[GRASS-dev] A0-A10 paper size calculations

[Moritz Lennert]

On 02/07/08 09:07, Hamish wrote:
> Hi,
> 
> I notice in ps/ps.map/paper.h that A0 is slightly different than other
> published values. Actually other values found on the 'net are all over
> the place.
> 
> In an effort to "do it right" I've recalculated them as follows, from the
> fundamental A0=1m^2; W:H = 1:sqrt(2) definition:
> 
> size|width_mm|height_mm|width_inch|height_inch
> A0|840.90|1189.21|33.106|46.819
> A1|594.60|840.90|23.410|33.106
> A2|420.45|594.60|16.553|23.410
> A3|297.30|420.45|11.705|16.553
> A4|210.22|297.30|8.277|11.705
> A5|148.65|210.22|5.852|8.277
> A6|105.11|148.65|4.138|5.852
> A7|74.33|105.11|2.926|4.138
> A8|52.56|74.33|2.069|2.926
> A9|37.16|52.56|1.463|2.069
> A10|26.28|37.16|1.035|1.463
> 
> the Matlab (should work in GNU Octave) script to do that is attached.
> 
> I notice these still don't line up exactly with the more trusted values I
> can find. Usually the error is minor rounding in the wrong direction on
> the least significant digit, or truncation. I assume cascading rounding
> or conversion errors are generally to blame. (width becomes length from
> the previous size, etc)
> 
> I'm not too concerned if ps.map paper sizes end up being off < 1/72"
> (1pt), but as there is so much noise out there on this I though it
> would be nice to use trusted values.
> 
> Before I commit the above values, does anyone see a problem in my script
> or know if the sizes are actually defined by ISO as "rounded" values in
> some unit, versus the mathematical approach that I've taken?
> 
> any requests on what sizes we should include? currently for ISO sizes we
> only define A0-A4.

I've never seen these sizes with decimals (my usual sources: [1],[2]), 
so I actually think that they are defined rounded to mm.

And quite honestly, I don't think that submillimetric measurements are 
really of any importance...


[1] http://www.cl.cam.ac.uk/~mgk25/iso-paper.html
[2] http://www.edsebooks.com/paper/papersize.html

Moritz