Side navigation
#3408 closed bug (fixed)
Opened September 25, 2008 12:25PM UTC
Closed April 17, 2011 05:40PM UTC
ajax accept quality score fix
Reported by: | jonrhall | Owned by: | flesler |
---|---|---|---|
Priority: | major | Milestone: | 1.5 |
Component: | ajax | Version: | 1.2.6 |
Keywords: | ajaxrewrite | Cc: | jaubourg |
Blocked by: | Blocking: |
Description
*/* is appended to the Accept header of an ajax request that uses dataType.
For "datatype: 'json'" the Accept header is:
"Accept: application/json, text/javascript, */*"
This means every mimetype has the same quality score of 1. Some servers may use order precedence but this is not within the spec (http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html).
The fix is simply to reduce the score of */* to below 1:
(25/09/2008 nightly - line 2730)
s.accepts[ s.dataType ] + ", */*" :
to
s.accepts[ s.dataType ] + ", */*;q=0.8" :
Attachments (0)
Change History (6)
Changed September 25, 2008 04:35PM UTC by comment:1
owner: | → flesler |
---|---|
status: | new → assigned |
Changed September 14, 2009 01:45PM UTC by comment:2
Not sure if I should raise this as another bug or not, but I would like the option to not send the "
*/*" at all.
On the subject of the ticket tho, is there a reason the code shouldn't just be
s.accepts[ s.dataType ] + ', ' + s.accepts._defaultinstead of using a hardcoded "
*/*".
I will prepare a patch to this effect tomorrow
Changed September 21, 2009 02:40PM UTC by comment:3
This is also a problem when trying to override the prefered type. I suggest the */* have a quality of 0.1 and the default jquery provided types have a quality of < 1 (say 0.2). This will allow the caller to override the prefered type with a higher quality.
XML has a lot of various types other than the generic "application/xml" and "text/xml", so these generic types should have a lower quality over more specific XML types provided by the calling function.
Changed December 27, 2010 10:36PM UTC by comment:4
keywords: | → ajaxrewrite |
---|
Changed April 17, 2011 05:19PM UTC by comment:5
cc: | → jaubourg |
---|
Please confirm this bug is still present in jQuery 1.6b1.
Changed April 17, 2011 05:40PM UTC by comment:6
milestone: | 1.3 → 1.5 |
---|---|
resolution: | → fixed |
status: | assigned → closed |
In fact, it doesn't seem to be - we have "q=0.01" in Ajax now. I'll mark this as fixed in 1.5.