Sending the escape after its target made things complicated for upcoming
stuff I want to add. Although it makes `recv.asm` slightly larger, it's really
worth it.

It seems that my little "@.c" trick only worked in BSD make...

Some compilers don't like assigning stdout statically.

Use it in shell instead of using popen()

Also, removed leftover in runbin.c (what did it do there??) that
could result in tests falsely passing.

Things are now much simpler.

HL, instead of being preserved, is set to the character following
the last read character.

Also, make parseHexadecimal "tail" (HL). Soon, all routines in lib/parse
will do that, making the life of lib/expr easier.

Instead of going left and right, finding operators chars and replacing them
with nulls, we parse expressions in a more orderly manner, one chunk at a
time. I think it qualifies as "recursive descent", but I'm not sure.

This allows us to preserve the string we parse and should also make the
implementation of parens much easier.

For some reason, I've mistakenly disabled tests in test_expr without noticing
and I also broke "-123" parsing. Fixed.

This should make tests a bit more convenient to write and debug.

Moreover, begin de de-IX-ization of parseExpr. I have, in a local WIP, a
parseExpr implemented using a recursive descent algo, it passes all tests, but
it unfortunately assembles a faulty zasm. I have to find the expressions that
it doesn't parse properly.

But before I do that, I prefer to commit these significant improvements I've
been making to tests harness in parallel of this development.

This completes instruction support for the Blink tn45 example.

Also, add a "real world" example in AVRA tests, a blink program on
a ATtiny45. Some instructions are commented out because they aren't
implemented yet, but not many.

The output of the program has been verified against AVRA's own
output.

Use IX directly for argspec rows instead of going through DE. It saves a bit
of processing. The code was this way because I initially didn't use IX at all,
so as code evolved, that DE translation stayed as an artifact.

We now treat the block device as fixed-size rather than trying to grow it in response
to kernel activity.

Previously, if you tried to create 2 files in a row, only the first 1 would actually appear,
because the device only ever got larger when a byte was written immediately past the end of
the device.

Now we just let the kernel write bytes to the disk anywhere, so new files can be created even
when the previous file is not completely full.

Also, fix buffer overrun in reading filesystem image, and use a more idiomatic fgetc loop.

It wasn't used much, so I replaced its use in the kernel with direct code
and moved the routine in apps/ed, the only other place where it was used.

The order in which cfspack includes its file depend on the platform. To have
tests that reliably pass, test.cfs has to be committed in.

Also, increase STDIO_BUFSIZE to 0x40 so that those while loops work.

Also, always end the CFS chain with a stop block.

fixes #55 #56 #57

ref #80