XML Considered Harmful

[Joe Pfeiffer]

Eli the Bearded <*@eli.users.panix.com> writes:

> In comp.lang.python, Michael F. Stemper <michael.stemper at gmail.com> wrote:
>> I've heard of JSON, but never done anything with it.
>
> You probably have used it inadvertantly on a regular basis over the
> past few years. Websites live on it.

If the user has any interaction whatever with the formats being used to
transfer data then something is very, very wrong.  Someone using a
website built on JSON isn't using JSON in any meaningful sense of the
term.

>> How does CSV handle hierarchical data? For instance, I have
>> generators[1], each of which has a name, a fuel and one or more
>> incremental heat rate curves. Each fuel has a name, UOM, heat content,
>> and price. Each incremental cost curve has a name, and a series of
>> ordered pairs (representing a piecewise linear curve).
>> 
>> Can CSV files model this sort of situation?
>
> Can a string of ones and zeros encode the sounds of Bach, the images
> of his sheet music, the details to reproduce his bust in melted plastic
> extruded from nozzle under the control of machines?
>
> Yes, CSV files can model that. But it would not be my first choice of
> data format. (Neither would JSON.) I'd probably use XML.
>
> I rather suspect that all (many) of those genomes that end up in
> Microsoft Excel files get there via a CSV export from a command line
> tool. Once you can model life in CSV, everything seems possible.

Whenever someone asks "can this be done?" in any sort of computer
related question, the real question is "is this practical?"  I have hazy
memories of seeing a Turing Machine implemented in an Excel spreadsheet,
so *anything* can, with sufficiently ridiculous amounts of work.  That's
not really helpful here.

>> [1] The kind made of tons of iron and copper, filled with oil, and
>> rotating at 1800 rpm.
>
> Those are rather hard to model in CSV, too, but I'm sure it could be
> done.

So let's try to point him at representations that are easy.