Joseph,
I'm still receiving updates with relative hub URLs. Here is an example I
received seconds ago (irrelevant tags removed):
<channel>
<title>Apron Strings for Emily » Proud Pet Owner</title>
<atom:link href="http://apronstringsemily.wordpress.com/wp-admin/post.php";
rel="self" type="application/rss+xml" />
<link>http://apronstringsemily.wordpress.com</link>
<atom:link rel="search" type="application/opensearchdescription+xml" href="
http://apronstringsemily.wordpress.com/osd.xml"; title="Apron Strings for
Emily" />
* **<atom:link rel='hub' href='/?pushpress=hub'/>*
This is what I get POSTed to my PuSH end point. But if you pull the
feed, you get the full hub URL.
Regards,
Waleed
On Wed, Mar 3, 2010 at 9:05 PM, Joseph Scott <[email protected]> wrote:
> Can you go back and look again. I noticed that some feeds weren't
> using the full URL for the hub and made changes to try harder to make
> sure it was always the full URL.
>
> I just looked at http://pinaysolobackpacker.wordpress.com/feed/ and it
> has:
>
> <atom:link rel='hub'
> href='http://pinaysolobackpacker.wordpress.com/?pushpress=hub'/>
>
> in it now.
>
> My intention is that the hub always list the full URL.
>
>
> On Wed, Mar 3, 2010 at 9:05 PM, Waleed Abdulla <[email protected]> wrote:
> > I'm using the Python universal feedparser to parse updates coming from
> > wordpress.com, but it's failing to extract the hub URL correctly. After
> > investigating the issue, I found out that when I pull the feed the first
> > time, I get a hub link correctly as follows:
> > <atom:link rel='hub'
> > href='http://pinaysolobackpacker.wordpress.com/?pushpress=hub'/>
> > Then later I receive a POST with updated content, but this time the hub
> link
> > is relative as follows:
> > <atom:link rel='hub' href='/?pushpress=hub'/>
> >
> > I've setup my system to check the hub URL even on POST updates to detect
> any
> > potential changes that might happen after the initial subscription, so I
> try
> > to extract the hub url. Technically, I should expect either an xml:base
> > attribute or a Location http header to use to resolve the relative
> path[1].
> > Neither is provided. I can, of course, do an exception for wordpress.com
> > updates and try to extract the domain name from my database or from other
> > elements in the feed, but I'd hate to add exceptions if at all possible.
> > Am I missing something, or is this a bug with Wordpress updates?
>
>
>
> --
> Joseph Scott
> [email protected]
> http://josephscott.org/
>
