Thanks for the link. Excellent comments there. Too bad that HN can't dedup canonical URLs properly. It wouldn't be that hard - just analyze the title, domain,

Thanks for the link. Excellent comments there. Too bad that HN can't dedup canonical URLs properly. It wouldn't be that hard - just analyze the title, domain, URL, samples of the article. pg: maybe crowdsource development and start accepting pull requests?