Embed this notice:p: (p@bae.st)'s status on Friday, 16-Feb-2024 16:06:23 JST
:p::rorshach: FSE went down on February 1. :manhattan: ... :rorshach2: SPC's been down all day. No explanation. :manhattan2: ... :rorshach: Someone's taking out fedi instances. :manhattan2: I wish you could perceive time as I perceive it. :rorshach: ... :manhattan3:
@NEETzsche@p@r000t@graf Server hardware failure right after replacing the seemingly defective parts. P gave up and claims to be working harder on his vaporware P2P backend instead.
I don't want to maintain hardware for this reason. But I also don't want to cuck to VPS hosts, since if they ban me I lose my data. I found the ideal compromise: daily backups.
I have a cron job that runs every day to pull all the shit updated since the last time it ran. I specify this because sometimes the computer running that cron job is off for whatever reason and so a day or three are skipped. But it pulls the changed data from all those days.
So if the VPS host jannies me I'll either make a new account on it or just find another VPS host, and do the same thing. I keep all my data and they have to maintain the hardware to make my shit run.
@NEETzsche@r000t@graf Box blew up, have replaced a large number of parts. Gonna get some new hardware, just bring it back as Revolver instead of trying to acquire another machine with enough RAM to handle FSE with Pleroma.
Don't get me wrong, backups are great, but the issue isn't the data. It's difficult to replace a machine big enough to run FSE. The box that blew up has 384GB RAM, Postgres on its own NVMe, and even then, FSE required a lot of tuning to keep up with the traffic. We were right about at 10m notifications, for example, DB is 500GB on-disk, I'd sloughed off all the media storage to Revolver.
Well, the backups are an anti-jannying measure. The point raised is that having to maintain the hardware can get expensive in terms of your time, whereas if you're a VPS fag you can throw money at the problem.
That being said, it sounds like the system you're running demands a lot more resources than mine, so maybe the VPS game gets untenable past a certain point (I'm nowhere near that point)
> having to maintain the hardware can get expensive in terms of your time
Except that I got a refurb that turned out to have a flaky mobo, the hardware has been less of a headache. VPSs get weirdness around I/O performance, network performance. VPS hosts get DDoS'd for reasons unrelated to anything you're doing, like when the Ethereum fork happened, people DDoS'd a lot of Ethereum nodes (for voting "wrong") on Frantech, and sharing a pipe with other tenants meant that was a problem. For example, FSE had split app server and DB server early on (the isolation was worth it almost immediately) but the DoS targeting Frantech meant that people were registering VMs just to DoS from the inter-VPS network and FSE couldn't get to its own DB server. All those headaches go away if you just have a computer and you run your software on a computer.
> maybe the VPS game gets untenable past a certain point
Yeah. At this point, system requirements exceed what I can throw at a VPS. It's hard to come by a dedi with enough RAM. Anyway, gradually acquiring replacement hardware, but with requirements designed around different software. :ocelot:
Plus a much nicer BMC (Linux-based, stable), better load-balancing setup, nodes instead of VMs, etc. It'll be a few weeks before I've got all the necessary gear together but I intend by then to have an alpha-quality build ready to run on it. If I'm lucky, I'll also have a couple of the nodes running Plan 9, which should eliminate a few headaches.
I'm just glad that I (mostly) stopped with the tech sperg shit for the time being because I no longer find it terribly fulfilling, and it's partly due to this kind of shit. Running a larger instance just sounds like it's all-pain, no-reward. At least, that's my impression from the outside looking in.
The last thing I did relating to that has all been about MUH SCROBBLEZ, which is a creature comfort and not that critical, and that was a pain in my ass for social rather than technical reasons. It's cool, but if it goes away it's not the end of the world, it's just dumb on the part of some project maintainer. I got native Pleroma support into WebScrobbler and PleromaFE to display the most recent one under people's usernames. Rich presence is neat, but that's all it is.
I've been working on my print magazine and the technicals of that have proven to be much cooler. The technical problems you run into with that tend to revolve around much "realer" things, in terms of physicality. Paper jams. Paper cutter blades getting dull. It's much more hands-on. On top of that, the act of generating actual content has proven much cooler to me than arguing with other tech spergs about non-issues like if a field should be url or externalLink.
Honestly I think my use of fedi might gradually shift to just running little single-user instances, or smallish ones that I just use to post about my other projects. And I'll never have to deal with needing a six million terabyte RAM computer, even if that accounts miraculously gets thousands of followers.
