[bseo]

When I read the title I thought you made something that found the original source of a syndicated/re-posted article.

Anyway, since you have a database of submissions to HN, Reddit and Digg, there's alot of interesting things to do.

Off the top of my head, how about a side-by-side view of parallel discussions?

EDIT, some more suggestions:

Make a downloadable dump of your data at the end of each day. Would be useful for all the people scraping HN for their own purposes/websites.

Twitter would be more useful than Digg, although all the short-urls do pose a problem.

Your top15 page (http://dexjra.com/row_view) is strange. Why isn't HN there except the "First at Hacker News view" part?

[jacktasia]

I used to have all three sources in the top15 and then some non-programmers said I should remove it...it would be trivial to add that back in. It's really about who I am trying to reach. Ideally I would have different types/sections for mainstream (reddit.com front page, digg.com,etc) and then computer people (proggit, HN, etc). I just wasn't sure which direction to take it. Thanks for your feedback.

I'll make a dump of the data...what format do you think would be best?

Thanks for your feedback.

[bseo]

I'd rather have tracking deeper than the first page of each site, than themed sections, not that it's a bad idea.

I wouldn't mind CSV or JSON data and I'm sure some people would prefer XML. Should be trivial to make dumps in each of those formats.