[microformats-discuss] Video Pop-up Link Maker

[Joshua Kinberg]

> Josh -- start simple with what people are doing in the videoblogging
> community. Exploit and encourage conformance to XHTML. Look at what
> people are actually doing is always the starting point.

Right, that's why the core of the suggestion is what I believe to be
the most useful (and would be the most widely used):

<a href=[url of video] title=[title of video] type=[mime-type] rel="enclosure">
<img src=[url of thumbnail image] class="thumbnail" />
</a>

This is basically just an image link to the video with some extra
optional attributes in the <a> tag, which is what is being generated
at  < http://joshkinberg.com/popupmaker/ > and for the most part is
intended to follow what is already common practice among
videobloggers.

The other metadata stuff in the example I sent previously would have
to be considered optional (even the thumbnail image could be
considered optional) and was pretty much derived from this table on
the Microformats Wiki:
< http://www.microformats.org/wiki/media-metadata-examples >

As for following what people really do in the wild, it doesn't really
work out when you begin adding the extra media metadata like Roles,
Date, Tags, Duration, etc... although when you begin to add this stuff
it starts looking closer to other exisiting schemas (which are really
theoretical constructs of what people *should* do -- not what they
already do on the Wild Wild Web).

> ... if you sit down and look at what the videoblogging community is
> currently doing, you'll see that in 95%+ of the cases, many of these
> elements are already covered -- by the enclosing blog.

You're right, much of that data may be contained in the enclosing blog
entry, though it might not necessarily pertain directly to the
enclosed media file. That's why I believe the extra, media specific
metadata would have to be optional.

But, I think it still might be useful to wrap multiple versions of the
same file within a <div> tag. That way it might more easily conform to
the <media:group> concept in MRSS, which I think is very useful.

There are several sites that offer multiple versions of the same video
(different formats or bitrates).
< http://rocketboom.com > and < http://revision3.com/systm/ > are a
couple examples of videoblogs that offer multiple versions of the same
video with every blog entry.

These sites also use a separate RSS feed for each media type. Part of
the promise of MRSS is the ability to combine these separate feeds
into one RSS feed that offers multiple versions of the file same file
by using <media:group>.

If you click "Download Episode" on an entry at Revision3, for example:
< http://revision3.com/systm/subsystm/ipodnano/media >
You will see a list of the media types available using <ul>.

This is probably not very common -- most videobloggers don't offer as
many versions of each file as Revision3 does. However, I think its a
nice structured way of doing it.

Still, the <ul> itself may not be necessary if you include multiple
relEnclosure links within a <div class="enclosure">, or something
similar -- or am I beginning to delve into the "what people *should*
do" realm at this point?


-josh



On 10/12/05, David Janes -- BlogMatrix <davidjanes at blogmatrix.com> wrote:
> Ryan King wrote:
>  > I think we're miscommunicating here.
>  >
>  > What you have above, Josh is some great analysis, a good conclusion.
>  >
>  > What we're missing is the evidence that led you to that conclusion.
>  >
>  > You know vlogging well, most of us don't- this is part of the reason
>  > why we advocate thorough documentation, because only then can non-
>  > domain-experts work with the format in an informed manner.
>  >
>  > David Janes (and others) are doing great work on http://
>  > microformats.org/wiki/blog-post-formats. We need similar work
>  > for media.
>
> Just to endorse this -- sitting down and spending the time actually
> systematically documenting and analyzing the cases "in the wild" -- is
> uncovering depths of information I had no idea were below the surface of
> the problem.
>
> Josh -- start simple with what people are doing in the videoblogging
> community. Exploit and encourage conformance to XHTML. Look at what
> people are actually doing is always the starting point.
>
> For example, you suggest ...
>
> Joshua Kinberg wrote:
>  > <div class="enclosure">
>  >     <a href=[URL of video] title=[title of video] type=[mime-type]
>  > rel="enclosure">
>  >         <img src=[URL of thumbnail image] class="thumbnail" />
>  >     </a>
>  >     <abbr class="duration" title="00:27:35">27 minutes, 35 seconds</abbr>
>  >     [hCard of XFN for roles of creators/authors]
>  >     [hCalendar for date of creation]
>  >     [relLicense for license info]
>  >     [relTag for category/taxonomy info]
>  >     [relPayment for payment info]
>  >     <blockquote>[description of video]</blockquote>
>  > </div>
>
> ... if you sit down and look at what the videoblogging community is
> currently doing, you'll see that in 95%+ of the cases, many of these
> elements are already covered -- by the enclosing blog.
>
> I can't speak for non-videoblog/porn sites. Mark's supposed to report
> back to us once he's finished "researching"....
>
> Regards, etc...
> David
> http://www.blogmatrix.com
>
> _______________________________________________
> microformats-discuss mailing list
> microformats-discuss at microformats.org
> http://microformats.org/mailman/listinfo/microformats-discuss
>