@r000t@NEETzsche@p@graf >Port 443 is only even opened up for the reverse proxies, *and* we require client certs for them Any particular reason for doing this instead of just setting up something like Wireguard and reverse proxying the plain HTTP port inside it?
@r000t@p@NEETzsche@graf Setting up certs and additional firewall rules sounds more complex to me than just slapping WG, binding nginx to a port inside the tunnel and calling it a day. Probably has less overhead, too, since it doesn't need to verify the certificate for each connection.
@r000t@p@NEETzsche@graf It's very lightweight and also fast thanks to being kernel-level, the whole config on both sides of the tunnel takes less than 10 lines.
> I (mostly) stopped with the tech sperg shit for the time being because I no longer find it terribly fulfilling,
Computer science only accidentally makes money for me. I like building out machines, reading research papers, making beautiful solutions, bashing out quick solutions, rattling off stunt-hacks, analyzing data from prototypes, diagnosing and fixing bugs. It's like music, but I can expect to make money doing hacker shit.
> Running a larger instance just sounds like it's all-pain, no-reward.
It's nice to give people a place to be, you know? Gives a fella the warm fuzzies. It is a huge pain in the ass and especially if you take a somewhat aggressive stance about one thing or another (e.g., freedom of speech) you kind of invite people that hate it or think you're doing it wrong, but definitely worth it. (I mean, still spending my money and time and effort on it, you can reason I think it's worth it.)
> I've been working on my print magazine and the technicals of that have proven to be much cooler.
That's cool shit. I hadn't heard of it.
> The technical problems you run into with that tend to revolve around much "realer" things, in terms of physicality. Paper jams. Paper cutter blades getting dull.
Oh, yeah, that's one of the things I like about not being on a VPS or a managed dedi, or doing RF stuff. But that's still circuits, it's not like printing or doing, like, woodworking or cooking or growing a field of corn. It's a completely different thing.
> much cooler to me than arguing with other tech spergs about non-issues like if a field should be url or externalLink.
The trick is to just not argue with them and if they insist on it anyway, have a robot disincentivize them. I like talking hackery but low-quality arguments about shit that doesn't matter are probably the most boring thing imaginable.
> And I'll never have to deal with needing a six million terabyte RAM computer, even if that accounts miraculously gets thousands of followers.
>Computer science only accidentally makes money for me.
I like building things with computers, and I'm much less ideological about it than your average tech sperg. I often joke that a lot of the software actually worth using and vidya worth playing is a bunch of dogshit spaghetti code written by some literal 12yo, and it's because said 12yo is motivated by accomplishing an outcome and not motivated by "clean" or "good" code. It's a perspective I stand by, not because my code is incoherent spaghetti garbage, but because I know the difference between inspired and uninspired work, and inspired work has a habit of not following rules.
>That's cool shit. I hadn't heard of it.
It's actually the original reason why I got this domain name. IDDQD was supposed to be a publishing house, but the early 2020s happened and I ended up just being here. It was initially a paid blog in 2019 but we shifted from that into print media, and only got our first real issue out in early 2023. I wanted my own instance not just to have one but as a stomping ground for my readership, who never got on fedi. They never got on fedi because they regard it as a Twitter clone, and they all hate Twitter. So my entrenchment in fedi is kind of an accident in other efforts.
I attached a few covers of the zine, which you can find here: https://iddqd.pub/
I'm considering doing a digital version because people are worried about doxing/opsec.
>The trick is to just not argue with them and if they insist on it anyway, have a robot disincentivize them.
Well, in this case, it was because they were a project maintainer and I was PRing them. I had the spec on my side but they had the merge button in theirs. I think you can see how this works out. Logos cedes to ethos in this fallen world. RIP in pepperonis mi amigo.
> it's because said 12yo is motivated by accomplishing an outcome and not motivated by "clean" or "good" code.
This is one of those things that has something to it as long as you are not taken literally, which you inevitably will be. (Cf., Moore's Law, Postel's Law, 90% of anything that Fred Brooks wrote.)
I've read what James Hague has to say about it, and he has a lot of very good points. (I have probably pasted a lot of URLs for his articles at you.) On the other hand, he's a game developer (and thus has a radically different lifecycle for his code than he would if he wrote distributed software) and I am old. (Maybe not as old as he is but old enough to know what I'm doing.) What constitutes "good" from my perspective is driven by practicality rather than some fantasy a dude has in his 20s about what constitutes "good" code. The young guy will write 10k lines of "clean" C++ to accomplish what an awk one-liner does, he's got no scar tissue and is prone to compulsive masturbation (technical and otherwise). VCs love this because in their view, either the market is cornered or the business is a failure, so they turn everything into a moonshot because a VC makes way more money with one unicorn and ninety-nine flops than with a hundred sensibly run modestly successful businesses.
