#ot1-perplexing-regexing
1 messages · Page 82 of 1
lol what
If you choose to be a given sexual orientation, it only means you have a predisposition for the opposite but you condition yourself to be a certain way.
Otherwise the statement is meaningless
My band's first single! (excuse the keyboard solo, i'm a keyboard virgin 🙂)
https://absentmindedband.bandcamp.com/track/absent-minded
hehe
old school and indie vibes at the same time lol
btw the stereo is not setup correctly
I get more volume from my left than my right ear
I reckon it's a bit more complicated, as it involves deep rooted emotions and complex social contexts.
Nice
Gg bro
keep going u got stuff
Yeah people change over time, I just think it's good for people to accept and love themselves instead of negating harmless feelings.
The notion that orientation can be changed and is a choice has been used to justify some pretty fucking awful things.
Orientation doesn't change.
Your understanding of it may.
Your choices may.
You may have shifts as to what you're into at any given point in time, but I submit that isn't a shift in orientation, but a shift abiding by it.
With that in mind, if you feel you do have a choice of orientation, you may have some self reflection to do.
The theory of evolution was used to justify a lot of the Nazi attrocities, doesn't mean that the theory of evolution is wrong, just means that people with no morals perpetrated evil
Which can be a shitty feeling, and for that feeling I am genuinely sorry and sympathetic.
Oh, my, no. I stand by what I said wholeheartedly.
Sexual orientation, preferences, gender roles, all of it can and does change because people are not static rocks, they are dynamic complex systems that adapt to new information and situations.
Just pick a person you like, its not that hard
Whats with all this orientation stuff
Thats just greedy
And greed is a sin
Mormons havent been involved in any wars
I've a poly friend, but he isn't a Mormon.
Not unless he's been exceptionally sneaky about it.
Why did you emoji that message
Have mormons been in wars
Oh
Kinda disappointed ngl
I agree that self reflection is good. You come to understand who you were, who you are, and the change that happened between.
Hi guys, I need some advice. I have a script running that generates a ton of data (over 60000 rows in pandas). I update this data every hour (this means I overwrite the old data completely). Right now I am doing everything in pandas. I want to setup an api where you will be able to request some of the data based on your request.
Does it make sense to convert this pandas DF to sql for the requests, or should I just keep it as a dataframe and call the data using pandas every time someone requests something.
If one choice is better than the other, can you explain why. Is it in terms of speed or accesability or smth else?
If you already have data in the one form and the API can access that data easily, leave it. Don't migrate the data unless you need to for accessibility.
but as soom as i promote json to store data im the bad guy
If you say so. I've never called you that.
i never said you did that
When you say "covert this pandas DF to sql" what do you mean by that?
But it's still in memory? I've never used pandas.
so the pandas dataframe is in memory
it takes up 3 gb of memory
when I call to_sql
it outputs as an sql script
write it to an sql database
So that you can populate a database?
DataFrame.to_sql(name, con, *, schema=None, if_exists='fail', index=True, index_label=None, chunksize=None, dtype=None, method=None)```
Write records stored in a DataFrame to a SQL database.
Databases supported by SQLAlchemy [[1]](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.to_sql.html#r689dfd12abe5-1) are supported. Tables can be newly created, appended to, or overwritten.
Okay. If you are hurting for memory then drop that into a database. Easy.
If it's running already and you are shrugging at cost of memory... use it as is. (right back to my first comment)
that seems to dump data in the db for you
oh nevermind
no sql script necessary
yeah i thought the same
but there are no performance benefits?
like I have a script that can quickly filter out a lot of entries in pandas
would this be faster/slower as a db
Memory will be faster read/write than the IO of a disk, always.
but I feel like if I would store it in a db it would take up less space
than I am using rn in memory
it would use less memory
60000 entries shouldnt be 3 gb in a db
It will take up just as much space (maybe more). But not your memory anymore.
that is strange though
All depends on what is in an entry.
14 columns, 3 of them are datetime objects, the others are floats and strings
and all strings contain less than 100 chars
Okay, I'm with psv. 3gb is outragious for that little data.
well I might be tweaking, because I just checked task manager
is there a way in pycharm
to check how much memory a dataframe takes up in debug mode?
DataFrame.memory_usage(index=True, deep=False)```
Return the memory usage of each column in bytes.
The memory usage can optionally include the contribution of the index and elements of object dtype.
This value is displayed in DataFrame.info by default. This can be suppressed by setting `pandas.options.display.memory_usage` to False.
might wanna use deep=True if you have objects
For 60k rows like that, I think you could store it any way you want. sqlite, .json file, in a dataframe, in an list of dicts, whatever. Are you not processing it at all? Just returning the data set?
so if you .sum() what does the total come up to
I am processing it
but I need the raw data too
so people can process it their own way in an api request
50682836
Oh it should be only 50 mb
I'd think your real question is: which path would allow you to easily build a query API against that set.
well my gut says pandas, but I would want to store the old data somewhere, to make a regression model of it over time
idk if storing that in csv would be optimal
Oh, well you never mentioned that! haha
no because my primary goal was the api
the model is something i would like to do at a later time
My approach to all problems is: 1. log to a file first, so I can do the inevitable database rebuild. In this case, write .json, .csv or .parquet files for each load.
