All boardsБлинчикиInternationalКрендельнаяЛес

#5 SciTech Thread


#1057809

>>1057580

>

tens of these in the wild

not in the wild, just for high-profile targets

you won't spend 0-day RCE on normies

#1057850

https://gizmodo.com/amazon-reportedly-ditches-just-walk-out-grocery-stores-1851381116

>

relied on more than 1,000 people in India watching and labeling videos

damn, if amazon gives up on it it means ai labeling has failed there will never be automated transcriptions except the current ones where a fifth of the translations are expected to be wrong (which kpop fans have been used to for 10 years theres no difference between whisper and most pre-ai fansubs)


#1057992
680×68071.37Kb

>>1057809

rces sure but supply chain attacks seem to affect a heckload of normies alongside high profile targets

#1057999

>>1057992

>

supply chain attacks seem to affect a heckload of normies

you just need to update very fast once those vulns go public

if you're late by few days/weeks then yeah, you're fucked


but in theory you might have no time to update and still be fucked, yeah...

#1058001

>>1057999

or perhaps update very very slowly >>1057289 ➡

security patches are on auto update but as a secondary safety measure i also have some internet money on this compoter on a dumbass browser extension, once that is gone i know im hakt and its time to go off grid

#1058010

>>1058001

>

as a secondary safety measure i also have some internet money on this compoter on a dumbass browser extension, once that is gone i know im hakt and its time to go off grid

#1059092
683×19524.76Kb
1080×19202.05Mb00:05

microsoft optics

#1062478

wait so you cant browse an account on twatter beyond 3000 posts or so but advanced search with date ranges work fine bimbambum i was dum

#1062840
750×75036.82Kb

https://crapts.org/2024/04/21/all-fritz-box-modems-have-been-hijacked/

>

DHCP server on these modems hands out leases with the DNS suffix fritz.box, which means that domains in DNS requests are appended with the suffix. Unfortunately, this setting cannot be modified. Normally this doesn't have to be a problem, but since the beginning of this year the fritz.box DNS suffix is an actual registered domain and the owner of this domain is not AVM or anyone affiliated to them.

#1064337

bought iphone 15 pro and disappointed with the battery life

I didn't charge it to 100% and used quite extensively but I've got a feeling it's not much better than my old iphone 12...


also original clear case is a bit slippery but why would you hide beautiful titanium under opaque case? makes no sense


maybe should have bought pro max instead but it's expensive and I don't like the size. but at least the battery should be ok


didn't notice the 120 hz difference at all tbh but always on display is very nice and the new dynamic island and action button (although very hard to press in that case )


also bought new airpods pro with usb c so I finally can get rid of stupid lightning but it's so expensive...

need work hard to afford those stupid apple prices...

#1064412

charged yesterday evening to 100%, in the morning it's 82% even though I almost didn't use it

it's over


I read factory reset might help, what if settings transfer from old iphone turned on some suboptimal setting on a new phone? but I don't want to setup everything from scratch...


so I've just bought expensive phone with shitty battery. please kill me

#1064477

36% already


wanted to bought 15 plus initially but non-pros are too limited because stupid apple

and pro max is expensive and big and heavier than even plus...

#1064538

imagine my face if 16 pro would have much better battery life tho

#1064554

the small pocket in New Jeans is too small, airpods case barely fits

I hope I won’t miss it

#1064596

>>1064412

>

charged yesterday evening to 100%, in the morning it's 82% even though I almost didn't use it

I think it got better this night

maybe it was indexing something first days

#1064602

>>1064337

i think the pro models always suffer from bad battery life because of 120hz but with a battery the size of the base model

>>1064412

>

charged yesterday evening to 100%, in the morning it's 82% even though I almost didn't use it

that does not sound normal at all though

>>1064596

that is also possible, true

>>1064571

on 14PM i only use heif max, it will shoot up to 48MP if it requires that much detail, less MP if not, but it will always be a small filesize, heif is good, raw is for nerds who need details, you'd only want to shoot in raw a couple pics ever, not a daily thing for sure