"Good" in my case is informed by having run Pleroma for years and knowing the shape of fedi traffic and what is fast and what is slow and what should be faster and how to handle that, knowing what sort of things I want the software to do, knowing the general shape of the unknown unknowns you run into with stuff like this, and knowing the process of long-term maintenance and debugging a network, which is different from spot-welding a feature.
> inspired work has a habit of not following rules.
You've got to master the rules before it's meaningful for you to break them. A kid shoves a goto into a function and the code is worse; Ken puts a goto into a function and the code is better, and there's a good reason why. Arthur Whitney dispenses with line breaks and his code is densely packed and nigh-unreadable even to an experienced hacker, but the reason he does that is different from the reason a clown does it. An idiot breaks the rules and you get Clerks, a genius breaks them and you get Waiting for Godot.
> IDDQD was supposed to be a publishing house, but the early 2020s happened and I ended up just being here. It was initially a paid blog in 2019 but we shifted from that into print media, and only got our first real issue out in early 2023.
Oh, that's cool shit. Congrats!
> They never got on fedi because they regard it as a Twitter clone, and they all hate Twitter. So my entrenchment in fedi is kind of an accident in other efforts.
Ah, yeah. I think most of the things that make Twitter hellish don't apply here, but I can see it leaving a bad taste in someone's mouth.
> I attached a few covers of the zine, which you can find here: https://iddqd.pub/
That is seriously awesome.
> Well, in this case, it was because they were a project maintainer and I was PRing them.
Yeah, usually when that kind of thing happens, I just leave it. There's enough work to do without wasting time: ultimately, they'll do what they want with the project and I've sent my PR and I'll make adjustments if they want adjustments but if they don't like it and we can't work together, I should probably do my hacking elsewhere. Sounds like you had the same kind of experience.
\>You've got to master the rules before it's meaningful for you to break them. A kid shoves a goto into a function and the code is worse; Ken puts a goto into a function and the code is better, and there's a good reason why.
I want to write something to the effect of "all control flow is goto in disguise" but I'll spare you that rant. So yes, you should master the rules before breaking them, but most importantly, you should get shit done, which is something I think we agree upon.
\>Ah, yeah. I think most of the things that make Twitter hellish don't apply here, but I can see it leaving a bad taste in someone's mouth.
I don't think it's necessarily a cultural problem, exactly. It's that they don't like microblogging. They want a chat room, which is a lot closer to a stream of consciousness in certain ways. Chat rooms are more conversational. Also we get on voice chat a lot, which is like a Zoom meeting except we're talking about cool shit and not gay shit.
\>That is seriously awesome.
Thanks fam. IDDQD Magazine is a zine for guys who either use the 4chan blue boards, or would if they knew about them. That's the culture that informs it. It's interest-based, but it's still filled with the right personality types, i.e., ours. Or rather, the parts of our personality types that overlap. There is a particular focus on /tg/, though.
Currently, this publication pays, but it's truly a pittance. An article buys you a nice lunch, basically. The "vision" I have for this is that it has a variety element. The readership should have an above average intelligence, but we put in stuff from all kinds of different backgrounds. So imagine this:
You are a tech sperg, a hacker in the old sense of the term. You might get an article pertaining to that, but when you turn the page, you'll get something else. You'll get an article about something like, I don't know, the Las Vegas underground rap scene. Or some weird math geek shit about dice pools in TTRPGs and the expected number of successes on X dice with Y sides with a success threshold of A and an explosion threshold of B. Then you turn the page and you get some political-theological philosophy shit.
So it's not your, the reader's, wheelhouse, and it's not supposed to be your wheelhouse, but it's supposed to be comprehensible to you in spite of that without insulting your intelligence. If you were to write tech sperg shit, I'd try to get you to write it in a way that a clever automechanic could read and comprehend it without it reading like Newsweek slop. And then when you turn the page you might get something about why Funkwhale is a piece of shit from the perspective of someone who wants to jam on their guitar and upload it without getting a computer science degree's worth of knowledge to do so. Or why Paladins past AD&D 2e are literally demonic and you don't realize it because you're not a theologian.
And none of this is made-up. We already ran all that shit. And who knows what next. I literally paid my mortgage shilling this shit mid-2023 out of Vegas punk rock shows.
\>Yeah, usually when that kind of thing happens, I just leave it. There's enough work to do without wasting time: ultimately, they'll do what they want with the project and I've sent my PR and I'll make adjustments if they want adjustments but if they don't like it and we can't work together, I should probably do my hacking elsewhere. Sounds like you had the same kind of experience.
