96579 - <object> with absolute URL and without MIME type doesn't work

Status: RESOLVED FIXED

(Core Graveyard :: Plug-ins, defect, P3)

Description

[Koike Kazuhiko]

<object> with absolute URL and without MIME type doesn't work.

Comment 1

[Koike Kazuhiko]

Attachment: Testcase

Comment 2

[Peter Lubczynski]

I think this is a simple fix if anyone wants to take a crack at it. 

We should be calling GetDocumentBase() instead of GetBaseURL() in
nsObjectFrame::IsSupportedDocument(). The only thing is that it returns a char *
string instead of a nsIURI so maybe we should get it directly from an nsIDocument.

Comment 3

[Peter Lubczynski]

Not gona make it--->mozilla0.9.5

Comment 4

[Heikki Toivonen (remove -bugzilla when emailing directly)]

SPAM. HTML Element component deprecated, changing component to Layout. See bug
88132 for details.

Comment 5

[Peter Lubczynski]

Taking a closer look, this is a different problem than I orriginally thought.
The problem seems to be that at the time we make the decision if this is going
to be a plugin, image, or document, we have not opened the stream to look at the
content-type header. Because of bug 95549 and/or bug 95548, I can not fix this
at this time.

The workaround at this time is to either force a document mime-type with an
attribute or be sure to include the document extension in the URL.

Comment 6

[Peter Lubczynski]

Attachment: patch to avoid crash

Comment 7

[av (gone)]

r=av

Comment 8

[Marc Attinasi]

Comment on attachment 49133 [details] [diff] [review]
patch to avoid crash

sr=attinasi on this patch (I guess GetTypeFromURI does not set the out param on FAILURE?)

Comment 9

[Peter Lubczynski]

Note: this bug is not about a crash. I happen to find this crash while using the
testcase to look at this bug. Fix for crash checked into trunk but it should
probably also go on the branch.

Comment 10

[Patrick C. Beard]

Marking nsbranch+.

Comment 11

[Peter Lubczynski]

orriginal issue not fixed yet

Comment 12

[Jaime Rodriguez, Jr.]

Peter - Is this something we should get on the branch?

Comment 13

[Peter Lubczynski]

general problem described in bug still going on but it's not serious. However,
the crash has gone (probably with another fix), so this is no longer a nsbranch
candidate.

Comment 14

[Asa Dotzler [:asa]]

Bugs targeted at mozilla1.0 without the mozilla1.0 keyword moved to mozilla1.0.1 
(you can query for this string to delete spam or retrieve the list of bugs I've 
moved)

Comment 15

[Jonas Jørgensen]

*** Bug 127963 has been marked as a duplicate of this bug. ***

Comment 16

[rubydoo123]

1.2

Comment 17

[Christopher Hoess (gone)]

*** Bug 153920 has been marked as a duplicate of this bug. ***

Comment 18

[Malcolm Rowe]

Note that this makes it impossible to use content negotiation (which provides
multiple representations of the same object).

For example:

The W3C 'valid XHTML 1.1' image is at http://www.w3.org/Icons/valid-xhtml11
Depending on your Accept header, you'll either get a PNG or a GIF image when you
request the image.

Obviously, if we need to specify the image type on the OBJECT element in
advance, we can't take advantage of CONNEG.

Comment 19

[Christian :Biesinger (don't email me, ping me on IRC)]

fixed by bug 1156 (I want to note that the relative URL only used to work
because it had an .html extension, which gecko mapped to text/html).

Comment 20

[Christian :Biesinger (don't email me, ping me on IRC)]

*** Bug 353119 has been marked as a duplicate of this bug. ***