#1064604

>>1064602

>

that does not sound normal at all though

>

bad battery life because of 120hz but with a battery the size of the base model

yeah

>

14PM

harambro


yeah, I think I will also use 24mp most and maybe 48mp heif sometimes

not sure about 28mm/38mm 1.2x/1.5x zooms tho, I will probably use just 1x

#1064605
1224×2029244.08Kb
378×1529.44Kb

>>1064604

15 has more and different options for MP and for zooms, i like that

the options they gave for 14PM are more than enough imo, i set it up like that in the camera app settings, and then when you open the camera app you either toggle HEIF max on, shoot up to 48mp if needed, if not less, or toggle it off and it shoots lower detail jpegs, but i stopped using that

you could also long press on it and toggle RAW max but i only used it on some vacation pics, it's really not worth it... heif max is really good and when you share to external apps it will convert it to some sort of jpg if heif is not supported there, e.g. discord

for zooms its a bit different, 14pro series have optical zoom at 3x, 15 pros have it at 5x

the 2x zoom on both should be pretty good, it's not a real zoom but a cropped down photo but it comes out looking fine

1.2 and 1.5x seem to be the same, although the 15 series has some different mechanisms behind it, it is still 3 cameras, so it surely uses the main 48mp camera for anything 1.0-4.99x on the 15pros, just probably does some clever processing behind the scenes

1.2 and 1.5x is probably more intended for PORTRAIT mode shots but i rarely ever used that

either way HEIF MAX is the way to go imo try it out

#1064606

>>1064605

>

either way HEIF MAX is the way to go imo

do you often take 48mp photos?

on my old iphone it was just 12mp and I think it was enough most of the time

now it's 24mp default on iphone 15 pro, I think should be good too

#1064607

>>1064605

>

15 pros have it at 5x

only pro max 15 pro is only 3x optical

but it's ok for me, I don't care about camera much


bought mostly because I thought it would be better battery, because usb-c, 2 esims, av1, always on, dynamic island, things like that

so far I don't think it's much better than my old one tbh, especially for that price. but most features I wanted there so I can live with that

should be enough for few years, I hope new ones would be better (especially in terms of battery)


iPhone 16 Pro 6.3" might be very good too but welp, too tired of waiting anyway

#1064608

>>1064606

it just decides on its own whether to register it as 48mp or lower

if using jpeg it will always be 12 on 14 series, 12mp is imo quite low... but i notice the pics at 48mp take up 5-7MB, the ones at 12mp take up around 1MB

i just left it at 48mp heif max for a while now and forgot about it but i have the 200gb icloud storage and im really barely above 55gbs in

>>1064607

>

only pro max 15 pro is only 3x optical

oh right, honestly i think the 3x is more practical most of the time, i use it a lot even indoors, 5x would be amazing for landscape shooting but there is only so much landscape you will see in a given year

>

av1

does it make a difference?

>

dynamic island

that is really cool tbh

>

iPhone 16 Pro 6.3" might be very good too but welp, too tired of waiting anyway

dont fall for the apple propaganda, half a year before the 15 series they hyped it up a lot and it only ended up being the 14 series but with usb-c and some minor changes...

we have reached the flat line of the S curve for innovation in terms of phones, i don't expect major changes going forward, just marginally better cameras, marginally better processors and what not... but iOS is well optimized and i imagine even 5 years old iphones run pretty smooth, as far as 60hz goes at least

#1064611
500×56158.05Kb

>>1064608

>

it just decides on its own whether to register it as 48mp or lower

don't you need to press "heif max" button in camera because it's disabled by default? or you always enable it?


>

48mp take up 5-7MB, the ones at 12mp take up around 1MB

yeah

>

i have the 200gb icloud storage and im really barely above 55gbs in

I have 50 and it's only 9 gigs and only because I started doing more videos recently. I don't takes much photos


>

does it make a difference?

to open av1 videos from kch and other nerd experiments. also futureproof haven't tried it yet tho


>

dont fall for the apple propaganda, half a year before the 15 series they hyped it up a lot and it only ended up being the 14 series but with usb-c and some minor changes...

I'm very unlucky, I'm afraid i16 pro would be amazing

#1064612

>>1064611

i think if you press it once it leaves it as default on? when you open the camera app the next time

unless thats another camera app setting like "preserve settings", i have that one on either way for things like exposure correction to remember it at -0.3 or whatever

i never manually enable it for sure, it's remembered as toggled on

>

to open av1 videos from kch and other nerd experiments. also futureproof haven't tried it yet tho

is that like webms then? i know they dont work for me unless i open them in VLC

i also still cant upload any sort of video here, no matter its encoding or format

>

I'm very unlucky, I'm afraid i16 pro would be amazing

fell for tim apple's propaganda-nim....

