On Wednesday, 14 January 2015 at 03:46:39 UTC, Philpax wrote:
Hey everyone,
I recently wrote a blog post about how I used D/vibe.d to help
find a new house. I haven't publicized it anywhere else yet, so
I'm looking forward to what the D community has to say! You can
check it out here:
http://philpax.me/blog/heady-house-hunting-with-d
D made it easy to model the problem and quickly crunch through
it; I'm pretty happy with how quickly I was able to get decent
results. It's not the most idiomatic of code, but D's
flexibility meant that I could concentrate on the concept
instead of the implementation details.
Heh, I wrote something similar (much much simpler) to find the
best hotel for DConf 2013. The program extracted the hotel's
geographical coordinates, and then queried the Google Maps API to
find the one that had the shortest public transit time assuming
we wanted to arrive at 8:30 AM.
Since then, I started doing it for all important purchases - TV,
AC unit, laptop... The laptop program used a score model like
yours. The biggest issue is that it's hard to find accurate
information on household hardware, especially since there are a
myriad models for each tiny European market.
Here's the hotel program (list.txt is a list of booking.com URLs):
http://dump.thecybershadow.net/2ebca3f47b801aed0104585f60b9587e/scan.d
By the way, does your blog have a RSS feed for D posts? Then I
could add it to Planet D.