Home > All Posts > Individual Post
Post #1494

Re: [videoblogging] vogfeed

By kenyatta cheese | kenyatta cheese <minitrue@...> | yattacheese
October 25, 2004 | Post #1494 | Topic #1469

Hey there. The vogfeed is just a reblog. It pulls RSS feeds from known videobloggers and republishes a single RSS feed with enclosure tags for each entry with a link to a media file. To handle the video file selection issue, I've come up with some highly effective code: "Hey, Josh, Jay... I need to look through these feeds and only select posts that have self-authored media files. Would you mind looking through them with me?" It's written in a highly inefficient language, I know, but it's worked so far. Andreas is right -- without sufficient metadata, it's nearly impossible to sort through the video automagically. It's a problem that can't be solved without setting up a format for media file metadata (XMLTV, anyone?) and building the tools for embedding that data in a file in a way that's readable by the client app or a server-side blog plugin. So I decided to not Solve The Problem in exchange for Getting Something Done. Vogfeed provides Josh with a second semi-regularly updated feed of posts with media enclosures with the help of all of you. (Demandmedia is the first.) He continues work on the application and we get to continue arguing about the proper and improper ways that people use their media. ;) -kc. > > On Fri, 22 Oct 2004 19:17:06 -0400, Jay dedman <jay.dedman@...> > wrote: > >> now, i dont know how he coded it...but as auser i see how it works. >> all you coders...check out the code. > > There are some logical problems with getting movies from any HTML page > (I've talked about these too many times to recount now), and I simply > don't believe that it can be done in a consistent and reliable manner > unless the videobloggers change the way they link to their own video files. > > That script may work today, for the entry you're adding today, but I > highly doubt that it's a solution that works for HTML documents in > general. And *that* is what we need. I can write something that grabs any > video link from a website in two minutes, but that's useless - we need to > get the *right* videos. We simply can't make that distinction today with > the way people are linking to their videos today. That's why I've been > talking about this rel="alternate" for so long. That's why I've been > trying to explain the different kinds of links. > > I can't check out the code used in the script. It gets executed on the > server so none of it reaches my browser. > > - Andreas