- Build a current table - current state / most recent. Don't try to live query against a history table (if you end up with one), because it's inevitably grows too large
- Keep a second history table, with the changes for each event..
- Compute some historical metrics, so you're not querying to ginormous raw data for each request.
something like that
For streaming data, things get a little more hairy, since you don't have discrete chunks of data
in what scenario would I want to stream it?
You described it as: you'd poll every hour and pull the entire dataset each hour. But, let's say the events were pushed to you... so at any time, you might receive an event that updates the current state. Then, you'd need to log that in the history table + apply it to the latest table. And, write to a file log so you can recreate the db.
not sure who needs to hear this but if you have WSL and Docker installed, there's a high chance WSL is running on boot and taking up like 4gb of RAM, even if Docker is set to not start
i couldnt find a good solution so i just uninstalled docker lol. wasnt using it anyway
my laptop would start helicopter mode if using more than 50% of the ram
why shouldn't n % 0 just be defined as n
what is the remainder of n / 0
by long division it's n
the quotient may not be defined which causes a problem if modulo relies on that number
but if modulo relies on the thing that if it appears twice, it's the remainder, it makes sense
or like https://math.stackexchange.com/questions/516251/why-is-n-mod-0-undefined
A typical first way of thinking about mods is to say that
a ≡ b (mod d)ifa = b + dkfor some integerk. In this sense, there is nothing wrong with sayinga ≡ b (mod 0), although this says nothing more thana = b.
why not that
A few rules that usually hold with modulo no longer do, then.
For example: for all x mod n, the result will be less than n.
It seems safer, in a programming language, to have the operation crash in this case than to potentially do something surprising.
it's rare you would have to mod by 0 for some reason
chatgpt debate of the ages
ages is quite an exaggeration
wasn't srs
debate is quite an exaggeration, you're arguing with something I didn't say
you responded to me respondging to somebody else that said it always makes mistakes.
A rule that applies to programming in general, and especially anything working with AI: there difference from 99.99% to 100% is a lot bigger than from 99% to 99.99%
so I continued my same line of reasoning that that makes no sense
to which you said something like, "yeah, but it's still unreliable"
👀
So you're being needlessly pedantic about it
to which my response is "so what?"
"needlessly pedantic"? Somewhat redundant, I think.
then why are you responding to me?
of course I'm gonna keep responding to the things you say
"it always makes mistakes" is commonly understood to mean "mistakes are frequent enough to be a significant problem", not "literally all of its output is wrong"
EDIT: OMG ELON MUSK TWEETED THIS VIDEO. I spent A LOT of time in undergrad and have tons of loans. I would basically procrastinate more than you can imagine and make stupid shit like this. If you can can you donate to me here: https://www.paypal.com/paypalme/trapped21 so I can pay off my loans
Argument by Monty Python with some scenes cut
shame about Cleese
What? He died?
sure, that clarification is fine. but the claim that it will introduce mistakes into any project that uses it is similarly absurd
Not that I am aware of
u killed him just now
he just turned out to be a bit of a git in his old age
? What'd he do?
not to mention that they responded saying they literally never had it produce correct output lol
which is why i asked for a simple example that could be tested. to which you butted in asking why it has to be simple
chatgpt is sometimes very cool i've got to say
went off on a rant defending Rowling and insisting she's not transphobic
I did not ask that, someone else did
Yes, the person I was responding to before you butted in
it's not a private conversation
1 or 2 days ago it taught me that < or > can use lt or gt even if the other one isn't implemented
eg < can resort to gt is false
you're right. you could have read those messages to understand why the example should be simple not complex
Well, I'm not the one who asked why it had to be simple
once again you're arguing with something I didn't say
ah, thought that was you
Maybe check your facts before going off on me
"get owned liberal"
jesus. I'm done.
alright i'm sorry maam i'll check the facts
not a sir
Might I suggest that this conversation has run its course? Seems like its getting personal.
that was unwarranted and frankly childish
I'm gonna go get a breakfast burrito
burrito 👍🏼
now that i can get behind.
i thought "check your facts" was a little unwarranted considering I admitted I was mistaken
burrito 👍🏼
🌯
there are so many emoji
Speaking of, I had a very unsatisfying Gyro yesterday. They wrapped it in a tortilla instead of a pita. Just was all wrong.
