I could imagine these could be useful to be defined by computer libraries, as (see article) it would help reduce large reductions in accuracy for some calculations (e.g. by subtracting similar numbers, or wanting to calculate sin^2).
1,634 karma · joined January 19, 2013
I am the upstream author of the Veusz scientific plotting package: https://veusz.github.io/
Github: https://github.com/jeremysanders and https://github.com/veusz/veusz
I could imagine these could be useful to be defined by computer libraries, as (see article) it would help reduce large reductions in accuracy for some calculations (e.g. by subtracting similar numbers, or wanting to calculate sin^2).
They led the teams that gave the decisive proof that there is black hole at the galactic centre. The teams developed techniques to precisely measure the orbits of the stars which orbit the central black hole. Looking at the speed of the stars, particularly as they pass close to the object, you can measure its mass and put a constraint on its size. The only thing which is small enough and heavy enough is a black hole. Doing this required lots of advances in data analysis and adaptive optics to be able to measure the stars orbits to enough precision.
It actually worked quite well, though was pretty memory hungry.
I think the biggest issue is the social problem of the long and arduous process to become a trusted Debian Developer. Although there's now an easier Debian Maintainer option for restricted access, it's still not easy to become one. Other people have to go through a developer to update or add packages. If you can't find a responsive developer, then package changes can sit for months or years in a broken state before they are applied (as I have found out). The whole packaging process is a frustrating exercise in bureaucracy compared to other non-Debian distributions (e.g. Fedora), which is no fun to be involved in for most people.
It's a shame they've exited when they used to have such innovative machines.