WinEdt 7: How has Unicode support changed from version 6?

The popular WinEdt editor has been just updated to version 7. It is claimed that now WinEdt supports Unicode. However it was also possible to compile Unicode texts in 6th version. I did not notice big difference between 6th and 7th version except that new version uses Unicode dictionaries. Could one explain item by item what has changed? The WinEdt website has only very brief statement on that issue.

Note that answers to XeLaTeX, WinEdt 6.0 and UTF-8 might be helpful but not exhaustive.

-

Well the answer you linked too already said it: Winedt 6 can handle utf8. But only if you need only chars from one codepage (e.g. if you are only using the chars from ansinew). When opening an utf8-file winedt 6 translates the chars to (e.g.) the ansinew codepage and all chars outside this codepage are shown as ??. It you change such a file and save it with winedt 6 all chars outside ansinew are lost. This means you can use at most 256 different chars and you can't write with winedt 6 a document with english and greek or english and cyrillic or a CJK-document. Winedt 7 now can handle all the thousands unicode chars.

-
 As a Chinese TeXer, I would point out that one can use WinEdt to write CJK document. Multibyte codepage support has been available long time ago, there were special options for CJK scripts in WinEdt 5.x. Although the inner working encoding was not Unicode, WinEdt 6.0 (and 5.x) can handle CJK scripts, and can read and write UTF-8 files since WinEdt 5.5. – Leo Liu Apr 10 '12 at 8:56 Yes, I see the difference in saving document! Does it mean that now one type e.g. greek letters in math if my source file loads unicode-math package? – Igor Kotelnikov Apr 10 '12 at 9:16 @Leo: are you sure that winedt 6 supports (that means can display and save) utf8-files which used chars from more than one code page? I never heard this and it certainly contradicts e.g. utf-8.tex in the winedt documentation. – Ulrike Fischer Apr 10 '12 at 9:48 @Igor: Yes if you want you can now type $a=αβγ$ in winedt (where "type" can mean that you have to click through a character map to insert the char). – Ulrike Fischer Apr 10 '12 at 9:52 @UlrikeFischer: No, I mean multi-byte codepage, but not multiple codepages. – Leo Liu Apr 10 '12 at 11:16

Before version 7, WinEdt do not use Unicode as internal working encoding. The UTF-8 support from WinEdt 5.5 to WinEdt 6.0 is partial: WinEdt can read a UTF-8 file and then convert it to local encoding, the editor works in the local non-unicode encoding, and then you can save the file in UTF-8 encoding.

For users, it means that, as Ulrike Fischer said, you can only use the characters from your own codepage. Say, I can use 中文 on my Windows system (codepage 936), but the same file may be broken on your Windows system (codepage 1251). That's the problem.

If you use only one input encoding, say, cp1251, cp866 or koi8-r, you don't need the full support of Unicode, WinEdt 6 is fine.

If you want to typeset this, you need new WinEdt 7:

% XeLaTeX
\documentclass{article}
\usepackage{xeCJK,arabxetex}
\setmainfont{Times New Roman}
\newfontfamily\arabicfont[Script=Arabic]{Times New Roman}
\setCJKmainfont{SimSun}
\begin{document}

English, la langue française, 中文, русский язык, Ελληνικά

\begin{arab}
اللغة العربية‎
\end{arab}
\end{document}

-
 It is pity that I accept both answers. – Igor Kotelnikov Apr 11 '12 at 3:17