View Full Version : Spinneret Problem with IE8?

01-23-2011, 07:45 PM
I'm having a problem opening my Spinneret web page with Internet Explorer 8. The laptop connects but, IE8 won't open the page stored in the Spinneret. IE6 opens the page just fine but, wife won't let me use her laptop. Can anyone provide a fix other than removing the wife?

Thanks - JL.

Mike G
01-23-2011, 08:24 PM
I have no problem with IE 6, 7, or 8. What does "Won't open the page" mean... Do you get an error? If so, what is the error. How about posting your code?

01-23-2011, 10:04 PM
Internet Explorer Cannot Display the Webpage is displayed by IE8. I didn't assemble the code but, will provide it should the problem appear in my other laptop that's using IE6.

Mike G
01-24-2011, 01:44 AM
So you don't need help?

01-24-2011, 05:08 PM
Here's the code that LamberJ is using. The webpage opened fine on my laptop using Firefox, I never tested it with IE8. Could this be a firewall/security issue?

He's using the same router as I was using. Since it did work with Firefox and older versions of IE then I would assume its a firewall/security issue. Any ideas?

Thank You,

01-24-2011, 06:13 PM
A few things you could check/try:

- To access the site are you using an IP address or do you have a DNS/HOST file setup. It would be good to confirm access via IP address to take DNS and name resolution out of the picture

- Do you have compatibility view turned on?

- In IE8 do you have the Smart filtering feature turned on to send sites to MS to check them?

- Can you add the address of your Spinneret as a "Trusted Site"

- Is there any weird proxy settings configured for IE8 than may not be in your other browsers?

- Also check any security Software McAfee, Norton, etc to ensure they aren't blocking anything

Mike G
01-24-2011, 08:34 PM
I'll run a test when I get home but at first glance I believe your HTML is not well formed. Secondly, your running processes like writing to an LCD and executing waitcnt(clkfreq / 1000 + cnt) while sending the page to the client. You should consider running those process after parsing the querystring and before sending the response header.

Check the HTML in firefox -> view -> page source

Mike G
01-24-2011, 11:39 PM
Well, the page open fine in IE8 for me at least. I commented the serial driver objects though.

Your HTML is not formed well. A non breaking is &nbsp; not &nbsp. The < form > tags do not have ending tags... but they all do the same thing. Your header request/response looks like...

User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv: Gecko/20101203 Firefox/3.6.13
Accept: text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8
Accept-Language: en-us,en;q=0.5
Accept-Encoding: gzip,deflate
Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7
Keep-Alive: 115
Connection: keep-alive

HTTP/1.1 200 OK
Server: Parallax Spinneret Web Server/demo 1Connection: closeContent-Type: text/html

The response header lines are run together because they are missing a line feed. Linux/Unix (Mac) look for line feed.

01-26-2011, 02:53 PM
Interesting. Thank you for the response and I'll work on my HTML (still a beginner at it).

Are you saying there should be a semi-colon after &nbsp? Is there a better way to "position" buttons on the web page?

Instead of starting the serial drivers at the beginning of PUB main, can I open a new cog and start them at the newly opened cog? The LCD screen will probably be taken off the robot, the serial object for the motors needs to be kept though.

What do you mean a line feed? Something like this <BR> ? I see that when I view page source most of the HTML is on one line.

Thank you for helping,

01-27-2011, 05:06 AM
Since you don't have access to IE8, you could always use http://browsershots.org/ (See bottom blue bar for the "Select None" button.) Of course, actually having the browser available is the best solution.

Mike G
01-27-2011, 12:49 PM
Yes, <br /> is a HTML break. I'm talking about the header response not the html. Some of the lines in your header response end with a CR (0x0D). While others end with CR, LF (0x0D 0x0A). The specification calls for header lines to be terminated with at least a line fed, 0x0A.