Gyro is so messy 😦
condolences
wouldn't that be a taco
taco arabe
I'm still upset that soft tacos are a thing. Tacos should be crunchy.
but that uses a type of bread more than a tortilla
Do what now
incorrect
Corn tortillas are amazing
tacos should be on soft corn tortillas made from masa
what in the taco bell
looks crunchy
i don't like this taco bell design cause your bite is either all meat or all salad
Well, you only get one bite. The rest falls apart
exactly, soft shell supremacy
if you really want, wrap a soft shell around the hard shell
how does a soft shell prevent this
it doesn't fall apart
hard shells crack and literally crumble
pita
thats my territory
in case we talk about the same thing
not pita
close enough 🙂
xD
🥙
Yah, but I need to drown it in hummus or sauce, so dry usually
hummus!
falafel is fire but i ate too much one time and now i don't want it anymore 😦
new england
:O
Good friend is Morrocan... cooks a lot of middle eastern food.
u need to visit israel one time
I have, in fact.
or i guess some arab states
u've had some good hummus?
Oh yah, and shawarma
xD thats neat
i'm almost vegan so i dont eat shawarma anymore
bbut shawarma is great
what does being an "almost vegan" mean?
well i eat honey
i dont eat eggs fish meat milk
I really like that garlic sauce. Toum, I think it's called.
no clue what that is
oh, honey counts as animal product
.wiki toum
Toum
Salsat toum or toumya (Arabic pronunciation of تُومْ 'garlic') is a garlic sauce common to the Levant. Similar to the Provençal aioli, there are many
Garlic sauce
refers to a tomato and garlic sauce, which is used on pizza, pasta and meats. Toum is a thick garlic sauce common to the Levant. It contains crushed garlic
gg
was she happy tho
she seemed really nice
i think it went pretty well
i texted her and said i had a good time and she said “that’s good me too!”
nice! planning the next one?
yep!
cool! hope it works out for you
thanks! i’m just happy i wasn’t catfished 😭
And she has Romanee Conti
i was worried i’d be catfished
but it’s hard to be catfished if they’re verified on tinder
so
The cheapest Romanee Conti is 5K dollars
Jesus
120K dollars
Why would anyone pay for that
Get this for your next date @sterile sapphire
i would rather shoot myself in the foot
"Hi are you willing to move out of the property in Freeport? I'm searching for a new place nearby and would love to discuss it. Up for a quick call? Ken from S&H Homes
Reply STOP to unsubscribe"
wtf
i don't live in freeport
lol
Hello! Can anyone tell me what is going on with my monitor screen? This "bleeding" is getting worse.
are you using VGA
yes
k, try it in a minute
seems to be it. reconnected my cable and tightened it. It got much better, but there are still some bleeding. Must be faulty monitor vga port, since everything works just fine on my other monitor with same cables
yeah i used to have this same issue with my old monitors
its just playing with the cable until it's manageable
long term solution is to just another monitor or cable tbh
Is there a judgement in this question?:) You are right tho. I have a video card with hdmi and dvi outputs, but still using an old 22" full HD monitor💀
no
I was just wondering
I havent seen VGA monitors a long time
is it a CRT monitor? @random ferry
i still have a CRT monitor
Some games are better on CRT.
Nah, a Samsung syncmaster 2253lw @civic pasture
I keep my CRT cuz nostalgia @fresh basalt
The old ones, yeah
Someone I knew used to have a huge widescreen CRT back when they first came out. Beast of a monitor.
this is the one I have
took up the entire top of a dresser
Damn, my grandpa had this guy
It had analog inputs tho, not vga
that's not the trinitron is it?
IT IS
That thing is a beast
I swear it's like 100lbs
Used to play Playstation 1 on this
don't sit too close - if it fell on you RIP
@random ferry 28"?
Wow they are so expensive on eBay.
Yep, thats the one
Bruh, I dont know. I once found an old ussr penny in the forest and sold it on auction for $40. Some people just like to collect old and useless stuff.
It is interesting that I remember the picture on these old TV's were considered very crisp. Though the resolution was something like 360p or something like that
Cant find the exact number of rgb pairs on trinitron screens, but refreshrate is surprisingly high - 100 hz
Well, there it is:
480p
Seemed pretty crisp back in the day
I can't be in the same building as a normal CRT
I can hear them being turned on through two floors and across the building
It's the only high pitched noise I can still hear with my hearing damage
And it drives me INSANE
Any bleach fans?
Are you taking a survey?
Yeah like wth
I quit last year
I don’t consume anything
Occasional weed
For anxiety
Prescribed by doctor
Edible form tho
I wouldn’t
Those are unhealthy
They wouldn’t help with productivity
Yes
Nicotine is no good for you
He is probably getting paid to advertise
How do you think he gets rich
Is he a some sort of YouTuber
Those influencers
They are idiots
Don’t get me wrong
They have 0 knowledge
Making money out of advertising
Or some other scheme
It can cause an increase in blood pressure, heart rate, flow of blood to the heart and a narrowing of the arteries (vessels that carry blood). Nicotine may also contribute to the hardening of the arterial walls, which in turn, may lead to a heart attack.
