I certainly wouldn't be one to say that I consider m$ *ANYTHING* to be "where it's at" standard wise, (I guess we can let them have the things that they own (like RTF) and I draw short of saying the things that they "invented" because most of us have read histories of where various m$ branded technologies originated!)

Unfortunately there is no "official standard" for CSV, which can be a major borker - maybe the different variants should be categorised and numbered like CSV1, CSV2, CSV3 etc so it could be agreed upon as to which variant was being used or referred to... :o)

Some kind people are (after MANY years of its being around, and at the point where lots of people say it's on the way out (being replaced by XML) trying to fix this problem by defining it...

• http://www.shaftek.org/publications/drafts/mime-csv/draft-shafranovich-mime-csv-02.html
• http://www.wotsit.org/download.asp?f=csv

I certainly can't speak for all platforms or environments, but one very common format follows the rules that these pages document:

• http://www.edoceo.com/utilis/csv-file-format.php
• http://www.creativyst.com/Doc/Articles/CSV/CSV01.htm

That was what I was referring to.

In this case Excel behaves according to those rules, and is thus an easy tool to use to show how CSV data (according to those rules) should be preserved when saved.

As the author/maintainer of Text::CSV_XS, DBD::CSV and several other CSV-related modules, I've seen quite a few definitions of CSV from users around the world. My definition: a text file that uses commas to separate fields and which can be parsed into fields and records. By this definition, Excel can read CSV but can only write CSV sometimes. :-).