Tinder-like application for gifs built with Elm | Hacker News Reader

Tinder-like application for gifs built with Elm

github.com

61 pointsmatthieu-b44 comments

You might want to add a demo / features page, that allows you to try the thing out, look at the main features, without having to login with Facebook.

Because I can tell you this now: No one in this crowd (tech/geek) will login with Facebook.

Exactly my first thought, I immediately bailed at the FaceBook login. Perhaps accounts outside of Facebook?

I quickly created a signup/login form, you should be able to try now

>Gipher has the main features we can find in any modern single page application:

>facebook authentication

What? Authentication with Facebook is a "main feature" of modern single page apps?

I'm not on Facebook and so there are all sorts of things I cant use that require auth with Facebook. This little app does not seem like something that needs Facebook auth to work.

Not the creator, but at a guess (from seeing a lot of apps that do this): the app doesn't strictly require Facebook in particular—but requiring a Facebook account is a popular and simple proxy-method of deduplicating your users so that people can't vote multiple times. Basically, it's a very cheap KYC solution, on about the same level as domain-verification for TLS certs.

I've been thinking for a while of creating a service that's specifically and only for this—an OAuth provider that gets your personal details the way Facebook does and enforces one-account-per-person, but then doesn't actually transmit your identiity to the site using it for OAuth. You, as an app developer, get to know each of your accounts represents a unique human, but don't get to know anything else about those humans.

Alternately, the OAuth provider could be set up so that each human can have multiple pseudonymous accounts on your service (picking at OAuth login time which one they want to use), allowing for creation of cheap "throwaway accounts"—while still giving the site the ability to ban an abusive human altogether, without knowing enough to link the accounts together. To the site, it'd simply look like they ban one account (as an OAuth provider API call), and then the other accounts just never sign on again.

The best way to prevent rigged voting is not to let the users select what's being voted on out of a large pool. Polling is better!

Clef does this. I implemented it in an app as a proof of concept.

Why the FB login?

I hit Command-w in the second I saw the FB login without any demo or feature section. Good luck with your bounce rate.

You can re-open a tab now an try it out, I just added a username/password login

I don't see any problem with required a Facebook login for this. It's supposed to be a toy "Tinder like" side project, and Tinder requires a Facebook login.

It also only requires access to your public profile information.

Great work, would love to try this out but I'm not going to link my Facebook to a random app.

Facebook will tell you what the app has access to before you give it permission. I tried it, and it only has access to your public profile information, which is already publicly available anyway.

Yea, but before there was nothing easy to tie this app and the public profile. Now there's no choice but to tie the two.

What are you attempting to avoid? What's the risk that you're trying to mitigate? Were you planning on signing up with an anonymous email?

As a non-facebook user, I would enjoy just being able to try the service. Email I have so that would be enough for me and I guess a lot of other people as well.

I wasn't planning on signing up at all—to my knowledge, nothing about this single-page-app requires a session at all.

The risk here is that I don't want the app creator to know my name. It's that simple—I don't know them, I don't trust them, I trust them less now that they ask for completely unnecessary information. Let's say traffic to this site is linked with lower credit scores: it's naturally absurd that the two are causally related, but it's a reality that I have zero control over how my traffic data relates to how companies use it. All I can do is reduce my traffic data from getting into their hands where I can help it.

>nothing about this single-page-app requires a session at all.

The app saves you the gifs you like/disliked that requires a session. Now you could say the app doesn't have to save that information, but then I'd say it's a toy app so it doesn't have to do anything. The creator wanted it to save your history, therefore it requires a session.

Let's say it requires a regular email login instead of facebook (in order to support password reset). If you're using an anonymous email that you've been careful never to accidentally associate with your name, then yeah you could keep the app creator from knowing your name.

If like most people you just use your regular email address, it's trivial to get your name with a quick google search.

>Let's say traffic to this site is linked with lower credit scores

If you're really worried about that, then you should probably have an entire fake virtual identity to handle things like this.

