ManMachine
@max@manmachine.me
the 555 timer is 55 years old, so if we celebrate on the 5th of May it can be the 555's 55th on 5/5.
#fossangel's favorite brand
Do you bring your phone to the shower?
| What? No???: | 235 |
| HELL YEAH SHOWER PHONE: | 46 |
| At least once, I have been in the shower for more than an hour, with my phone: | 27 |
| My water bill is crying: | 17 |
Closed
@ShadowJonathan @patterfloof I have no need to test the waterproofing claims of my phone's vender inside the house.
me: "I'm sending you a link to something I translated."
my husband: "is it more Confucius?"
me: "it is not Confucius. It is the antithesis of Confucius. It would strike Confucius dead if he saw it"
@0xabad1dea I dunno, this looks like a harmonious social relationship. Plus order is being restored!
@0xabad1dea he who fart in church sit in own pew
he who stand on toilet, high on pot
if a clown farts does it smell funny?
@terribletowelie I agree that would strike Confucius dead if you subjected him to it, but I'm gonna encourage you to check out an idiomatic translation of Confucius (hey I have one right here, that I wrote) and to steer away from associating him with what happens when Chinese is translated hyper-literally with a dictionary https://xn--hmr.net/classicalchinese/analects/
@0xabad1dea Nice.
Until the last panel, I thought it might have been the Ken Liu story *Good Hunting*. (recommended)
@max @0xabad1dea smilarly, it reminds me of SCP-953. Although, being an SCP, it takes a more horrific tone.
@max @0xabad1dea Jesus that Wikipedia page is terrible. It's not actually a proper article, it's a high-school 1-page essay about the book.
@0xabad1dea what’s the original comic called?
@twinkle "Demoness in Disguise" (妖精的伪装) is the name of this specific short; there doesn't appear to be an overall series name for the artist's many shorts, they're just labeled as being by hands2.
@0xabad1dea is there a page 7 for another day?
@gkrnours this author does one-shots about supernatural hijinks, but sometimes characters recur, maybe they’ll pop up later
It's common for network services to squash 404 (not found) and 403 (permission denied) responses into the same response, so as to avoid giving away whether a thing exists at all that you're not allowed to see.
You can see why they do it, but it is annoying, especially if you're not alert to the possibility. You might start trying to debug the wrong problem: checking and double-checking the URL spelling when you should have checked your credentials. Or vice versa, depending on whether the site is pretending everything is 404, or pretending everything is 403.
(Worse, the confusion spreads to sites which _do_ separate the two errors. Once you've started mentally conflating the two, you waste time checking the wrong things even when the error message truthfully told you which thing to check.)
But which is _more_ annoying, of the two?
| 404 → 403: pretend nonexistent things are secret: | 211 |
| 403 → 404: pretend secret things don't exist: | 112 |
@simontatham to me 404 means only "either you messed up the url or we messed up the website", for everything else there's 410.
If i'm not allowed to know whether the URL exists or not, that's clearly a 403. 403 implies it's not final and can turn into something else with the right credentials. Sometimes the something else is a 404 and/or the right credentials don't exist, and I don't need to know that, I just need to know mine aren't it.
So yeah, fake 404 is the most annoying thing and should never be done. And Github of all places should have known better. (I wouldn't expect today's Github to make the right call on anything, but it's been like that forever)
Edit: meant 410 not 405
… actually, today's case of this turned out to be neither.
I was inspired to post the poll by what I thought was a 403 annoyingly misrepresented as 404. But it turns out it was some kind of 5xx transient failure misrepresented as 404. OK, _that_ is worse!
@simontatham better yet, the 200 OK response with 403 Permission Denied as the body text, so beloved of IBM Tivoli access control suite
@simontatham
The company I work for exposes an API that returns 'error 500' when your query doesn't return any data (even if the call was valid).
@simontatham had a live production webservice reporting 4xx errors via a redirected page *saying* 404 or whatever, but with a 200-OK response code.
loadbalancer healthchecks hate this one simple trick
@simontatham I'd prefer making everything a 419.
