home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Games of Daze
/
Infomagic - Games of Daze (Summer 1995) (Disc 1 of 2).iso
/
x2ftp
/
msdos
/
source
/
snip9503
/
snippets.rat
< prev
next >
Wrap
Text File
|
1995-03-14
|
3KB
|
53 lines
SNIPPETS Rationale
------------------
Where do SNIPPETS come from?
I'm constantly on the lookout for new SNIPPETS material. The goal of SNIPPETS
has been to collect useful pieces of code that would be free for reuse. This
limits consideration to code that is either explicitly in the public domain
or which carries a free use license with few or no restrictions which would
prevent its use. Current SNIPPETS material has been collected from the
FidoNet C_Echo, comp.lang.c on the Internet, various public domain archives,
and my own commercial library (MFL).
How are SNIPPETS selected?
As the collection has evolved, it's taken on distinguishing characteristics
based largely, I'm afraid, on what sorts of things interest me. A primary
focus of SNIPPETS is portability - first of all between DOS compilers, and
secondly with the larger world of Unix and other platforms. I tend to be
interested in the same problem domain as Posix - system and utility
functions. What I'm not interested in trying to include are things which
really require massive amounts of code to do right (e.g. GUIs, databases,
interrupt-driven comm I/O, game-quality graphics, multimedia, etc.) and are
therefore better addressed by large single-purpose packages, whether
commercial or otherwise.
In addition to the utility stuff, I also snag anything that seems moderately
obscure (e.g. ISBN validation, wind chill, sunrise/set time, etc.) Whether I
need to know these things or not, someone sometime will need to be able to do
these and I like to be able to provide the option of a simple canned solution
rather than sending them on a research spree.
What makes one SNIPPETS entry "better" than other candidates?
Anything that enhances portability or usability is better in my book. Note
that I'm not necessarily soliciting completion to full library standards. As
the name SNIPPETS implies, many of the entries knowingly have minimal
bullet-proofing. Most of the entries I've excerpted from MFL differ only in
that my commercial code provides bullet-proof error trapping at the cost of
slightly obfuscating the algorithms. In cases where error trapping would, in
my opinion, make the algorithm harder to follow, I've generally opted for
clarity over safety. This was a conscious decision since the target audience
for SNIPPETS includes both newbies and experienced programmers. The pros will
recognize the shortcomings and know to add the error trapping, but the tyros
can tend to get lost. My yardstick on this has been feedback. If too many
tyros have to ask how it works, then I try to simplify and/or make the
underlying algorithm stand out. If too many folks complain about it being
unsafe, I add more bullet-proofing. When I hear as little from both sides as
possible, I figure I've achieved the desired balance.