Re: The illusion of simplicity and low cost in data design and computing

If files were layered the way that networks are layered, the data for each
level would commence with the information needed to announce what was
immediately contained within. There would be no resorting to mythical
outside systems that somehow know things about the document.

Each layer need 1) a convention to say what it is by its metadata, and 2) a
series of conventions for how to look in its data to find enough
information about the next layer to be able to read its metadata.

XML provided the first (headers, doctyoe, namespaces, etc) but kinda fails
to provide the second (standard inline conventions to introduce embedded
notations). Which is why it struggles whenever it has to contain non-xml
data inside: hence see ludicrous approaches like requiring an external
schema  to say some element value is bin64.

Regards
Rick


On Sun, 14 Aug 2022, 8:56 pm Michael Kay, <[email protected]> wrote:

> >So perhaps the idea that the best place for the metadata of a file is IN
> the file (e.g.  as magic number, as XML header, etc) not WITH the file, is
> not hacky but the most unified and best approach?
>
> If it was layered the way protocols are layered, then there would be one
> layer at which the metadata is visible as part of the file content, and
> another layer at which you only see the content.
>
> Michael Kay
> Saxonica