That's exactly the kind of experience I had, but I will admit I was trying to overcome it in this case specifically so I can get scrobbles off the ground on instances other than my own. And it worked. I won that battle. I'm not sure what client you're using, but if you're using the Baest website, you'll see my scrobbles on SJW's fork of PleromaFE.
>Oh, right, I had been meaning to ask you about Paranoia. It seemed like a lot of fun but there was some soulless corporate interference in a couple of editions and then the original team bought the rights back but then licensing issues made the entire thing impossible to get again. Apparently it didn't get popular enough for the novels to arrive on TPB or Libgen, but it's one of the more credible dystopias and actually dystopian.
I found every edition of it on MyAnonamouse (deliberatley misspelled for some reason), so yeah. It's out there. How do you like to transfer files? You got an FTP server somewhere?
>Ha, sounds like an entertaining zine.
It is, and we take submissions from basically any down-the-rabbit hole shit that belongs on a 4chan blue board, so take that into consideration.
>Oh, yeah, it's bloat. (Local on-device bloat; I use FSE's bloat when FSE's up as a dogfooding measure but for the moment, if I'm not hosting bloat, I can just use it locally.)
>I can figure out how to get at the metadata from the client side if it comes down when you fetch a profile; what's it look like on the other end? Something cmus-fm might be able to do (or get hacked to do)?
Unfortunately, the scrobble info comes from an entirely separate API call. I put in a PR to make it return the most recent 5 scrobbles as part of the account information but they basically closed the PR spuriously. If you want the endpoint, though, it's /api/v1/pleroma/accounts/$account_id/scrobbles. Here's an example of the output, when the limit param is set to 1:
> I want to write something to the effect of "all control flow is goto in disguise"
Well, sure, it always boils down to "jump" or "jump and keep track of return address" or "pop return address", but language constraints can be helpful. But I think you understand the point.
> most importantly, you should get shit done, which is something I think we agree upon.
"Get it working, get it working correctly, get it working fast." or "Real artists ship." or "A good plan violently executed now whichever guy you like, there's a quote. This is one of the commonalities you can get from any of the really good hackers from this field's brief history, and is a rare case of the really good business guys and product guys also saying the same thing.
> It's that they don't like microblogging.
Fedi feels more like mailing lists used to feel; "microblogging" is more of a UI concern than an interaction style. (Just thinking aloud, not trying to get you to drag people here.)
(I still think an IMAP/SMTP interface to fedi would be workable. It might end up much more usable.)
> Thanks fam. IDDQD Magazine is a zine for guys who either use the 4chan blue boards, or would if they knew about them. That's the culture that informs it. It's interest-based, but it's still filled with the right personality types, i.e., ours. Or rather, the parts of our personality types that overlap. There is a particular focus on /tg/, though.
Oh, right, I had been meaning to ask you about Paranoia. It seemed like a lot of fun but there was some soulless corporate interference in a couple of editions and then the original team bought the rights back but then licensing issues made the entire thing impossible to get again. Apparently it didn't get popular enough for the novels to arrive on TPB or Libgen, but it's one of the more credible dystopias and actually dystopian.
> get you to write it in a way that a clever automechanic could read and comprehend it without it reading like Newsweek slop.
Oh, yeah, I've had a lot of non-technical clients. I think I get what you mean.
> Or why Paladins past AD&D 2e are literally demonic and you don't realize it because you're not a theologian.
Ha, sounds like an entertaining zine.
> And none of this is made-up. We already ran all that shit. And who knows what next. I literally paid my mortgage shilling this shit mid-2023 out of Vegas punk rock shows.
That is awesome!
> That's exactly the kind of experience I had, but I will admit I was trying to overcome it in this case specifically so I can get scrobbles off the ground on instances other than my own. And it worked. I won that battle.
Nice. I think I've seen them around, ryona.agency had them, I think SPC had them.
> I'm not sure what client you're using, but if you're using the Baest website, you'll see my scrobbles on SJW's fork of PleromaFE.
Oh, yeah, it's bloat. (Local on-device bloat; I use FSE's bloat when FSE's up as a dogfooding measure but for the moment, if I'm not hosting bloat, I can just use it locally.)
I can figure out how to get at the metadata from the client side if it comes down when you fetch a profile; what's it look like on the other end? Something cmus-fm might be able to do (or get hacked to do)?
I added WebScrobbler support but that's really not the same thing as a desktop program. So yeah. Maybe I'll have to write one for that. Which player do you like these days?
