I want to use the WordPress APML Plugin to provide all APML files a
user have. For example the native WordPress APML, several tastebroker
APMLs and some Engagd APMLs. There are several possibilities to do so:

* Merging the files: To merge all files to one big APML file.
Advantage: You have only one file to provide. Disadvantage: The file
is blown-up very fast.

* Diverting the source attribute from its intended use: <Source
key="http://research.sun.com:8080/AttentionProfile/apml/last.fm/
pfefferle" name="tastebroker.org" value="1.00" type="application/rss
+xml" from="tastebroker.org" updated="2007-03-11T01:55:00Z"><Author
key="Sample" value="0.5" from="tastebroker.org"
updated="2007-03-11T01:55:00Z" />
Advantage: You only have to provide one file. Disadvantage: No
determination between apml files you like and own apml files you
provide.

* Using OPML: Providing your APML files through one OPML file:
<outline type="category" title="APML">
  <outline text="last.fm" type="apml" xmlUrl="http://research.sun.com:
8080/AttentionProfile/apml/last.fm/pfefferle" htmlUrl="" updated=""/>
  <outline text="notizblog.org" type="apml" xmlUrl="http://
notizblog.org/apml/" htmlUrl="" updated=""/>
</outline>
Advantage: Providing only one file. Disadvantage: The parser has to
parse the OPML first.

Any other suggestions, or what do you think is the best way?

This would be my preferred option, as it allows APML consumers to not have
to understand the concept of retrieving multiple APML files.

I'd get the feeling this could be a bit confusing. And the source attribute
doesn't actually imply that you want to draw in the given content, which
this strategy would require.

This is an option, but again I'd still prefer just to merge the APML files.

No others that I can think of at the moment. All good ideas to mull over
though. Thanks!

Paul.

Hi,

On Dec 28, 2:33 pm, "Paul Jones" <pauljone...@gmail.com> wrote:

This is the approach I'm following with Midgard:

* Gather attention data from all over (external APML files, web2.0
services, local Midgard storage) and store in a central DB
* Split the attention items to profiles as needed
* Serve everything in one big APML file (or a profile-specific file if
requested)

/Henri

Progressing via baby steps is a good way to go.  Let's first figure
out how to make use of APML (meaning creating relevant content based
on APML).  After we can do this, it makes more sense on figuring out
what aggregating algorithms work best.  So let's crawl before we run.

I think both are fairly small steps - no harm in two different people/groups
attempting to solve the two given problems :)

Chris

On Dec 31, 2007 5:08 AM, Mike Reynolds <reynm...@gmail.com> wrote:

--
Chris Saad
FaradayMedia.com - For Audiences of One
Particls.com - Are You Paying Attention?
Engagd.com - The Open Attention Platform
Media2.0Workgroup.org - Social, Democratic, Distributed
APML.org - The OPML of Attention
DataPortability.org - The next frontier; Data