Show HN: Boxy SVG editor

If I recall correctly, the simplest algorithm was as follows:

1. When user begins scaling an object, compute the initial bounding box (bbox) of the object.

2. On each pointer move determine the scaled bbox into which the object should fit. If user holds down the shift key, adjust the scaled bbox so that it has the same aspect ratio as the initial bbox.

3. Finally, compute the transformation matrix that transforms the initial bbox into the scaled bbox and apply it as a "transform" attribute on the object.

The current implementation in Boxy SVG is a bit more complicated because it has to handle all the edge cases like multiple selected objects that belong to different groups or objects that have zero width or height (e.g. straight horizontal or vertical line).