HN user

divad12

591 karma

I'm a student at the University of Waterloo for Software Engineering.

Also a former two-time intern at Sift Science, Khan Academy (past work: http://david-hu.com/ka-ml.html), and Google.

Blog: http://david-hu.com

Posts10
Comments20
View on HN

You're right that GitHub stars probably isn't the best metric of how good a plugin is. We try to use a more unbiased metric of relative # of actual users, and this we get by scraping dotfiles repos on GitHub, looking for references to Vim plugins. That's shown as the # users count to the left of stars on GitHub. It's also not perfect, but it's a useful source of relative usage statistics.

Hi! Glad you like the j/k scrolling.

All it is is a call to `window.scrollTo()` with some logic to determine the position of the element we want to scroll to. See [1] for the code of the actual function that does the scrolling. Is there acceleration? There doesn't seem to be on Chrome for Mac. :/

Oh, if you meant the little arrow ("»") that slides to the right on focus of a plugin, that's just the CSS 3 transition `all 0.1s ease-out` applied to a changing `left` and `opacity`. See [2] for code.

Thanks for checking out the site!

[1] https://github.com/divad12/vim-awesome/blob/13d576664aa88fe0... [2]: https://github.com/divad12/vim-awesome/blob/13d576664aa88fe0...

In fact, this was the first React project that he worked on -- he rewrote our old angular app within a week of when React first came out, and had been hooked since.

Hi! Thank you for checking this out!

The reason for the mis-match between # of plugins and tags count is the tags count only counts plugins with that tag in the category the tag is nested under. I agree this is quite confusing, and should just make the two views consistent. I'll add an issue to GitHub about it.

Each category has a distinct color.

Other->'manager x 3' shows no results

Oh, we hide plugin mangers right now, so no results are shown. :( Will fix this to be nicer, thanks.

We have seen fraudsters reinstall an app over and over, each time signing up with a new email and a new stolen credit card after their previous account was shut down. A persistent device ID allows our clients to block a new signup from a fraudster that is from the same physical phone the fraudster has used before. (This greatly increases the cost of committing fraud, by forcing the fraudster to root their phone or use a new phone each time.)

(I work at Sift Science too.)

Just to clarify, we charge 10 cents per unique user that you query in a month, not per query. So, you can query the same user as many times as you like in a month and the cost will not exceed 10 cents for that user.

Thanks for pointing that out. The ticker came from some work I did while interning at Khan Academy, and was not originally meant for negative numbers. If I get some time, I may make it into a jQuery plugin and make it work for more cases and be configurable.