@simontatham
There's an even worse 404 error: where the httpd responds OK, but the framework that's supposed to display content decides to put the message "404 Error" in the page instead of a useful message.
End user: "I'm getting a 404 Error!"
Sysadmin: quickly checks server logs, sees a 200 result.
@simontatham So, the sane story, IMHO:
* If the resource type requires authentication, and you're not authenticated (no/invalid token) you get blanket 403.
* If you're authenticated, your view is functionally scoped to your authorization: listings include only things you can see, trying to access any other URL gives 404.
* If you're trying unallowed things on resources you can see, that's a 403 (may be 405, if split by method). Doing an allowed operation referencing unallowed data is 400.
@henryk I think you're disagreeing with @virtulis in another subthread. Perhaps you should argue that one out with each other!
I was thinking that Unix filesystems will sometimes return EACCES, analogous to 403, for a nonexistent file, but only if you don't have access to some on the path to it. If you have x permissions on the directory, you know for sure what files do and don't exist inside it, independently of whether you can read each of those files. And when you're local, it's cheap to make that extra query and get more information to resolve the confusion.
@simontatham @henryk @virtulis Actually, I think they’re both saying the same thing, just approached from a different angle.
@jornane do you think so? I read @henryk as saying that it's correct to return 404 for a thing that exists but you aren't allowed to see it, and @virtulis as saying it's wrong and that should be 403.
When I made the original post, I was thinking of the question "which is more annoying?" in terms of the practical consequences. For example, if in some particular case it's relatively easy to check a URL's spelling (maybe it's linked from lots of places) and hard to check your credentials (maybe they're stored on some CI server and you don't have a local copy), then it's more annoying to make people check the credentials unnecessarily than to make them check their spelling.
So it's nice to see the question being looked at from this completely different angle as well.
@simontatham @jornane @henryk fwiw I think in most cases the answer is use long random IDs and don't conflate statuses.
But yes my response was mostly about human browsing, and how e.g. Github will log you out and then tell you your own private repo no longer exists, which is just asshole behavior, isn't it.
APIs are always a mess and if the response doesn't contain anything specific then the status code is unlikely to help much either, IME. But yes, if the response is literally "resource does not exist" while it does exist, that's much worse than "access denied" when it doesn't. Because one of these is blatantly a lie.
@simontatham Also, please return 401 if the authentication is correct.
If you return 403, I’ll assume it’s a permission problem, not a token problem.
401: I don’t know who you are, and anonymous access is disallowed here.
403: I know enough to identify you; you can’t access this URL
404: I know enough to identify you; this URL doesn’t exist
405: You’re using the API very wrong
@simontatham I can understand presenting forbidden things as non-existent, to prevent an unauthorised user determining the existence of private resources.
But presenting non-existent things as forbidden seems completely wrong. Implies a correctly-authorised user is being given authentication errors, and presumably then going to try debugging their authentication.
@simontatham Also, when everything is 400 .. "you made a boo-boo", opaque non-descript error messages, such as "something went wrong".
@simontatham I strongly believe there should be a 40x NOT TELLING response to make this more correct.
@Floppy 403½ :-)
@simontatham I smell an RFC
@floppy @simontatham 403½ Glomar: Server can neither confirm nor deny the requested resource exists
So, something that's been bugging the shit out of me?
These fucking assholes who let LLMs run rampant and delete prod?
They query the LLM for "why" it did that.
This is delusional behavior.
LLMs do not have a concept of 'why': they assemble a response based on a statistical sampling of likely continuations of the original prompt in their database.
LLMs do not have the ability to have motivation. It is a machine.
LLMs, further, function by instantiating a new runtime -for each query- that reads the prompt and any cache, if they exist, from prior sessions:
which means, fundamentally, "asking" the LLM to explain "why" "it" did a thing is thrice-divorced from reality:
It cannot have a why;
It cannot have a self to have motivations;
And the LLM you ask is not the one that did it, but is a new instance reading from its predecessors notes.
Treating it as tho it is an entity with continuity of existence is fucking delusional and I am fucking sick of pandering to this horseshit.
Touch some grass and get a fucking therapist.
