Case sensitive problem in Set-Analysis

[lukaspuschner]

Hello everyone,
i have a question to the logic that is used in Set-Analysis. Set-Analysis is in some cases casesensetive(CS) in some cases caseinsensetive(IS). It looks like a two stage logic, first QV compares with a = and if no data matches he uses the like. But the = is CS and the like is IS. The selection/data-model logic in QV is CS, so it is a problem if i use Set-Analysis to simulate a users selection. Is there a way to use a command like the SQL Collate-Statement, to force the Like-Statement to be CS.
Please see the attatched file,where you can see the Problem.
I would be very thankful some good solutions.
Lukas

[tresesco]

Really strange! And a good notice. I never noticed specially this:

[juleshartley]

I'm afraid I haven't got a solution for you - but that's a very interesting analysis! Thanks.

[Gysbert_Wassenaar]

It's a bug (see here), but I don't know of any workaround.

[juleshartley]

I can only think of something really horrible like switching out lower case letters for an integer... and then using that in the set analysis. ouch!