[wp-xmlrpc] APP and XMLRPC Ticket questions

[Peter Westwood]

> On Aug 21, 2009, at 11:53 AM, Peter Westwood wrote:
>>
>> This is about changing the response of WP when APP disabled from  
>> the current plain text response to a wp_die formatted one.
>>


On 21 Aug 2009, at 19:06, Daniel Jalkut wrote:

> MarsEdit has no dependency on the text-based failure messages.
>
> Would wp_die mean an HTML formatted failure?   I think it would be  
> even better if the APP and XMLRPC endpoints could fail with XML  
> formatted failures. This would automatically allow for error  
> presentation in MarsEdit, and I suspect would do the same for other  
> clients.

On 21 Aug 2009, at 19:19, Joseph Scott wrote:

>
> I looked at the current situation versus using wp_die.  The wp_die  
> function is really targeted at providing HTML pages to users, so I  
> don't think that it's a good fit when talking to other software. And  
> I think the other things WP does to work around browser bugs doesn't  
> help.  Here is what we current send back (tested with curl -v):
...
>
> I think the patched version is extra work with no reward.
>

Yeah the HTML version for APP does have a large overhead.

I like Daniels suggestion of an XML response.

Reading the APP spec is doesn't seem like there is any standard for  
this - maybe this is something we can pioneer for APP failure responses?

>
>> Secondly: <http://core.trac.wordpress.org/ticket/10658>
>>
>> This is about improving the feedback on GET requests to xmlrpc.php  
>> endpoint.
>>
>> I like the basic idea of the ticket but I think we can go one  
>> better and, using wp_die again, provide feedback on enablement  
>> status and a like to documentation on the codex about XMLRPC.
>
>
> Agreed, in this case we are returning content to a browser for the  
> end user, so it should be HTML.  One more additional step would be  
> return a more useful message for GET requests even when XML-RPC is  
> enabled.
>

Yeah a better get when enabled as well was what I was trying to imply.

Something along the lines of a short message about what the endpoint  
is for (links to the codex?) and then a statement about if it is/isn't  
enabled.

-- 
Peter Westwood
http://blog.ftwr.co.uk | http://westi.wordpress.com
C53C F8FC 8796 8508 88D6 C950 54F4 5DCD A834 01C5