@munin How did we get to the point of _asking_ the computer? You don't ask a computer, you tell it. You give it a command and it either succeeds or it fails or or it is broken. It's a complicated box of sand. There's no awareness, no spark, just an odd arrangement of doped silicon and metal. Believing there's more than that is deeply deeply delusional, like believing socks are sentient because you made a sock puppet once.
Intelligent
@falseknees The beak is the pinnacle of evolution. The ancestors of birds, the dinosaurs, had mouths, but only those with beaks survived.
https://phys.org/news/2022-09-bird-neurons-glucose-mammalian.html
1. bird brains (no pun intended) use 1/3 the glucose mammalian brains do
2. the neurons are packed tighter, are smaller
"the neurons consume less glucose—this could have been expected by differences in the size of their neurons," says Kaya von Eugen of Ruhr University Bochum, Germany. "But the magnitude of difference is so large that the size difference cannot be the only contributing factor"
birds are our cognitive superiors
they have more advanced CPUs
@falseknees wanted to introduce my little niece to the geniality of False Knees, but to my great horror your website falseknees dot com shows up as "temporarily disabled"
Any other link apart from the mastodon bot that is not a social danger to an early teen?
@falseknees I saw a post on FB recently from a "science" page that shared about how whales aren't just singing, they are utilizing vibrations in their voices to forecast ocean currents 6 months in advance telling [us humans] that maybe they know more than "we" thought. I was offended at the notion that only humans understand seasons and that the post suggests whales can only predict up to 6 months from now.
@falseknees This reminds me of a fiction story I am writing about corvids that form communities and villages 😅
@falseknees I have always adored the skills of my brain, which does my thinking, when it concluded it is the most important organ of my organism.
Will you be buying the new Steam Controller from Valve?
| Yes: | 224 |
| No: | 201 |
| Waiting on reviews: | 237 |
Closed
@gamingonlinux I'm torn, initially I wanted to but at a leaked $99 it's a tad more than I'd want to.
I've actually started to use my Dualsense a tad more now too as finding more and more games support the haptic triggers which I do like.
@gamingonlinux with the caveat that I bought the last one and still have it somewhere and might have to justify the purchase in that knowledge
@gamingonlinux I know this an unpopular opinion™ but I got 2 of the originals, and they were and still are a big let-down
@gamingonlinux Without question. I love playing my Steam Deck docked on my living room TV, but always sad to lose the touchpads. I’ve got the OG Steam Controller, but it doesn’t have enough buttons, and the shape is funky (tho more comfortable than you’d think). This feels like the perfect evolution.
@gamingonlinux I thought I would but at $99... I think I'll save the money for the Frame which is what I'm most excited for.
@gamingonlinux Secret fourth option:
Maybe when my DS4s crap out because I'm very happy with those otherwise
Kinda waiting on reviews tho, I guess. I have a pair of the old Steam Controller and it was a pretty mixed experience with them. The new one seems to address some of my gripes but a big one was that the build quality on the old controller was not great. It just doesn't feel super good to hold etc...
Optimistic, still, really want these to be great
@gamingonlinux Got two Steam Links, an OG Steam Controller and a Steam Deck. Gotta keep the family together 🤣
@gamingonlinux No, I'm not buying anything from Steam/Valve anymore. I could consider GOG-controller if they make one. 😉
My "good" controller seems to have gone missing in the last month, so it's a well-timed replacement.
In the long run, I expect to have two, as the Steam Machine is also supposed to come with one - which is ok, I like some two-player games. But if I can get one now and they offer the Steam machine without, I'll likely go without.
@gamingonlinux I answered yes, but it's more of a "very likely." I'm in no rush, but my kid adores the original Steam Controller, and theirs isn't always reliable after ten years of being dropped on the floor and handled with sticky kid fingers. I think we'll get at least one of the new ones to replace the old.
My PC gaming is 99.99% RetroArch, and at the moment, my old PS3 controller works fine. I know it won't last forever though, so when I need a new one I'll likely be mulling it over between a Steam controller or something from 8bitdo.
