I am trying to promote the use of svg-exported
draw.io diagrams for the standard image format for technical documentation at my company. These images are stored in a git (Bitbucket Server) repo, alongside the source code they describe, and are referred to by relative paths in markdown image tags. SVG are generally smaller and render better in a browser bc of their inherent scaling design. In my experiments with png and svg, however, I've encountered some problems with using svg in this way, that I seek help and clarity on:
1. If my goal is to keep svg diagram size to a minimum,
must I refrain from using raster/bitmap type Shapes from draw.io shape panel? As soon as I add a raster/bitmap shape to my diagram, like a picture of a PC or a Equipment Rack, the size of the exported svg balloons to greater than the size of the corresponding bmp.
2. When exporting the .svg, if I don't select "Embed Images", the file is of course smaller, but these raster images don't render, but more importantly, the svg doesn't render at all in Bitbucket markdown.
Are there ways to filter
draw.io shapes to get vector-only shapes, so that when I insert them into my diagram, and export as svg
without Embedding Images, the diagram will be as small as possible?