Rich Wales <wales@CS.UCLA.EDU> (02/23/88)
With the aid of several friends across Canada, I have managed to figure
out the essentials of the "yellow bar code" system used to encode the
postal codes on Canadian mail. I have all the digits and more than half
of the letters figured out.
To finish the job, though, I need to get my hands on more data -- some-
thing that's a bit hard to do down here in the States. So . . .
I would like to ask as many people as possible to send me information
about the bar codes on mail they have received. I will explain a sys-
tem for describing a bar code via e-mail -- though if people want to
just mail me old envelopes, that would be fine too.
When you reply, also let me know if you're interested in knowing how
the system works. If enough people ask me, I'll arrange to have the
scheme posted to the net (with "spoiler warnings" in case others want
to try their hand at it and don't want me to ruin the challenge of
discovery :-}). If only a few want to know, I'll reply via e-mail.
If you want to try your hand at the system, let me know and I'll send
you a copy of my raw data which you can play with.
If you decide to reply to me via e-mail, the following explanation
should give you enough of an understanding of the bar code system to
be able to describe a bar pattern (without actually giving the whole
system away!):
(1) The possible bar positions are spaced 3 mm apart. There are a
total of 27 possible places were a bar can be. There is only one
kind of bar (i.e., no "long" or "short" bars as in the US mail's
bar code scheme); each position either has a bar or is blank.
(2) The rightmost bar (closest to the lower right-hand corner of the
face of the envelope) is always there; it serves as an alignment
indicator for the rest of the code. The leftmost position (78 mm
to the left of the rightmost bar) may or may not have a bar; this
is a "parity bar" whose sole function is to make sure there is an
*odd* number of bars in the code. The remaining 25 bar positions
represent the postal code of the letter's destination address, via
a *very* weird binary coding scheme that took me several weeks to
even get a general picture of.
(3) Usually the bars are stamped in fluorescent yellow-green ink, but
sometimes they are sprayed on by an ink-jet printer in orange- or
salmon-coloured ink. The coding scheme is the same in any case.
(4) There are often several bars off to the left of the main bar code.
These seem always to be the "orange ink jet" type, even if the main
code is in yellow. For the time being, *ignore* these extra bars;
the only ones I am interested in are in a field starting with the
rightmost bar and extending 78 mm to the left.
(5) To transcribe a bar code, lay down a millimetre ruler by the code
and look every 3 mm to see whether there is a bar or not. If there
is a bar, write a vertical bar; if there is no bar, write a period.
Please transcribe the *entire* code as it appears on the envelope
(except for any stray orange ink-jet bars off to the left). Please
do *not* omit the rightmost "alignment" bar, or the leftmost "par-
ity" bar. However, don't worry about the extra "no bar" positions
which may be at the left end of the code (assuming there is not a
"parity" bar); I can extend the code out myself with "no-bars" to
the left to get the full 27 positions.
(6) Send me the bar code (transcribed as explained above), along with
the postal code in the *destination* address. If the postal code is
not *clearly* written/typed, and you feel it could have been mis-
read at the post office, please mention which part of the code may
have been misread.
For example: V3N 4L9 |...|..|||.||.|.||..|.|||.|
K1P 5W9 ||..|.|||...|.||||.|...|
I would welcome any and all bar codes. However, I am particularly
interested at this time in codes for which the corresponding postal code
contains one or more of the following letters: A, N, P, R, T, V, X, Y,
or Z. I am especially interested in codes where the last letter (fifth
character) is an X; the data I have so far with an X as the last letter
doesn't appear to make any sense.
I should be reachable via e-mail in one or more of the following ways:
wales@CS.UCLA.EDU (preferred)
wales%cs.ucla.edu@RELAY.CS.NET
...!uunet!cs.ucla.edu!wales
...!rutgers!ucla-cs!wales
...!ucbvax!ucla-cs!wales
If you prefer to send me one or more envelopes via Canada Post, my
address is:
Rich Wales
UCLA Computer Science Department
3531 Boelter Hall
Los Angeles, CA 90024-1596
(USA)
Thanks very much for helping me out in this little bit of "intriguing
trivia" research.
(Why, you might ask, don't I just ask Canada Post how the system works?
Well, first of all, that would spoil all the fun of figuring it out for
myself. :-} Secondly, I actually *did* try, but without success. While
on a trip to Vancouver last August, I talked with a senior postal offi-
cial who took my name and address and promised to send me the info.
Maybe he had other things on his mind -- such as the rolling strike that
plagued the system shortly after I talked with him -- but, for whatever
reason, I never heard back.)
-- Rich Wales // UCLA CS Dept // wales@CS.UCLA.EDU // +1 (213) 825-5683
3531 Boelter Hall // Los Angeles, California 90024-1596 // USA
...!(ucbvax,rutgers)!ucla-cs!wales ...!uunet!cs.ucla.edu!wales
"Sir, there is a multilegged creature crawling on your shoulder."