[MetaCRS] Specifying Coordinate Order in Test Data File...Again

[Sunburned Surveyor]

Martin and I had a short discussion on the Proj4J mailing list about
the need to specify the ordinate type in the metacrs test data file.
It seems I added some test data northing first - easting second, which
subsequently broke our code that parsed and executed the test data
file. This backwards entry of coordinates comes from my background as
a surveyor. That got Martin and I thinking it might be a good idea to
reconsider introducing the ordinate type back into the the test data
file. So instead of something like this:

2225563.01, 6352118.95, 10.21

we would have something like this:

y:2225563.01, x:6352118.95, z:10.21

This change allows makes the format of the test data file a little
looser by specifying the ordinate type with the ordinate value,
instead of by convention in the order of the "columns" of the CSV
file.

What are the disadvantages of this approach? Does it allow us to code
more robust test data file parsers? Does it make parsing more
complicated?

Are there any objections to this suggestion? Martin, did I communicate
this idea correctly?

Landon