W3C home > Mailing lists > Public > www-math@w3.org > May 2007

RE: MathML won't display (or: what triggers mathplayer behaviour?)

From: Richard Kaye <R.W.Kaye@bham.ac.uk>
Date: Thu, 24 May 2007 10:51:56 +0100
To: Robert Miner <robertm@dessci.com>
Cc: www-math@w3.org
Message-Id: <1180000316.2185.6.camel@mat140.bham.ac.uk>

I really think that I did get the Content-Type as emitted 
by the server exactly as in the registry key names.  Did you 
have any success?  I also read some microsoft "documentation" 
web pages... dreary stuff about how internet explorer can guess 
the type of document you are serving much better than you
could ever possibly specify it, but not once mentioning 
charsets.  At least I know now why a batch file I serve on
one machine as text/plain is executed immediately in MS-windows
rather than just displayed.

Richard



$ wget -S
http://mat140.bham.ac.uk/~richard/blmath/index-uhtml-mathml.uhtml

--10:43:40--
http://mat140.bham.ac.uk/~richard/blmath/index-uhtml-mathml.uhtml
           => `index-uhtml-mathml.uhtml.1'
Resolving mat140.bham.ac.uk... 147.188.55.51
Connecting to mat140.bham.ac.uk|147.188.55.51|:80... connected.
HTTP request sent, awaiting response...
  HTTP/1.1 200 OK
  Date: Thu, 24 May 2007 09:43:40 GMT
  Server: Apache/2.2.3 (Debian) mod_jk/1.2.18 mod_python/3.2.10
Python/2.4.4 PHP/4.4.4-8+etch3 mod_perl/2.0.2 Perl/v5.8.8
  Last-Modified: Sat, 19 May 2007 10:56:01 GMT
  ETag: "29524d-465c-8ba4b240"
  Accept-Ranges: bytes
  Content-Length: 18012
  Keep-Alive: timeout=15, max=100
  Connection: Keep-Alive
  Content-Type: application/xhtml+xml; charset=utf-8
Length: 18,012 (18K) [application/xhtml+xml]

100%[====================================>] 18,012        --.--K/s

10:43:40 (64.77 MB/s) - `index-uhtml-mathml.uhtml.1' saved [18012/18012]