#1064615

>>1064612

>

unless thats another camera app setting like "preserve settings"

>

it's remembered as toggled on

yeah, you probably enabled that


>

is that like webms then? i know they dont work for me unless i open them in VLC

>

i also still cant upload any sort of video here, no matter its encoding or format

yeah, you can put AV1 video format into WebM (and also MP4) container, I was uploading many of them here

they work in Chrome, Firefox, Edge, but on mac/ios it requires hwdec so only i15 pro+ at the moment

#1064617
800×80058.26Kb

I think I can buy this one

problem solved...


new iphones support Qi 2 15W so I think it should be enough...

just need to find good model of those magsafe powerbanks

#1064620

>>1064617

Yeah but power banks and just overly charging your phone in general will always damage the battery rapidly

#1064622

>>1064620

I mean if I outside and forgot/didn't have time to charge my phone enough and it's dying at 5-10%, I can just attach powerbank and keep using it


I have a normal 10Ah powerbank but it's very inconvenient to attach with a cable while you're outside

#1064640

>>1064622

>>1064623

yeah just charge when it drops to 20% or something, dont be like those people that keep the phone taped to the powerbank even at 100%

i used a powerbank on vacation too

#1065441

>>1062478

some tweets are just not found from search what else is there, scraping googol results from site:twitter.com/bvndit_official?

#1065600

>>1064962

I can just charge over night. problem solved


but I've watched few reviews of 15 pro and everyone seems to be complaining about the battery, so it's probably the way it's supposed to be. very sad....


I could have checked it before buying but I didn't have much options anyway: pro max is $200 more which I don't want to pay and I was too tired of my old one so waiting 6 more months would have been painful...

maybe I can sell this one when 16 comes out, if it's really good

#1065687

>>1065674

>

rust without all the yikes

I can't see what's wrong with the rust

#1065888

>>1064477

should have bought plus

the only major feature I would have missed is always on but better that than feeling anxious about battery life 24/7...


realistically, 15 pro should be enough for 1 day if charged 100% before leaving home

I just expected more

#1065896

>>1064611

>

haven't tried it yet tho

tried

AV1 in WebM doesn't work

AV1 in MP4 works


but MP4+SVT-AV1 had some strange issues last time I tried

welp


maybe even related to mpv encoding mode but nobody fixes bugs there anymore...

#1065899

>>1064596

>

I think it got better this night

https://www.reddit.com/r/applesucks/comments/1b939lr/wife_just_got_iphone_15_pro_battery_seems_terrible/

>

I have a 15 pro max. The battery is REALLY bad for like the first week. Then it optimizes or something and isn't too bad.

>

I've noticed this as well. The battery is fucking shit the first week or 2 lasting me like 5 hours. Then suddenly after a month or so the battery seems to last me double that.

>

It gets better, trust. When I got my 15 it was dog shit for like 2 weeks. Saw others with the same experience on here as well


maybe I just need to wait a bit

#1065955

>>1065904

indexing your phone, trying to find something interesting

#1065970

https://leetcode.com/problems/single-number/description/

made simple fixed array solution in raw Python: works but much slower than others

welp, obviously that's because arrays/loops in Python are too slow. rewrote to a nice native optimized numpy primitives. and... it's still too slow


read the discussion: actually you have to use the math trick to make it fast

#1066198
1440×1800152.72Kb

>>1066143

bit craycray

if i remember right 10-15km was the annoying part with legs hurting but after that things got eezy breezy

#1066205

lost 30% of battery today doing almost nothing, just occasional apps check and a bit of photos (no music)

I guess it's almost ok, comparable to my previous phone

#1066593
1332×62532.89Kb

>>1066591

without modifying array and checking only first/last in sorted it's actually ok

with 2 people it seems to be actually very simple

but with N people...

#1066594
1298×60858.17Kb

I'm so fucking fast

#1066598

>>1066596

why is it easy???? I'm thinking for like 15 minutes already and only have small chunks of bad code...

#1066607

>>1066596

>>1066598

1 hour for easy... ok...

am I stupid and there should be easier solution?


class Solution:
    def readBinaryWatch(self, turnedOn: int) -> List[str]:
        if turnedOn == 0: return ['0:00']
        if turnedOn >= 9: return []
        return list(calc(turnedOn))

def combs0(arr):
    yield 0
def combs1(arr):
    for el1 in arr:
        yield el1
def combs2(arr):
    for el1 in arr:
        for el2 in arr:
            if el1 != el2:
                yield el1+el2
def combs3(arr):
    for el1 in arr:
        for el2 in arr:
            for el3 in arr:
                if len(set([el1,el2,el3]))==3:
                    yield el1+el2+el3
def combs4(arr):
    for el1 in arr:
        for el2 in arr:
            for el3 in arr:
                for el4 in arr:
                    if len(set([el1,el2,el3,el4]))==4:
                        yield el1+el2+el3+el4
def combs5(arr):
    for el1 in arr:
        for el2 in arr:
            for el3 in arr:
                for el4 in arr:
                    for el5 in arr:
                        if len(set([el1,el2,el3,el4,el5]))==5:
                            yield el1+el2+el3+el4+el5

