On Fri, 18 Sep 2020 18:59:20 +0200
Hans Hagen
On 9/18/2020 5:56 PM, Tomas Hala wrote:
Hi Hans,
# We also considered that in order to enforce the use of si units the # 'in' should go away but that has been postponed (it is anyway last # in the checked keywords so has no overhead).
I do not remember details of the discussion about 'in' but I would like to say that 'in' is quite important unit, especially for Americans. We have to take into consideration that paper and other things are sold in inches in the USA and it is not very comfortable to recalculate it to the metric system.
We can let it depend on the outcome of the elections and foreign politics ... we can use the unit as bargain option (and Alan uses thise french units anyway).
So I vote for keeping it forever, eg. till Americans will completely convert to SI. It was a joke (seems you didn't pay attention when arthur and I discussed it the break ... we even mentioned 'meters' as possible units and dropping cm and mm). Dropping nd and nc is because they were never accepted as standard.
An inch is official tex and we're tex aren't we? But in luametatex the inch is at the bottom of the scanned unit list .. you can guess the order that we like most; scanning for in takes 25% more time than for mm but only Idris will notice that (and when he used "truein" it will be twice as slow).
I had a colleague in France who liked doing all of his calculations in meters. The problem arose when he wrote a program that needed to take high powers of quantities whose natural units were rather nanometers or even Angstroms. This quickly led to mathematical underflows, solved by putting everything into double precision or better (this was many years ago, when processors and compilers had more limitations). I vote to eliminate the pt and the bp. We could introduce the Covid, the diameter of the viral capsid. Alan