« first day (4691 days earlier)      last day (323 days later) » 
02:00 - 18:0018:00 - 00:00

2:09 AM
@ShadowWizardStrikesBack Ladysir
Or you can just combine them, and say "lasyr".
And when you're too tired to use that many letters, you just use "lazy"-
 
2:34 AM
@Andreasdetestscensorship some days I resemble that remark
 
@JourneymanGeek Not all days? mhm, I beat you!
 
3:18 AM
@Andreasdetestscensorship beating me requires effort
 
3:31 AM
I want to ask this as a question and couldn't find it on any searches but whats the life of a question? ask, gets traction, then like 3 days later nothing. is this normal?
 
3:50 AM
It depends
But sounds about right
Buuuutt now isn't normal times
 
4:06 AM
@BigJoe depends on the site, number of active users, number of new questions, and how many users are really interested in the question
 
 
3 hours later…
6:48 AM
@Andreasdetestscensorship sounds more like an upgraded version of "ladyboy" which better not be used formally... ;(
Not sure if it's considered offensive to anyone, but pretty sure it's vulgar slang.
throwing slang on @M.A.R
@Andreasdetestscensorship lazyboy is missile name I think... let me see...
 
Isn't that a couch thing, lazyboy?
 
oh I was close.
The Lazy Dog (sometimes called a Red Dot Bomb or Yellow Dog Bomb) was a type of small, unguided kinetic projectile used by the U.S. Air Force. It measured about 1.75 inches (44 mm) in length, 0.5 inches (13 mm) in diameter, and weighed about 0.7 ounces (20 g).The weapons were designed to be dropped from an aircraft. They contained no explosive charge but as they fell they would develop significant kinetic energy making them lethal and able to easily penetrate soft cover such as jungle canopy, several inches of sand, or light armor. Lazy Dog munitions were simple and relatively cheap; they could...
:P
 
La-Z-Boy Inc. (pronounced "lazy boy") is an American furniture manufacturer based in Monroe, Michigan, United States, that makes home furniture, including upholstered recliners, sofas, stationary chairs, lift chairs and sleeper sofas. The company employs more than 11,000 people. La-Z-Boy furniture is sold in retail residential outlets in the United States and Canada and is manufactured and distributed under license in other countries including the United Kingdom, Australia, Germany, Indonesia, Italy, Japan, Mexico, New Zealand, Turkey and South Africa. La-Z-Boy holds US and international patents...
 
Don't get those mixed up; your enemy would be able relax, and you'd be very uncomfortable.
 
@Tinkeringbell lol!
@ElementsinSpace but... but... it's so happy.... :D
@GcL heh didn't hear about this one. Might check one day! ;)
@ElementsinSpace more than 1000 now! 1033, still not matching my unread inbox items. ;) (1173)
@ElementsinSpace it does!
 
7:09 AM
@ShadowWizardStrikesBack DON'T remind me of that thing.
 
@MetaAndrewT. Thanks for that info
 
@ElementsinSpace which part? The food or the dancing? ;)
 
@ShadowWizardStrikesBack The aggressive dancing, and the accompanying music I'll never get out of my head.
 
@ElementsinSpace huh? Gif has no music. I think... o.O
Well, there is a long time and very popular request to make gifs animate only on demand, which does make sense. ;(
31
Q: Disable GIF Auto Play in chat

user214961I think that having the user to click the GIF to play it in chat when oneboxed, would make the experience more user-friendly and bandwidth-friendly. Can this be implemented here? I saw it on a few websites (well... 9gag). This would also make all the chat users more tolerant on GIF's and not dis...

Bribe a developer, maybe @balpha and it might be done. :P
 
7:25 AM
@ShadowWizardStrikesBack <https://youtu.be/sbJoaVhjpkk> Warning, do not click this link.
 
Would have started a bounty on it but on a strike lol
@ElementsinSpace that's exactly the way to make me click anything you want. :D :D
(well, copy and go to link in this case lol)
ohh so that's where it came from. Thanks!
Also... might it be that Harry Styles was inspired by it?
 
@ShadowWizardStrikesBack You can't trick me into looking that up; I'm disciplined.
 
