Browser showing source instead of rendering page

foosh101 · Post by **foosh101** » Tue Apr 01, 2003 9:46 am

A very strange thing is happening. I user executes a search, and everything works fine. If they then click on a link to another page, and then hit back on their browser, bringing them to the search results they were just on, they see the actual source of the html, not the rendered page. Its the strangest thing. Its a valid url, and the page that the vortex script creates is valid html, yet for some reason, the browser is not rendering it. Has anyone heard of anything like that and maybe knows why it could be happening. It seems to happen with much more frequency in the AOL Client.

foosh101 · Post by **foosh101** » Tue Apr 01, 2003 9:57 am

NOTE: if I refresh the page, it comes up fine.

Post by **mark** » Tue Apr 01, 2003 10:02 am

Never heard of that. Not much the webserver can do about such a badly broken client. The webserver's not involved when the user hits the back button. Time for a new browser.

Post by **mark** » Thu Apr 03, 2003 11:42 am

We did just recently hear about a case where some old http server software (ncsa/1.5.1) was confused by http/1.1 requests and returned an "HTTP" (instead if "HTTP/1.1") response which the latest version of IE 6.0 didn't understand. This is unrelated to Texis. It was a problem with the webserver software not supporting http 1.1 properly and the web browser (or proxy server between them) no longer supporting the oldest version of HTTP.

You can test your server by telnet'ing to port 80 and entering
HEAD / HTTP/1.1
followed by a blank line. You should get back something like
HTTP/1.1 200 Document follows
or
HTTP/1.0 200 Document follows
The problem occurs when you get back
HTTP 200 Document follows

foosh101 · Post by **foosh101** » Thu Apr 03, 2003 11:48 am

It was partially luck, but I figured out what it was. Prior to actually outputting anything to the screen, I have a whole bunch of vortex script that is executed. For readability purposes, there are several blank lines in that script. Those blank lines result in blank lines at the top of the html source. Aparently, I must have reached some threshhold at which IE/the AOL Client decided it was not an HTML document, and rather rendered it as a straight text file, or something like that. It must look in the first X number of lines or first X number of characters, and if it doesn't see a document type declaration or and HTML tag, it renders it as text. I went into my source and took out all the blank lines, and everything works fine.

Post by **mark** » Thu Apr 03, 2003 12:54 pm

Vortex will issue a content-type header before any of the text. If the browser doesn't obey that it's broken. I know IE 5 tended to ignore the content-type header thinking that it knew better, which it didn't.

kedlin · Post by **kedlin** » Sun Feb 18, 2007 12:01 pm

Help me someone! I search for a name, word, etc and the results appear numerically. I need to be able to click thru from the results page to each source page. If I click on the empty space to the right of each number - the source comes up. How can clients find this empty space to click thru to the search source page? The results are great - now I need to be able to have my clients click that empty space? Any advice?
www.NeedFREEestimate.com
Ken

Post by **John** » Mon Feb 19, 2007 9:43 am

It looks as if you are using the Master.com service, and have the look and feel set to show links as white text on a white background, so they aren't visible. You should either change the background or link color.

Post by **mark** » Mon Feb 19, 2007 10:47 am

Questions about master.com services should posted to the master.com message board at http://www.master.com/texis/master/search/msgboard.html