June 20, 2005, 10:56 AM — I'll come straight to the point this week (for a change).
XML is not - repeat NOT - a 'file format' in the sense that most people use the phrase 'file format'.
In order to back up this strong assertion, we need to agree on some basic terms. Never an easy thing to do in this business. For the moment, let us pretend that internationalization is not a concern in this global economy we live in. For the moment, let us treat the term 'ASCII' as synonymous with plain text. This little ruse will help keep things simple without adversely impacting the core of my argument.
When we say some data is saved out of an application in ASCII format, what do we mean? We mean that if you open up the file in a simple text editor you will see letters and numbers and punctuation. It may or may not be possible to understand what the letters and numbers and punctuation mean but given enough information, it would be possible to create a file by hand in the text editor that was indistinguishable from the original. It might also be possible to edit the file with a text editor in such a way that the edited file will still work with its intended destination application.
Now, imagine in your mind's eye an application that gives you these File-Save options:
Save file as CSV
Save file as RTF
Save file as ASCII
But hang on, leaving internationalization aside as we agreed to, these are all ASCII files. It doesn't really make sense to classify them this way, any more than this makes sense:
Save file as Chevy
Save file as Harley
Save file as Motor powered vehicle
How to make sense of all this? Here is what I think "File-Save As ASCII" really means to most people. It means "Save the information in my file in a form that preserves my letters and numbers but strips out any application specific information."
That works in the example used here. CSV is ASCII but it is ASCII specifically constructed to preserve application-specific information. Namely, the row and column structure of tabular data. RTF is ASCII but it is ASCII specifically constructed to preserve application-specific information. Namely, the formatting of text into tables, paragraphs, pages and so on.
Now let's turn to XML. What might a set of save options like this mean?
Save file as XHTML
Save file as RSS
Save file as XML
D
