/MathJax/jax/output/HTML-CSS/jax.js fails to load

[Peter Cao]

In IE8,

File failed to load: http://www.mathannotated.com/MathJax/jax/output/HTML-CSS/jax.js

page can be found at http://www.mathannotated.com/page.html?topic=probability-1&&subtopic=simple-or-elementary-events

[Sean Hogan]

I couldn't duplicate this error.
Could you clear your browser cache and try again?

A couple of unrelated observations about that page:
1. You might want to add a meta tag at the top of the head block to
force IE8 into IE7 mode, otherwise MathJax processing will be slow, see
http://www.mathjax.org/resources/faqs/#ie8-slow

2. Does there need to be a double ampersand (&&) in that page url? Not
that it makes a difference.

Sean

[Peter Cao]

Hi Sean,

I cleared the IE8 cache, assuming this is the correct way to do it:

Ctrl-Shift-Delete
Check box next to temporary files
Delete

It made no difference...


I don't know why the results are so variable.

1) From the beginning, it worked on my XP machine - albeit took ages,
even after I applied to metatag that forces the browser into IE7 mode
- can you check my source to see that I implemented this properly?
2) On my sister's Windows 7 machine, it is producing "[Math Processing
Error]"s all over, telling me that "File failed to load:
http://www.mathannotated.com/MathJax/jax/output/HTML-CSS/jax.js".
3) On my Windows 7 Netbook, it's failing to load Italics and a
completely ugly, sleazy font is being displayed. To quote MathJax
"Can't load web font TEX/Math/Italic", though it went ahead and
processed with very ugly results.

Thanks for alerting me to the double &&. I'm not sure why, but I was
under the impression you needed two &s to separate variable-value
pairs... I'm now being led to think you only really need one & after
all, is that right?

Please help me see this through. I want to convert all my pages to
MathJax, if only I can be sure of it's cross-machine compatibility.

Peter

> > page can be found athttp://www.mathannotated.com/page.html?topic=probability-1&&subtopic=...- Hide quoted text -
>
> - Show quoted text -

[Peter Cao]

Sean, I found you some more information which may help you to diagnose
the problem:

Webpage error details

User Agent: Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 6.1; WOW64;
Trident/4.0; SLCC2; .NET CLR 2.0.50727; .NET CLR 3.5.30729; .NET CLR
3.0.30729; Media Center PC 6.0; InfoPath.2)
Timestamp: Tue, 18 Jan 2011 09:27:12 UTC


Message: 'this.div.firstChild' is null or not an object
Line: 35
Char: 3
Code: 0
URI: http://www.mathannotated.com/MathJax/MathJax.js

On Jan 18, 6:49 pm, Sean Hogan <shogu...@westnet.com.au> wrote:

> I couldn't duplicate this error.
> Could you clear your browser cache and try again?
>
> A couple of unrelated observations about that page:
> 1. You might want to add a meta tag at the top of the head block to
> force IE8 into IE7 mode, otherwise MathJax processing will be slow, see
>      http://www.mathjax.org/resources/faqs/#ie8-slow
>
> 2. Does there need to be a double ampersand (&&) in that page url? Not
> that it makes a difference.
>
> Sean
>
> On 18/01/11 6:08 PM, Peter Cao wrote:
>
>
>
> > In IE8,
>
> > File failed to load:http://www.mathannotated.com/MathJax/jax/output/HTML-CSS/jax.js
>

[Sean Hogan]

On 18/01/11 7:54 PM, Peter Cao wrote:

Hi Sean,

I cleared the IE8 cache, assuming this is the correct way to do it:

Ctrl-Shift-Delete
Check box next to temporary files
Delete

It made no difference...

OK. Did you try this (plus restarting IE) in all the following problem reports?
Also, would you be able to report the IE version number reported by Help -> About.

I don't know why the results are so variable.

1) From the beginning, it worked on my XP machine - albeit took ages,
even after I applied to metatag that forces the browser into IE7 mode
- can you check my source to see that I implemented this properly?

The meta tag needs to be at the top of the head tag, see
http://msdn.microsoft.com/en-us/library/cc288325(v=vs.85).aspx#SetMode

2) On my sister's Windows 7 machine, it is producing "[Math Processing
Error]"s all over, telling me that "File failed to load:
http://www.mathannotated.com/MathJax/jax/output/HTML-CSS/jax.js".

In the browser address bar can you enter the following url:
    javascript:alert(MathJax.Message.Log())
and report the result.

3) On my Windows 7 Netbook, it's failing to load Italics and a
completely ugly, sleazy font is being displayed.  To quote MathJax
"Can't load web font TEX/Math/Italic", though it went ahead and
processed with very ugly results.

Presumably it is falling back to image fonts. Try clearing the cache and restarting IE8.
Alternatively I suppose this might be due to security settings in the browser.
Using this machine can you try the page at http://playground.meekostuff.net/MathJax-fonts/webfont-redirect.html and report whether all three paragraphs are rendered with "fantasy" or "MathJax" fonts.

