Why use gfycat?

[ddevault]

You should consider MediaCrush instead, which has a better API and is open source. It also does more than just GIFs, has albums, etc, it's a lot better.

[adamkdean]

Until recently, I hadn't even heard of MediaCrush. Gfycat on the other hand has been providing an excellent service recently, and is admired by many. As such, it was chosen. IIRC, MediaCrush already has a host of extensions/addons for browsers?

[ddevault]

It doesn't actually, just a Chrome/Firefox extension for rehosting images on the service. Nothing like this. MediaCrush has 1000 github stars, it's not entirely unheard of :P I'd support the superior and open source alternative (though I'm obviously biased). MediaCrush is quite a bit more polished and quite a bit nicer.

[adamkdean]

Does MediaCrush have the ability to prepend URLs with a domain in order to seamlessly replace them?

Example: http://i.imgur.com/test.gif -> http://mediacru.sh/gif/http://i.imgur.com/test.gif etc?

[ddevault]

It's not quite that seamless, but a browser extension should be able to do it easily.

sircmpwn@kyouko ~> curl -F "url=http://i.imgur.com/Ni18onN.gif" https://mediacru.sh/api/upload/url
{
  "hash": "GnMDN6gPkZ03"
}

https://mediacru.sh/GnMDN6gPkZ03

Just a quick POST. You can also hit Ctrl+V on the home page to upload a URL.

API docs: https://mediacru.sh/docs/API

[jdiez17]

Also, you can check if a URL has been uploaded to MediaCrush without actually uploading it:

POST /api/url/info
list=http://i.imgur.com/rctIj1M.jpg,http://does.not/exist.gif

{
  "http://does.not/exist.gif": null, 
  "http://i.imgur.com/rctIj1M.jpg": {
    "blob_type": "image", 
    "compression": 1.0, 
    "extras": [], 
    "files": [
      {
        "file": "/4Gt0YcGMPA7S.jpg", 
        "type": "image/jpeg"
      }
    ], 
    "flags": {},
    "hash": "4Gt0YcGMPA7S", 
    "original": "/4Gt0YcGMPA7S.jpg", 
    "type": "image/jpeg"
  }
}

[sergiotapia]

I think it's not as useful if it doesn't allow you just append the URL to the service.

Gfycat is massively useful because you just go: http://gfycat.com/fetch/[some-gif-url] and it works seamlessly. If the image already exists you are just taken to the page, if not, the service fetches it taking just a few seconds longer.

[ddevault]

There's really little difference. You have the URL on your clipboard, you go to https://mediacru.sh, and you push Ctrl+V. It works exactly the same, MediaCrush just offers a GUI instead of asking users to tap out http://gfycat.com/fetch/ first. There's also the browser extensions that let you right click a GIF and upload it immediately, and that beats everything else mentioned so far in terms of usability.

None of that matters, though, because we're talking about it from the perspective of a browser extension, which can do either method with pretty much exactly the same amount of effort.

[jdiez17]

There are two endpoints: one uploads a URL and one checks whether the URL had already been uploaded, but doesn't upload it.

Also, you should consider API stability. The gfycat API is notorious for being rather ad-hoc and put together just for the purposes of supporting the frontend, and they make no guarantees of stability. The MediaCrush API, however, takes backward compatibility very seriously and is specifically designed for ease of use.

This is very simple: compare http://gfycat.com/api and https://mediacru.sh/docs/api side by side and decide which one to use.

[sergiotapia]

@SirCmpwn but it doesn't beat it in terms of usability. You can't suggest pasting a URL is simpler than just clicking a link created by this extension.

[ddevault]

Like I said, from the perspective of a browser extension, it adds trivial complexity, and does so transparently to the user. The user experience here does not change, and the extension is made only slightly different than it is now.

[adamkdean]

But it adds a layer of complexity that is not actually required. Being able to scan through and prepend anchors with a URL is far easier than having any action occur upon clicking of a link. If mediacrush were to bring out a version that allowed that function, I would be interested, but otherwise, I do not honestly see the need to switch.

I may add an option to configure the URL to use another service when I work on the preferences.

[jdiez17]

@Imdsm - you can do that. I'm not sure if you're misunderstanding MediaCrush's API capabilities but give the documentation a chance: https://mediacru.sh/docs/api

[ddevault]

Seriously? It's less than 10 minutes of work. Do you want a pull request?

Maybe if you had any actual reasons to use gfycat instead of a clearly superior, open source alternative, it wouldn't be so absurd.

[sergiotapia]

@jdiez17 I'm not seeing where you can do something like:

http://www.mediacrush.com/fetch/[url]

[adamkdean]

Maybe the question here is whether there is a demand great enough to migrate this from gfycat to mediacrush?

I do not think an extension named AutoGfy should be based on MediaCrush. I think an extension named AutoMediaCrush should be based on MediaCrush. This extension provides a basic function and on top of that, there doesn't really need to be anything more substantial.

[sergiotapia]

LOL MediaCrush sucks!

Their video controls are terrible, the website loads much slower than Gfycat and it's harder to use (well, more steps to use). Don't be butthurt about an extension called AutoGfy not using a totally different service. Have a :pill: and calm down.

[adamkdean]

@SirCmpwn: The problem I have has migrated from the lack of accessibility the MediaCrush service has, to the utter rudeness of it's developer. Your service maybe be, as you say, "open source, clearly superior and pro-privacy", but the reason you're being ignored is because you're a narrow minded, rude, immature child who isn't ready to play with the grown ups yet.

For the sake of the archives, this is how you react to suggestions: http://pastebin.com/znQQGEHt.

[sergiotapia]

^ Literally first time in the years I've been here that I've seen someone act this way on Github. Act more professional.