combs = [combs0,combs1,combs2,combs3,combs4,combs5]
hours = [1,2,4,8]
mins = [1,2,4,8,16,32]

def calc(n):
    combs_hours = [[h for h in combs[i](hours) if h <= 11] for i in range(5)] # 4 hour bits
    combs_mins  = [[m for m in combs[j](mins)  if m <= 59] for j in range(6)] # 6 min bits but only 5 max possible
    seen = set()
    for i in range(5):
        for j in range(6):
            if i+j == n:
                for h in combs_hours[i]:
                    for m in combs_mins[j]:
                        if (h,m) in seen: continue
                        seen.add((h,m))
                        yield f'{h}:{m:02}'
#1066753
368×44834.32Kb

>>1066143

today 11km but I've tired even more

#1066936
1294×59859.00Kb

why am I so fucking fast

#1067015

>>1066384

>

bruteforced my code

most of the time it’s not brute force, you understand the general idea

but it’s too hard to check all invariants in your head, so you just hit “submit” without being completely sure about corner cases

is it ok

#1067055

talking with chatgpt as with my personal business coach and psychologist

#1067435

>>1067420

>

[100000,99999,99998,99997,99996,99995,...]

but how can I solve it in general case

100000 should be kept, but I have no idea if 99999 should be kept based on that information

stupid challenge...

#1067441

>>1067435

and of course tons of comments in discussion about how easy it is

I'm stupid

I should quit programming, I can't solve even fucking simplest problem

#1067445

>>1067441

how can I even solve shit like this


google solutions and remember? but then I won't be able to solve general case problem

just google tons of solutions to problems like these so you would be able to combine them? then why is it even intellectual if you just copy other people

#1067489

>>1067353

Assuming you're exceeding the time limit because you're going through the list too many times, you don't have to compare every pair of nodes in the list. How could you solve this efficiently if you were told to remove every node with a greater value to its left side rather than its right side?


>>1067445

Speaking about puzzles in general, I wouldn't recommend directly looking up the solution, but rather putting it aside and coming back to it after solving other stuff. If it's a real life problem that needs to be solved in a timely manner, then you often do have to just look it up or ask others for help and then try to remember it for the future.


In this particular case, I think studying sorting algorithms would help you immediately think of possible ways to approach the problem. It would probably also help with a lot of other puzzles related to ordering numbers.

#1067505

>>1067489

>

How could you solve this efficiently if you were told to remove every node with a greater value to its left side rather than its right side?

I read in comments that you need to reverse the list and compare to the left, yeah

now I'm not sure if I should solve this at all because I couldn't solve it by myself

#1067513

>>1067489

>>1067505

with the help of Copilot and comments I made working solution very quickly

but I didn't come to it by myself (even revlist is generated by Copilot ) and don't understand why is it the correct one

I'm too stupid...


I guess I need to sit with a paper on it and analyze, maybe I will be able to solve tasks like that better in the future


class Solution:
    def removeNodes(self, head: Optional[ListNode]) -> Optional[ListNode]:
        headRev = revlist(head)
        node = headRev
        maxval = 0
        headNew = ListNode()
        nodeNew = headNew
        while node:
            if node.val >= maxval:
                nodeNew.next = ListNode(node.val)
                nodeNew = nodeNew.next
                maxval = node.val
            node = node.next
        return revlist(headNew.next)
def revlist(node):
    prev = None
    while node:
        node.next, prev, node = prev, node, node.next
    return prev
#1067514

>>1067513

>

and don't understand why is it the correct one

I had experience like that in the past too tbh

like I can copy some algorithm from other piece of code and made it work correctly in my program

but I don't fully understand the algorithm

#1067515

>>1067513

>

I guess I need to sit with a paper on it and analyze

oh yeah, on paper it makes sense. I guess I shouldn't have try to solve it only in my head

also the hint actually helps but I understood it wrongly so took the wrong path


welp, need more practice I guess

#1067524

>>1067513

>

Copilot

I'm biased (because I hate AI), but I heavily recommend against this. It's like practicing the multiplication table using a calculator. You should use it to save time doing things you already know how to do


>>1067515

I didn't realize those problems had hints. If you had already read about operating in reverse, I think you just missed the idea that if you know the biggest number in a set, then you don't have to compare against every other number in that set. On top of that, you're comparing N nodes against N sets, but each set is built as a superset of the previous one with the addition of the node you just checked. So the result of that comparison tells you what's the biggest number in the new set.


