I wonder why everyone seems to get what this does from reading the description. When I read "eat and shit", I assume that the "food" is "transformed" into ... youtube videos. So I thought this would generate new videos and upload them to youtube automatically, maybe showing the txt files as a huge StarWars scroller, or something.
This is also exactly what I thougth (with the starwars scroller and all!), but the website is pretty self explainatory once you visited it and cliqued "enter txt" link.
Yeah, I had uploaded the first 10 chapters of Moby Dick, all along thinking "what is this annoying video playing," before realizing it doesn't do a StarWars scroller.
For this particular link I'm not sure I would have clicked it with the text you suggest. And I'm glad I did click it. But I agree that in the general case, titles on HN should be accurate rather than catchy.
WOW! I had exactly this idea at a Hackaton festival called Codebits in Portugal in November last year and presented it[1] to the audience but didn't receive that much excitement from them about my implementation done in 12 hours using JS and Rails3.
Now I see I should have stick to it. Deep inside me I knew if the service would be useful to me, it would be to some other folks, but I was a bit disappointed and didn't correct the bugs that the project had.
It is buggy, sometimes doesn't change to the next video and it lacks the controls and the display of the playlist as this site does and it is a lot slower. It also lacks the upload of a text file. But I had other ideas in mind, the mode implemented was supposed to be the "feeling lucky" mode, where you just typed each video one per line and hoped it would match to the video you wanted, but you'd also have a way to create playlist more carefully by passing the exact urls of the videos you wanted, like this site does.
How do you overcome this feeling when you know you should have sticked to your guts?
Anyway, good work guys! This is exactly what I wanted.
Never feel like "they got there first" - when someone does a similar idea to your own, consider it validation that you had a good idea!
If it's something you really love, and feel your implementation works in a way you prefer to the other(s), just keep building it.
There's room for more than one version of everything. What if the iPhone was the only smartphone, or Ivory was the only soap?
If it's not too near and dear to your heart, and you like what the other guy has done, just go ahead and use their version and have fun with it, and move on to your next project idea, with the confidence of knowing that your ideas are good, and you should keep believing in them and making more.
Here's an idea of mine for free (I came up with this as I was reading your comment, so feel free to critique it):
Instead of uploading a text file, a user can type in an artist and song title on their smartphone. This gets added to a playlist that's displaying on a large monitor/TV/projector screen. This could pick songs from YouTube/iTunes/Winamp etc.
Uses:
- Nightclubs. People can offer track selections for the DJ. These are displayed on the wall. Or maybe there is no DJ at all. Users can vote on their track choice that are coming up next on the screen, as well as offering their own track pick - like HN voting/article submissions. (This would be the geekiest nightclub of all time.)
Once I thought of something like that too, but in the environment of a park with speakers and wifi (yeah, far fetched, I know). People could go to a website available on the wireless network with their smartphones and pick a track to play (from a library such as Spotify). Then everybody on the park would be able to up or down the track using their smartphones. If the track had enough up votes, that person would be able to pick another track to play.
It would be the interactive/social/sound park. :P
But since I don't know that many parks with speakers, on clubs it would probably work better. They could do it as an experiment on a tent of some festival like they did with the concept of the silent disco (http://en.wikipedia.org/wiki/Silent_disco).
I am totally into this idea and I'm looking right now (experienced javascript, python, java + MBA). If anyone is serious about doing this, let me know. I can code and raise cash.
I was thinking of building something like this, too.
I was going to start in cafes, and monetize by charging the cafe for the service. The patrons get the jukebox-ey service for free, and the cafe pays a small monthly subscription.
it needs a serious development time (and probably investment cause you will need hardware, space and time) but i think this could be the future of public music.
Regarding the _USEFULNESS_ of this: low. Regarding the entertainment factor: pretty high. I could use a m3u-> txt parser to dump a huge load of youtube videos and make a "music video playlist" -- kind of cool.
That is a good idea! Like Google advertising, where the ads are selected based on the textual context -- words in your search string or email or website -- but instead of ads as the output, short video clips. For long text blocks, run the videos alongside a scrolling reader such that the video and text stay synchronized. The word "respect" prompts an Aretha Franklin clip, maybe. Appeal to the [people who think they are] multitaskers! You'll have nerd users searching for which Shakespearean passages lead to the most amusing clip playlist!
If the video bandwidth is all Youtube's, and the UI is Javascript, your ongoing cost is mostly the server time to run the text against your corpus and spit out a playlist.
I love when startup ideas spring from a misinterpretation of how someone else summarizes their product.
I second this. It's crass and unnecessary and just makes me look at it.. i dunno... differently.
And because of that -- and only that -- I would never see myself sharing this with people or using it to send video playlists to people aside from some anonymous web forum.
Seriously, though, I'm not one to be easily offended -- and I think fake umbrage in our society has been taken to a ridiculous level (especially in politics). But the reason I like good design is because when something is well designed it gives me a sense of calm and peace. Having "I eat text and shit out whatever" is just very bad design to me.
