thank you very much for your support and thank you for your new API
release.

We're happy to announce now that the SlideShare API has been
integrated successfully in Agglom.com.

We use currently the Slideshare API to get embed code of Slideshare
presentations using the relative URLs: when a user saves Web search
results in Agglom (either with the extension or copying / pasting the
link), if one link is referred to a SlideShare presentation page,
automatically the SlideShare presentation embed code is retrieved and
displayed in the Agglom generated multimedia page.

For this purpose, we call the get_slideshow_info api passing as
parameter slideshare_url and not slideshare_id.
We used most of the code distributed with SlideShare official C#
DotNetUtil solution. That has been really useful.

Now, talking technically, I would show the pattern we used, as well as
it may be useful to other developers as well.

1. User insert a Slideshare URL
2. We store in our DBs the URL in one field (fieldA) and set an empty
value for another one (fieldB)
3. a cron service in the Web Server is running every 280 seconds and
checks if there are slideshare links in the DB that have an empty
fieldB
4. If records are found, it takes the first one
5. It starts a new thread with a call to the slideshare API and update
fieldB with the value of the .swf (or "/" if some error is returned)
and a timeout of 1 minute

NB: of course if later on the slideshare link is changed, fieldB is
emptied, so the new link will be caught automatically by the cron
service.

When displaying the multimedia page (the one with all the results
organized by type), for every SlideShare saved page we check if fieldB
is empty or not. If it is, we display only the link, otherwise we
display the full SWF + the link.
Usually we don't have SlideShare links with empty fieldB for more than
20 minutes.

Why don't we request the embed code to SlideShare as soon as the user
enters the link?
Because we would like to control SlideShare calls and not to make more
than one call at a time (we don't want to overload SlideShare servers
and we don't want too many Web connection opened in ours).

Why on a new thread?
Because the Website is not on a Web Garden (we have only 1 worker
process for the whole website) so, if we call the SlideShare API in
the same thread, the whole Website back-end would be idle until it
would finish (and, if SlideShare servers are busy, it may take minutes
before timing out)

Suggested improvement:
When calling the "get_slideshow_info" with "slideshare_url" parameter,
it should automatically strip the eventual extra suffix to the url
(starts from "?" and "#"). In this way, it would get the same results
with:
http://www.slideshare.net/cathina/beijing-2008-in-color
and
http://www.slideshare.net/cathina/beijing-2008-in-color?srcembed=true
and
http://www.slideshare.net/cathina/beijing-2008-in-color#anchor
and
http://www.slideshare.net/cathina/beijing-2008-in-color?srcembed=true...

Conclusion:

This is an example page that was build with the firefox extension and
13 tabs opened (11 on Slideshare presentation pages and 2 on normal
links):
http://www.agglom.com/agglom/475/Doing_Internet_Web_business_in_China...

At last, just to say we were greatly satisfied by the support offered
by SlideShare team, by Gee Chuang in particular.

Hope this helps.

Thanks,
Enrico Foschi
http://www.agglom.com