I mean, it takes some amount of knowledge to be able to make it as an influencer
I think the hate on them can get kind of irrational
I mean he is promoting nicotine @low chasm
Someone with that kind of influence shouldn’t make such statements tho
ah, you weren't speaking in general? In that case, sure, it's pretty bad.
Before researching
What state do you live? @heavy trench
I use marijuana for improved focus
And anxiety
In edible form
What does?
how about just not taking anything
Ciggarettes for kids 🗣️🗣️🗣️🔥🔥🔥💯
fr tho, cigerattes used to be promoted as safe and even healthy
vapes are viewed as not as bad as cigarettes or not too dangerous evem though they are
So did a lot of unhealthy things.
and why are vapes promoted with flavors??? Cotton candy, sour apple
Either due to inadequate science or misinformation or influence from lobbyists.
Or lack of regulation.
cuz the KIDS buy them. Fr they just gather in school bathrooks and vape between periods
You just have to look back and think "at least we're not like that anymore."
We still are
vapes are relatively new so we dont know the full impacts they can have on humans
And they are way too accessible for teens
6th graders do it
not to mention the use of KENT here
Shockingly, as an adult, I enjoy flavored products.
I don't think vapes should be accessible for teens, but cigarettes were too.
Cigarette smoking is down; and while we don't have conclusive evidence on the long term effects of chronic e-cigarette use, the tangible evidence for short-term effects of vaping for the average person health-wise are significantly less impactful than smoking cigarettes.
While I'm not going to advocate for vaping as a 'safe alternative to cigarettes', medical studies have shown that vaping is significantly less impactful on your health, so if you're already smoking, then vaping may impose less known and standing health risks.
This is something echoed by John Hopkins which concludes that “there’s almost no doubt that vaping exposes you to fewer toxic chemicals than smoking traditional cigarettes.” Furthermore, Hopkins goes on to postulate the cause for significant lung injuries is largely associated with "people who modify their vaping devices or use black market modified e-liquids."
This touches on a factor that I believe is glossed over far too frequently, which is that we can regulate the chemicals present in e-cigarettes rather than outright banning them. This is a major movement of the ENDS regulation via the FDA, which attempts to control the sale, manufacturing, and modification of e-cigarette liquids.
Given that we still have not abolished the existence of cigarettes, over-control of vaping may have negative health effects if continued research goes on to show that vaping is actually a healthier alternative to consuming cigarettes.
Many of the health effects associated with e-cigarettes are instead associated with the consumption of nicotine. The granular control of nicotine content in the substances your consuming is an appealing portion of vaping, and often, these studies do not target what I believe to be the 'average consumer', where an individual may be using 3mg juice but the subject of a study may be correlating the amount of nicotine with 'the average cigarette', which is substantially more.
While I don't think advocating for chemical dependancy is a good thing, anecdotally I've noted significant improvements to my personal health by quitting smoking and switching to vaping instead.
</soapbox>

what do you think about marijuana in edible form
I don't have opinions; I don't use marijuana. I've considered THC for anxiety control. It has implications to my hirability that prevent me from using it. If it becomes federally legal, that discussion may change.
thank you very much
and thank you for joining our podcast sir
Next week we will discuss global warming
time for sleep
thanks discord now i can voluntarily lag my whole laptop
since when is promoting nicotine a bad thing
1 drag off a cigarette takes 5 minutes off of your total lifespan - it simply can't get any better than that
Only 5?
Sadly yes
There are strains of weed that make anxiety go away. That’s why I’m on medical marijuana
Anxiety reduction via cannabis is a subjective experience.
.8ball will teams web version work today?
Yes definitely
.... some doubts about that

LinusTechTips did a WAN show awhile back remarking that Discord would be a reasonably solid contender to enter the enterprise communication space considering a lot of the functionality is already built in.
I'm inclined to agree, despite my woes with Discord.
GOD the way I fucking wiiiiish other apps implemented the client-side per-person volume controls
Discord multi-person voice/video group chats are honestly better than anything enterprise out there
One person too quiet, one person too loud, and there's no happy middle ground to set it locally.
Yeah... That reminds me of my Teams meetings at my last job.
And the Discord Stage feature is just the best execution of any 'host'/'audience' kind of call I've seen yet.
Especially since they added video/screen sharing
Kind of nutty that Discord is so widely adopted yet none of these other orgs have gone "Yknow what, that's actually a pretty good idea," and ran with it.
!e @gaunt warren replying here to not clog up #internals-and-peps, after some digging it looks like gc does the job, since obviously to prevent the thing being iterated from being garbage collected the iterator has to store a reference to it.
import gc
print(gc.get_referents(iter([1,2,3])))
print(gc.get_referents(iter(lambda:1,0)))```
@outer sundial :white_check_mark: Your 3.12 eval job has completed with return code 0.
