Closed
Bug 145546
Opened 22 years ago
Closed 22 years ago
<embed> type attribute not passed to plugin
Categories
(Core Graveyard :: Plug-ins, defect, P3)
Tracking
(Not tracked)
VERIFIED
INVALID
mozilla1.2beta
People
(Reporter: nephtes, Assigned: srgchrpv)
References
()
Details
(Whiteboard: [plugger])
Reproducible always, seen on Linux 1.0RC2: On the page referred to above (see URL) there's an <embed> tag in it with a type=audio/x-zipped-mod attribute, *but* when the file is requested the web server it goes out with an application/octet-stream Content-Type: header. The result, when I open the page using Mozilla, is a Plugger error window which says that it can't find an appropriate type for application/octet-stream. So presumably the HTML "type" attribute is being used to select the plugin, but the HTTP header is being passed to the plugin when it's instantiated. I don't know where the correct behaviour is defined but I'm assuming this inconsistency is a bug.
when testing without the MOD plugin installed, i too was prompted do download the file. After installing the plugin, it simply played the sound. Testing with current trunk CVS, Linux. (Crossover w/Windows version of MOD plugin) Reporter: Have you configured plugger with something that can play .mdz files?
Reporter | ||
Comment 2•22 years ago
|
||
It works with Crossover for me as well, but not Plugger. I suspect the dedicated .mod plugin just assumes that whatever data it gets is the correct type without checking it -- I tested this by creating a page which embeds an object with "type=audio/x-mod" but with the src attribute pointing to a .png file. The .mod plugin tried to load it and just played garbage noise. Of course this could mean my bug is invalid and the plugin is being passed the MIME type in the <embed> tag's "type" attribute, but Plugger's behaviour seems to contradict this. To answer your question, I'm pretty sure I've got Plugger configured properly -- it wouldn't have registered itself to handle the audio/x-zipped-mod MIME type otherwise.
Reporter | ||
Comment 3•22 years ago
|
||
More supporting evidence: If I go try using Plugger on my test page, and set the <embed> tag's "src" attribute to point to a .gif file, the error message I get is "Plugger: No approperiate application for type image/gif found!" ("approperiate" is not my typo, incidentally) Here's my test page, if it helps: http://www.hasc.com/~nephtes/embed1.html Now using 2002051909 (branch), incidentally, no change in behaviour that I can see.
Comment 6•22 years ago
|
||
--->INVALID We need to pass the HTTP header's mime-type to the plugin rather than the EMBED's because the plugin needs to know what kind of mime the data will be and what to do with it. Removing the TYPE attribute should cause us to use a plugin as defined by the HTTP header. This currently has some problems, see bug 157554.
Status: UNCONFIRMED → RESOLVED
Closed: 22 years ago
Resolution: --- → INVALID
Reporter | ||
Comment 7•22 years ago
|
||
Hmmmm. HTML 4.01 seems to agree (this is for <object>, there is no <embed>): type = content-type [CI] This attribute specifies the content type for the data specified by data. This attribute is optional but recommended when data is specified since it allows the user agent to avoid loading information for unsupported content types. If the value of this attribute differs from the HTTP Content-Type returned by the server when the object is retrieved, the HTTP Content-Type takes precedence. I guess there's no arguing with the spec, but it seems to me that in practice the "type" attribute is used in exactly the opposite way -- to override the server's Content-Type: header when it's known not to recognize the file properly. (Besides, wouldn't the above make the attribute only relevant at all when there *isn't* a Content-Type header?) In any case, this still leaves the issue that the MIME type used to choose the plugin and the MIME type passed to the chosen plugin are different. Is this reasonable, and is this bug still invalid?
Updated•2 years ago
|
Product: Core → Core Graveyard
You need to log in
before you can comment on or make changes to this bug.
Description
•