The current release includes a two set of roads: a "skeleton"
set that covers the interstates and major cities, and a "detailed"
set with a LOT more information. Much of the detailed stuff comes
from other users; I've cut it back a little and tried to make the
various roads compatible with each other.
My original data was created from data relating to days when a
speed limit of 55 mph was universal across the USA. Since then,
some states have increased the limit, often "only on interstate
sections not near major cities". Some have not changed from 55.
I rehacked that data to fit the revised situation.
Then I needed to try to rationalize my data with that supplied by
others. Some of the western highways had you smokin' along at
70 and over. I tried to tone that down, but also to speed up
some of my own original links for those long runs over empty roads.
I don't plan to work much on RoadRoute over the next while, unless
somebody turns up bugs...
But these are some of the items that have crossed my mind as
to what might be done.
Feel free to take any of them on as projects if you wish.
3. (Not too hard from program standpoint; some adjustment of file
Roads called for). Roads could be marked for their scenic value
by prefixing the line with 0 to 3 asterisk characters. Thus,
**Santa Cruz,San Francisco,85,1:30,101. The computer would build
a third road parameter: "scenically adjusted" mileage, and
calculate a third route based on this. You might then be asked,
SHORTEST, FASTEST, SCENIC OR ALL? One asterisk might mean "pleasant
scenery", three "breathtaking" .. each asterisk might reduce the
scenic mileage by 25%.
2. (A little tricky coding). TWO city files: MAJOR and MINOR cities
.. MINOR points would only be involved if they were named as FROM,
TO, or VIA points. This would avoid cluttering up the output with
too many small places. The coding idea: only roads between MAJOR
points would be recognized by the system; if a MINOR city is named,
its roads would be linked in for the run. All cities would be
available for user prompts.
3. Regional files (an expansion of MINOR files, above). Called in
only when you name a state. Could allow massive files on disk,
while not overloading RAM.
4. (Opens up a rather complex area). For computer fairs, it would
be ideal to have a "learning" program. If a user asked for mileage
between DETROIT and AURORA, the computer would start a dialog: "I
DON'T KNOW AURORA; WHAT STATE IS IT IN?" .. "WHAT CITIES ARE NEAR
AURORA?" .. etc. At the end of this interaction, Cities and Roads
files would be automatically updated. At the end of a computer fair,
you'd have a LOT of new data. Some of it wrong.
5. Maps? It still seems to me that it would place a burden on the
file writer. Somebody would need to type coordinates.
--Jim Butterfield
