Yes this works now as expected.
Did you consider point 3. from the list in#12391? (If you think this doesn't apply, that's fine.)

Maybe the text should clearly state here that in order to see the effect one would need to compare the shown (addplt.show() at the end?) with the saved figure? Maybe the linewanttoprint = False can be appended with a comment# set to True and compare saved figure or similar?

Can we include images in the tutorials? I guess I could just include the png.

WRT item 3 I was t sure it was necessary but happy to do it if you feel it helps

I think some of the confusion comes from looking at the produced images, so if at least the code has the savefig command in it, it would point more towards where to look for the feature.

Ithink you can use rst inside the examples, so.. image:: some.png might just work.

The manually added figures are a bit... HUGE.

But appart this looks good. With both figures shown like this I think the whole section becomes really clear.

Seems better now, maybe a little small, but I think thats fine:

https://14431-1385122-gh.circle-artifacts.com/0/home/circleci/project/doc/build/html/tutorials/intermediate/constrainedlayout_guide.html#sphx-glr-tutorials-intermediate-constrainedlayout-guide-py

What's the reason for the difference? Shouldn't it save with the same dpi as it shows?

Ah, you changed the width... Why not produce the figure withsavefig(..., dpi=100) locally?

I had to hard-set the width to 300 px in therst. We are only trying to show what the layout looks like, not what the exact sizing of the png is.

Squashed; final product athttps://14460-1385122-gh.circle-artifacts.com/0/home/circleci/project/doc/build/html/tutorials/intermediate/constrainedlayout_guide.html#sphx-glr-tutorials-intermediate-constrainedlayout-guide-py

Thanks for the help@ImportanceOfBeingErnest

It seems one drawback of this is that now every time you build the docs it'll create two png images in the tutorials folder, which will then end up in any commit (if one doesn't pay attention).