@tevo I'll support you on that one.
And even as annoying/verbose as XML is, at least it's not YAML ☺
@tevo I'll support you on that one.
And even as annoying/verbose as XML is, at least it's not YAML ☺
@teajaygrey
amusingly, my brain first tried to parse your curse-word abbreviation as a Usenet group hierarchy 🤦
I regularly hit CloudFlare-fronted services that refuse to return (even static) pages because I'm browsing from lynx(1). Really irksome. And yes, changing the UA-string to something Chrome-ish/Firefoxish merrily bypasses the braindead check.
@feld @Tubsta I'm unaware of any such coops, but there aren't that many folks in my circles who fall in the intersection of technical competence and familial trust. Fortunately, the VPS requires minimal upkeep and I try to keep the domain-name registered as far out as I can, so, while my demise might mean she has to switch to something like gmail or a different mail provider, she at least has some time to do it.
Of a "vintage" age myself, I remember when file-servers (good ol' Novell Networks), email servers, FTP-servers, etc, were all measured in *mega*bytes of RAM. The first machine I had with >1GB of RAM blew my mind that software could be so profligate. And yet here we are
:glares at Chrome & Firefox:
RA RA ah-ah-ah, roma, roma-ma…
@Tubsta As much of a pain as running your own mail-server is with regards to successfully delivering, everything else is pure win. Your mail provider dropped POP3 or IMAP support? Bummer, mine still works without issue. Oh, they added OAUTH2 requirements and your preferred local mail client can no longer connect? Sad to hear, but mine doesn't. Your Exchange server requires client-application verification and no longer allows you to connect from mutt? Sorry, but I can still connect to my mail just fine even with `openssl s_client` if I need to.
(having experienced all of the above with connecting to $DAYJOB mail, I feel the pain acutely and appreciate that my personal mailserver doesn't manifest any of that grief)
ActivityPub/ActivityStreams allows for instances to cache media from external instances, so depending on whether instances choose to do that (some do, some don't), you could either get media requests per-server (scales roughly with a nebulous decentralization-factor) or per-user (which can drag a machine to its knees for popular posts)
(and please no blockchain! 😂)
Having been around the internet since pre-web days, there was very little paying others for services in the early days—the only paying involved setting up your own (obtaining your own hardware, paying for connectivity & electricity, etc), Or paying tuition to a university that gave you internet access 😉
@ItsThatDeafGuy 👋 notes that "free" is largely part of the past, and I think there's some truth in that:
It could mean users paya huge central host to keep it afloat as it tries to manage bajillions of users uploading umptygazillion media files, and any failure gets felt by all users.
Or it could be that more folks obtain their own domain, set up their own $5/month VPS, and run their own instances independent of the problems at the central site. Maybe they avail those services to friends/family allowing free participation for some folks.
But the AP model scales fairly nicely as the decentralization grows as @WiteWulf opens with.
impossible: when you have the ability to turn an imp to bone
It's not DNS! (it was DNS)
@stefano
It's also part of what seems broken over in the Bluesky world: sure, it's "federated", but only if you have a huge investment in infrastructure.
From Bryan Newbold's defense of atproto¹, "Our design goal is not to run the entire network on small instances…This architecture facilitates big-world indices…Such indices are not cheap at scale!"
But I want my federation to be so affordable that anybody can get a domain-name, spin up a $1/month VPS instance², and participate in the fediverse. Things like snac make the completely feasible. So thanks, @grunfink!
⸻
¹ https://whtwnd.com/bnewbold.net/3lbvbtqrg5t2t
² https://lowendbox.com/blog/1-vps-1-usd-vps-per-month/
Recently saw an interesting comparison between modern LLM prompt-injection and old-school phreaking—interpreting data *and commands* over the same channel leads to arbitrary users being able to send commands.
https://lobste.rs/s/bbrgdy/lessons_from_red_teaming_100_generative#c_d0tdc1
I'm sure I knew this at one point, but (re)learned today that a DNS TXT record has a limit of 255 characters.
(he says noting that it's roughly the length of a Tweet and pondering the ramifications of microblogging-over-DNS which is absolutely horrible…hey, brain, no…stop it, STOP it!)
@brynet So this crossed my feed and made me think you'd get a smile
Looks like the full list of those friends can be found here:
https://github.com/so-okada/toXiv?tab=readme-ov-file#list-of-bots
Thanks, @nemobis, for highlighting those!
@ltning @rubenerd
Those hardware specs are delightful (and it says a lot about @grunfink's work that snac is svelte enough to run in such a constrained environment; I can't fathom attempting to run full Mastodon on such specs)
@freedosproject Congrats to all involved!
Need a 5-minute count-down timer in the terminal?
$ sleep $(( 60 * 5)) ; play ding.wav
(or whatever notification you want like inotify or mpg123 or xeyes(1) etc). If you run a BSD that respects SIGINFO, you can even see how much time has elapsed.
Need a stopwatch instead?
$ time cat
Use control+d to stop. Again, access to SIGINFO can give you a lap-timer for free.
GNU social JP is a social network, courtesy of GNU social JP管理人. It runs on GNU social, version 2.0.2-dev, available under the GNU Affero General Public License.
All GNU social JP content and data are available under the Creative Commons Attribution 3.0 license.