Form controls inside the <object> element should not send data to the server if the object is rendered correctly. Content inside the <object> element is alternate content not additonal content to the <object> element. In this example, data from the <textarea> should not be sent to the server if the <object> element is rendered correctly: <object data="logo.gif" type="image/gif" width="180" height="60"> <textarea id="abc" name="abc" rows="5" cols="70">Hello World!</textarea> </object> Test case at: http://misc.xstandard.com/apple/alternate-content1.asp We are a browser plug-in vendor and would very much like to use the above construct with our plug-in called XStandard. IE and Opera already support correct use of alternate content in the <object> element.
To further clarify the request for this bug fix, when our plug-in is not installed on the client, we need to provide a "textarea" in its place. To do this, we use the fallback mechanism built into the "object" element. For example: <object type="application/x-xstandard" name="editor1"> <textarea name="editor1"></textarea> </object> Both "object" and "textarea" need to have the same value in the "name" attribute because our plug-in is used in existing content management systems (CMS) where it is very difficult or impossible to change the server-side code to received HTTP POST data from 2 form controls instead of one for what is really one control. Since both controls have the same "name", the server will received double the data.
Vlad filed this in Radar as: <rdar://problem/4268878>
I tested the problem in the latest WebKit and it looks that the bug is already fixed. Any confirmations?
I'm still seeing this issue with WebKit r34513 (i.e., I'm seeing "abc = Hello World!" in the result). Firefox behaves the same.
(In reply to comment #4) > I'm still seeing this issue with WebKit r34513 (i.e., I'm seeing "abc = Hello > World!" in the result). Firefox behaves the same. > Argh... You are right. Description of the bug was unclear for me, so I incorrectly interpret result of the test case :(
this is invalid per html5
Created attachment 70300 [details] Patch
Comment on attachment 70300 [details] Patch Clearing flags on attachment: 70300 Committed r69720: <http://trac.webkit.org/changeset/69720>
All reviewed patches have been landed. Closing bug.