>

welp, need more practice I guess

I reiterate my suggestion of studying sorting algorithms

#1067530

>>1067524

>

then you don't have to compare against every other number in that set

in my first try I build separate array with values and indexes and sorted it in reverse, and was comparing against it. but it was too slow

in my second try I build array as reversed linked list and was walking it in reverse order, remembering index with last big value or something like that

it wasn't obvious that if I go from right to left in original list I have all the information in just single variable


>

suggestion of studying sorting algorithms

sure. also lots of other algorithms because I'm very bad at it


>

It's like practicing the multiplication table using a calculator. You should use it to save time doing things you already know how to do

yeah, it's just always enabled in VSCode and it just suggested me the code and I hit TAB

writing code by yourself is of course much harder and better practice

#1067536

>>1067530

Maybe try rewriting the solution from scratch without copilot now that you already understand the ideas behind it. You could also try it in a different language

#1067538

>>1067536

>

You could also try it in a different language

oh yeah, must be fun in C/C++

Python is too easy for implementing algorithms, it's almost pseudocode

#1067544

>>1067538

Python does abstract away a whole lot when it comes to data structures. I think C is a great idea here

#1067566

>>1067562

oh, it's actually very simple

for some reason I'm very bad when I need to backtrack previous choices

#1067640

>>1067573

lol, it's actually very simple to do recursively (with memo)


I was trying to calculate all combinations manually for an hour, but didn't come with correct formula

I guess it's possible, just too stupid to do it right

#1067985

>>1067961

>

Testcases passed, but took too long.

all combinations working

now need memo, but not sure how

#1068019

>>1068011

seems like cache is working now, but it's slow in another place: factorization

should I come with good factorization algorithm by myself or just google it

#1068041
650×60034.92Kb
658×61042.15Kb

>>1068021

https://stackoverflow.com/questions/46841968/fastest-way-of-testing-if-a-number-is-prime

oh, yeah, my prime check was wrong, you need to go +2 and compare against square on each step


it's accepted, but even with combs cache it's very slow for some reason

probably I'm missing some simple solution


from math import prod
class Solution:
    def checkPerfectNumber(self, num: int) -> bool:
        cache.clear()
        if num == 1: return False
        divs = []
        n = num
        while n & 1 == 0:
            divs.append(2)
            n = n >> 1
        d = 3
        while d * d <= n:
            while n % d == 0:
                divs.append(d)
                n = n // d
            d = d + 2
        if n != num:
            divs.append(n)
        unique = [div for div in set(combs(divs)) if num % div == 0]
        return sum(unique)+1 == num
def combs(divs):
    for n in range(1,len(divs)): # take 1 number, take 2 numbers... (no need ALL)
        for div_arr in cached_n(divs, n):
            yield prod(div_arr)
def combs_n(divs, n): # take N numbers from array
    if n == 0: yield []; return
    if n == len(divs): yield divs; return
    for div_arr in cached_n(divs[1:], n-1):
        yield [divs[0]] + div_arr # selected first, combs without first
    for div_arr in cached_n(divs[1:], n):
        yield div_arr # not selected first, combs without first
cache={}
def cached_n(divs, n):
    key = (len(divs), n)
    try:
        return cache[key]
    except KeyError:
        cache[key] = list(combs_n(divs, n))
        return cache[key]
#1068056
1352×522107.14Kb

>>1068041

wtf, fucking cheaters

do you even know how many time I spent on this??????

#1068071
>

It's not much about coding , just take a look at list of perfect numbers

fucking cheaters...

>

Euclid, over two thousand years ago, showed that all even perfect numbers can be represented by, N = 2p-1(2p -1) where p is a prime for which 2p -1 is a Mersenne prime. That is, we have an even Perfect Number of the form N whenever the Mersenne Number 2p -1 is a prime number.

oh, so you just need to know math

I'm not good at math either...

#1068086

>>1068041

>

d = d + 2

we can also skip all non prime d's

but not sure if that would be faster

#1068136

What is the best way to stream stuff from PC to TV without using a direct HDMI cable?

