By drazin on

When i add the feed url to Google Desktop 2.0 webclips module, it says unrecongnized webclip stream. It says it should work with any RSS/Atom feed. Anyone know anything about why my drupal site wont work with this?

Does anyone know if Drupal's RSS feed will be "webclip" compatible?

Comments

drazin’s picture

drazin commented

anyone?

robjohnston’s picture

robjohnston commented

I cannot answer the question, but I can confirm that the problem exists. I know the Google prefers Atom, but they have no information in their Desktop help about what defines an "unsupported web clip version."

drazin’s picture

drazin commented

OK thanks for looking into it for us. Atleast we know now....

Steve Dondley’s picture

Steve Dondley commented

I'm running a site with an older version of Drupal 4.5 and Google picks up the RSS feed just fine. It is not doing so with the 4.6 site I'm running, however.

--
Get better help from Drupal's forums and read this.

drazin’s picture

drazin commented

any update on this? i really want to see updates on my site though google desktop....

sethcohn’s picture

sethcohn commented

+1, This is a real problem, and it's annoying that our RSS feeds are flagged as invalid by Google Desktop. What's wrong with them?

Stefanos Karagos’s picture

Stefanos Karagos commented

Same problem!
Yes, it's annoying that our RSS feeds are flagged as invalid by Google Desktop!

any help pls???

-------------
the art of war is to know the weakness of your enemy ;)

danubetech’s picture

danubetech commented

This is a real problem for us too. I see there is an active bug open though:
http://drupal.org/node/31032

Edward C. Zimmermann’s picture

Edward C. Zimmermann commented

I don't wish to comment on G*ggle but the RSS produced by Drupal is hardly conformant to any of the "standards" but is naively produced right out of the data structures used for the blog. People use images and links inline then they are just copied right over into a "template" as RSS.

Lets look at what I just got from THIS Drupal (this site).... I see inlined links in the description. Other sites I see inlined images. I see all kinds of HTML, formatting information and other assorted bits copied verbatim out of the entry. They do not belong in this element. YES, one can use extended tagging in RSS via namespaces but I don't see them. There are attributes in RSS for images and links and much of what one wants. These, when I see them, are ill-used to point back to the blog rather, as intended, to point to the link or image or XYZ that's been intended in the newsitem.

In collecting RSS feeds from many Drupals (in the context of what we have called RAIN, our embeded search service module into Drupal) we have found not only are the RSS feeds syntactically a mess but also that character set encoding was often bogus.

To do it right, however, I think would demand some re-thinking and a design of the means for metadata and content collection. The ability, for instance, to type HTML into the body of articles might appeal to some people sense of aesthetics but is counterproductive .

--
Edward C. Zimmermann <edz@nonmonotonic.net>
BSn's Nonmonotonic Lab

Zethris1’s picture

Zethris1 commented

Why not just put filters in the feed creator that takes out all that crap? It would also check for modules which have a variable set for what else to take out before the feed is created. Modules would need to be updated, but this feed issue should be fixed in the core so it should be a requirement for existing and new modules for those of use who want clean rss

--------------------
Visit http://www.thinkingpost.net to add your thoughts and ideas to the community.

Edward C. Zimmermann’s picture

Edward C. Zimmermann commented

"Why not just put filters in the feed creator that takes out all that crap? "

That's, I'd suggest, easier said than done. The current model takes what people type in as content and if the content is a picture and a link? Or multiple pictures and multiple links?
Sure one can have pictures and links but they don't belong in the textual body as description.

What do we do with tables?
OK.. One solution would be to rewrite all the HTML as XHTML namespace but it does not improve the utility, only fixed the basic syntactical failing.
Better would be to design the data organization with RSS and blogging as the base model and built on top as extensions. Links and images then would be specified as metadata. User would be prohibited to wildly type in, as now, whatever they wish. One should also start off with a vocabulary of description concepts (tags)--- a DTD and schema-- for the articles and do away with the ability to type HTML. None of the <h1> ... etc. stuff.. Instead RSS as RSS with XML in namespace and CSS etc. to render.

--
Edward C. Zimmermann <edz@nonmonotonic.net>
BSn's Nonmonotonic Lab

handelaar’s picture

handelaar commented

I'd say there's a half decent chance that Google Desktop isn't actually retrieving the feed.

I have several sites where Google Desktop is asking for countless hundreds of ever-more-ridiculous URLs on every page view. So much so that I'm actively considering blocking its client string from the server altogether.

The process appears to be:

a) Ask for a feed in a subdirectory which doesn't exist
b) Get a 404, but delivered from the URL it asked for
c) Try again deeper still; repeat until URL is >1024 characters