I've made an A4 poster and want to scale it down to A5. I made sure the proportion padlock by the width and height boxes was turned on, changed the width to 148mm but the height didn't change to 210mm. It's less than 1mm out, at 209.314, but it shouldn't be out at all, right?
I'm probably doing something stupid, so I'm sorry, but any help would be most appreciated.

A4: 210 mm/297 mm
A5: 148 mm/210 mm

BUT
210/297=0,7070707
148/210=0,7047619

while in theory they are both (2^(0,5))/2=0,7071068

so if you resize by the width,
the height will be 148*297/210=209,3142857 mm by calculation.

Would add a right sized rectangle to the bottom and group it with the scaled poster (after center alignment) as a workaround.

So I'm not doing anything wrong?
But if A5 is half A4 then shouldn't the width be 148.5mm?
And even changing the width to 148.5 produced a slightly off height at 210.021.
Thanks for the work-around, though I'm still concerned about why I even need one.

If you have objects with an outline, and if that outline doesn't change width when rescaling the object, then there will be some minor errors to final size calculations.

If outline change size as the total size changes, or if you have no outlines - then in theory the size calculation should match the object size repported by Inkscape.

I hadn't thought of that, but I don't think it is applicable; I have a background behind everything which is the size of an A4 page with no stroke, so surely that shouldn't be causing the issue. And I have many objects but I grouped them before scaling, or is that what could be causing the minor error?

A0 is the base of the system -iso 216-.

Theoretically A0 should have an area of 1 m^2, with a width/height ratio of 1/2^0,5.
Which would call for an A0 that is [2^(-0,25)]*1000 mm / (2^0,25)*1000 mm =
840,8964153 mm / 1189,2071150 mm width/height ratio;

an A4 that is [2^(-2,25)]*1000 mm / [2^(-1,75)]*1000 mm =
210,2241038 mm / 297,3017788 mm,
and
an A5 that is [2^(-2,75)]*1000 mm / [2^(-2,25)]*1000 mm=
148,6508894 mm / 210,2241038 mm.


If you check the original iso 216 and the older din equivalent, you will find the maximum +-errors that the size of the paper can differ from the optimal rounded values.

Then again, each trimming uses up a bit of the paper size, as there is no zero width trimming.
And paper size differs with air humidity and temperature, fibers used etc.


You shouldn't be much concerned about that.
Even if your image is not exactly of the rounded values, if you print from pdf, you can chose actual size.


If you need the print to go to the paper edge, then you will need to crop the paper after printing -which would need a different preparation.