I've worked out a quick and dirty solution that is not
perfect but works
quite well.
Heres a sample of my data.
Postcode, Suburb, State, ID
1445 ROSEBERY NSW 146
1450 CAMPERDOWN NSW 147
1455 BOTANY NSW 148
1460 MASCOT NSW 149
1465 KENSINGTON NSW 150
1466 UNSW SYDNEY NSW 151
1470 DRUMMOYNE NSW 152
1475 MARRICKVILLE NSW 153
1480 KINGSGROVE NSW 154
1481 HURSTVILLE BC NSW 155
1485 KOGARAH NSW 156
1490 MIRANDA NSW 157
1495 CARINGBAH NSW 158
1499 SUTHERLAND NSW 159
1515 WEST CHATSWOOD NSW 160
1560 NORTHBRIDGE NSW 161
1565 MILSONS POINT NSW 162
1570 ARTARMON NSW 163
1585 CROWS NEST NSW 164
Basically i get the users postcode and take 20 from it to
give me a
minimum_postcode
I then get the users postcode and add 20 to the value to give
me a
maximum_postcode
I then return all results between min and max postcodes.
If recordsset =0
I add and take 50 to the postcode and search again.
if recordset still = 0 I increase to 100.
Im lucky enough in Australia that the postcodes run in quite
a ordered
geographical manner.
From my testing it works well.
"dempster" <webforumsuser@macromedia.com> wrote in
message
news:e4iao5$79v$1@forums.macromedia.com...
> In the US, there are databases available that provide
the latitude and
> longitude of the geographic center of zip codes (our
postal codes). With
> that
> data, you could calcuate distances to get the closest
locations (there are
> a
> few different complex formulas to do this).
>
> If you have that kind of data available in Australia,
that's the way to
> go.
>
>
>