001 | [[1, 2, 3]]
002 | [<function <lambda> at 0x7f296c496a20>, 0]
(note that as levithan said, I could have a class return a stupid iterator as the result of __iter__, so this is only guaranteed in the stop iteration case, and with builtin classes.)
!e py import gc class A: def __iter__(self): return iter([1]) print(gc.get_referents(iter(A())))
Notice how there is no reference to class A
@outer sundial :white_check_mark: Your 3.12 eval job has completed with return code 0.
[[1]]
The correct thing to say is it gives the contents of the iterator, not what actually constructed them.
Yeah, and there's no one universal iterator, it's just anything that implements __iter__ and __next__. I'm only concerned with the two builtin iterators returned from iter() in this case.
for what it's worth I think this would have been fine to keep in #internals-and-peps, it's what that channel is about. Though I guess you're fast approaching #esoteric-python 🙂
What? Python in an offtopic channel? I say...
That does look great
What's the white stuff? Sour cream?
potatoes, onion, beans and sour cream or creme fraiche?
tahini
back home we have this pastry stuffed with tahini mixed with honey and brown sugar
its incredible when fresh
I think I heard the name but I've never had it.
oh, sesame
I love sesame
Those beans look good too.
love beans
Its the main ingredient in hummus
Why are you pinging random people
in multiple channels no less
Dont have time to read fully, but conceptually looks great... and this is a great way to learn (at a deeper level)
what's that cream?
omg
i'm from israel, and in israel tahina and hummus are quite popular
i think u should really try either of them!
hmm is it like a little bit salty?
yea
do you mix it withhttps://mydearkitcheninhelsinki.com/2015/11/25/tahini-molasses-tahin-pekmez-turkish-dessert-spread-this-is-sweet/ @dusky turret this is how we eat it
with molasses
Its sesame paste
idk then
nope
but i've had some sort of middle easter cream that looks like that
yea it's like between a cream to a paste
Buy Online Tahini Halva Chocolate Flavored from Turkey. Koska tahini halva is a traditional healthy Turkish dessert made from tahini and white sugar prepared by crushing the best quality sesame in special stone mills, with the experience of craftsmanship since 1907. Fast shipping with DHL worldwide.
this is so yummy
Tzatziki?
no clue
halva yea
turkish delicacy
Tzatziki is yogurt and diced cucumbers
then no
hmm in the mood for some tiramisu
have you ever had baklava? @spare oriole
i think so
oh yea thats good
idk the name of most of the stuff i eat
u should try tahina though
i think for someone who never ate tahina it can be life changing
well, since i eat it almost every day i guess
only if i can stay at your mansion
lahmacun is yummy too
hey if I go there in the summer visit me there @spare oriole
Pupusas look like arepas
you can stay at our summer house
And arepas good
they're similar to arepa but different
arepas are mid compared to pupusas
@granite tree can probably attest
No I dont think so but looks great
at first glance i thought it's hummus
where is it
Turkish is favourite food and greek
heres your lahmacun @near gull
looks like an island
yea i know this
dont know the name in english though
"grape leaves"
o.0
but baklava is probably the best thing ive ever tasted
stuffed grape leaves
this looks like tamales
grape leaves sounds weird but sure
o.0
i see that u mention a lot of mexican food
every search in google tells me mexican food
its called simit
isn't it bagel
Bro 😍
is that top one
the one hamster mentioned
Look weird
I guess
thats grape leaves man
Grape leaves or banana leaves or whatever, its all good
i mean the version of grape leaves i know, is with rice inside
very nice
I also find turkish people make best rice
Its not too sticky
But not dry
bye

it's not all mexican lol, pupusas are salvadorean
me eating veggies
part of latin america
not really
turkish food sucks, authentic mexican food is awesome
o.o
do you live close to London?
Yes
noice
Close enough to be annihilated by a russian meganuke i guess
LOL
Or at least radiation poisoned
London is gorgeous
parts are
Kensington
of course
Sohooo
Peckham 😫
in Philly Kensington is a druggie alley
Oof
Barking 😩😩😩
Hackney😩😩😩😩😩😩
the ritz mayfair 😩 what a dump
i mean madonna lives in london it must be better than nyc
Central isnt safe either
nothing has ever happened to me anywhere in london
Westminster highest density of crime
i doubt violent crime
I thought Peckham sucked @near gull
Mostly theft and robbery
It was worse 20 years ago but its still not great
This is marketing
uh yea my bad
Also lot people there so theres prob gona be more crime
u would be fine
should I be prepared not to walk on the streets after a certain hour?
If you dont move to a crappy neighbourhood
Yes in some places
Yep
Its a city of 10mil ofc theres crime
what kind?
U would need to be paid a lot of money to move there for a job tho otherwise not worth it
a lot of homicide?
mexican food is often very spicy
Theft most commonly
oh okay
Robberies
Decent chance of getting robbed but getting assaulted/murdered highly unlikely
better than the US then
Theres a bunch of violent crime but its between "gangs"
not a lot of shooters?
