Re: [xml-dev] The illusion of simplicity and low cost in datadesign and

["Liam R. E. Quin" <liam@fromoldbooks.org>]

On Thu, 2022-08-11 at 22:05 +0100, Michael Kay wrote:
> 
> Eric Raymond's quote would be valid if the operating system
> constrained all files to be XML, say. But it's not valid for an
> operating system that can support any file format, but can also
> reliably tell the application that it's reading an XML file rather
> than a JSON file (or more precisely, one that can identity the
> application that wrote the file and tell you what claims that
> application made about the file format).
> 

MacOS (the pre-Unix OS) could do this (to some extent). It was a pain,
and made the system massively less flexible than Unix. For example, if
i make an SVG file in Inkscape (to use today's application names), i
might then want to look at it in Inkview or in a Web browser, and i
might want to edit it in Oxygen XML Editor.

Putting the applications before the documents is the wrong way round.

However, i do agree that file extensions suck, and the Unix "file"
command (Ian Darwin of SoftQuad wrote the public domain version of
that) is incredibly useful but limited and not 100% reliable. Of
course, nothing is 100% reliable except the statement that noting is
100% reliable...

"Open" being able to offer a choices of likely tools seems more useful
than identifying a single application, to me at least.

liam

-- 
Liam Quin, https://www.delightfulcomputing.com/
Available for XML/Document/Information Architecture/XSLT/
XSL/XQuery/Web/Text Processing/A11Y training, work & consulting.
Barefoot Web-slave, antique illustrations:  http://www.fromoldbooks.org