On Sun, 2007-05-20 at 08:30 -0700, Robert Miner wrote:
> Hi Richard,
> 
> No, I'm not sure at all that the MIME types with appended charsets
> work properly.  I know that other MIME type parameters, such as q=.99,
> etc. are not recognized, and just like you, I've assumed they were
> silently stripped or something.  But it isn't obvious, since the base
> MIME type doesn't match either in these cases.
> 
> We were confident at the time we released MathPlayer that the utf-8
> and iso-8859-1 charsets did work. But I haven't tried it lately, and
> everytime I get an automatic IE update, I cringe for fear they've
> changed something with the MIME handling. 
> 
> I'll try to remotely log into my development machine and try your test
> page under the debugger and see what happens.  I do recall it has
> always been incredibly picky about spaces, etc. but I'm sure you've
> eliminated those sources of error.
> 
> --Robert
> 
>  
> 
> Robert,
> 
> I notice Mathplayer adds some registry entries
> for other mimetypes other than application/xhtml+xml
> For example, I noticed one for "application/xhtml+xml; charset=utf-8"
> (I typed that exactly, including getting the case right and the
> space after the ;)
> 
> Are you sure that these registry entries are actually working?
> 
> I ask because I just set up a document at
> http://mat140.bham.ac.uk/~richard/blmath/index-uhtml-mathml.uhtml
> which is served statically by apache with header containing
> Content-Type: application/xhtml+xml; charset=utf-8
> and this doesn't display in my IE7+MathPlayer
> 
> I am wondering if IE does something strange with such
> mime types, perhaps modifying or removing the charset before
> matching against the registry keys, or perhaps there is some
> invisible character in there, or carriage-return or line-feed.
> 
> As for my Tomcat and the java servlets, I am stuck.  I can
> reproduce the problem with a "hello world" type of servelet at
> http://mat140.bham.ac.uk:8180/sample/hello
> but I cannot reprogram this servelet to omit the charset.
> It seems that tomcat's programmers read the specs as saying
> that charset is manditory and the implementation always adds
> a default.
> 
> Richard
> 
> On Fri, 2007-05-18 at 10:26 -0700, Robert Miner wrote:
> > Well, I just spent some time running MathPlayer under a debugger,
> trying
> > to see why MathPlayer doesn't fire on
> >
> >
> http://mat140.bham.ac.uk:8180/blojsom/blog/default/?flavor=xhtmlplusmath
> >
> > It seems like you've been experimenting a bit with MIME types and I
> was
> > working, but I've been using "Charles", an HTTP header sniffer, and
> > adding registry keys by hand under HKCR\PROTOCOLS\Filter for all the
> > various MIME types I see for that page coming from Charles. So I'm
> > fairly certain my debug copy of MathPlayer is configured to fire for
> the
> > declared MIME type for the content coming from that URL.
> >
> > Unfortunately, what seems to be the case is that whenever content is
> > coming out of Tomcat (or any other dynamic source) and the URL
> doesn't
> > correspond to a filename, IE just doesn't honor the MIME type in the
> > HTTP Response header, and instead treats it as HTML.  If I set a
> > breakpoint in the code that is notified when IE thinks an
> > application/xhtml+xml document is coming down the wire, before any
> of
> > MathPlayer's code is executed, it simply isn't hit for
> >
> >      
> >
> http://mat140.bham.ac.uk:8180/blojsom/blog/default/?flavor=xhtmlplusmath
> >
> > but is hit for
> >
> >      
> > http://mat140.bham.ac.uk/~richard/blmath/index-xhtml-mathml.xhtml
> >
> > Very frustrating. 
> >
> > --Robert
> >
> > Robert Miner
> > Director, New Product Development
> >
> > Design Science, Inc.
> > 140 Pine Avenue, 4th Floor
> > Long Beach, California  90802
> > USA
> > Tel:  (651) 223-2883
> > Fax:  (651) 292-0014
> > robertm@dessci.com
> > www.dessci.com
> > ~ Makers of MathType, MathFlow, MathPlayer, WebEQ, Equation Editor,
> > TexAide ~
> >
> >
> > > -----Original Message-----
> > > From: www-math-request@w3.org [mailto:www-math-request@w3.org] On
> > Behalf
> > > Of Richard Kaye
> > > Sent: Friday, May 18, 2007 7:54 AM
> > > To: www-math@w3.org
> > > Subject: MathML won't display (or: what triggers mathplayer
> > behaviour?)
> > >
> > >
> > > Dear all
> > >
> > > I have an annoying problem, and would be grateful for any help.
> > >
> > > I have been trying to MathML-enable my blog, and I have
> > > the usual result: it works fine in Firefox but I can't
> > > get IE7 to behave.
> > >
> > > The blog is at
> > > http://mat140.bham.ac.uk:8180/blojsom/blog/default/
> > > This address serves the main file as text/html and
> > > has incorrect headers, etc.  The version served
> > > as application/xhtml+xml with proper !DOCTYPE and
> > > so on is
> > >
> >
> http://mat140.bham.ac.uk:8180/blojsom/blog/default/?flavor=xhtmlplusmath
> > > ml
> > > As mentioned, it works as expected in firefox.
> > >
> > > It doesn't work with IE7+MathPlayer however.
> > > The text/html version looks as before (maths incorrectly
> > > rendered) but the application/xhtml+xml version is just
> > > rendered as an XML listing, exactly as you would expect in