Edit: forgot to say "reviews will play into that decision" 📊
@gamingonlinux either when my current controller dies (which may be soon sadly) or whenever it gets on sale. Whatever comes first.
@gamingonlinux It definitely costs more than I would like, but it really looks to be the ultimate controller for the steam deck, especially with the dual touch pads.
I usually use an Xbox controller but occasionally have to swap to a PS4 controller due to it having a touchpad.
I really think the steam controller will take the place of them both.
@gamingonlinux Not yet, maybe if it was 20€ cheaper or I needed a new controller. But I don't need one right now, and the price although maybe fair considering the specs, is higher than I expected.
Tired: “Dogfooding”: Using the software you work on with unbridled happiness and glee, no consideration what it is
Wired: “Catfooding”: Looking at the software and saying “Fuck that shit. I ain’t using that”, and then going for a nap.
half the point of programming-tool design is to reduce the need for hypervigilance on the user.
if we're designing tools that require you to be *more* hypervigilant, legitimately what use are they?
Speaking of Unicode, I think many people would be surprised just how strong the pushback against it was even in the early to mid 2000s. UTF-8 adoption in Linux was a fight.
@mjg59 at least around here, people still do not seem to fully understand it which then ends up in, hey try not to use any Umlauts in the masterdata which makes me go gnarrr all the time.
@mjg59
The opposition in Poland was crazy, especially given that we suffered 3 competing codepages not that long before unicode arrived. The fights on the newsgroups were massive!
@mjg59 i know backwards-compatability but it's always striking that perl still needs you to do a whole dance for it not to break everything that is utf8 which is basically everything.
@mjg59 More features means more things that can break. Things are mostly fixed now, not perfect, but most of this renders correctly for me: https://antofthy.gitlab.io/info/data/utf8_demo.txt
@mjg59 Ah, I remember running webservers on 4 different ports with different Cyrillic encodings. Why would anyone want to stop doing that...
And I can still "read" the 7bit version of koi8-r, which obviously made it the best encoding.
@mjg59 I recall getting into a mailing list argument argument about supporting it properly in the in-kernel terminal emulator. It forgot UTF-8ness after a hard reset sequence and had to be configured again using (if I recall correctly) an `ioctl` to set it back again. I felt a detail like that should be remembered.
@mjg59 I can still recognize Shift-JIS decoded as Windows 1252 on sight from how much time I spent on Japanese sites in the early 00s
please don't tell me to read that great essay on why AI is so bad, when it features:
> I use AI tools sparingly for assistance while refactoring code in languages I understand. I occasionally use it to help compose command line arguments for tools like ffmpeg.
straight after why not to use it
YOU CAN LITERALLY NOT DO ANY OF THAT AND NOT MISS A THING
and directly before
> The explosion of AI has played a significant role in my own burnout.
i mean,
We know we have posted this comic a couple of days ago. We're just curious what you prefer: *4 panels in one image* or *panel by panel" posts?
?
| 4 panels in one image: | 1177 |
| panel by panel: | 400 |
Closed
@warandpeas I personally prefer the 1 panel for your comics since they're short, but the alt-text in the panel by panel makes it a lot more accessible, imo.
So panel by panel.
@warandpeas why not both? That worked fine here.
edit: you can include any text you want to along with the all in 1 image reply message, such as the usual call to action for paid supporters.
@warandpeas The individual panels are cut off in previews and have to be opened separately. One picture is more likely shown in full. (Screenshot from Phanpy web client)
@warandpeas
Have you seen @davidrevoy's take on this subject recently ?
@Zekovski @davidrevoy Yes! His experiment is the reason we are doing this!
@warandpeas @Zekovski @davidrevoy
Voted 4 panels 1 image for you, and the reverse for Revoy. I think one panel works better for your work.
@warandpeas 4 panels in the same image please.
Allows one to download the comic for future reference and sharing it with friends.
@warandpeas I don't have much of an opinion. I will expand the image anyway; all together lets me glance before expending, and one by one lets me look better at each panel. Four panels in a picture would allow you to play with the panels' borders, if you decide to
@warandpeas panel by panel is easier to read here; but I wanted to shared it and I didn’t because I’m not gonna download 4 images, etc…
@warandpeas I interpreted "panel by panel" add one panel per post and not "all four panels as a picture on it's own but in a single post". And choose the other option... I would prefer 4 individual panels in a single post.
