> Yeah, I know. In regards to thumbnails you can say 'if you want your
> thumbnail on fireant you will have to use the image as a link to the
> video'. Then people who wants thumbnails will have to make image links,
> and you will hopefully get a slow adoption.

Yep, that's exactly what we've done. This is explained in our FAQ.
We also get some thumbnails from MeFeedia through their API, but that
is often less successful than parsing the HTML for images.

> If you guys ever want to share any research I'm sure it would be of huge
> value to the microformats community.

The main thing I've learned is that there is literally a cacophony of
markup styles, formats, and standards such that its very difficult to
really nail things down in practice, even when it comes to relatively
vanilla HTML or RSS.

The shoddy XML support in iTunes has definitely not helped matters
when it comes to RSS standardization. Also, there's are a number of
methods people use to try to track downloads with redirect scripts
which can botch things up in innumerable ways.

You're right though, once we get things settled in a little further we
should run some reports and create some sort of "state of the
vlogosphere".

-Josh



On 5/31/06, Andreas Haugstrup <[EMAIL PROTECTED]> wrote:
> On Wed, 31 May 2006 02:16:53 +0200, Joshua Kinberg <[EMAIL PROTECTED]>
> wrote:
>
> >> It is unnecessary. I have never met a blogger who have put up a
> >> thumbnail
> >> and then not used it as a link to the video. All you need is
> >> rel="enclosure" - if that link points to a video resource then you can
> >> safely assume that the first image contained in the link is the
> >> thumbnail.
> >>
> >> e.g. <a href="" rel="enclosure"><img src=""> > >> /></a>
> >
> > I only wish people were so consistent and reliable. Truth is that
> > there is vast inconsistency in the wild.
> >
> > Many people use text links to point to videos, sometimes they use
> > special "play button" icons to link to videos (not exactly a
> > thumbnail). Sometimes people link to an HTML page with the media
> > embedded and somtimes they just embed the media and do not use an
> > image to link to it. In practice relEnclosure is rarely, if ever used.
>
> Yeah, I know. In regards to thumbnails you can say 'if you want your
> thumbnail on fireant you will have to use the image as a link to the
> video'. Then people who wants thumbnails will have to make image links,
> and you will hopefully get a slow adoption.
>
> > We run into all sorts of problems like this trying to parse HTML
> > descriptions when we aggregate feeds on FireAnt.tv.
>
> If you guys ever want to share any research I'm sure it would be of huge
> value to the microformats community.
>
> > I don't think adding additional "semantic" HTML attributes would help
> > much either. Often the bloggers who post with strange HTML conventions
> > aren't the type of people who know much about HTML. I certainly
> > wouldn't expect them to grok a Microformat like rel="media:thumbnail"
> > ... but then perhaps broad acceptance is not really the goal of this
> > proposal anyway.
>
> +1
>
> --
> Andreas Haugstrup Pedersen
> <URL: http://www.solitude.dk/ >
> Commentary on media, communication, culture and technology.
>
>
>
>
> Yahoo! Groups Links
>
>
>
>
>
>
>


SPONSORED LINKS
Fireant Individual Typepad
Use Explains


YAHOO! GROUPS LINKS




Reply via email to