ha yes :-) Carnot died of cholera aged 36, and Rudolf Diesel committed suicide too. Maybe I should add an appendix about mental health at the end!

Following up with actual numbers for the project, from Lulu, in Euros: List Price 43 Print cost 11 Distribution fees (read: Amazon, 50% of selling price): 21.5 Lulu share of profit : 2 Rest to author: 8,5

Because of the different prices on different locales in different currencies the actual share I receive averages 7€ (gross revenue before income tax, although in my case the yearly income is too small to trigger it where I live).

For books sold directly on Lulu List price 43 Print cost 11 Lulu share of profit: 6.5 Rest to author: 25.5

The mindset should not be "this is all that’s left for me", however: a book is many things at once and for better or for worse, Amazon creates a big part of it. Kevin kelly has some excellent advice at https://kk.org/thetechnium/everything-i-know-about-self-publ...

Amazon takes the lion’s share, and then the rest of the pie looks very different depending on which route you go. Big publishers print in batches and have very low print/distribution costs. I ended up on the other end of the spectrum, self-publishing with Lulu (print-on-demand, so much higher costs). I wrote an article in French on exploring the economics of textbooks, from the open-source point of view, a few years back: https://framablog.org/2022/01/20/mais-ou-sont-les-livres-uni...

Glad the book can help! Regarding image size, it’s on my todo list now (see thread further down the page).

I am using Payhip for the transaction, and then Paypal. Fees end up averaging 0,5€. Calculating in terms of percentage is not really helpful here at such low prices: there is an incompressible flat rate pretty much anywhere you go. See it differently: these intermediary fees go towards making the UX smooth, the payment secure, reducing fraud etc. I want to spend my energy on the content, not making a credit card form :-)

It’s accessible with end-of-high school math: you need to be okay-ish with integration and derivation. Practicing engineers might find it a bit slow, but that’s for each to decide :-)

Thank you for the compliment! I am touched by the generosity of the HN people today.

Since you mention VAEs: you might be interested in Active Inference, which is built on minimizing variational free energy. Same E−TS form as the thermodynamic kind (Helmholtz free energy), so it does fit your thermo/info/NN question. And the negative free energy is equivalent to the evidence lower bound used in generative models like VAEs. Lots of really fascinating connections between all of these topics :)

There's a free 2022 textbook: https://direct.mit.edu/books/oa-monograph/5299/Active-Infere...

And a brand new one (though it's a bit pricey): https://mitpress.mit.edu/9780262050951/fundamentals-of-activ...

Thank you! To be honest I think there is not much there because when I started 17 years ago my technical skills were much shallower, plotting a diagram to visualize/compare two models would have been quite a lot of work. But if I can be frank, at undergraduate level playing with equations of state is not expected nor a requirement. I’m happy to just say "there are others and they all have some common things". I‘d much rather the students and I spend our energy wrapping our heads around say the concept of entropy and what it allows us to do. Regardless, I appreciate your comment muchly (and shall look up CoolProp tomorrow).

Thank you! ^^

The payment for the paid-for PDF is processed through Payhip. As far as I can test here in Germany no cell number is required. Maybe you trigger some extra verification that I don’t?

As for the email, I require it so that I can subscribe all of my readers to a weekly newsl… haha just joking. Juuuuust joking :-) If you try it, you will find that a bogus email address works just fine.

Thanks! I think most of the weight comes from the PNG diagrams, but I don’t actually know: I’ll put it on my to-do list to investigate, maybe there are some easy wins here.

Does it matter?

I used to think that way about the publishing industry too, but I now think that this description is not deserved. Just observing it from an operational perspective: a laptop to type, software to edit, book printers that will talk to you, payment rails, and even access to world-class distribution: it’s all available to us now. On the other side of the market, students and teachers are not locked in either. Barriers to entry are nonexistent. For context, see the links I posted in my other comments on this page. It’s all in our hands!

I would bet that you can find used calculus textbooks quite easily.

I did chemical engineering long ago (how long? Computerized process control was just being introduced into the curriculum, and we learned to program in Fortran). I did several calculus courses, but it was always just a matter of memorizing techniques.

Much later, I came across Calculus Made Easy on Project Gutenberg, a textbook from 1914 that actually helped me to understand why calculus works, instead of just treating it like magic.

https://www.gutenberg.org/files/33283/33283-pdf.pdf

Calculus hasn't changed a whole lot. There are probably better books for learning than I used in the 1970s, but I have to believe that you can find pretty decent older calculus texts for not a lot.

All textbooks are free if you know where to look

A scan of contents list suggests that it is mostly about heat engines. No mention of chemistry. Chemical reactions are mentioned in passing in the text but with no detail. Also no obvious signs of much interest in conduction, convection, or radiation.

So it's fairly narrow focus, but that is not necessarily a bad thing.

I suggest clicking on the link and easily discover for yourself.