7:48 AM
:D
Well I mean only the song name... "Watermelon Sugar" vs. "Peanut Butter Jelly". ;)
@starball do you like dancing bananas?
 
@ShadowWizardStrikesBack I'm neutral.
 
@ShadowWizardStrikesBack to the tune of do you like Pina colada?
 
@starball aka on the fence, so careful not to fall! :D
@JourneymanGeek in general, tune can vary. ;)
 
my head is hot from thinking about the recent posts on MSE from SE.
 
8:11 AM
@starball hot? Try some ice then... ;)
I stopped getting excited/annoyed from any blip of SE though.
Be it bugs, or meaningless posts.
If SE has their way to detect AI generated posts, just give it to mods as well. Problem solved. We don't need all those numbers and shiny marketing presentations they show to stock holders to justify their actions, along with "It's the end of SE" alarms to justify instant one sided action.
 
@ShadowWizardStrikesBack that's a good idea. actually, I've just realized that the actual temperature is a little high where I am right now. that probably hasn't been helping
 
@starball little high.... 40 degrees? (Celsius ;))
 
8:33 AM
@ShadowWizardStrikesBack nah. way lower I think. but pretty low temperatures are too hot for myself.
 
8:43 AM
@starball pity! ;)
 
8:53 AM
@Tinkeringbell Their store near us is shutting down
 
GcL
@ShadowWizardStrikesBack I suspect their heuristics for AI generated posts are data driven and not nearly as good as they report. I would take their numbers as a lower bound given how many syntactically correct wrong answers I've seen on bountied questions.
 
9:07 AM
@SonictheAnonymousHedgehog why?
 
No idea.
It's retail apocalypse v4
 
Well they appear to have enough other locations...
 
 
2 hours later…
11:18 AM
Hello meta friends!
How’s it going?
 
Metatastic
2
 
@EleezatheOtherWorldWizard slowly but surely off the cliff. ;(
 
Metaokay
@EleezatheOtherWorldWizard Have you ever knitted dogs before?
^ that's probably my best one.
 
@ElementsinSpace nice! What about puppies? ;)
 
11:34 AM
 
11:50 AM
@ShadowWizardStrikesBack Oh dear ;-;
@ElementsinSpace This is so cool :0 and adorable!
 
12:10 PM
@ElementsinSpace huh! Awesomeness. :D
@EleezatheOtherWorldWizard yeah, but we have fun here until then. ;)
Like EOWP, End Of (the) World Party. :P
 
12:25 PM
@ShadowWizardStrikesBack Aaaaw, cute ladyboys.
 
@Andreasdetestscensorship I'm sure they are!
 
1:04 PM
Eira May on June 08, 2023
If you’re thinking about rolling out a new tool to your team, you should also be thinking about how to get colleagues and management on board, how to embed that tool in your everyday workflows, and how to assess whether it’s working as it should. Tech that solves human problems needs humans to participate in those solutions.
 
1:16 PM
@Feeds onebox isn't on a strike?
 
@ShadowWizardStrikesBack I mean that could be arranged
 
ohy.
just a question, maybe I am blind
Do anyone know what The Company refers to when mentioning the "HuggingFace GTP Detector"?
Because... well... HuggingFace is a community and not every sample posted there is the same sample. And I could not find any official one by the HuggingFace user (huggingface.co/huggingface).
 
1:32 PM
@JourneymanGeek ban Feeds? ;)
@SPArcheon if you mean Philippe latest question, then he refers to Stack Overflow Inc. or whatever the company is called. Can't see any reason or way he'll refer to any other company.
HuggingFace isn't a company, and he tried to prove how wrong their detector is, that's all.
 
@ShadowWizardStrikesBack You didn't got what I meant.
> HuggingFace’s GPT detector assigns a “threshold” score from 0 to 1 of the post being authored by GPT.
 
To my knowledge there is no official HuggingFace made detector, just detectors that happen to be hosted on HuggingFace.
@Andreasdetestscensorship yep, and that is no HuggingFace. That is OpenAI.
I am not trying to nitpick, but it is kinda important to be precise about WHAT detector was tested.
 
That's what everyone refers to when they say "HF".
 
