I've implemented the core of what's to be the next search program, but it's turning out to be more difficult than I expected. In fact, I even had to write pseudo code before I got it right, and some of you know how often that happens (i.e, never).

As if that wasn't bad enough, I ran it and... what do you mean did it work? of course it did... its memory usage went peaked at 46MB. Here are the global variables I'm using:

vector <string> 		blog_vector;
map<string, int, ltstr> 	blog_map;
vector<string>			link_vector;
map<string, int, ltstr> 	link_map;

map<string, vector<int>, ltstr>  link_blog;
map<string, vector<int>, ltstr>  blog_link;

I don't have the exact numbers right now, but there are nearly 10,000 blogs, and roughly 1 million links. So, yeah, maybe it's not all that surprising.

What's Plan B, you ask? I modified wordex to extract all links that appear in at least two blogs, so I'm going go use those links and ignore the rest. That should cut down the number of links to <500k.