Re: [cgmo-webcgm] Fwd: RE: script types...

[Lofton Henderson <lofton@rockynet.com>]

To finish this conversation, per this morning's telecon...

All of the executable examples in Ch.5 use text/ecmascript, except 5.2.5 
which used application/ecmascript.  None used language="JavaScript".

5.2.5 appears to work in IE.  That is because the underlying code in the 
linked ex_Picture.html actually has text/ecmascript.  Only the code listing 
in the HTML text of the spec has application/ecmascript (obviously an 
editor's oversight from the past, when the underlying code was corrected so 
that it would work.)

I tested:  changed underlying code to application/ecmascript, and indeed 
the example failed to execute correctly in IE.  So I changed it back, and 
will "proposed change" to the code listing in the EXAMPLE block of 5.2.5.

CONCLUSION (empirical):  all examples and all test cases should use 
text/ecmascript.

-Lofton.

At 02:04 PM 5/24/2006 -0600, Lofton Henderson wrote:
>I thought I forwarded this to the TC list, but can't find it...
>
>>From: Robert Orosz <roboro@AUTO-TROL.com>
>>To: 'Lofton Henderson' <lofton@rockynet.com>,
>>         "Galt, Stuart A" <stuart.a.galt@boeing.com>
>>Cc: "Cruikshank, David W" <david.w.cruikshank@boeing.com>
>>Subject: RE: script types...
>>Date: Thu, 18 May 2006 17:45:05 -0600
>>X-Mailer: Internet Mail Service (5.5.2653.19)
>>Mailarmory-Level: *
>>Mailarmory-Category: clean (1)
>>Mailarmory-Filter-Date: Thu, 18 May 2006 17:45:10 -0600 (MDT)
>>Mailarmory-Details: 
>>UmFuZG9tSVYbSS8Uf8Lequ2gMJdov1B6fyqplElRsSl9q2vW2GIsjINlaa4kwxiUC4gUQDeJrigb62YK6WiKjg==
>>X-RCPT-TO: <lofton@rockynet.com>
>>X-SpamCatcher-Score: 1
>>X-SpamCatcher-IP: 127.0.0.1
>>X-SpamCatcher-1: a49dfe7e7fb53a086c93788b1811c2b4
>>
>>Hi guys,
>>
>>Yes, I did look at this issue last summer and found that text/ecmascript was
>>obsolete.  I remember somebody at the time (probably Stuart) stating that
>>"application/ecmascript" did not work in Internet Explorer, and I mentioned
>>that at the last telecon.
>>
>>Last summer the RFC registering the text/ecmascript and
>>application/ecmascript media types was not yet published (it was in the RFC
>>Editor's queue).  It is now published as RFC 4329
>>(ftp://ftp.rfc-editor.org/in-notes/rfc4329.txt).
>>
>>I did some further research today to try and determine exactly what is meant
>>when a media type is labeled "OBSOLETE."  The current media type
>>registration procedure (RFC 4288) states, "Media type registrations may not
>>be deleted; media types that are no longer believed appropriate for use can
>>be declared OBSOLETE by a change to their "intended use" field; ...."
>>
>>So, text/ecmascript is a stillborn media type.  RFC 4329 points out that in
>>the past, ECMAScript and JavaScript have been exchanged with many
>>unregistered media types.  This is bad.  Using text/ecmascript is better
>>because it is registered, but not ideal.  However, for the sake of backward
>>compatibility with older applications, I think using a media type declared
>>OBSOLETE is perfectly acceptable.
>>
>>Chapter 5 of the CS text uses text/ecmascript throughout except for the
>>example in 5.7.5  (WebCGMPicture) which uses application/ecmascript.  I
>>think Lofton was testing examples before placing them in the text.  Somehow,
>>that one example slipped past the goalie.  It will eventually work one of
>>these days.  Until then, I think an informative note somewhere describing
>>the application/ecmascript versus text/ecmascript media type would be
>>helpful.
>>
>>Stuart also asked a question about the <script language="JavaScript"> tag
>>"that is all over the test suite."  This is bad because the language
>>attribute is deprecated in HTML.  The type attribute is the correct
>>attribute to use for specifying the scripting language in HTML.
>>
>>Regards,
>>
>>Rob
>>
>>-----Original Message-----
>>From: Lofton Henderson [mailto:lofton@rockynet.com]
>>Sent: Thursday, May 18, 2006 3:43 PM
>>To: Galt, Stuart A
>>Cc: Robert Orosz; Cruikshank, David W
>>Subject: Re: script types...
>>
>>
>>I'm copying Rob, as he had some definitive research...
>>
>>At 01:52 PM 5/18/2006 -0700, Galt, Stuart A wrote:
>> >Hello,
>> >
>> >I must have not been paying attention when we made the decision on
>> >what we wanted the scripts to be or at least the answer that I though
>> >was correct does not seem to work with my browser.
>> >
>> >I thought that we wanted something of the form...
>> ><script type="application/ecmascript">
>>
>>I think Rob said that that was the preferred IANA mime type designation.
>>
>>
>> >However this does not work with IE and he Itedo viewer (the only
>> >one I have available).
>> >
>> ><script type="text/ecmascript">
>> >
>> >does seem to work however.
>>
>>Someone mentioned:
>>
>>1.) text/emcascript is deprecated but allowed by IANA (probably because of
>>next...)
>>2.) IE doesn't handle application/ecmascript -- but no one seemed to have
>>definitive proof.  I guess you do now.
>>
>> >And I believe it is more correct than the
>> >
>> ><script language="JavaScript">
>> >
>> >that is all over the test suite.
>>
>>Remembering that this stuff is supposed to be ECMAScript, then
>>language="JavaScript" is probably invalid, right?  (Or ... since this is on
>>the HTML 'script' tag, it is at least undesirable, if not invalid.)
>>
>> >Unless I hear otherwise I think
>> >that I will change the script to "text/ecmascript" as I edit/touch
>> >them for other reasons.
>> >
>> >Comments?
>>
>>1.) ...after making sure that all changed tests still work, of course.
>>2.) if you edit the tests, and replace them on FTP, then please send a
>>message to cgmo-webcgm with a list of the files that you changed and the
>>nature of the change.  ("files", not "tests" -- you should only need to
>>fuss with the HTML files for this fix, right?).
>>
>>Lacking CVS (and its auto-list-notify), #2 will help us preserve some
>>sanity.
>>
>>-Lofton.
>
>
>
>