DDoS is a bit more troublesome for UDP protocols, though smaller players will get smacked down by any DDoS anyway. I hope to see more applications and games wrap their traffic in QUIC/HTTP3 to "defeat" firewall filtering.

A lot of english-advertising hosts from the regions in my comment are going to advertise DMCA-ignore, freedom of speech, and/or privacy. And you basically can't go wrong with any of them. Seedbox-as-a-service providers specifically are going to be generally be fine too although many of those are hosted in Netherlands which is slowly becoming less copyright-infringement -friendly.

One specific host I'll mention is vsys.host (UA) (which anna's archive uses too) but they're not going to be the cheapest option.

Being DDoS'd because you joined IRC and didn't read a manual on how to use it so you didn't know all the cool kids have bouncers or that the server may or may not have "cloaks" that you had to opt-in to. Discord is/was definitely a much better nicer experience than this.

Here's some of my pet-peeves and experiences from the webapp, the webapp wrapping desktop app, and the iPhone app. Maybe they'll qualify for awful usability, maybe they won't.

There's been an issue open for Push-to-talk in element for 6 years. Insane.

The voice and video calls/conferences aren't very good interface-wise. You could defer to this Jitsi being bad though. Screen-sharing is just broken so your alternative is to use OBS and its Virtual Camera, which means audio-sharing doesn't work. You won't actually see the video until you find the quality option somewhere to slide it all the way up.

The whole screen-sharing or conference thing remaining visible, after everyone has disconnected, until the "starter" closes it? I'm not sure.

Element X is mobile only (correct me if I'm wrong) so it's irrelevant to many users.

You can't reorder your "favourites", only sort by Activity or A-Z (and showing recent first). "Favourites" being non-American English was something I noticed right now which doesn't matter but is worth noting for i18n.

The lines after a `>quote` also get swallowed up into a quote block which I'm not a fan of and requires you to have an empty line after. Github does something similar afair.

The inexplicable Unable to decrypt messages that happen sometimes. Can likely be deferred as user-error.

Trying to sync or transfer keys across devices and not knowing if it worked so you just sit there and wait to see if the Unable to decrypt messages resolve.

"Use the Desktop app to search encrypted messages" (from the webapp)

And finally, one of the most annoying things: shit scrollback for messages. It loves to jump to the very bottom/latest messages.

folks who refuse to use a system which doesn’t have total parity with every feature Discord has”, or who are annoyed at having to run more than one chat app; Discord is where my friends are.

Well, they are on Discord. And Discord has PTT(LOL!!!) and working screen-sharing (very important!). And Element has a bunch of these random things that make it annoying to use in comparison that makes everyone call it shit and then only use it for text chat and maybe Mumble for voice, when not using Discord of course.

One more bug: some combination of right-clicking on the Element tray icon causes it to be covered by Windows' task-bar right-click menu.

That's interesting. Cloudflare tunnels do a few things that I expected to to make it perform better in general: obviously TLS termination on CF's side where they likely have faster hardware doing that (at least faster than many customers), then the keep-alive sockets for tunnel<->CF, and I think they use UDP/QUIC for the tunnel<->CF connection[0] which I figure could remove some latency.

[0]: `lsof -i | grep cloudfl` shows me 4 UDP connections & 1 TCP