@warandpeas I prefer panel by panel because it makes it less likely I read the punchline too early by accident.
@warandpeas For people with bad eyes, rhe four picture solution is very welcome and more comfortable to read. On a smartphone
@warandpeas I personally prefer 4 in 1 but I understand why people would prefer four separate ones so as to get a better resolution
@warandpeas I voted for for 4-in-1 and then realised the 4 separate panels look exactly the same except for better resolution. Obviously this is all dependent, I'm using Pachli, but some apps may not display the separate panels as well. So, for me, 4 separate is actually better.
4 panels in one image.
Also, artifact.
@VedaDalsette Depends on where you live.
@warandpeas 1 image rather than 4 images. Contrary to what one would think, 4 images is worse for mobile, because the controls to change images cover parts of the image and sometimes text is unreadable. Every time someone posts 4 images, I email myself a link so I can read it at home on desktop, because it's unreadable on mobile.
@warandpeas In this case, *definitely* the all-in-one, because it seems you've accidentally reversed the first two images in the split-out version. :-)
(But yeah, I also generally prefer all-in-one. I'm a sighted user on a laptop and this is closest to how I read comics normally.)
Should I be shitposting more on here?
I mostly think of Mastodon as a lower-rate platform, where I try not to flood people, but iunno, thoughts? Poll!
(your own experiences are quite welcome too)
| shitpost it up: | 57 |
| nah keep it chill: | 7 |
Crypto? Yeah I have some, in /usr/src/crypto, like krb4 and ssh, why are you asking
@nina_kali_nina you're forgetting about the endless source of crypto in /dev/urandom
Kind of kills the market, I think 🤔
(This toot might contain traces of a slight stretch of the word "crypto")
have you beaten the ender dragon? (in survival)
| i've not/barely played minecraft: | 95 |
| i haven't beaten the ender dragon: | 96 |
| i've beaten the ender dragon: | 104 |
| other (comment): | 4 |
Closed
@eniko Once, with a some friends on a server we had running for a few months. Most of my Minecraft time over the years has been spent just building and exploring though.
Also, maybe this is unusual, but I rarely bother with mods. The vanilla game has been plenty entertaining and most mods or (worse) big 'ol mod packs seem to just pile on needless complexity.
Which is weird because apart from the portal storms I've enjoyed Vintage Story which appears to be forged from weapons-grade complexium.
Two thousand year old #opsec advice:
Hoping that others will not hear is not the same as not speaking; hoping others won’t find out is not the same as not doing.
(from the Book of Han, a history book authored by a brother-sister pair)
#classicalchinese #fountainpen
@0xabad1dea Stop, my freude can only get so schaden.
I'm a full-time professional novelist. Have been for 25 years. Before that I was a software dev. From the inside, the cognitive experiences of writing prose fiction and writing software *feel identical*. The creativity exists outside the words, and most of the phrases and grammar I use are unoriginal.
Ball's back in your court.
like... the cost of a single "state-of-the-art" llm training run would be a permanently institution-changing amount of money for your average computer science department (to say nothing of depts in CS-adjacent fields, especially in the arts and humanities). the claude mythos result, even if you take anthropopo's claims at face value (which: don't) seems indicative of this, like, yeah if you toss nine figures at security research, you're gonna find some exploits?
“it’s just a column in a database” said presumably a full grown adult whose ability to live under capitalism is a column in their bank’s database
If you buy a chainsaw or a bicycle or a power drill, and you don’t spend time learning how to operate it properly, harm may come to you. If you give your child a chainsaw or a bicycle or a power drill without imparting said understanding, great harm may come to them too.
Somehow we’ve decided computers are different and this is a problem we need the government to fix by breaking computers for people who do know how to use them.
it seems someone decided to prove you really can just publish any nonsense protocol draft with the IETF https://www.ietf.org/archive/id/draft-meow-mrrp-00.html