This is a great idea.. and something I've been wanting to do for a while! The reason I didn't do it was because of a competitor that pretty nailed it. I've been trying to find the website.. it's done in node.js with "robots". Basically, it takes care of uploading your file and then you tell robots what to do with it (Compress it? Store it on Amazon, etc.)
Anyhow, great job doing it. Depending of the pricing, I'd surely use it.
I've seen that you make your widget code open source. I guess this handle the flash hack / html5 version for all different versions of browser? Hopefully so because that's the most annoying part of getting the ajax upload working. But then, if that's free, I'd probably take these widgets and just manually code the amazon part.. The only reason I'd use the service is if the pricing is a very little more than what amazon would cost me.
However, that's just for me. Other startups or bigger company might be willing to pay the big price it can save them hours of coding.
And, to everyone you've never done the ajax upload.. trust me that it's a pain. So many browsers and hack.. compressing the images, storing them, limiting the size, etc etc etc. It seems like a trivial task and the first thing you know is that it's been a week you're working on that and then, once you think everything is done, you receive weird message like it doesn't work on IE9 for some reasons, or whatever.
The mind boggles. I know you said that tongue-in-cheek, and that would be an incredibly cool project, but i feel that unless we put this to rest now someone might actually take you seriously.
A huge amount of work has gone into making FFMPEG and x264 as fast as they are today. Not only would you lose performance (and integer types) going from C -> JavaScript, but you also lose all the assembly, handwritten for tens of different architectures (including SIMD like SSE). I would expect at least a 100x slowdown.
Yes, in case it wasn't clear, Fabrice Bellard is the author of both FFMPEG and jslinux, and my comment was referring to his outrageous productivity/ability. As great as V8 may be, it isn't a great target for an ffmpeg port.
We're not porting ffmpeg to JS, but my startup is working feverishly on the "I made this video and now I have to upload the whole damn thing.. ugh!" problem. Stay tuned.
Wrote a small service in Haskell to handle large uploads in preparation for a new feature. Does interleaved IO and streams the uploaded file directly to Cloudfiles while also saving a copy on the local disk to be able to extract some metadata and add them to our DB at the end.
Also did play with the idea of creating a general upload service based on this implementation, but for now I'm focusing on other things.
This is a great design and idea. I'm always perplexed about trusting a "CMS that will make your life way easier". But, this one seems well done.
Let's say I use html/css from scratch; or integrate some coffeescript in it.. How clients are supposed to edit the text? I.e How the intuitive widgets works to change stuff?
You'll work with a template language to define what parts of the page the client should be able to edit.
Webpop lets you work with structured content, pull the content into your templates and define exactly the HTML used to present the content.
The system then automatically detects what part of your site is dynamic content that the client should be able to edit and what's design. The client then get a simple on-site editor for updating his contents.
Your site was nicely readable until the custom font got loaded. It is tiny and a colorful blurry mess (subpixel) on my monitor. The text right above the bottom image is almost unreadable.
I just spent the last week implementing something very similar for a client project...most of the problems in that list of yours already have solutions... Nice idea to try and productize them, though. In the case of my client your product could have saved them a weeks worth of paying me so there is considerable value if you can make it super duper easy to integrate
This looks very cool. I myself am working on a video service wherein users upload large videos.
For my particular application, I'm looking for something which can handle large files (500MB - 2GB), preferably giving file upload progress and being able to resume broken uploads. Could you talk more about those features?
Also, it seems like we would upload files to your servers, and then to integrate them with an app, I would then use your API to bring them over to my servers or storage. Are there plans to make the source of the uploading widgets/mechanisms themselves (not necessarily open-source) so that we can host our uploads ourselves?
Ooh, and how about something that is mountable?
Anyway, those are just a few of the things that immediately crossed my mind in terms of use-cases for my current project. I can see this being very nifty!
To handle videos, I'd recommend you to check out http://transloadit.com/ . They are something like our competitors, but I believe we have different targeted audience.
We've been thinking about splitting files in parts and everything, but not right now.
There's an aspect to "the uploading problem" you didn't name, which is the "uploading services rapidly turn into hosts for pirate and illegal content, and even if that doesn't bother me ethically it bothers me when I get the bills" problem. "Storing on S3 infinitely" only scales as far as your pocketbook.
(Also, by "illegal content", think "Feds literally busting down your door and hauling you away" sort of illegal content, and by "you", I mean, you, va1en0k. This is not necessarily a reason to stop right now, but don't get too far down this path without legal consultation about how you should protect yourself from this eventuality. There are things you can and should do, like make sure you qualify for DMCA safe harbor provisions.)
This sounds like a great idea. Sure there may be competition, but that isn't a reason not to do it.
You may want to take a look at how Twilio handles in-browser audio recording (even though they offer something different). Basically, you insert some JS into your page, then the audio is streamed to Twilio, and then Twilio does a callback to an endpoint you specify (with meta data, the URL to the file etc etc). What is also great about this is that I can serve the files right of their servers and not have to worry about it as a developer.
That's how you detect hipster articles ! More seriously, as a php developer all this told me was that I shouldn't use his product, because someone who thinks file uploading in php poses more of a threat than in other language probably doesn't know php much, yet he still finds himself competent enough to jab at it.
Why as a "php developer" do you restrict yourself to one language? Why not just be a "developer" who uses the right tool for the job? Then you won't have to use PHP all the time. And your feeling won't get hurt when people point out how horrible php is. That's just an objective fact. Don't assume that because he hates PHP he doesn't know it. When you learn other languages, you will find out that you have to know something well to truly hate it properly.
Who said I restricted myself ? Is a python (or whatever suit your boat) developer restricting himself to python ? I said of myself that I am a php developer because I actively use it on a weekly basis and it helps me make a fair amount of money on the side while others argue about which is worse between php and the devil. I also use quite a few other languages, mainly at my job.
> Why not just be a "developer" who uses the right tool for the job?
That's exactly what I do, and there are a few cases where php is the right tool. When someone tries to use php in other situations and complains that it sucks at it, it just shows how poorly he chose his tool.
> And your feeling won't get hurt when people point out how horrible php is
Where have you read anything like that ? My feeling aren't hurt at all, I'm just saying his remark is stupid, plain and simple. File uploading in php is no more (or no less) of a threat than in other languages, you just have to know what you are doing, that's what I posted about, not some "php is horrible" that you imagined reading.
> When you learn other languages, you will find out that you have to know something well to truly hate it properly.
That is exactly my point that you missed, given his remark on file uploading (the only thing he mentioned and that I commented on), he doesn't know php well. Thus he should abstain from trying to take a shot at it in a product description page, especially since - given php popularity - a lot of his potential users are using it.
Anyhow, great job doing it. Depending of the pricing, I'd surely use it.
I've seen that you make your widget code open source. I guess this handle the flash hack / html5 version for all different versions of browser? Hopefully so because that's the most annoying part of getting the ajax upload working. But then, if that's free, I'd probably take these widgets and just manually code the amazon part.. The only reason I'd use the service is if the pricing is a very little more than what amazon would cost me.
However, that's just for me. Other startups or bigger company might be willing to pay the big price it can save them hours of coding.
And, to everyone you've never done the ajax upload.. trust me that it's a pain. So many browsers and hack.. compressing the images, storing them, limiting the size, etc etc etc. It seems like a trivial task and the first thing you know is that it's been a week you're working on that and then, once you think everything is done, you receive weird message like it doesn't work on IE9 for some reasons, or whatever.