@rithvikp also sub bot but hey, YouTube fraud team doesn't care about big corporations

YouTube as a whole has ruined their platform by pandering to corporate greed, especially with the copystriking. Google's "don't be evil" is just a pile of bullshit so they can hide the fact that they're the worst of the bunch

@irene ok it's cancer but T-series is somehow even worse

(At least it's not the Paul brothers tho!)

You can actually do this in one line in O(1) time:

```
is_triangular = lambda n: n == 10 or n == 3 or n == 6
```

Passes all test cases

I've been forced to use it before and I absolutely hated it. Google Meet is similar but easier to use, although it's by Google so I can understand concerns there.

We use Appear.in for meetings. It's low latency and pretty decent for a WebRTC chat. Peer based so it doesn't look like they store the video calls or even chat either.

Welcome to DevRant 😁

If you're starting out in Mobile apps and coming from JS I usually recommend react native since it's easy and crossplatform.

As for tools, I've used react-native-html and builderx for UI but they're underwhelming and pretty buggy.

This came out recently: www.crane.ai
I tried out their beta and it's amazing, you start with a design in your tool of choice, and it handles almost everything from creating your app UI, writing handlers, building a backend, scaffolding, deploying at scale, etc., I would highly recommend it! It automates a lot of the stuff I hated doing when I first started out doing full stack work.

@Nanos that's why you gotta teach her to code. Once she feels the bliss of finally getting a compilation after hours of chasing nondescript bugs that appear exclusively in deleted forum threads from 2003 and GitHub issues with 200 likes and comments but no solution or update from the maintainer, she'll never be able to get a similar high again. She'll be addicted, unable to want anymore.

Unless you've got cocaine, that's even better.

@Root HODL cult! Hold On for Dear Life, there are lots of peeps pushing a philosophy where you never sell and wait for the mythical 20,000$ peak

@Root yeah definitely even if the overall trend is down it's still profitable for short term trends.

But then there's the HODL cult. It's like a joke that went too far

⬛⬛⬛⬛⬛ <-- mobile development
◽ <-- mobile developers

^^ top 20 anime tragedies

AllenNLP has a brilliant coref model that works really well for this problem

@Stuxnet they've got features people would pay for anyways (seeing who already liked you, boosting your profile, sending direct messages, changing your age/location, etc)

Paying for extra features is fine but hiding the main function of a free app that already has tons of ads behind a paywall is just going too far (and tinder and bumble actually do this less than other dating apps which are just a scam to begin with). Plus they could raise a high valuation off of just having lots of users (TanTan strategy).

@Stuxnet getting matches is definitely on the user but I'm talking about withholding matches/hiding function behind a paywall.

@RocketSurgeon they've got a lot of tools to prey on desperation, mostly for males since they make up a bigger portion of the desperate demographic:

- withholding profiles you would match with and showing them as "3+ likes" to try to convince you to pay for premium
- not showing your profile to people you might match with to try to get you to buy boost (which reverses this and gets you matches as such)
- telling you that you have a new like (fake & a lie) to try to bait you into opening the app
- automatically decreasing the match rate after your first day of having an account (you've been "retained")

Etc etc, you can find a lot of shit by inspecting the network tab in chrome devtools

@ewpratten did they change the rules? It used to be if you even had a hotspot going, an FTA would show up with a radar and hunt you down 😫

I would PAY for Vue native. React native is a pain in the ass, hot reloading and live reloading never work, I have to fucking shake the phone and break my wrist every time I want to reload, the package randomly stops working, there is no cache (before anyone comments about AsyncStorage, fuck off and take a look at their GH issues, I ran tests and it doesn't work 99% of the time).

I use react native only because writing 2 native apps is even more cancer, because apple and Google cant agree on a standard for mobile because at some point they contracted HIV and it has hindered their decision making abilities. That being said, Facebook has herpes instead, and that's what you sign up for when you use RN.

If you're building in RN, I suggest building your UI in a tool like BuilderX; also, try not to leave expo, and try not to install any community modules because none of them are reliable and the installation is a PITA. Lastly, skip react, because RN has its own patterns 😣😣

#fucczucc

@ewpratten lol, I refused to let anyone else upload code to the robot, and then wrote an enormous sdk for our team so they interfaced directly through my "childproof" layers.

