Just a few first impressions during and after the initial installation of 2020 Technical Preview;

The installation incl IIS (checking) worked perfectly.

Question: I noticed the when the Studio was started at the end of the installation, that it was started as [Administrator]. Was that intentional?

I fully understand that not all things are ready yet but I just report what I found, and perhaps you just want to make notes for later.

A tiny thing during the installation phase;
Installation Wizard - End-User license agreement last row:
"Revised: March 2019" (Maybe update the date later on)

Edited some view labels to other languages
Compiled and run the OE as 64-bit and also entered data in other languages.
- It all worked exactly as outlined in the documentation.

Successfully Converted OE to MS-SQL
- Pasted the Chinese equivalent of the "Access Miles" to the Customer view and saved. Could save & re-find records just fine.
- Still had the record active in the Customer view and then closed the program by clicking the applications upper-right "x".
- "Unhandled Exception in Debugged Application" and the debugger points to "Start_UI" after that.

Noticed that the Order.src psVersion was still set to "19.1" and so was the Activate_About (Yeah, I understand it is still early in the process, but still wanted to point it out so it doesn't go under the radar)

During the conversion of OE to SQL I saw that the wizard "ANSI" data format was checked. That was of course a bit confusing. But this raises a more important question and that is that the collation sequence for a particular database will be very important when using Unicode in SQL.

What I do in DUF (The Database Update Framework) is to present the user with a drop-down box to select an SQL collation (I enumerate all current collations currently on the SQL server) to the user to select from on the "Create Database" dialog. In DUF there is also a "psCollation" property that can be set when an SQL database is created with DUF code. Maybe this not the best way to handle collation in DataFlex, but selecting the correct collation will become increasingly more important when using Unicode.

All in all it has been a very successful and impressive first test(s) with the DataFlex 2020 Technical Preview, including the excellent documentation. I very much looking forward to make more in depth tests with my own applications.

I'm very excited about the new DataFlex 2020 version. Well done and congratulations to the whole development team!