41–50 of 160 posts
I think it's most likely done with the API. From what I remember from looking at it a while ago you can create the entry before, then it'll give you the right endpoint where you post your actual file to later on in the process. You just have to do that quickly enough so the endpoint doesn't expire.
It's apparently even been done before, see here at around the three minute mark: https://youtube.com/watch?v=UMYTSJTU0Vk
Maybe 1) start upload till URL appears 2) go offline 3) replace file 4) go online
1) Make an educated guess about what the URL will be and put it in the video.
2) Upload.
3) If you were wrong, delete it and go to 1).
I'm half joking, but that's basically what they did for a git utility that would let you edit histories so that messages could refer (by partial hash) to later commits. I forget the name, though, and couldn't find it on a quick search.
Done more than once before, this one even has an explanation, which matches some of the guesses people already posted I think https://www.youtube.com/watch?v=ufq2Eb78kSU
1. create a resumable session (here we specify video length in bytes)
2. YouTube returns a session URI which contains the upload ID
3. create a video with the upload ID baked in (edit video to match length from step 1)
4. upload video to the session URI
https://developers.google.com/youtube/v3/guides/using_resuma...
Earlier quoted context omitted.
Start uploading a video, pause the upload, you could already get the ID of the vid, render rest of the video and then continue the upload that now contains the ID of the video :D
How to pause? and re upload pending part with new video?
I tried it with a named pipe ("mknod /tmp/foo.mp4 p") and it gave me an error, but before the error, the web page blocked during upload, but it showed me the URL that will be used (before I gave it a single byte of data).
Earlier quoted context omitted.
Reminds me of the brilliant New York Times crossword which "predicted" the winner of the US election in 1996: http://www.alaricstephen.com/main-featured/2017/7/3/the-clin...
This page seems to have zero text or image content unless I let SquareSpace run Javascript (and ostensibly track me).
Interestingly, there's a frame or two at the very beginning that shows a document open that could suggest an exploit of some kind. Screenshot: https://i.imgur.com/3oTXTDK.png
I believe that’s just part of the joke. It’s a mock email from the future, telling him what the URL will be and how to make the video.
The document is covering the sign at the start of the video, so the uploader could have uploaded that section (without any text on the sign in-game), then "resumed" the upload with a different video that contains the URL on the sign and starts after he switches to the Minecraft window.
(And as we all know Twitter has no edit button, no post-upload shenanigans are possible here, just some clever guessing and reverse-engineering Twitter's ID generation algorithm)
Earlier quoted context omitted.
Reminds me of the brilliant New York Times crossword which "predicted" the winner of the US election in 1996: http://www.alaricstephen.com/main-featured/2017/7/3/the-clin...
This page seems to have zero text or image content unless I let SquareSpace run Javascript (and ostensibly track me).
I imagine that website is doing something that I wouldn't approve of.
---
It's actually a bit of a long story, but it's kind of an interesting part of reddit's history. Some of my details are probably off, but this is how I understand it.
A long, long time ago, reddit only supported submitting links. Link submissions were pretty much exactly the same as they still are now, including having a comments page that you could go to with an address like the page we're currently on: http://www.reddit.com/r/help/comments/2bmy3l/what_does_the_s...
If you look at that address, one thing to notice is that it includes the submission's ID in base-36, which for this post is "2bmy3l". If you go to http://www.reddit.com/r/all/new and look at the links to the comments pages of the newest posts, you'll notice that their IDs are increasing. For example, at the time I'm writing this, the newest IDs are 2bo3uw, 2bo3ux, 2bo3uy.
Since the IDs are increasing, you can predict which ones are coming up. You know that the next ID after those ones should be 2bo3uz. So some person decided it would be funny if they pre-constructed a link like this: /this_post_will_link_to_itself/" rel="nofollow">http://www.reddit.com/r/reddit.com/comments//this_post_w... and then checked which IDs were coming up, filled in the "" spot with one that should be used soon, and submitted it.
If they got their timing right, they'd end up with a link post that actually went nowhere. Clicking it would just take you to its own comments page, since they had managed to predict the url that the comments page would have. This was a "self post", a post that linked to itself.
So this was a pretty neat trick, and when it was successfully pulled off for the first time, it got a whole bunch of attention. Unfortunately, reddit's never been very good at just seeing some new popular thing as a novelty and moving on. No, of course pretty much everybody wanted to get their own "self post". The majority of the new submissions to the site were suddenly just people trying to make a self post, completely drowning out all the real submissions.
It was causing a gigantic mess, so one of the admins at the time decided to get people to stop by taking all the fun out of the game. They made it so that you could just choose to make a self post. You didn't have to guess the ID or anything, you just selected that you wanted to make a self post and automatically got one. It didn't have the option to add additional text or anything yet (that came later), it was just a title that linked to its own comments page.
So that's how self posts came about, and where the name originated. A quick solution to a mess being made by users that eventually turned into one of the most important pieces of reddit.