reusing generated images *many* times

[kcri...@gmail.com]

With the CLI mechanism (as opposed to the old "use pretext script and they all go in the same place" mechanism) of managed directories, it seems that the easiest way to reuse generated images is to do the hackish "make a symlink to them in the external directory", which is obviously not ideal for several reasons.    

In any case, from reading the following:

* https://groups.google.com/g/pretext-support/c/Ndt91leW42w

* https://groups.google.com/g/pretext-support/c/p_3xmYf2Qh8/m/v8tVcSI9AQAJ

* https://pretextbook.org/doc/guide/html/topic-xinclude.html

it seems that the "official" way to reuse a generated image is to instead extract the code for the image and then include that code at each location.

However, doesn't that make a new copy of the figure for each location (or rather xml:id)?  In one of my existing projects which I'm trying to convert to CLI (primarily to test slides) I have quite a few figures appearing a dozen or more times.  (Yes, for legitimate reasons.)  This seems wasteful in terms of both time and space (the latter less of a concern, but maybe not if one has to pay hosting costs, or in epub?).

Any thoughts on this would be helpful, even if it's to say "out of luck, either do the extraction or hackish symlink".  Thanks.