> then you should probably have an entire fake virtual identity to handle things like this.

I don't believe facebook allows this. Hence the problem with a facebook-only login.

I hope you realize the irony of "protecting your identity" when you link your Github on HN (that has your real name) and which, in turn, links to your resume, portfolio, personal website, and (presumably) private email address.

Stop being a dick.

My email address isn't private, and I am OK with these comments being linked to my real identity. But—and this is important—I elected to do that.

It's far more of a dick to force your users to reveal their identity.

EDIT: to be clear, OP is not a dick, but there should be a non-session or non-sso signin option (as there is now) if I want to preserve anonymity for what is essentially a (polished) toy.

Why would anyone want to trust their precious personal information including real name, photograph, and friend list to some random guy from the internet? Would you publish it on reddit? On NH? On shady anonymous forums?

All of that information is already public. Anyone can get it from facebook (not just your friends), why would you care if already public information gets reposted?

Also the app didn't request access to your friends list.

How would the app know which profile to look at?

If the app requires an email login, they could search for the email and for the vast majority of people it would return enough information to locate their facebook profile.

That's beside the point though. You're worried that a random person on the internet has access to publicly available information that the entire internet already has access to.

The only extra information you're giving out to the owner of the site by logging in with facebook is that you are a user of this site. The worse thing the owner can do is publish that fact.

That is the only foreseeable risk. I can't imagine enough people are concerned with that risk to make it worth most developer's time to address.

you can try it out, I just added a username/password login

"Login with Facebook"

No thanks.

#1 huge fan of Elm -- was this build with start-app??

#2 i used the app and noticed you can collect data to build discriminant classifier. Even simpler - a Naive Bayes classifier since outcome is binary - Like / Dislike

#3 (continuation of #2) how many users do you have? to estimate data

No clue about elm - could you please share why it is good?

Wow, there are so many JS libraries today - even just making a list is a big task in itself :(

Elms isn't a JS library, it's a completely separate language that happens to compile to JS.

Well, good is a subjective term, but this demo of its debugger [1] certainly peaked my interest :) I even has live examples with hot-swapable code (that is opensource if you'd like touse it.)

But it is not a library. So if you would want a list of pros/cons you'd need to compare it to other compile-to-js languages.

Its purpose seems to be creating reactive web-based UI. Basic building block is converting event-stream of user input into an event-stream of html, simplest example probably were [2]

Ok, it might be web-gl as well :) And it is type-checked if that is your jam, but has decent type inference if you don't like writing types explicitly.

[1] http://elm-lang.org/blog/time-travel-made-easy [2] http://elm-lang.org/examples/mouse-position

Sounds like a fun little side project, would appreciate a screenshot (or a gif of how it works) in your readme

You should allow people to delete their profile.

http://itsthisforthat.com/

It's like tinder for gifs

wow, literally every comment is about facebook login

seriously guys come on

I really like this! It seems your "mobile first" design is "mobile only", you might want to make gifs show up as their actual size on bigger screens, the square chops them off at the sides sometimes

Are you going to use the likes to show me personalised gifs?

I would love to give feedback but I can't even try the application, maybe that's why the comment section here is filled with comments related to Facebook... I'm not the only one who wants to try it out.

Apparently the guy didn't want to go through the hassle of implementing a login system and the dependencies that involves (ability to send password reset emails mainly).

If I were you, I'd make a fake facebook profile just for situations like this.

You should be able to try it out now!

Thanks for the feedbacks, just added a login with username/password

Interface would probably be nice for a Reddit client.

looks good! huge fan of elm. might want to fix the flickering in the gifs though. they appear to load twice.

Thanks! Indeed there are a lot of things to fix/improve! It was only supposed to be a simple project to build something with elm, but it looks like people actually like it. I'll try to fix all these things really soon!

page is empty (Iceweasel, Linux)

Page is also empty for me, vanilla Firefox on Linux

so...imgur?

Reply on news.ycombinator.com