Is Unicode conversion REQUIRED for 8.11??

All,

I'm trying to find out if the Unicode conversion is REQUIRED for 8.11. I am not looking for reasons as to WHY I should do it, I have been asked to find out if it is absolutely REQUIRED. Also, I have been asked to find an official document as to Oracle's stance on this... I have searched all over the Knowledge Zoo, but I can't find any official doc on Oracle's stance on this..

With new installation your db will be UNICODE (mandatory).
With an upgrade all DB EXCEPT Business Data and Control Table will be UNICODE. After upgrade you can decide if convert Business Data and Control Table to UNICODE. If you decide do not convert, no problem, this solution is supported
Gigi

My memory of the official stance (in webcasts and powerpoints I had seen) was that you had to convert your business data in 8.11 for performance reasons. It was supposed to be exponentially better or something, than if you didn't convert. Don't have any saved documents to send you, unfortunately...
I know you are not looking for reasons why, but they may throw this in your face if you call them with any performance support case later, even if it may be officially supported not to convert...

Gigi's post is right on. All DB's must be in unicode format for all E1 releases 8.9 and above, EXCEPT for business data and control tables. You can NOT perform a unicode conversion of BD and CT at the point of upgrade. Your BD and CT WILL be in unicode, however, if you perform a new install. You WILL experience a performance hit of 25 to 45 percent if your BD and CT are NOT in unicode format, as E1 will perform a non-unicode to unicode to non-unicode conversion internally (all E1 code is written to only accept unicode). So, the answer is, if your business is willing to deal with the performance hit, your data and control tables do NOT have to be in unicode.

True, except for one detail... you only have the option to not upgrade to Unicode if you use a western european (ie windows codepage 1252) language, if I recall correctly.

The upgrade mechanism does not have the capacity to perform a non-unicode to unicode conversion. In my experience, if your data and or control tables, in any code page, are non-unicode, during an upgrade you will not automatically convert. You may HAVE to manually convert afterwards (though I've not run into this, not going to say it doesn't exist), but again, the upgrade conversion programs don't have the capacity to perform the convos at the same time as the upgrade is running.

And realistically, it makes sense. After an upgrade, if you have issues and the upgrade process did both TC's AND unicode conversions, tracking down what the actual issue was caused by would be extremely difficult.

Plus, it's always nice to perform an upgrade, go-live, and know you've got the performance enhancing unicode upgrade in your pocket to make the end users believe that you are the keeper of the "pixie dust"!