IRC channel logs
2024-11-21.log
back to list of logs
<Zarutian_iPad>well, yet another eruption has started here in Iceland, ...again. <jfred>Now I just need to decide if I want to do this enough to switch to another web host so I can set the right Content-Types ^^; <jfred>the inability to set the content-type header for specific files in a static site is something I hadn't considered before... I feel like there should be some standard way to customize HTTP metadata for specific files (but I guess there isn't) <dthompson>and yeah, the server is in total control of the response headers <dthompson>if you don't have control over the nginx config or whatever then you're at the mercy of the provider <jfred>I could switch back to running my website off a VM but running servers is my day job etc etc :P <dthompson>jfred: yeah I hear you. I run my own vm but I have some strict rules I follow like "no database servers" because I simply do not have time/energy to deal with complex server stuff in my free time <jfred>I've got some stuff with DBs that I'm stuck with at this point, but mostly I'm in the same boat these days <jfred>with the exception of anything running in sandstorm. those can have DBs, sandstorm makes running those easy peasy <jfred>(this has got me thinking about headless home servers running on pis and such as an always-on companion for some spritely apps) <jfred>with libp2p potentially helping out with nat traversal and giving the home server a stable identifier you could have the initial pairing happen via a sturdyref QR code, if the server has a little display <dthompson>I wish libp2p helped with NAT traversal... so far we have not had a great experience with that <dthompson>we've done internal testing of a chat app and rotated who was hosting the chat room each day. we found that the libp2p connections worked well when the home router had upnp enabled but couldn't break through the NAT otherwise <dthompson>perhaps there's additional config we need to apply to libp2p or something but it didn't "just work" unfortunately <dthompson>some of us run openwrt on our routers and the default config disables upnp so we pretty quickly discovered that was the major difference between success and failure <jfred>it's a bit funny that tor seems like one of the more reliable options for establishing a (slow) connection between arbitrary devices <jfred>provided they both have internet connections, anyway <dthompson>jfred: oh cool link... will try to read sometime <dthompson>I see "simultaneous connect" mentioned, that's one of the techniques I'm aware of for trying to do NAT hole punching <jfred>cwebber: your post yesterday planted it as a brainworm in my head and I just had to try it XD <cwebber>guess what just got a big ol' revamp <cwebber>but we figured time to bring it live <cwebber>we're merging the best of the old and new sites <cwebber>the fun and whimsy of spritelyproject.org, the professional and research stuff of the previous spritely.institute <makx>I like the whimsy playfulness <dpk>dthompson: ah, thanks. it went live while i was having dinner just now <dpk>and now cwebber can see just how dorky i can make an example :D <dpk>nice, new Spritely website! <cwebber> (make-manga (make-book "That Time I Got Reincarnated as a Slime" <cwebber>ok I'm gonna go on a walk and then it's time to Update Our Social Media Presence (TM) <cwebber>we're gonna have a light theme option too <cwebber>and there are a billion tweaks we wanna do still <jfred>cwebber: oh, I just noticed that the fediverse/twitter links on the homepage go to the older spritelyproject accounts rather than the spritely institute accounts <bryce_m>I notice in Hoot that procedure->external is currently undocumented. Intentional because it's probably going to change, or unintentional? Seems useful is all.