search

vectorgraphics / asymptote

2D & 3D TeX-Aware Vector Graphics Language
https://asymptote.sourceforge.io/
GNU General Public License v3.0
542 stars 90 forks source link

Consistency of coordinate systems of returned `point`s of inversion-related functions in `geometry.asy` #355

Closed yarusome closed 1 year ago

yarusome commented 1 year ago

I am working on a new PR to fix all the bugs in the inversion section of geometry.sty, but one issue holds me back for now.

Since the returned values of many functions in this section are points or inversions, which has points as members, and each point may in a different coordinate system, it seems necessary to decide what convention to follow when fixing the bugs to achieve consistency of the returned points. Here are a few proposals:

  1. All the returned points must be in currentcoordsys.
  2. All the returned points must be in defaultcoordsys.
  3. Decide case by case which coordinate system the returned point must be in for each function, and when there is no obviously rational choice, choose currentcoordsys (or defaultcoordsys).

This might also affect consistency with other sections of geometry.asy, and the documentation needs fixes to reflect this. (Incidentally, the TeX source file of the doc is nowhere to be found now since https://www.piprime.fr seems to have shut down.)

johncbowman commented 1 year ago

Fortunately I found a local copy of the documentation that I have now uploaded here:

https://github.com/vectorgraphics/asymptote-geometry-documentation

johncbowman commented 1 year ago

You may need to apply these two pull requests: https://github.com/pivaldi/asymptote-geometry-documentation/pulls I can add you as a developer on the above fork of this repo on request.

yarusome commented 1 year ago

Then I'd go with the first proposal for now. After the patches have been merged, the doc fixes will be in a different PR.