[Peter Cao]

On Jan 18, 11:04 pm, Sean Hogan <shogu...@westnet.com.au> wrote:
> On 18/01/11 7:54 PM, Peter Cao wrote:
>
> > Hi Sean,
>
> > I cleared the IE8 cache, assuming this is the correct way to do it:
>
> > Ctrl-Shift-Delete
> > Check box next to temporary files
> > Delete
>
> > It made no difference...
>
> OK. Did you try this (plus restarting IE) in all the following problem
> reports?

Case (1) Solved now - metatag relocation to the top of the head space
worked
Case (2) Yes, unsuccessful.
Case (2) Not yet, my sister's gone to sleep and the computer's in her
room; I'll have to try tomorrow.

> Also, would you be able to report the IE version number reported by Help
> -> About.

All of the IE instances mentioned are version 8.

Case 1) IE v8.0.6001.18702
Case 2) Will post this detail tomorrow when I have access to the
computer.
Case 3) IE v8.0.7600.16385

Case (1) Solved now

> > I don't know why the results are so variable.
>
> > 1) From the beginning, it worked on my XP machine - albeit took ages,
> > even after I applied to metatag that forces the browser into IE7 mode
> > - can you check my source to see that I implemented this properly?
>
> The meta tag needs to be at the top of the head tag, seehttp://msdn.microsoft.com/en-us/library/cc288325(v=vs.85).aspx#SetMode

You appear to be right; loading sped up significantly after I
relocated the meta tag right to the very top of <head>...</head>.
You'll noticed this caused layout issues, but that's to be expected
since I designed the pages for IE8. But that's not for you to worry
about now.
Just double check the source to make sure you're satisfied I did
exactly as you suggested:
http://www.mathannotated.com/page.html?topic=probability-1&subtopic=simple-or-elementary-events

> > 2) On my sister's Windows 7 machine, it is producing "[Math Processing
> > Error]"s all over, telling me that "File failed to load:
> >http://www.mathannotated.com/MathJax/jax/output/HTML-CSS/jax.js".
>
> In the browser address bar can you enter the following url:
>      javascript:alert(MathJax.Message.Log())
> and report the result.

Will post results tomorrow when I have access to the computer.

> > 3) On my Windows 7 Netbook, it's failing to load Italics and a
> > completely ugly, sleazy font is being displayed.  To quote MathJax
> > "Can't load web font TEX/Math/Italic", though it went ahead and
> > processed with very ugly results.
>
> Presumably it is falling back to image fonts. Try clearing the cache and
> restarting IE8.

Tried. Unsuccessful. I'm not sure if it is falling back on images, I
can highlight individual characters. Though the behaviour is quite
unusual upon highlighting, characters variably disappear.

See the results of the test at http://www.mathannotated.com/tests/mathjax-font-issues.html

> Alternatively I suppose this might be due to security settings in the
> browser.
> Using this machine can you try the page athttp://playground.meekostuff.net/MathJax-fonts/webfont-redirect.htmland
> report whether all three paragraphs are rendered with "fantasy" or
> "MathJax" fonts.

See the results of the test at http://www.mathannotated.com/tests/mathjax-font-issues.html

[Davide P. Cervone]

I notice that you are running version 1.0.1, which is pretty old at
this point. We are readying v1.1, but you might want to try out v1.1-
beta available from the MathJax download page at

http://www.mathjax.org/download/

Davide

[Paul Topping]

I am not sure that clearing "temporary files" is really clearing IE's cache. I thought the right way to clear IE's cache is to use Tools/Developer Tools (which can also be brought up with F12) to bring up the Developer Tools window and then choose Cache/Clear Cache. There is also an Always Refresh from Server command on the same menu that can be handy when debugging. You may have to enable Developer Tools somewhere in IE for this to work.

Paul

[Peter Cao]

Tried that, no luck. I'm not sure what the importance of refreshing
cache is?

> >http://www.mathannotated.com/tests/mathjax-font-issues.html- Hide quoted text -

[Paul Topping]

The browser's cache helps performance in general as the browser can get the file more quickly out of its cache than if it must request it from the web server. Unfortunately this can cause problems if the version of the file in the cache is different than the one on the server. This is particularly a problem with files that are still under development like MathJax's files. You mentioned that your results are so variable and this is often a symptom of caching problems.

[Peter Cao]

When I enter javascript:alert(MathJax.Message.Log()) I get the
following error

Webpage error details

User Agent: Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 6.1; WOW64;
Trident/4.0; SLCC2; .NET CLR 2.0.50727; .NET CLR 3.5.30729; .NET CLR
3.0.30729; Media Center PC 6.0; InfoPath.2)

Timestamp: Wed, 19 Jan 2011 06:20:19 UTC


Message: 'MathJax' is undefined
Line: 1
Char: 1
Code: 0
URI: res://ieframe.dll/tabswelcome.htm


IE v8.0.7600.16385, Windows 7

Does this tell any of you anything useful? I mean so far MathJax's
cross-machine compatibility is really patchy. I've only got one last
suggestion to try: you Beta test version. It'll only help assuming
you've addressed these issues in development before, but it doesn't
seem to be ringing a bell for any of you...?

[Sean Hogan]

Based on the URI, that's the message you would get in a new tab. You
need to try this in the page where MathJax is running.

Sean

[Peter Cao]

See screenshot at http://www.mathannotated.com/tests/error.PNG

> > seem to be ringing a bell for any of you...?- Hide quoted text -