As I have the data sheet for the PIC16F877 in the form of
the file 30292c.pdf, I am starting out with the same raw material
everybody else is probably using.

	Just this part of the knowledge acquisition process used
to be a royal pain for anybody who is blind since you first had
to obtain the paper copy and find somebody with the patience to
read it to you.  If they knew something about electronics, this
was a boon as they weren't confused as to what they were
reading and everything went along more smoothly.

	When one got to the pinout, the practice was usually to
read that in to a tape recorder and then the person needing the
information could write it down in Braille.  It usually ends up
in columns starting with a pin number followed by enough text to
identify the purpose of that pin.

	Nowadays, if one is lucky, the data sheet or book is a
portable document format or pdf file that is not a scanned image
but a file containing the actual text formatted so that a printer
will produce the book with all the pictures in the right spots
and everybody gets what they need.

	In this case, Microchip has really done a great job and
seems to have done everything one would hope so that we can wring
every drop of information out of the file and nothing is wasted.

	I have a notebook full of Braille pages full of pinouts
to various chips I have wire-wrapped over the years in to
projects so the first thing I will do to use the PIC16F877 is to
write such a chart for the 40-pin DIP packages I have.

	When I got to the pinout which is Table 1-2, I realized
there is a problem.

	The software we all use for reading PDF files may differ.
Some may be using Adobe Acrobat and I use something called
pdftotext which extracts ASCII text in reading order which is
actually the desired format most of the time as it makes the most
sense.

	Table 1-2 might not be a problem if there was only 1
40-pin package but there are 3 footprint styles and only one is
the traditional DIP package which, if you look from the bottom is
1-20 with 1 in the upper right and 20, lower right.  Pin 21 is
lower left and you count up to 40 which is across the chip from
1.

	I need to know it that way with the other 40 and 44-pin
footprints being a couple of surface-mount patterns that one
absolutely does not want to get mixed up with what is the correct
pattern in this case.  I don't know a better smoke generator than
wrong pin numbers which are about as good as AC line voltage on
the VCC pin.

	When you look at a proper screen rendering of Table 1-2,
how do you know for sure which Pin belongs to which footprint? Is
it color or do all the 40 pins for the DIP package show up in a
pattern resembling the solder side of the board?

	These markup languages like XML, html or others embed
codes that dictate the position on the page for every drop of ink
plus what color it should be.

	One can run pdf files through various filters that
perform different functions.  If I knew what is done to
differentiate all the 40-pin DIP listings from the one remaining
40-pin pattern and a 44-pin footprint, I might be able to sort
this out more easily.

	This practice is common to most of the PIC data sheets so
I want to see if it can be mechanized so one gets only the
pattern they need without accidentally including wrong pinouts
from other patterns.

	For those who mess around with text processing, perl is
great for writing helper programs for situations like this as
long as you know what it is you are looking for.

	On a general note, when it comes to making information
accessible to all, we generally have many more tools these days
than used to exist so what I just described is not a complaint as
much as it is a hunt for ideas to fix the problem.  Not being
able to get the information at all used to be the norm and there
were lots of complaints about that so we definitely do have
progress.

Martin McCormick
--=20
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.