root/trunk/Notes.txt

$Id$

Credits
=======

The vast majority of PDCLib is original work by me. I felt it was the only way
to ensure that the code was indeed free of third-party rights, and thus free to
be released into the Public Domain.

Another issue was that of coding style, quality and stability of the code, and
the urge to really understand every miniscule part of the code as to be able to
maintain it well once v1.0 has been released.

That is not to say there are no credits to be given. To the contrary:

Paul Edwards (author of the PDPCLIB), for inspirational code fragments, thanks.

P.J. Plauger (author of "The Standard C Library"), for a book without which I
would not have been able to create PDCLib at this quality, thanks.

Paul Bourke (author of mathlib), for allowing me access to a complete math
library under public domain terms so I could use it as a reference, thanks.

Peter ("Candy") Bindels (netizen of osdev.org), who located a copy of Cody
& Waite's "Software Manual for the Elementary Functions" for me and saved me
serious cash in the process, thanks.

Michael Moody, who contributed the generic implementation for <stdarg.h> to
the Public Domain which can now be found in <_PDCLIB_config.h>, thanks.

Rod Pemberton, for pointing out several flaws in early versions of PDCLib and
giving other valuable hints, thanks.

Brian Damgaard, for a very friendly exchange over the fine details of the
Quicksort algorithm and its implementation in PDCLib, thanks.

Everyone involved in the first, "public" attempt at PDCLib, for bearing with me
when I restarted from scratch, thanks.

Everyone bearing with me during the "stdio block", a period of many years in
which PDCLib received not a single update because I was stuck and could not
find the time and energy to work it out.

Lennart Frid�n and Sammy Nordstr�m, who have been great pals even after I sunk
some other project that had eaten countless hours of work between the three of
us, thanks.

My wife, daughter, and son for sharing husband and daddy with this strange
machine, thanks.


Style
=====

I followed a set of guidelines in creating PDCLib. If you find some piece that
does not adhere to them, that's a bug worth reporting. I mean it. I am a bit
obsessive when it comes to coding style. ;-)

- All the stuff that is not part of the standard specification is "hidden" in
  the _PDCLIB_* namespace - functions, variables, macros, files, directories.
  This is to make it easier to distinguish between what the standard dictates
  and what I added to make PDCLib work.

- Any internal includes (i.e. those not specified by the standard) have their
  header guards defined in the *including* file, for a tiny bit of compile-time
  performance.

- I always try to minimize the use of local variables. Wherever possible I used
  parameters passed by-value directly, and deferred declaration of locals to the
  innermost block of statements, in hopes that it might reduce memory footprint
  when the library is compiled with a compiler that is not that great at
  optimization.

- Every function, every static data item that could possibly be shared, got its
  own implementation file. This means the library itself is probably larger than
  strictly necessary, and might take a couple of clock cycles longer to link,
  but it reduces size of object files and executables.

- Where possible, I tried to share functionality between similar functions (as
  can be seen in the atoi() and strtol() function families). This means one or
  two additional function calls, but again reduces memory footprint and eases
  maintenance of the library.

- Function arguments are named exactly as in the standard document.

- The standard is taken quite literally in places. For example, the default
  implementations of memcpy() really copies char-wise. This runs contrary to
  earlier claims of performance, but is consistent with the *letter* of the
  standard, and you will probably use your compiler builtins (through a platform
  overlay) anyhow.

- Every file has an Id tag, so that it is on file for *every* code file.

- PDCLib code has no bias towards POSIX; indeed the absence of POSIX tidbits is
  one of its hallmarks. However, PDCLib also has no bias *against* POSIX, and
  when one platform abstraction is as good as another, I chose the POSIX one for
  sheer familiarity. (This is mainly referring to naming and parameter lists of
  OS "glue" functions.)

- Identifiers are tersely named, but not cryptically abbreviated, and should be
  intuitive enough to allow easy understanding of PDCLib inner workings.

- I disagree with the notion that good code needs no comments. Code tells you
  *how*, but not the *why*, and you have to figure out the *what* yourself. So
  I added comments to every nontrivial piece of code explaining my motives and
  hopefully keeping overly ambitious editors from repeating my mistakes. The
  header files especially should be self-documenting to the point of being a
  suitable replacement for any library reference book you might be using.