Everything seems too convoluted or user unfriendly or straight up doesn't work

Like there isn't one solution that would be universally recommended so I am kinda at a loss

#1068154

>>1068136

>

stream stuff from PC to TV without using a direct HDMI cable?

wifi direct or whatever it's called

windows/macos can stream to modern TVs out of the box


but I think it's better to connect USB stick (e.g. xiaomi) to make your TV smart because controlling TV screen with keyboard/mouse isn't user friendly at all. it's much better to control nice adapted UI via TV remote control

#1068155

>>1068154

yea wifi streaming is superbad as explained by all the airplay/google cast etc things actually playing directly from the internet source instead of streaming the screen locally. kinda funny that Lan streaming loses to Wan play here, whats that about


personnelly i ended up buying a minipc to attach to tv and using mpv with a bluetooth keyboard, and then further down the line bought apple tv to control via remote for its a bit more comfy. sadly of course lost all the screencap/clip functionality of mpv

#1068156

>>1068155

>

actually playing directly from the internet source instead of streaming the screen locally

I doubt, delay would be unmanageable. wifi direct works via direct wifi connection (obviously lol)

>

minipc to attach to tv and using mpv with a bluetooth keyboard

yeah I've recently used bluetooth mouse to control youtube on a notebook attached to TV via short cable (notebook and TV near the wall, mouse and I on the bed)

but it's way too inconvenient... any kind of smart TV usb stick with proper TV remote control would be miles better

#1068160

>>1068156

i mean if you airplay a youtube video to your tv it plays it directly from youtube, maybe i mispoke

i definitely didnt try wifi direct (possibly requires hardware support?) tho maybe that might work for screen mirroring but idk

#1068194

I'll just have to rely on good old copper cables until TV manufacturers and Windows devs get their shit together

#1068203

>>1068202

As I said, LG needs to get their shit together

I expect better from a company that gets shilled by my waifus

#1068208

>>1068207

Makes it comfy wor watching Youtube and Crunchyroll

#1068211

>>1068210

Gimme a YT tutorial hoe to set it up and I cab guarantee that if I follow it step by step it won't work

#1068213

i kno someone who bought a 1200 dollar tv and now has ads on it

smart tv? more liek customers an idiot tv

#1068219
992×374223.17Kb

>>1068211

dunno, I just open youtube in chrome, youtube app on tv and in chrome it shows me this icon

also in macos it just lets me to add my TV as display when it's turned on

(both mac and tv are connected to the same wifi network of course)


it just works

#1068220

>>1068213

Legit seems that just plugging it into a PC and having it as massive monitor is the best way to use a TV still

Just wish I didn't have to fuck around with the cable


>>1068219

I am on Windows 10

#1068230

>>1068224

Yeah well...


This just goes to show, you can't beat the good old copper

As long as all the development is outsourced to Pakis, wireless will be useless

#1068234

>>1068230

>

This just goes to show, you can't beat the good old copper

nah, you just bad with electronics

#1068236

>>1068234

As I said

Gimme a Yt turorial of something that works

#1068240

screen 1: vscode

screen 2: documentation

screen 3: terminal


that's still not enough because can't see webpage with the app

#1068361

>>1068355

>

becoming a chef?

putting alcohol into food, making girl drunk and then

#1068384

learning next.js, it's kinda interesting

client and server share everything so code reuse (like data types, common validation, common functions, etc) works like a charm


I guess the only downside is that JS on the server isn't very fast/effecient, also only single core

but writing all API methods by hand (e.g. in golang) seems too tedious. unless it's some solution like PocketBase/Hasura, but I haven't tried it much

#1068393

>>1068384

>

JS on the server isn't very fast/effecient

but I guess everyone just swallow the pill, because it's very very comfy and productive to share code

even apple uses it https://twitter.com/rauchg/status/1286769353027801088


I spent some time in the past copying TypeScript type declarations to Golang and it's horrible exprience tbh...

#1068417

it's like the next generation of ruby on rails/django, but for heavy JS apps

previously you didn't need that much JS on frontend so you could code 90% of the code in comfy backend framework like RoR and add a tiny bit of JS for the client


but now you need shit tons of client JS so duplicating everything isn't cool

I wonder why lots of people still continue to use non-fullstack frameworks

#1068523
742×32435.22Kb

#1068524
>

react tutorial

>

added 1488 packages, and audited 1489 packages in 51s

are you crazy, oppa

#1068536

would it be hard to become a chef tho it seems like a good idea

working in a restaurant could be meh but going through some training might be nic e still

#1068577

>>1068536

giving up on programming because too much fear of AI?

#1068736
566×52849.89Kb

why did I start at 2nd of May

I can't have cool badge now...