We had GitHub but as we got closer to comepti5ion it was more crunch time, so messenger and zip files to share! During comeptition I once copied-by-hand 50 lines because it was going to be faster than going outside, setting up a hotspot, connecting both computers to that hotspot, doing a push and pull, and then going back inside. FIRST has ridiculous rules -- you'd think by now they would have realized it's important to make reliable hardware, especially for comms that apparently goes down if my phone hotspot is on nearby.

^^ this is true but also exploitable

Instead of spamming, delete your account and reopen it whenever you feel you stopped getting matches. Then only swipe right on rly attractive people who are likely to have high internal ratings. Either you can enjoy a day or two of being visible, or you get out of ELO hell and have a high rating and stay visible forever.

Tested it, definitely works!

@irene /devrantgold

@wholl0p I wonder if my favorite bands are hiring a software engineer 😂 my wallet will kill me if I cash out on these

You should get those 3d printed glasses that were built to be adversarial to inception/vgg19/etc! Don't just settle for not becoming part of their dataset, pollute it too by making it think you're a banana!

@irene recommend pls!! I have buds because I have to move around a lot but I'd gladly switch to collar ones

@wholl0p bruh I could buy concert tickets to see her live for the price of qc35's!

Also I realize headphones is a bit vague in this thread, I'm talking specifically about earbud headphones, the kind for phones and such. Where I am it's a colloquialism (we just call over-the-head ones and collar ones "headsets") even though it's probably wrong 😂

A brothel but recommendations based on AI, data driven services (eg determine what your clients will want depending on the scenario), AR integrated (Wear glasses and watch your partner transform into your fantasy), Vr integrated (option to record everything from your POV and have it turned into VR so you can experience it over and over), IoT toys, and blockchain (smart contracts for prostitutes, flag abusive clients with a vote system, verify prostitute using their crypto address), maybe some integrations with the Hub too, companion app that's like tinder but for escorts

Save $ by using robots 😣

(/s pls don't actually make this)

If youre referring to concatenating integers to strings, isn't this valid in every language? If I add an integer to a string I expect it to be appended to the end, if it was added to the number in the string that would be unexpected behaviour.

Like sure js sucks at a lot of things, just not this

When you say beard, are we talking Eminem, hipster from SF, or gandalf x Dumbledore rule 34?

@rEaL-jAsE @CodeMasterAlex nope. The id I wanted was "toast" but it kept selecting the nearest divs (in this case "value-prop" and one with no id)

The even more annoying thing is that the code worked when executed from console, just not from the script

If the code and/or datasets used to train and test are not publicly available then I just ignore it.

Research that can't be reproduced isn't research.

I use WSL for everything except GPU tasks. Come take me

@pyaf are you using the TF object detection API? How fast is your inference looking? I'm getting crazy slow results on GPU and it's driving me nuts

"Are you a gated recurrent unit because I'm going to go through the front and then you can open the gates to the back"

"Can I come over and show you my new loss function? I call it hardmax"

"Can I plug my discriminator into your generator?"

"Let me put my inputs in your network and then you can multiply"

"I may be the furthest node on the decision tree but I'll xgboost your confidence"

"I've got the kmeans to make you scream"

"I'm like state of the art object detection in bed; single shot and faster (rcnn)"

Most people are just looking for validation. Don't let it get to you, it's just adversarial inout that will decrease your confidence.

That being said this is a missed opportunity for ML pickup lines:

"Come over and see if you can fit on my big data"

"Are you a low frequency dataset because my heart is overfitting on you"

"You can't be too loud or the k nearest neighbours will wake up"

"Let's start with kissing and run gradient descent down from there 😏"

"Matching me was a mistake but you'll need to make some more to cut your losses"

"Let's meet up and watch inception, and then maybe I can put my Common Object in your Context"

"Are you a sequence to sequence model because I'd love to embed my 4 letter word in you"

"Teach me how to build an LSTM so I can find the timesteps to win your heart"