Hi Tim,

This is a follow-up to our email exchange. I'm doing this on the newsgroup in case
anyone else out there is struggling with the Unicode version.

Delphi 2007
ElevateDB 1.06 build 1 (Unicode)

These are the steps which we did and problems we met during our attempt
to use ElevateDB Unicode and open converted DBISAM tables:

First attempt

1. Uninstall the earlier version of ElevateDB Unicode.
2. Reinstall ElevateDB Unicode from: edb_vcll_d2007_106b1.exe
3. After reinstall, open Delphi 2007: ElevateDB was present on the palette.

4. Try to open the project: C:\Program Files\ElevateDB VCL Standard with Source\Version
1\Delphi (Unicode) 2007\examples\demos\createdbddl\createdbddldemo.dpr
In theory that should create the tables for the samples.

5. Run and press the button Create Tables, and this is the message that I received:

 'An error was found in the statement at line 4 and column 32
  (ElevateDB Error #401 The collation ANSI does not exist)'

Presumably my DatabaseServer is missing something, because I assume that the sample project
should already have everything that it needs?

Second attempt

6. Then I tried the project:
C:\Program Files\ElevateDB VCL Standard with Source\Version 1\Delphi (Unicode)
2007\examples\cdcollector.
That is a sample project that comes complete with data.

7. Ran the project and received this error message:

'ElevateDB Error #100 There is an error in the metadata for the catalog CDCollector
(Signature, password, character set (ANSI/Unicode), or version number mismatch)'

8. Also on the CDCollector project. If I click on the property Database
on the EDBDatabase component, Delphi closes without any error message.

Third attempt

9. When I tried to open my own tables converted from DBISAM with the migrator,
I received exactly the same error:
'ElevateDB Error #100 There is an error in the metadata for the catalog CDCollector
(Signature, password, character set (ANSI/Unicode), or version number mismatch)'


If you can point me to the possible problem, I would be really grateful.

Regards,
Dan

Lexicon Software Ltd, UK

Correction to para 9:

9. When I tried to open my own tables converted from DBISAM with the migrator,
I received exactly the same error:
'ElevateDB Error #100 There is an error in the metadata for the catalog [catalog_name]
(Signature, password, character set (ANSI/Unicode), or version number mismatch)'

Sorry!

Dan

Dan,

<< 4. Try to open the project: C:\Program Files\ElevateDB VCL Standard with
Source\Version 1\Delphi (Unicode)
2007\examples\demos\createdbddl\createdbddldemo.dpr In theory that should
create the tables for the samples.  >>

Sorry about that - the examples are currently only for creating ANSI
databases.  They will need to be modified to use the 'UNI' collation instead
for the tables, etc.

<< 8. Also on the CDCollector project. If I click on the property Database
on the EDBDatabase component, Delphi closes without any error message. >>

I'll look into this, because it should just ignore any character
set/collation issues and not show any databases.

<< 9. When I tried to open my own tables converted from DBISAM with the
migrator, I received exactly the same error: >>

How did you do the migration ?   Did you use the unicode migrator and the
unicode EDB Manager ?

--
Tim Young
Elevate Software
www.elevatesoft.com

Tim,

<< the examples are currently only for creating ANSI
databases.  They will need to be modified to use the 'UNI' collation instead
for the tables, etc. >>

I should have spotted that difference in the COLLATE: many thanks for pointing it out.

<< Did you use the unicode migrator and the unicode EDB Manager ? >>

What ever I did previously was obviously wrong. I think that possibly I picked
the wrong migrator. Having carefully stepped through the Migration Guide,
the tables have now converted to Unicode just fine.

Thanks for your patience!

Regards,
Dan

Dan,

<< What ever I did previously was obviously wrong. I think that possibly I
picked the wrong migrator. Having carefully stepped through the Migration
Guide, the tables have now converted to Unicode just fine. >>

Well, there were a few bugs with the Unicode implementation initially, so
it's possible that you might have run into one of those also.

I'm glad you've got it working okay now.

--
Tim Young
Elevate Software
www.elevatesoft.com