@Andreasdetestscensorship possible. But that is pretty wrong.
HuggingFace: a community for AI related technologies that also provides hosting.
OpenAi: the ones who made ChatGTP
that is why I was wondering.
 
1:45 PM
We have links to many OpenAI detectors. HF is just a shorthand for the one hosted by HF.
Not sure if I should share a lot in an open chat.
 
Possible, that is why I was asking for clarifications.
 
We've been relying on security through obscurity.
 
They mentioned them as if there was ONE detector MADE by Hugging Face
 
As far as I'm concerned, we've only been using a single detector hosted at HF.
 
@SPArcheon Sure, but we've never mentioned HF in a context where it mattered
 
1:48 PM
Oh well, to be hones I have used multiple samples on HF, but all of them where Stable Diffusion demo
 
Most detectors realistically run OpenAI's code in one way or another
 
@ZoestandswithUkraine not sure. It is like everyone thinking that ChatGTP is the only one demo.
 
@SPArcheon isn't it?!
:D
 
@SPArcheon At least the more active AI hunters are well aware there's more than just ChatGPT.
I haven't been particularly active at hunting it, but I've been keeping up, taking note of the important things. Did flag a few posts.
 
@ShadowWizardStrikesBack Since I'm not that sure that I want to post the name here...
Copied a question from SO
this is not ChatGPT
 
1:55 PM
gpt
 
@PeterKolosov yep, you are right
 
It doesn't matter which model was used to generate the text; what matters is that it was not generated by a human, and that we can recognize it. It's not long-term sustainable; once we get to a point where the AI models still generate garbage, but we can no longer reasonably recognize it, we can no longer enforce a ban on AI content; we will need to have developed other means of protecting us by then.
Not that SE helps us do so. They just ruined everything instead.
 
@SPArcheon how can you tell?
 
@ShadowWizardStrikesBack because I used a tool that is not ChatGPT.
@Andreasdetestscensorship I think you lack some context. If you see above I was asking this because the company mentioned a (single) "HuggingFace GPT Detector" and I was wondering if they had a specific one in mind or if that should be reworded to be more generic
 
@SPArcheon I'll just assume that they mean the one we've been using, as the company did use to cooperate with the company at one point, and they've been a part of the private chat room we used for the purpose of handling AI content. There's a bunch of links there to detectors, one of which is "HF".
There's also a bunch of messages there, saying not to rely on the detectors, because they're inaccurate.
 
2:09 PM
@SPArcheon pretty sure they use it just as example. The post does use "detectors", plural, including "human detector" aka the moderator own common sense, which they don't trust of course. Only senior staff of SE can detect AI content with no false positives.
 
So the company is clearly and blatantly lying when they tell the media, and post on MSE, claiming that we've been relying on the detectors. If you flag AI content with the sole reason that the post gave a high score in a detector, you're likely to have the flag declined.
 
@Andreasdetestscensorship yep, that does indeed clarify how the wide blanket term came to be
 
Or well, speaking of what used to be the case before this strike.
@SPArcheon But we've repeatedly pointed out that we cannot rely on the detectors, and these messages are starred and pinned.
 
@Andreasdetestscensorship were, bet the room is long nuked to oblivion by the staff.
 
Or, well, the messages aren't pinned now, because we replaced the pinned messages with stuff more related to the strike.
Anyway. I think that's enough talk about a private room in the public.
 
2:13 PM
Can't see any harm talking about private room, mods talk a lot about TL for example. Just can't share contents from a private room.
But yeah, the whole subject can be very simply concluded, SE management does not trust any detector, and insist on banning them all, including mods own judgement.
 
Just Irminsul it.
 
That's a lot of Nicks.
 
2:35 PM
Sign of strike #158: off topic lost soul questions stay for hours. :D
 
> So I googled “how to post on stackoverflow with gpt”, and looked for relevant results dating from after the initial rush in December 2022. The first relevant result starting from February 2023, on page 2 of the results, is a Chrome extension, released on 28 February 2023. Going by the demo video, this extension posts the answer a few characters at a time, at the rate a human would type. Someone using this extension would generate a similar number of drafts as a human.
 