And it's offensive enough that why would I risk offending somebody? Why do I care enough about the site to take on that risk?
I'm not one to be easily offended either, but I am also less likely to share this with my 12 year old daughter who really likes to explore youtube. Of course all she plays is that bieber kid, some kid named cody and lady gaga.
Do you think this would also work with Grooveshark? The problem with YouTube is that most music videos are geo-blocked outside the US, or at least for my country they are. (So are Pandora, Spotify, etc.) Furthermore I believe that Grooveshark's song library is bigger anyway, so it would be an improvement for everyone.
I really like this, I had a similar idea a few years ago with 'YouLink' - essentially solving the playlist problem of YouTube - but I didn't think of making such a simple and clean execution. Mine was more Youtube meets iTunes style - pre Spotify days.
Plus back then, bandwidth was not so great / YouTube often buffered causing breaks in the music. So I abandoned the project.
However Roll.io is a great solution, often I have the situation where I have a track listing for something that I can find individual songs for on youtube, but need to quickly make a playlist out of it, but don't due to the horrible youtube playlist interface.
This is especially true when it comes to 'Remixed' music where the songs have particular unique names. e.g. http://roll.io/#4fvjsj!0
Thanks for making it. The only request I would have would be to automatically remove or recognise track listing numbers at the start of the tracks.
E.g. if I copy a track listing I often have:
1. SongA
2. SongB
3. SongC
4. SongD
- It is a slight pain to manually remove the 1., 2., 3., 4. from the front and may cause incorrect matching on the songs if I leave them on.
Mobile version please!!! (for my flash enabled android phone)
For some reason the page is very slow to interact with and type on while using my Motorola Atrix (supposedly a fast device). Ideally, the home page would have just a textarea and no video playing by default.
Feature request: show the current playlist entry in the page title. I have your site running in a background tab and I want to quickly glance up and see the song name without switching tabs.
This is a great idea and your elevator pitch on hacker news is brilliant, except you should have said "text" rather than txt.
I assumed you meant you took a .txt file of text and extracted any and all videos from it.
"Show HN: roll.io eats text and shits relevant youtube videos. Useful?"
The proposition is so obvious you should just spell it out on your home page, like, have a text box with some pre-populated text and a big arrow that says "becomes" and then a playlist of youtube videos.
Holy shit, this is amazing. I'm really into electronic music, and when I am listening to new music online (youtube, music blogs, etc.) I don't always take the time to download songs right then. Instead, I simply record the artist and title in a txt file. I uploaded my electro.txt file to roll.io and out came a perfect playlist for all the music (~500 songs) I liked, but haven't downloaded.
A similar idea to the linked text file (not to roll.io itself) is http://www.tuneset.com/ (disclosure: made by a friend of mine).
roll.io is awesome though. Just making and sharing lists of talks broken up in several videos, things like this.. I wanted this exact thing just yesterday without knowing it.
Very useful. I always wished for something like this.
I was a little disappointed to see that it didn't work on the iPad, though. The page just tells me (in grey text on a black background btw) that the flash player is needed, although embedded YouTube videos usually work without the need for flash on this thing.
It's most useful if a) Always (settable option?) HD videos, b) uses less cpu time/battery/memory than playing the video (when hiding the video) c) youtube playlists.
Mostly b. Even on an imac youtube can take 10% of cpu I'd rather be used for something else, like a lagging game.
I wasn't sure about this when I read your tag line here but I have to say that is pretty nifty. I think its the sheer simplicity that makes it a pleasure to use, typing a bunch of song titles seems to actually be quite a natural way to get a playlist together.
Great idea and very promising application. I've exported my favourite tracks via foobar2000 and last.fm statistics via softplaylist plugin (it creates XPSF playlist) et voila! my videoplaylist was created en instant.
Actually, what would be very cool is a chrome plugin that allows you to play the songs listed on any wiki album page, with a click of a button. That would be incredibly cool.
This is a great idea, and amazing for previewing new albums! However, sometimes when it fails to search and substitutes, it would be nice to be able to hand edit the playlist.
used this last night on a whim to paste in the contents of a tracklist from a radio show. Great way to flip through some of the tracks the djs played without having the actual show recording on hand. Handy for those into electronic/dance stuff where there's lots of livesets and tracklists floating around. I'd say it was 80% accurate in picking up the right track from youtube. Amazing how much stuff is on youtube of course.
i'd like to see a list of most common songs. which is probably not simple to do per se, based on how this probably works, but i think it'd be neat to see, at least
not what i was expecting (i thought it'd be some sort of programmatic generation of presentations that were then converted to video) but a very neat idea indeed.
just thought of something else. It would be quite cool if we could direct roll.io to play files via url parameter. Say we want to play 3 songs. We could type something like this in a forum:
very nice idea! would be nice if we could type out playlist in textboxes and then open them via roll.io so that those boxes are easy to emded in forums when someone wants to list out a few viddys without going to youtube to expressedly look & link for them.
