Re: Restructuring program to cope with modern database layouts

Hello John,

Am 21.05.2025 um 07:26 schrieb John Yeung <gallium.arsenide@xxxxxxxxx>:

On Mon, May 19, 2025 at 7:35 AM cesco via RPG400-L
<rpg400-l@xxxxxxxxxxxxxxxxxx> wrote:

But, at the end of the day, you need to store the thing somewhere, print somewhere etc. (despite the "infinite string" illusion of some languages).

Yes, but if you're not bound by the fixed-byte-size storage model imposed by physical files, the "illusion" can be much more easily preserved. If you're working exclusively with stream files, for example, it's somewhere between trivial and automatic to just use more bytes as needed, and always present the programmer or user the full number of characters they expected and asked for.

I wonder how SQLite does this under the covers. They mapped different data types to storage classes. As I understand, things like e. g. varchar are an internal alias to "text". But then, SQLite doesn't bring with it a decades long heritage.

https://www.sqlite.org/datatype3.html

Maybe IBM would consider to come up with a new object type, *EFILE, or *SQLDTA along this thinking? A new internal format better fitting modern requirements such as dynamic byte lengths due to UTF-8, etc. I'd even suggest to not create a compatibility layer for accessing those objects from the classic positional APIs. Handling of data is done purely with SQL. However, there should be a migration tool, akin to cvtrpgsrc, and some classic green screen applications to interact with those extended functionality files.

Just an idea worth carving out and passing on as an idea to IBM ideas website?

:wq! PoC