I initially had a lot of headaches with my Portuguese characters (e.g. ç), but eventually figured out that I had to use charset ISO-8859-1 for them to be represented properly. After downloading the 2009 editor, I now find that the program insists on using the default "charset unicode". When I run the code cleaner and specify "Latin 1" for the charset, I get the ISO-8859 and everything is fine, except that after a switch from the code to the visual tabs and back (without any changes) the charset has magically changed back to unicode. I then have to change back to the ISO-set and quickly save the page before it switches again. Very frustrating! This only happens with one of my pages (index.htlm) and the only relevant difference I see in the code for that page is the line:
<meta name="GENERATOR" content="MSHTML 8.00.6001.18812">
On the pages that that do not have this funny behavior, the corresponding line is:
<meta name="generator" content="CoffeeCup HTML Editor 2008 - www.coffeecup.com">
Is this an issue related to 2009 vs. 2008, or am I doing something wrong?
My URL is www.mingas.com and, yes, I am a novice at all this!

I'm just a user, but since that generator line isn't necessary, just remove it as it's causing the problem.

I had tried removing the line:
<meta name="GENERATOR" content="MSHTML 8.00.6001.18812">
but the program insists on putting it right back in the next time I switch from the visual tab to the code tab, and at the same moment the charset=ISO-8859-1 is replaced by charset=unicode.

Just noticed one more item that is changed in the same way as the "GENERATOR" and the "charset"-items: In the very first (<!DOCTYPE...) line, the text "...HTML 4.01 Transitional..." is changed to "...HTML 4.0 Transitional...". This change is also persistent in the same way and goes back to "4.0" when I switch from visual to code editor