"Writing a spellchecker in the mid-1980s was a hard problem. Programmers came up with some impressive data compression methods in response to the spellchecker c

"Writing a spellchecker in the mid-1980s was a hard problem. Programmers came up with some impressive data compression methods in response to the spellchecker challenge." It's still a hard problem. Most working programmers today couldn't write a good spellchecker if their life depended on it. Sure, your average coder can load the dictionary file into RAM in a hash table (built by someone else), and a few of those could maybe use a library implementing a Trie or a DAWG (built by someone else)...and a percentage of those might even understand what they're doing well enough to know how to use those tools to make things better. But once you've got the data into memory, what do you do with it? About the most anyone can do is point to Peter Norvig's blog post. Sadly, very few people understand the theory behind it well enough to make improvements. So for everyone else, it's a major feat of software engineering -- on a bigger computer.