#1068803

solving those requires too much energy

I can just do 2-3 in a row and then I'm exhausted

#1068934

>>1068838

>

it's not difficult

what's even difficult then

#1068957

>>1068936

>

In the YouTube video "it begins... users leave Stack Overflow," the host voices his disappointment and frustration over Stack Overflow's partnership with OpenAI, which involves using user-generated content without permission or compensation. This decision has led to strong reactions from users, with some deleting answers and being banned for trying to prevent their content from being used. The host emphasizes the importance of ethical standards and user consent in handling user-generated content. Some viewers have shared their own frustrations with Stack Overflow, including the prevalence of duplicate questions and condescending behavior towards newcomers. The speaker, who has relied on the platform to solve complex issues, expresses his gratitude but criticizes the company for making changes without consulting users. He implies that this may be his last video on the platform and invites viewers to watch his previous experiences with Stack Overflow in a related video.

#1068958

>>1068838

idk if kqts care if you actually work as a chef

you just need to be able to make nice looking byte sized foods and boom instant kqts


so just thinking about how to train into a chef

#1068981

>>1068957

>

He implies that this may be his last video on the platform

very wrong

he said it may be last video about stackoverflow

#1068984

>>1068958

>

you just need to be able to make nice looking byte sized foods and boom instant kqts

you wish

#1069176

>>1068958

>

you just need to be able to make nice looking byte sized foods and boom instant kqts

thats not how it works

#1069550
1179×2556540.66Kb

>>1066143

small improvement

18k steps

16km distance


I want 20km tbh, maybe a bit later

#1070697
>

Update your Chrome browser ASAP. Google has confirmed a zero-day exploited in the wild

RIP

#1070698
>

Exploit code for critical "use-after-free" bug is circulating in the wild.

how many of us have hidden malicious code in the system at this point

#1070710

>>1070697

at least on chrome you need to go to a page that includes ebil code

imessage exploits can be wirelessly sent to you unprompted

#1070731

time to update 4 apple devices at once again

#1070898
>

1st Single Album Hello, World!

#1071091

just took 6 hours of very lazy attempts to do a smal thing in go ling

the hard part is over, think i deserve at least an hour break

#1071270

>>1071091

>

6 hours

>

to do a smal thing

>

i deserve at least an hour break

are you me

#1071577

>>1071571

>

evidence that we may have reached a phase of diminishing returns

developers are safu

#1071634
1180×1080252.21Kb

#1072183

Is twitter now officially only X when it comes to the domain?

My twitter download addon stopped working and I think it's because of that

#1072184

>>1072183

twitter.com domain still works for me

the only difference I noticed is that x.com doesn't redirect to twitter.com anymore, it works on its own

#1072194
>

Let's be honest, guys, Twitter's died a long time ago. You can call it whatever you want but it's not Twitter anymore. He killed it out of stupidity and now begs every person in the room to have pity for this dying app

those redditors

#1072195

>>1072192

I hope the dev for the extension that I am using fixes it up

I wanted to send him a message on twitter, but twitter locked me out and says I require authentication to send messages

Modern internet was a mistake

#1072196

>>1072194

He says as he buys twitter premium

These fuckers are too terminally online and addicted to attention

they will never leave

#1072198

>>1072197

I don't know, I closed the window when that popped up

Could be just a phone number or something

#1072201
1379×1931248.08Kb
1475×2065415.23Kb

>>1072199

was considering getting disposable sim card for cases like that

you can get one for 6 bucks

#1072683

>>1072680

I guess the problem is that it's painful to see lots of stupid people on the internet

not like you're too smart, you just can see them behaving in a stupid way and it's horrible experience...

#1072684
>

YT channels are a lot like subreddits where you have to conform to the echo chamber, or get downvoted into oblivion.

#1072700
>

What annoys me the most is seeing something like an animal video and EVERY TIME (literally, you can count on it) you see "Oh, this precious baby! This magnificent creature!" I'm so f*cking sick of it!

#1072702
>

I'm sick of the comments starting with "as a ..." Nobody cares who you are.

#1072706
1280×1280107.85Kb

#1072707
>

I wish, I even have friends who do that and brag about their comment likes that do not translate to subscribers or follows

#1072804

if I don't like kpop stage mixes, how to tell youtube's recommendation algorithm that


if I click dislike on them, it may come to conclusion that I don't like group in the video or song or something like that

I think we should have some type of input form near dislike so you can hit dislike and write "don't like stage mixes" so that algorithm can easily understand that


otherwise if you just hit tons of dislikes on stage mixes, you may break something in your recommendations

I hate non-transparent systems...