And by gangs i mean children on mopeds
i read somewhere that stabbing was a big deal in london
They prefer knives and swords
An elegant weapon for a more sophisticated criminal
The knives are crazy now
do they just randomly stab someone?
Like samurai swords
Lmao
okay
Or they are very desperate
mfs pulling out katanas out of their 5 layers of pants but i cant even carry pepper spray
the mayor is such a failure as well
No taser
You can only carry a fixed or manually folded blade of up to 2.5 inches i think
Can barely scratch anything
Whatever u call it
i mean pepper spray is not even a weapon
Better than nothing
why cant you carry a pepper spray
🤷
i need to get some pepper spary myself...
uk is very against citizens carrying weapons
culture i guess
It hasnt really been an issue until recently
pepper spray up to some density is usually legal
Just get yourself a long and illegally powerful flashlight
Best weapon
You can blind them and smack them
Not kidding get a 10-20 watt long steel flashlight u will blind everyone
Lmao
Especially in the dark
My gf has one that looks like a cute cat head thing like so 😺
But has slots to put your knuckle through
knuckleduster type things are kinda useless
they hurt your hand more
you're a very sexy man
what's ur msc in
unfortunately its not me its kimi
Discord usage
Very hard to do
If they have a knife its useless
aint nothing better than corroding your stomach lining and shitting it out
for that sure
xD
pepper spray will instantly disable any attacker
E20
i have yet to see someone take it to the face and not react to it unlike tazers
I guess 2, hackney downs is getting better
mainly west
People complain about gentrification and then get on jubilee and go to canary wharf 💀
that's just not true
i've been sprayed before
bbut i dont remember the feeling
Actually I want london more crime so I can become batman 💀
I dont have the backstory to be batman
I gotchu
i have a better chance to become the joker
🤨🤨🤨
I'll hook you up
*theater
😩
🤔
You also need to inherit a multi billion dollar weapons company
That is part of character development
theatre disctrit/
I dont have an inheritance
Playing life on hard mode 😔
i'd say normal mode
New BMTH is so good 🙏😊
whats that? @honest star
A band, Bring Me The Horizon. Their new single is very good
𝙋𝙊𝙎𝙏 𝙃𝙐𝙈𝘼𝙉: 𝙉e𝙓 𝙂𝙀n: https://bmth.co/nexgenAY
Tickets for our 2024 tour dates: http://bmthofficial.com/live/
Bring Me The Horizon – Kool-Aid (Lyric Video)
Listen to ‘Kool-Aid’ here: https://bmth.co/Kool-AidAY
Follow bring me the horizon:
instagram: https://bmth.co/socialsAY/instagram
twitter: https://bmth.co/socialsAY/twitter
facebook: http...
enjoy
JIGGLY
yes?
Whatcha doin?
lol
.....
drugged attacker:
Mein gott pattern matching enums can be pretty tedious
https://drewdevault.com/2021/12/28/Dont-use-Discord-for-FOSS.html# wonder what people here'd think about this
he makes some decent points
but
If you use GitHub, consider SourceHut[3] or Codeberg. If you use Twitter, consider Mastodon instead. If you use YouTube, try PeerTube. If you use Facebook… don’t.
that doesnt even load
this reads like bitter ranting
.randomcase When you choose Discord, you are legitimizing their platform and divesting from FOSS platforms
WHen you CHOOse dIsCord, YOU ArE LegITiMIziNg ThEir pLaTfOrM aND DIVeSTINg FrOM FOSs PLatfoRMS
womp womp, do better if you dont like it
does it even matter for FOSS projects to micromanage what we use to communicate between members
incredible cope
who has even heard of sourcehut or codeberg
wtf is matrix and zulip even
matrix I've heard
The issue with FOSS is that it's almost always dogshit.
Nobody wants to use LibreOffice or whatever it's called.
real
Why is FOSS dogshit @foggy jungle ?
Conceptually it's good.
In reality, FOSS almost always lags behind in user experience functionality.
Why though?
No one is paying these people to make good products
Lack of money, lack of adoption, lack of contribution
And what projects do you have in mind when you consider this?
I can agree with those points.
Like when you're at a family dinner and you want a little bit of everything but by the time you get 2/3 through your plate you realize you took too much food.
Federated social media has almost entirely fallen flat after the slight uproar after Twitter stuff.
It's more like...
"Mom I want Discord."
'We have Discord at home.'
Discord at home: 
Most FOSS enthusiasts are heavily on the cope train rationalizing a lack of featuresets while they chase ideals.
Which isn't a bad thing; like I said, I'm a huge fan of FOSS as a concept.
But when it comes to execution FOSS almost always misses the mark.
I'm a huge fan of world peace as a concept.
World peace, like the FOSS movement, likely won't ever be a reality though. Good metaphor.
