fixes #80

That's my mega-commit you've all been waiting for.

The code for the shell share more routines with userspace apps than with kernel
units, because, well, its behavior is that of a userspace app, not a device
driver.

This created a weird situation with libraries and jump tables. Some routine
belonging to the `kernel/` directory felt weird there.

And then comes `apps/basic`, which will likely share even more code with the
shell. I was seeing myself creating huge jump tables to reuse code from the
shell. It didn't feel right.

Moreover, we'll probably want basic-like apps to optionnally replace the shell.

So here I am with this huge change in the project structure. I didn't test all
recipes on hardware yet, I will do later. I might have broken some...

But now, the structure feels better and the line between what belongs to
`kernel` and what belongs to `apps` feels clearer.

Instead of expecting a `USER_CODE` symbol to be set, we expect `.org` to be
set in all userspace glue code. This gives us more flexibility with regards to
how we manage that.

Moreover, instead of making `USER_RAMSTART` mandatory, we make it default to
the end of the binary, which is adequate in a majority of cases.

Will be useful for my upcoming mega-commit... :)

Most of register fiddling routines (which is now the only thing contained
in care.asm) are used by almost all userspace apps, often in inner loops.

That makes the penalty of using jump tables for those a bit too high.
Moreover, it burdens jump tables needlessly.

Because this unit is very small (now that string routines are out), it makes
sense to always include it in binaries.

The goal is to avoid mixing those routines with "character devices"
(acia, vpd, kbd) which aren't block devices and have routines that
have different expectations.

This is a first step to fixing #64.

This hard-binds ed to the filesystem (I liked the idea of working
only with blockdevs though...), but this is necessary for the
upcoming `w` command. We need some kind of way to tell the
destination to write to truncate itself.

This only has a meaning in the filesystem, but it's necessary to
let the file know that its registered file size has possibly
shrunk.

I thought of alternatives that would have allowed me to keep ed
blkdev-centered, but they were all too hackish to my own taste.

Hence, this new hard-bind on files.

Previous refacoring broke all seek/tell within fs. fs handles now lost
the responsibility to keep track of current position. It's blkdev's job.

This huge refactoring remove the Seek and Tell routine from blockdev
implementation requirements and change GetC and PutC's API so that they
take an address to read and write (through HL/DE) at each call.

The "PTR" approach in blockdev implementation was very redundant from
device to device and it made more sense to generalize. It's possible
that future device aren't "random access", but we'll be able to add more
device types later.

Another important change in this commit is that the "blockdev handle" is
now opaque. Previously, consumers of the API would happily call routines
directly from one of the 4 offsets. We can't do that any more. This
makes the API more solid for future improvements.

This change forced me to change a lot of things in fs, but overall,
things are now simpler. No more `FS_PTR`: the "device handle" now holds
the active pointer.

Lots, lots of changes, but it also feels a lot cleaner and solid.

What used to be `tools/emul/user.h` was in fact specific to zasm, so I
moved it there.

To avoid name confusion, I renamed what used to be kernel.h and user.h
to kernel-bin.h and user-bin.h.

Allows for the ".fill x-$" pattern.

It was a bad idea to remove it. Now that I'm introducing the concept of
a per-app glue file, it becomes much easier to build emulated zasm as a
userspace app.