Just realized you can build top ten lists with SQL. Here's one for the top ten artists of all time:
pleroma=> SELECT COUNT(*) "count", data->>'artist' "artist" FROM objects WHERE data->>'type' = 'Audio' AND LENGTH(TRIM(COALESCE(data->>'artist', ''))) > 0 GROUP BY data->>'artist' ORDER BY COUNT(*) DESC LIMIT 10;
count | artist
-------+------------------
469 | OT the Real
440 | James Kibbie
278 | Jedi Mind Tricks
247 | Vinnie Paz
240 | e00
220 | Rippingtons
209 | Aesop Rock
204 | 2023
193 | Taylor Swift
193 | Outerspace
(10 rows)
>Appreciated! I was mainly curious, since you're plugged into the tabletop crowd, does anyone play it? Has anyone even heard of it? I randomly found a forum post by one of the authors complaining about licensing issues (it is actually fitting that the game should end up in a Kafka-esque nightmare situation) preventing the novels or the game from being sold at the moment.
Paranoia is one of those games everybody in the TTRPG world has heard of but few have played. It's also one of the few I haven't played. One of my friends from middle/high school has, though. He sung its praises for years, and it was off of a really short campaign that lasted like three sessions.
Paranoia doesn't lend itself to long campaigns and in the last fifteen years or so, games that do have taken over in the form of living world, online games. These kinds of games pop up all over the place but the people administering them don't have much technical background and end up using ad-supported meme hosts like freeforums.net, which aren't catalogued that well by Google etc. One such game I've been on for like three years is Dying Every Day, which is found at https://dyingeveryday.freeforums.net/. They tend to also rely on commercial chat services like Discord and, before that, Skype, and most of the real decisions in those circles are made over those. This makes most of the content effectively obscured. Hilariously enough, things that want to be very public end up existing only in whisper networks of TTRPG spergs, creating an accidental opsec that's surprisingly effective.
I've spent dozens of hours trying to convince these people to get off of centralized services like Twitter and Discord and what-not, but it was to no avail because the motivations of TTRPG spergs are not the same as the motivations of FOSS spergs. If you rub shoulders with any of the serious TTRPG spergs you will figure that out near-instantaneously.
I built a Telegram <--> Matrix and Discord <--> Matrix bridge, and might build a fedi <--> Matrix bridge, but I'm really wary of the latter. Some of my guys (@sundowner and @Wednesday) occasionally do post on fedi but, occasionally is the key word here. I was in a voice call with these two just today, though, so they're active, just... not in the same spheres as tech spergs.
I kind of wish we could bridge that gap, though, seriously. There would be a lot of benefit if different spergarinos had better networking.
>The way I usually transfer files is fedi attachments or IPFS or I have a million webservers or I just upload it to Revolver (which, as of last year, also handled FSE's attachments), but usually it's me sending files. bae.st has a filebin at files.shittyurl.org .
It expires in 24hrs though so you should just grab it before it autodeletes.
>Seems ideal to lurk some.
That's fair. Like I said, I've been more seriously considering a digital edition because a LOT of guys have been worried about me doxing them or something since I have to mail print to a physical location.
>Excellent. Where's the endpoint to send stuff to, though? You post there?
You POST to /api/v1/pleroma/scrobble, and pass it a title, album, artist, and length as JSON. I've been using length to be an integer that is equal to the number of seconds, but I think they may have wanted to use milliseconds originally.
> I found every edition of it on MyAnonamouse (deliberatley misspelled for some reason),
AHA
> How do you like to transfer files? You got an FTP server somewhere?
Appreciated! I was mainly curious, since you're plugged into the tabletop crowd, does anyone play it? Has anyone even heard of it? I randomly found a forum post by one of the authors complaining about licensing issues (it is actually fitting that the game should end up in a Kafka-esque nightmare situation) preventing the novels or the game from being sold at the moment.
The way I usually transfer files is fedi attachments or IPFS or I have a million webservers or I just upload it to Revolver (which, as of last year, also handled FSE's attachments), but usually it's me sending files. bae.st has a filebin at files.shittyurl.org .
> It is, and we take submissions from basically any down-the-rabbit hole shit that belongs on a 4chan blue board, so take that into consideration.
:bigbosssalute: Seems ideal to lurk some.
> /api/v1/pleroma/accounts/$account_id/scrobbles
Excellent. Where's the endpoint to send stuff to, though? You post there?
> closed the PR spuriously.
kaniini was working on scrobbles before leaving. There may be some things going on (due to projects of a certain age often having accumulated some interpersonal historical cruft) that weren't apparent on the surface; I don't know.