The first step to improvement is not to come across as a spoiled, idealistic, elitist child
Yeah-- I have a lot of hot takes regarding... FOSS, Linux, etc.
Discord is truly an awful alternative to a wiki, I am annoyed that it's becoming the norm.
i dont understand how this even happened in the first place
they arent even in the same class of software
As it turns out, most people want to reap the benefits of a system without contributing to it.
Wiki's almost always rely on a small subset of individuals maintaining a knowledgebase.
Writing a comprehensive explanation in discord is easier than creating an account on a wiki and writing sth
It's a less formal process, which helps as well
its such an awful website
Yeah, fandom is also a part of the problem
I know in the videogame world, wiki's are still mostly manually maintained, which breaks my brain.
You could easily export that data to the wiki and format it properly utilizing an API.
But it doesn't happen.
Cant you contribute anonymously on wiki? Does it also need an account?
It tracks contributions by IP address if you're trying to contribute anonymously (on most Wiki-engines.)
making an account takes a trivial amount of time, i dont see how thats an issue
Wiki's aren't designed for user experience.
It's more that it's a series of steps

Whereas on discord it's just start writing
Wiki user and discussion pages are a very legacy concept.
And it grows into an article on its own
does wikipedia count
A wiki is more " I will now write the article on ...", whereas discord is "I started talking about it and now I have 2k characters and this is a useful resource to refer to"
and the latter is much more accessible
writing articles is hard
i dont have enough exp with wikis to talk more about this tbh
i only have a few tens of edits on wikis
This might be one of the areas where NLP actually has some rational application.
Synthesis of Discord discussion topics into concise Wikipedia articles.
one of the things im most excited about in LLMs is how they will help search engines
not that you need an LLM for it, this has existed for years and probably doesnt use an LLM
Yeah....
Incoming wall.
The provided transcript is a conversation on Discord discussing the challenges and perceived issues with contributing to wikis, particularly focusing on FandomWiki and the difficulty of creating and formatting content on traditional wikis. The users express frustration with the formalities, user experience, and steps involved in contributing to wikis, contrasting it with the more informal and accessible nature of Discord discussions.
The conversation touches on the manual maintenance of wikis, the tracking of contributions by IP addresses for anonymous contributors, and the legacy concepts of user and discussion pages on wikis. There is also a mention of the potential application of NLP (Natural Language Processing) for synthesizing Discord discussions into concise Wikipedia articles.
Based on this information, a Wikipedia article could be created with the following outline:
Introduction
Brief overview of the conversation on Discord regarding challenges in contributing to wikis.
Mention of specific concerns about FandomWiki and the perceived issues with traditional wikis.
Contributing to Wikis
Discussion on the less formal process of writing on Discord compared to creating an account on a wiki.
User opinions on the accessibility and user experience of contributing to wikis.
FandomWiki Critique
User dissatisfaction with FandomWiki and its perceived shortcomings.
Comments on the overuse and negative opinions about the website.
Wiki Maintenance
Insights into the manual maintenance of wikis, especially in the videogame world.
Mention of the possibility of exporting data to wikis and utilizing APIs for proper formatting.
Anonymous Contribution
Discussion on contributing anonymously to wikis and the tracking of contributions by IP addresses.
User opinions on the ease of making an account on wikis.
User Experience on Wikis
User perspectives on the design and user experience of traditional wikis.
Comments on the series of steps involved in contributing to wikis compared to the more accessible nature of Discord discussions.
NLP and Wikipedia Synthesis
Brief exploration of the idea of applying NLP for synthesizing Discord discussions into Wikipedia articles.
User opinions on the potential rational application of NLP in this context.
Conclusion
Summary of key points discussed in the Discord conversation.
Final thoughts on the challenges and potential improvements in contributing to wikis.
It's important to note that the Wikipedia article should adhere to Wikipedia's content guidelines and policies, ensuring neutrality, verifiability, and reliable sourcing of information.
bro who uses a conclusion in a wiki article
Probably could've retooled that prompt a little bit.
i dont think ive ever seen such a thing in an encyclopedia
My point moreso is that if the knowledge exists, it can probably be restructured via NLP into some Wikipedia-friendly format.
wasnt the semantic web about this btw?
The Semantic Web, sometimes known as Web 3.0 (not to be confused with Web3), is an extension of the World Wide Web through standards set by the World Wide Web Consortium (W3C). The goal of the Semantic Web is to make Internet data machine-readable.
To enable the encoding of semantics with the data, technologies such as Resource Description Frame...
the cooler web3
Hire a nerd to constantly take minutes down in the discord channels
this one with a .0
No nlp needed
Isn't Discord starting to do like... NLP summaries of conversations?
but muh effective accelerationist agenda
I swear I saw that feature for awhile.
clyde? shitass feature
wait actually
!rule 10
yt made a really cool use of NLP on thge mobile app
It's not an answer.
isn't it?
