Since I - and a few others - liked
my org-mode SDL2, this post was The only thing I am unhappy with is
intended to be about that. I how we can get org-mode tables into
formalized (heh. [sic]) the headings the code tangle (as numeric C
of my sdl2 C org template and arrays, generally). This consists of
rejigged it a bit. I made it so a making a table with a #+name:
code block named main looks like orgtable and filling it with
this: numbers, using either TBLFM
calc/spreadsheet domain specific
#+begin_src C language or elisp. You could also
generate it with any code having
<<declare-variables>> (and doing) a :output table (which
is default for anything org-mode
<<start-sdl2-etc>> thinks looks table-like).
<<after-start>> My problem is that I am putting
named tables in the intimidatingly
<<loop-switch>> named Symbol macro tangles, but then
the main code block header :var line
<<stop-sdl2-etc>> needs somethng like
c2darrayname=orgtablename appended
<<after-stop>> to it, which is the only case of
modifying the main block and is
#+end_src no-where near the logic the table is
involved in. Further, a variable has
Though I refer you to 0eg.org in my been introduced far away from the
1org-drafts for a complete example. Variable declaration block. If I
Please change could, I would actually like to move
all variables out of C and into
#+HEADER: :includes orgmode (not in the main header),
"/usr/local/include/SDL2/SDL.h" but I don't think this will
particularly work for SDL2 structs
to meet your needs (probably to or maybe structs generally. There
are a few ways this could go, but
#+HEADER: :includes <SDL2/SDL.h> none of them appealled to me yet.
). I love how this completely Since everything structural and
functionless SDL2 template works, organizational has been referred
though a little bit of it bears into orgmode and out of C, having
exposition and I will gradually make absolutely no functions is nice.
my own game arcade to work on the After all, you would have been
details. pragma inlining them anyway, right?
But there are two things I don't
Making the template into a game like. Annoyingly, SDL2's timers
involves declaring your variables in feature works by function callbacks,
declare-variables, initializing them so I guess we're stuck with adding
(or something) in after-start, and functions somewhere. More
then in the loop heading clone the precipitously, this highly organized
**loop***Basic section for each game and organizable tangle is not at all
mode you will have. Inside Basic object oriented. In C we would by
there is a basic-event-pump and a convention sometimes prefix
basic-action. Fill those in with the functions with the name of the sort
inputs you want to have in the of struct they act upon, yielding
your-event-pump, and fill in something like a namespace and
your-action with whatever you want (probably non-generic) methods. Alan
to happen in this game mode. Put Kay's
your mode into the loop-switch mode all-you-can-do-is-send-a-message is
enum/switch like BASIC. From there, not part of my org template.
all of everything about your program
is absorbed into org-mode
tangling/export/execution/debugging.
------------------------------end-page-1----------------------------------
My probable solution to both of
those is to just tangle the template
inside of ECL's sffi (see: Other
stuff I've programmed) which will
work extremely well. And then it's a
lisp program and has CLOS with
closures and classes and so forth.
The problem is that this is a
titanic and intrusive dependency,
implying Boehm garbage collection
and libecl and so forth. And the
snappy C slows down to that sort of
heavier pace. Further, non-elisp
lisps basically don't play well in
orgmode or emacs at all;they have a
mixture of redundancy and rivalry.
ECL especially I would have to write
a new ob-ecl for to handle its use
of common lisp's #'compile-file (and
C).
Dear oh dear, I didn't even get to
tables yet which were all I wanted
to talk about. I will table the
tables for today.
In other vanes
Other than my SDL2 C template and a
tiny bit of existing work on my own
i2p gopher usage, I am going to
start paying attention to the great
stuff everyone else is making and
has made - so I will try to
investigate more about almost
everyone else's signature gopher
clients, for example. cside's bbc
news gopher. Actually thinking about
jns' justify.
Snipped somehow, really should get a newer version of justify:
On the off chance C. is reading this, ed is a positively magnificent editor
as is its extremely oppositely-idiomed spiritual successor (plan 9) sam -d.
I also came to understand much more about vi and ex after getting into ed.
Seeing what you decide to show yourself works really well, which is odd
considering the absolute prevalance of screen editors that inundate you with
visual information you often don't want or care about.