Random thought... @mousetail do you have any checks for suspended users? e.g. a user with network wide suspension wants to sign the open letter, can they? Guess it won't do any harm to let it, but it's quite pointless. No?
@SPArcheon you formatted this as a quote, where is it from? Context?
 
108
A: GPT on the platform: Data, actions, and outcomes

Gilles 'SO- stop being evil'Thank you for posting the methodology! That's a step forward. Unfortunately, there's a critical flaw in your methodology: you assume that In principle, if people are copying and pasting answers out of services like GPT, then they won’t save as many drafts as people who write answers within Stack...

Gilles actually providing solid proof of what I had guessed - the expected weapon race with generated content mean that the users trying to bypass check would try to remove the blatant tell-by, with the usual "too fast for an human" posting speed being the most obvious one.
as for why this is relevant, see the post itself.
Right now I am mostly baffled that a browser extension exists
 
2:52 PM
@SPArcheon also, many use external editors to formulate their answers, then just copy and paste from there. And SE now counts them as AI generated answers. lol
SE editor isn't ideal, to say the least.
And.... they revealed their method so super easy to bypass it for anyone who want to get easy rep, even without extension.
 
Wait, just noticed this. It gets even better:
2
Q: Chrome Extension with Chat GPT-3.5 - "you must provide a model parameter"

Russell HertelI am making a chrome extension that uses Chat GPT 3.5 and have coded a simple prompt to send to the API using openai api and returns a value in the console. I have my code below and keep getting this error... error: code: null message: "you must provide a model parameter" param: null...

Someone asked for help to make a Chrome extension to post generated content to Stack ON STACK
 
@SPArcheon and?
 
@ShadowWizardStrikesBack and... nothing, it is just funny. In a weird way it seem like going to ask how to crack a game to the devs of said game.
Not the same actually, but funny enough.
 
Not so different from calling for a strike on the platform you're striking against, just from the opposite side, in a way. ;)
 
@ShadowWizardStrikesBack Correct, but in that case the tag was removed.
Alas, that not the point, I just find it quite ironic.
and a tad hypocrite.
 
3:01 PM
@SPArcheon and the SO question isn't featured either...
 
If you think that the content you are going to post using that extension is worth posting, just ask GPT for you answer instead of asking the inferior humans at SO :P
But on second though, maybe they did want to use that as a base for starting writing an actual answer.
 
3:30 PM
TTGH
@Journeyman your task is to decipher this ^
:D
Out of unicoins to give as prize but can give something else. ;) (gave all my 100 unicoins to @Eleeza for her drawings ;-))
 
Thank THANATOS GOOD HUMMUS
 
Time to grow horseradish?
 
@JourneymanGeek half way there!
 
Time to go home?
 
3:33 PM
yes!!! dang you're good. :P
I'm really bad at those things lol
 
@JourneymanGeek Hey! What does my switching from a pony avatar to a radish one have to do with this?
 
I was gonna say "Table Top Goat Herding" But I'm not a collie
 
Baffled every time you're using something I'm not familiar with, lol.
 
@SPArcheon Many many things. Some of them Wasabe, some willsabe
 
GcL
3:55 PM
@ShadowWizardStrikesBack It's not vim, but it is easier to exit.
 
At least all of Vim's shortcuts work :p
 
4:14 PM
asked for fun to starchat if "It is safe to parse HTML using RegEX?". It said that yes, RegEX is a good tool to parse HTML.
something something hE C0m€s
 
@SPArcheon You can actually do it in recursive PCRE
 
4408
A: RegEx match open tags except XHTML self-contained tags

bobinceYou can't parse [X]HTML with regex. Because HTML can't be parsed by regex. Regex is not a tool that can be used to correctly parse HTML. As I have answered in HTML-and-regex questions here so many times before, the use of regex will not allow you to consume HTML. Regular expressions are a tool th...

