Each sample was appended to a log file. Once per hour, the last week's worth of samples was taken out of the log file and used to plot a graph of temperature for the last week. The plot was generated using a PostScript program. GhostScript was used to run the PostScript program and produce an image, which was then run through conversion tools to produce a GIF file suitable for display on a web page.
If the temperature in the house rose above or fell below a certain temperature, the computer sent email to my cell phone informing me of the condition. That way, I was able to call someone to fix the air conditioner or heat pump before my pipes froze or my computers melted, even when I was out of town. Yes, this actually happened.
The temperature graph was also used to convince the air conditioner's manufacturer to replace the compressor under warranty - since it clearly showed that the system was consistently failing when the outside temperature rose above a certain point. Once while away on travel, I looked at this graph and noticed that I'd left the basement heater on.
I suppose this might also be able to let me know that the house was on fire. Fortunately, this didn't happen.