> > > IE7 when it downloads an XHTML document and no helper or plugin
> > > is triggered.
> > >
> > > This is strange, as there should be enough stuff there to
> > > trigger Mathplayer into action. It's just the same as a hundred
> > > other pages I have written.  Hence the subtitle of this
> > > mail: is there somewhere definitive information on what
> > > in an xhtml document triggers the mathplayer plugin to do
> > > something?  (As an aside, I would also quite like to use
> > > mathplayer's behaviour - even if there is no visible mathematics
> > > in the webpage - to avoid these lengthy XML listings of XHTML.)
> > >
> > > It gets stranger.  To investigate the problem further
> > > I wget'ed all the files and put them up on my apache server
> > > (the blog itself is served directly from Tomcat5.5).
> > > The results are all in
> > > http://mat140.bham.ac.uk/~richard/blmath/
> > > In this directory, the document
> > > index-default-orig.html
> > > is the original default text/html document
> > > and index-xhtml-mathml-orig.xhtml is the original
> > > application/xhtml+xml document, exactly as I got from wget.
> > > I also put in there index-default.html and
> > > index-xhtml-mathml.xhtml are the same except I edited
> > > these files by hand to make sure that the associated
> > > files (css, jpg, js) are all served from the same
> > > directory rather than from tomcat.  All the names are
> > > different, so I don't think there is a content-negotiation
> > > issue.
> > >
> > > As you'd hope, everything looks identical under firefox.
> > > But this time with IE7, Mathplayer is triggered and both
> > > index-xhtml-mathml-orig.xhtml and index-xhtml-mathml.xhtml
> > > display mathematics.
> > >
> > > So how does IE7/Mathplayer manage to display identical
> > > documents served from different servers in such different
> > > ways?
> > >
> > > There are also some problems with the stylesheets and the
> > > pages with mathematics aren't actually rendered correctly
> > > in IE7/Mathplayer, though they are at least legible.
> > > This is something else I need to fix too, in time.
> > > Unfortunately, this example is not the simplest
> > > (since it comes from the blog software I am using) and
> > > there are a number of CSS and JS things in it I need to
> > > understand, but I ***am*** worried that MathPlayer and
> > > the CSS/JS or whatever it is aren't living together nicely.
> > >
> > > For the record, I append some of the details of the servers
> > > and the HTTP headers they send when I do a wget.  (Sorry,
> > > the wget agent isn't quite the same as IE7 - but I don't
> > > think it really matters. If you want I can do other experiments.)
> > >
> > > Thanks for reading this far, and (in advance) for any help you
> > > might have.
> > >
> > > Richard
> > >
> > >
> > >
> > >
> > > richard@mat140:~/tmp$ wget -S
> > >
> http://mat140.bham.ac.uk/~richard/blmath/index-xhtml-mathml-orig.xhtml
> > >
> > > --13:28:48--
> > >
> http://mat140.bham.ac.uk/~richard/blmath/index-xhtml-mathml-orig.xhtml
> > >            => `index-xhtml-mathml-orig.xhtml.1'
> > > Resolving mat140.bham.ac.uk... 147.188.55.51
> > > Connecting to mat140.bham.ac.uk|147.188.55.51|:80... connected.
> > > HTTP request sent, awaiting response...
> > >   HTTP/1.1 200 OK
> > >   Date: Fri, 18 May 2007 12:28:48 GMT
> > >   Server: Apache/2.2.3 (Debian) mod_jk/1.2.18 mod_python/3.2.10
> > > Python/2.4.4 PHP/4.4.4-8+etch2 mod_perl/2.0.2 Perl/v5.8.8
> > >   Last-Modified: Fri, 18 May 2007 11:44:31 GMT
> > >   ETag: "29524e-47e4-1b4065c0"
> > >   Accept-Ranges: bytes
> > >   Content-Length: 18404
> > >   Keep-Alive: timeout=15, max=100
> > >   Connection: Keep-Alive
> > >   Content-Type: application/xhtml+xml
> > > Length: 18,404 (18K) [application/xhtml+xml]
> > >
> > >
> >
> 100%[===================================================================
> > > =============>] 18,404        --.--K/s
> > >
> > > 13:28:48 (58.20 MB/s) - `index-xhtml-mathml-orig.xhtml.1' saved
> > > [18404/18404]
> > >
> > > richard@mat140:~/tmp$ wget -S
> > >
> >
> http://mat140.bham.ac.uk:8180/blojsom/blog/default/?flavor=xhtmlplusmath
> > > ml
> > >
> > > --13:29:18--
> > >
> >
> http://mat140.bham.ac.uk:8180/blojsom/blog/default/?flavor=xhtmlplusmath
> > > ml
> > >            => `index.html?flavor=xhtmlplusmathml'
> > > Resolving mat140.bham.ac.uk... 147.188.55.51
> > > Connecting to mat140.bham.ac.uk|147.188.55.51|:8180... connected.
> > > HTTP request sent, awaiting response...
> > >   HTTP/1.1 200 OK
> > >   Server: Apache-Coyote/1.1
> > >   ETag: "3605938c728a24802ed11501eaa2860f"
> > >   Last-Modified: Thu, 17 May 2007 09:43:55 GMT
> > >   Content-Type: application/xhtml+xml;charset=UTF-8
> > >   Date: Fri, 18 May 2007 12:29:19 GMT
> > >   Connection: close
> > > Length: unspecified [application/xhtml+xml]
> > >
> > >
> > > [ <=>
> > >           ] 18,404        --.--K/s
> > >
> > > 13:29:19 (19.49 MB/s) - `index.html?flavor=xhtmlplusmathml' saved
> > > [18404]
> > >
> > >
> > >
> > >
> > >
> > >
> >
> >
> >
> 
> 
> 
> 
Received on Thursday, 24 May 2007 09:52:07 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Saturday, 20 February 2010 06:12:59 GMT