If you're able to input invalid URL's I guess I didn't test the regex thoroughly enough ;)
I'll look into it.
The prepending "http" is just a lazy hack to handle the case where they enter "google.com" instead of "http://www.google.com, which would fail the regex, but is a valid URL. This would be deprecated by improving the regex validation.
Keep in mind I threw this together in an afternoon.... But thanks for the critique ;)
Good point. Yes, in theory they could be duplicated. I chose 6 characters firstly to keep the url as short as possible (I thought 5 chars was too short, 7 too long).
I am planning on introducing a background process to clean up shortened URL's that haven't been accessed in a given time frame (still deciding but I think around 7 days). While it still wouldn't solve the problem you mentioned at a certain scale, it would help. But true, I very much doubt it will ever reach that scale.
"Duck extends Bird" is not the wrong thing at all. As other comments have pointed out, the justifications in the article don't apply to someone who is just starting out learning the basics of OO.