#1073595

got my first captcha from simply going to a youtube link vpn future not looking very bright

#1073954
1600×24002.11Mb

snake case is beyond better than stupid camelcase

#1075271

>>1073595

fourchen now has some weird cloudflare click through cptcha on both main site and media server but those are separate so i cant browse site and also see images because those are 403d

my daily tech news...

#1077218

>>1069751

>>1069769

I change my mind, Thais are silver+ now

Malaysians on the other hand muslim country... 0/10

#1077220

>>1073954

>

stupid camelcase

yeah, because you never know, should you write isValidId or isValidID and multiple uppercased acronyms together look ugly like XMLHTTPRequest

just use whatever existing code uses anyway tho OCD is a bitch but what can you do in this unorganized world...

#1079450
686×103680.97Kb

>>1075271

we're back

guess something with firefox tracking protection now fixed by cf

#1086651
>

Omakub also installs…

>

WhatsApp + Signal for messaging.

>

Xournal++ for filling out and signing PDF forms.

>

Flameshot for more powerful screenshotting (using the Ctrl+Print Screen hotkey).

>

VLC for playing videos.

>

WHAT’S THE DRESS CODE

>

By default, Omakub ships with a Tokyo Night theme that looks ace with everything.


just a bunch of highly opinionated shit by random person which you will spend more time learning/adopting to than instead of configuring your own system. but I guess 100-500 random people will still star that shit and say how "awesome" it is

#1086653

Apple Intelligence is only on iPhone 15 Pro+

finally I've made a good choice


only for US so far tho


also seems like the problem with the battery is magically solved, I have ~80% by the end of the day now. or maybe just use it much less lol


but I have more problems with the new airpods (pro gen2 usb-c) even brought them to the apple shop, they reset it, seems like it helped, but dunno. it's the first time I have massive issues with the apple device...


also I won't have new watchOS 11 on my Apple Watch SE 2020 maybe time to buy a new watch?


also won't have Apple Intelligence on my iPad Air Gen 4 (because not M1) but I don't use it anyway so whatever...

#1086669

>>1086647

https://news.ycombinator.com/item?id=40630546

this ones interesting tho perhaps be amazing to just have a set of configuration files and reboot after security updoots and everything would work, instead of mucking about with apt and sudo nanoing files on /etc/

i assume thats what it means, wish i knew how to do it

#1086670

>>1086653

did they say there were three levels of ai: on-device, contact apple headquarters and contact openai


apple intelligence was the middle one i think? i wonder why they need specific hardware for contacting the cloud but maybe they preprocess stuff before sending it out or something

#1086731

wanna try linux stuff on spare computer but would have to switch plugs each time, kvm switch not big enough

#1087023

>>1086670

>

apple intelligence was the middle one i think?

I think it’s all 3, but they want to support on-device AI, so the entire thing would only be enabled on a device with fast NE

#1087079
399×54936.08Kb

bankerino informed me ios15 wont be supported next year

the small phone dream is officially, unironically dead

its over

#1087608
1440×1800256.23Kb
1440×1800331.57Kb

tested gnome on vm while waiting for the bigger switch and holy shit cjk just works? all without messing about with terminal or weird settings. can't believe i fell for the kde meme all this time

cjk: macos = chromebook = linux > windows


rdp on gnome also came out just a few months ago... and it works

remote desktop: windows = linux >>>> macos = chromebook


once i go from vm to metal there is still the mouse acceleration stage maybe i'll finally pass it and get away from the shackles of windoos once and for all

#1089142

>>1088873

>

thumbs for command until i started thinking this cant be good for my hands

>

how the heck do i learn to use my middle finger or somhthing

been using swapcaps for years and can’t live without it

dunno why would people use standard control location, so inconvenient

#1089259
1200×1800335.92Kb

fell for the 4o scam

i was just doing one of the example questions about fruit and picking ripe fruits and it decided to generate the most disgusting fruit with like a hole in the middle with weird gunk like spiderwebs inside


worst thing ive ever seen and i've used the internet.

fuck image generation

#1091213

https://www.xreal.com/air2

>

Level 1 is configured at 0% and slightly shaded, ideal for when users need to interact with the real world or in a darker environment.

>

Level 2 is set at 35%, reducing the amount of light passing through the lenses. This is ideal for a more immersive experience while still being aware of your surroundings.

idk about input yet but usb-c connected/powered monitor-glasses for walking while computing

#1093161

>>1093160

mpv is shit without wm4 anyway tho...

can just use VLC as well at this point