 
It's actually possible though
 
@mousetail I was trying to make it output some part of this, which would have proved that the training set included SO content :P
 
@SPArcheon You can do it as mousetail mentioned. It just depends on your actual goal. Do you want a full parser or just get a few links?: unix.stackexchange.com/questions/388991/…
and I'm aware what you were trying to do, basically making the model output part of that one popular question: stackoverflow.com/questions/1732348/… ?
there also a good reverse version of that post that I liked: stackoverflow.com/questions/4231382/…
@mousetail there a few libraries/tools that use regex to parse html, eg: html2text (the original py2 version) did
@ShadowWizardStrikesBack wait, so you're telling me I can't ever use my favorite editor to craft an answer and then copy paste it as answer on SO/SE?
I always did that for some answers :/
 
GcL
4:32 PM
@SPArcheon I used to work with a front end dev who pointed out the regex for valid html is .* because browsers do their darndest to render something and keep on chugging. It was a tongue in cheek remark, but also a funny way of remembering that browsers will happily consume garbage.
 
@GcL See: Chuck Norris is a valid color.
 
GcL
@NordTheStarWizard It's how I roll. Vim to hack together some markdown then paste into a question or answer. I still usually end up with some edits for longer answers. Not so much for shorter ones.
 
yep, same here
 
Bit misleading, any string is a valid color
 
GcL
@SPArcheon Ha! The explanation was an entertaining and brief read.
 
4:35 PM
although I tend to hop between editors these days (now stuck between Geany/Pycharm)
@mousetail any string made of 3 byte or so, so ffffff?
 
1 byte is also valid I think
 
really
I guess it could work if you count ff0000 as a color hmm
 
JUSTIN BAILEY
------ ------
If you know you know.
 
@SPArcheon metroid :o
 
@NordTheStarWizard more specifically, something that is a funny coincidence despite not being deliberate nor the only words that produce that result (only NARPASSWORD bypasses the password normal logic, everything else is a coincidence)
 
4:54 PM
IF i submitted a bug and it had status-review attached to it but has been fixed now and there is no status-complete on it is there something i should do ? Question - meta.stackexchange.com/questions/389879/…
 
@Sonic What happened to your "Den"? Why it is inactive now?

Sonic's Den

If you have time to worry, then run!
 
@PeterKolosov I stopped posting things there?
 
@SonictheAnonymousHedgehog Yes
 
@PeterKolosov Wasn't safe anymore after Amy found out about it.
 
Err... I do not know much... I registered here only about a month ago so I know so few things
Don't know about Amy, or are this a joke
@Sonic Or you just don't have time to chat as ShadowWizard have
 
5:03 PM
@PeterKolosov Yeah, my personal circumstances changed
 
@Sonic Huh. Okay, I'm not disappointed. ;)
 
Hey, did you ever wonder how 'git' handles symlinks?
 
@Son Hmmmmmmmm.... Confused...
 
@Exampleperson not really
 
5:08 PM
Symlinks cannot be opened like normal files, but then how does it store the symlink?
 
@GcL sure, I'm using it fine, but still, many are probably used to better editors, so using what they know and copy paste when it's ready.
 
see here, this is how they explain it. But I am still in doubt, because I have tried to use even very low-level functions to open them and view their contents, but they seem like a special file whose contents cannot, and absolutely cannot be viewed. The answer in that link isn't good enough (but the number of upvotes though!)
 
@NordTheStarWizard sure you can, but it will be marked as AI generated by SE internal tracking system. No worry since SE praise AI, you might get a badge for that at some point. :D
 
that's lose-lose for me since I would hate that badge (given my stance on "AI")
 
@SonictheAnonymousHedgehog yeah, the fate of most rooms. /cc @Pet
My Den is the oldest user made room, there was room by Pekka and someone else I forgot about, which lasted longer, think they abandoned it 2-3 years ago.
 
GcL
5:14 PM
@Exampleperson No. We avoid them because of the associated problems that come with committing symlinks.
 
@Tinkeringbell can you mark meta.stackexchange.com/questions/389879/… as status-completed everything is working now
 
Hmm... what do you think happens when you commit a symlink?
@GcL any explanation about the issues? (I am curious)
 
@NordTheStarWizard pity! ;)
@BigJoe but might be temporary fix, maybe wait a while to see if it sticks?
Maybe like electricity, when there's blackout, it sometimes return fast, just to fall down again for longer time.
 
