Richard Heathfield said:
Writing a book is not the hard part. Getting a publisher to pay /you/
(rather than vice versa) is the hard part.
hadn't thought of publisher...
would have assumed just making it available online, but I guess this is more
of a paper...
Ecclesiastes 12:12
<snip>
yes, ok.
I guess in a way, books are more the organized form of the otherwise
disorganized information trudging...
one of the ideas I had in the past had been a 'vectorized' physics book
(AKA: physics reworked to largely replace all this dirty old trigonometry
with vectors), but then I have disovered, apparently newer college-level
physics books have already done this...
the only real next step along this road would then be to explain how to do
real-time computer simulations of most of this stuff, but this would be
about like an annotated physics engine in book form.
the question then is that of it having a point...
it is unclear if there is much other topic at present, since most other
topics which come to mind either already have books, or don't need them.
VMs are another possibility, but this topic is very fluid and would be
difficult to give adequate coverage.
and, I can't probably say much, having essentially created something almost
a sort of monstrosity (need a bytecode? how about interpreted x86?...). it
works, but would seem at the outset to be a distasteful design, almost an
ad-hoc chimera of different technologies. it is not clear that this would be
a good direction to present (as opposed to the more clean/sterile designs
pursued by many other VMs...).
many people would probably stop just as soon as x86 were proposed as a
bytecode (or, even likely an x86 derived bytecode).
(horrid as it is though, I am left thinking x86 may actually be a worthwhile
model for a low-level bytecode...).
there are a few possible tweaks, but as noted, this proposition is sort of a
"house of cards" (it may be tempting to "clean house" in a few "innocent"
ways, but in so doing destroy any real merit of following this path, as one
may find something "x86-like but not binary compatible" to be an almost
worthless prospect...).