No?
the issue is that this requires a lot of knowledge
I wonder if you could embed the entire message log of this server into an LLM and have it become a python expert
it sorts comments into topics
basically
i want to see NLP in things like that
Considering the amount of blatantly incorrect code/information posted, volumetrically, I'm sure we'd actually just create the dumbest model known to man.
I would expect it mostly gets corrected often enough that it could theoretically figure it out
it might average out
It gets corrected, but I'm not sure that it all gets corrected in a manner that I think some sort of LLM could comprehend.
so far i feel like LLMs worked best to complement search engines, for datamining and for ad hoc text/NLP tasks e.g. spreadsheets 
automated wikis sounds cool but i worry it might end up like those crappy LLM generated blogs
but well, you can't write an algorithm for the truth
i'm fairly certain the average pydis message is not great
Better than going over half a dozen channels worth of pins which are 70% memes
offtopic but spreadsheet AI is very good
of course a human written wiki is better, but clearly that ship has sailed
like having summaries of text cells, automatic translations, etc
oh yeah, I could see spreadsheets being a good fit
i use it mostly for learning new languages, but i can see it being good for adhoc text processing that may take too long to write code for if that was possible in the first place

I think my only real use for chatGPT so far has been reformatting notes.
If I mess up indentation in Obsidian or something or want to restructure my notes to be a certain way.
i do actually use chatGPT a lot, handy tool
I've found using it to turn a bulleted list of subjects into something like a markdown table is convenient.
along with bard, apparently bard is goated for translations, friend asked me to help him find the best AI for translating a novel and he found bard to be best
oh right theres also helping you practice learning a language
thats a pretty nice use for LLMs
for some reason, i decided to learn german recently
i dont know why
LLMs helped with that 
hi
go away, only chatGPT fans allowed
😭
foss is really jst free labor for companies and researchers
I often use copilot to convert tables of instructions into some programming-language-compatable format
it sometimes even implements the simple ones for you
e.g. when I was writing my .class file generator
copilot does seem nice
i have friends with disabilities who cant type fast and it seems great for them 
I just copy pasted the list of bytecodes and it vomited out the correct classes
I haven't used copilot at all.
That's actually a cool use for that.
I found it unreliable when it gets too "bold", trying to invent entire implementations and such
but it is very good at speeding up typing
i had it for a few months cause students and honestly forgot i had it
I type fast enough that I'm not sure Copilot would save me much time 
probably not, yeah
LLMs in general are really nice for people who cant type fast, i knew quite a few people who have that issue, disabilities and age generally
I am pretty close to the point where reading the suggestion and assessing its correctness takes longer than typing the right thing
Yeah I can see someone like my elderly father using LLM's to interface with technology in a more idiomatic way.
which is also why I stopped using it recently
this is why i stopped using LLM generated code
except for scripts that are only like a 100 lines
LLMs goated for that actually
It really depends on the size of the LLM.
I've use Mistral OpenOrca 7B in GPT4All and I've had to correct it a lot.
like i just have an algorithm or forumla that i dont want to install a library for and implementing it takes too long, so i just ask an LLM to generate it for me, i dont have to worry much about correctness because i can just throw solved examples at it
I'd say somebody not checking the GPT-generated regex crashed your infrastructure.
to be fair, i had my co workers do that with regex too
always need to sanity check
Yeah, I often just use it for like... opencv/pyautogui stuff
I am not reading your awful unpythonic APIs
git commit -am "Sending it."
git push origin HEAD
not only did it open a DDoS opportunity
yeah bud
but it also crashed the whole service on invalid regexes
To be fair, it looked correct.
because someone hasnt learned about error handling
all regexes look wrong
Everything is cake @foggy jungle
The specific feature chatGPT tried to use in the regex wasn't supported with what we were doing.
YARA's regex engine does not support look-ahead/look-behind.
oh, different ragox flavor?
Which chatGPT had utilized.
human failure
regex how the hell did i typo that in
get rid of the weak link
ragox
sounds like a caveman
protip: you can ask the LLM to generate code that tests its own regex
I asked an LLM to generate the steps we need to research in order to get to the Alcubierre drive.
That was interesting.
I like asking LLM fantastical questions and seeing what it makes up.
ngl half of my chats with LLMs are just me trying to figure out a hyperspecfic term for something because i despise ambiguity
gotta use the corrects words for everything
I wonder how many parameters Bing Chat has.
isn't it just lobotomised GPT4
I don't know.
I don't know how big GPT4 is either and OpenAI won't say because sEcReTs
The largest I've gotten to use on my machine is 13B with koboldcpp but it's slow
I want Jarvis.
whats your machine?
specs?
r9 5900hs, 16GB RAM, RTX 3060 6GB
it's a zephyrus
One of the coolest machines I've ever owned.
first time I've gotten >15fps in a game
well regularly at least
oh those are strong specs lol



