Re: [Docutils-users] OT: DSV vs CSV

[Alan G I. <ai...@am...>]

On 10/17/2011 11:09 PM, Paul Tremblay wrote:
> So then I suppose python language got it wrong, and so did the developers at restructuredtext.

Again, you are conflating two things:
the existence of functionality with a correct default,
and the appended flexibility to handle related DSV formats.
I already explained this.

Again, in the context of scientific data, if you sent
someone a ``.csv`` file that used any delimiter except
a comma, it would be like playing a prank.

Similarly, if you wrote a CSV parser that could not
handle comma separaters, it would be completely baffling.
But csv parsers that require commas exist.

The core meaning of "comma-separated values" is that there
is a comma separator. Initially, that was its unambiguous meaning.
But because many CSV parsers have appended the ability
to accept other delimiters (especially whitespace),
there has been a small drift in meaning.  But not to the
extent you imply.  For example, it would still be very
puzzling behavior to send someone a file with a .csv
extension that uses white-space delimiters.

Can we at least agree that would be bad behavior?

Cheers,
Alan