GcL
@Exampleperson Doesn't play nice with collaborators developing on windows machines. Doesn't play well if the path to thing doesn't exist on the machine pulling the commits. Doesn't work well if some collaborators have the save links as text files and some do not. Requires extra care when restructuring a project.
A topic of conversation I'd expect you'd get more out of in an SO room. Or at least a more comprehensive conversation.
 
@GcL I understand that it would be better on SO, but, here goes: if you do know, can you tell me, does git treat symlinks as normal files? (I am not gonna ask anymore here, just this one pls)
 
5:20 PM
@Exampleperson tried asking GPT? ;)
 
Should I? lol. It would waste my time, for such a simple Q.
 
@Exampleperson if simple, why you ask? :D
 
GcL
@Exampleperson Depends on your settings. I don't know the internals of git very well to say how it is handling the data. If you're a solo developer with the default settings and you commit a symlink, you'll get the symlink back when you clone or pull. Symlinks aren't regular files to begin with though.
 
Well dunno, Google gives back 12 years old question from SO. lol
 
14 yrs old infact that question is
 
5:22 PM
@Exampleperson nope
40
Q: git commit symlink as a regular file

hasenSuppose I have a file fname which is a symlink to a file from some other repository/project, say ../../proj2/fname. Is there a way to add/commit fname as a regular file? It seems that, by default, git gives the file mode 120000 and sets the path to the linked file as the blob content. I know t...

That was first result, didn't bother to check others. ;)
 
I was talking about this
 
@Exampleperson well that wasn't first Google result for me, probably somewhere down. Google show different results to each user, based on trillion things probably.
Maybe you viewed it before, or something Google marked as related, so they gave it bigger weight.
 
The answers have thousand votes and stuff, would google actually consider those things?
 
no, not usually
 
That question I mentioned seems to have answers that imply that git treats them as regular files
 
5:25 PM
(Google knows exactly what questions each user without adblock ever visited)
@Exampleperson nah they don't care about it, or quality of the posts.
 
@ShadowWizardStrikesBack You mean, through the ads on SO?
 
@Exampleperson no, via trackers put in place by SE in the core of the sites.
to get "stats", ya know. lol
 
Trackers? How do those get sent to google?
 
@Exampleperson JS, sometimes hidden deep sometimes plainly on the surface.
 
How do adblockers block them if they are not ads (or from ad related sites?)
 
5:27 PM
Or tracking beacon used as image.
(empty image, not displayed of course just loaded to send the data)
@Exampleperson it blocks any request to the blacklisted domains, or to all domains except those in whilelist.
It can't really know if something is ad or legit content of a site.
 
@ShadowWizardStrikesBack, I never realized there is a comment archive in meta and SO meta, what purpose does it serve? How on earth does it fetch new comments? (pls tell me)
 
@Exampleperson yeah well it's pretty new, couple of weeks perhaps. How it works @Zoe can tell, probably using the API.
 
Ohh... That's why I never saw it before. Nice.
 
As for purpose, many comments are deleted, sometimes people claim they posted something they didn't, or just forget. This way, we can know for sure what was going on.
 
What if personal info is posted (by accident ofcourse)? then what?
 
5:31 PM
The comments in chat stay forever, unless of course mod removes them.
@Exampleperson should be flagged, and mod will delete.
Probably same for NSFW stuff etc.
 
What if they don't know about that chat? It is very common for new users to make such mistakes
 
@Exampleperson it's not for new users, it's for "us", e.g. when user claim comment was deleted, we can look it up and help them.
So it's a tool for curators, in a way.
 
I mean, when a user posts sensitive info, and doesn't know about that chat, what will happen to the sensitive info in the chat?
 
@Exampleperson well those who view it should flag, the user indeed can't know about it.
 
Is there one for SO? or only for metas?
 
5:35 PM
But you raise a good point, that's probably a downside of this tool. (leak of stuff that shouldn't leak)
@Exampleperson dunno, probably on-demand, not on all sites.
 
@ShadowWizardStrikesBack, I wonder how much data can SE servers hold
 
@Exampleperson a lot. ;)
 
@ShadowWizardStrikesBack Thanks for understanding. What happened to the strike now? Still going on?
 
