String and Special String migration from 8 to 9

Author: jim.e.mitchell@oregon.gov (mitcheje)

Hi all,

In migrating from version 8 to 9, I'm trying to decide which migration utility option to choose:

1 - Allow modifications to enforce V8 compatible behavior
or
2 - Do nothing to achieve V9 default behavior

We don't use Unicode (and don't plan to) but what would be the harm in going with version 9 defaults (option 2)? It looks like it just expands the character set and would require less changes to the repository.

On the other hand, would it cause unwanted behavior in our applications? We just use basic Uniface client/server (no web applications).

Any advice would be appreciated.

Jim

4 Comments

  1. Good Morning,

    we have migrated all of our Application with option 2, to get U9 behavior.

    The first migration we have done in 2009, since we faced no problem with setting string (special string) to all fields.

    So in my opinion it is okay to get U9 behavior.

    And you use only Unicode if you also set your Entity-Interface definition to W*. If you don't do it, you will get "normal" character set.

    Best regards

    Thomas


    Author: Thomas.Young (thomas.young@young-consulting.de)
  2. Thanks Thomas. That's what I needed to know.

    Jim


    Author: mitcheje (jim.e.mitchell@oregon.gov)
  3. Hi,

    one think I have forgot to mention.

    Have a look at the interface definitions.

    Maybe you have to set MUL to FUL to avoid wrong characterset transfer to the database.

    If I remember correct Uniface set all empty Interface definition when datatype is string to MUL automatically.

    We use a simple form update this direct in the repos.

    Best regards

    Thomas


    Author: Thomas.Young (thomas.young@young-consulting.de)
  4. Thomas,

    Thank you for the additional information. I will look at our Interface definitions and make changes if needed.

    Jim


    Author: mitcheje (jim.e.mitchell@oregon.gov)