mixing Unicode with non-unicode

[former_member204746]

Hi, Wwe have ECC 6.0 and BI 7.0 running on Windows 2003 and Oracle 10.2.0.4 all on non-unicode

We are about to install another BI 7.0 landscape.

If I install this new BI 7.0 on unicode, should I expect issues when communicating between systems such as BI loads?

Thanks.

[markus_doehr2 · Accepted solution]

If the ERP does not use other codepages than Latin-1 you won't have a problem.

Be aware of the fact, that if you currently connect a portal (BI-Java) to the non-Unicode BI-backend, you are running an unsupported combination (Note 975768 - Deprecation of Java features with non-Unicode Backend).

Markus

[former_member204746 · Accepted solution]

Very interesting.

BI-Java are in the works for 2009, so, I will need to convert my first BW instance to unicode... ouch, that part was not planned. so, I guess I will have a lot of work in 2009.

so, I will install the new BI landscape using unicode, that will save me from converting this system later.

thanks Markus.

[markus_doehr2 · Accepted solution]

BI-Java are in the works for 2009, so, I will need to convert my first BW instance to unicode... ouch, that part was not planned. so, I guess I will have a lot of work in 2009.

Technically it might work with a non-Unicode backend but you never know...

If your BI system is a single codepage system the UC conversion is not that big. The tools are meanwhile so good you shouldn´t have any big issues. Make sure you are on the latest BASIS support package, this will help a lot.

so, I will install the new BI landscape using unicode, that will save me from converting this system later.

yes - definitely.

And if the connected ERP (or CRM or whatever) backend is only single codepaged there won´t be any problems when you exchange data.

Markus

[former_member204746 · Accepted solution]

thanks again for your help!