And it's easy to just bring another HD, or server, when they need.
 
And google on the other hand has infinite storage
 
5:39 PM
@Exampleperson yeah, there's Discord group you can join to get real time updates.
Should be mentioned in the open letter.
@Exampleperson no such thing infinite, but yeah they surely make it look this way.
 
:P 1/0
 
I read once that they have hundreds of employees whose single job is to replace failing drives and install new drives.
In huge data centers.
Every bit of data has two backups, etc.
 
They give about 20 GB of free space for every user's google drive. How can they handle 20 * 1 billion users, atleast
 
Might look for it again, was interesting to read. ;)
 
@Exampleperson compression ;)
 
5:42 PM
Maybe they have internal techniques for using 'symlinks' to the same file if two users upload the same file?
 
@Exampleperson I don't think so
They probably use RAIDs
 
Of course that could, in theory, be a privacy risk, but they can significantly minimize storage for files containing the same data
 
GcL
@Exampleperson If you're looking for strategies that deduplicate data even between different files, check out IPFS.
 
I am not running any servers lol
@PeterKolosov, compression? To such great extent?
Wow.
I can't even imagine how they do these things
 
@Exampleperson lossy compression ;)
 
5:47 PM
@ShadowWizardStrikesBack It's using the API, yeah
 
Hmm...? How, wouldn't that cause data loss?
 
For the more technically inclined, the core code for it is fairly straight-forward: github.com/LunarWatcher/boson-light.cpp/blob/…
 
@ZoestandswithUkraine, what do you think about the privacy issue from the chat that archives comments?
 
Considering the license of all content on the platform, I don't really see the problem. It also doesn't leave the platform, but there are definitely scrapers that do precisely that
 
5:50 PM
@ZoestandswithUkraine hmm...
"but there are definitely scrapers that do precisely that" - think about it.
 
In general, don't post stuff to the internet that you wouldn't want to be picked up elsewhere
Because on any sufficiently popular platform, it can and will be picked up
 
@ZoestandswithUkraine Don't accidents happen? (especially to new users)
 
@ZoestandswithUkraine figures, you won't go and scrape all posts to read comments. ;)
 
@ShadowWizardStrikesBack wow. Nice logic.
 
@Exampleperson sure, good thing we have exception handlers.
 
5:51 PM
^
If it happens on the platform, a mod can just remove it
 
Well, what if the user deletes the comment thinking it is gone? It is extremely common from my experience.
 
But I guarantee you there are currently AI projects scraping the API for comments to dump into their dataset
 
then it'll stay in the archive until someone flags it.
 
@ZoestandswithUkraine, there should be some kind of barrier to avoid privacy issues, don't you think?
 
5:53 PM
@Exampleperson haha, reminds me of "accidents" happening to my kids when they were younger. :D
 
@ZoestandswithUkraine I respect your opinion
 
@ZoestandswithUkraine sadly so.
 
@Exampleperson like pasting the whole shrek script into the comment textfield?
 
like I said, when it's posted on the internet, it's going to stay on the internet
Stuff almost always finds a way to anyway
Any deletions or redactions following that are damage control at best
 
5:55 PM
That's a sad truth many don't learn right away.
 
@Exampleperson or ever.
 
@ShadowWizardStrikesBack or just don't care
 
@ShadowWizardStrikesBack yeah...
 
There is no way we can teach them though. That's not our responsibility either. Google can teach if they want to. (but I can't tell that for sure, you know)
 
You can still teach peopleif it's done in the right context
But indeed, not our problem
it's ultimately up to education systems and, at least for kids, parents to at least reduce the rate
 
5:58 PM
@RocketNikita depends, I don't care to share some private info, but got a limit.
My real name is easy to find, think I even posted address once, somewhere.
Still waiting for someone to arrive saying "it's you from Stack Overflow!". :D
 
@ShadowWizardStrikesBack Still remember my fake hand?
 
@RocketNikita can't get it out of my memories! :D
 
@ShadowWizardStrikesBack (pls dont post it)
 
Such a hand...
 
02:00 - 18:0018:00 - 00:00

« first day (4691 days earlier)      last day (323 days later) »