RubyGems.org redesign mockup: Focus on Search, News, Fancy Graphs

[Kurtis Rainbolt-Greene]

While working on upgrading to 3.2 (via the rails-3.1 branch) I was moving around the styles and was thinking about how the front page was designed.

After thinking about it I came up with this quick mockup: http://i.imgur.com/zKr79.png

I think it focuses more on the core values than the current website: Searching for gems, News about gems, and of course fancy graphs. In retrospect the fancy graphs would probably be replaced with learning material (Like how to download a gem, or links to guides?)

Since we download and edit our gems via cli and files the real reason I think most people use the website is to find new things, find old things they half-remember, and look at version crap. The vein (myself included) sometimes look at their gem's download count.

Gem specs happen to have a lot of metadata that I don't think is accessible via the current search, thus the "Advanced Search" button on the mockup.

[Nick Quaranto]

Yeah, agreed! We need it for sure. Here's my thoughts previously about doing a homepage redesign:

https://github.com/rubygems/rubygems.org/issues/217

Some thoughts re: this mockup...

1) The homepage is our #1 most popular/hit page, followed by the downloads page. The homepage *needs* to have install/help stuff on it.

Here's our gaug.es for today…I can't do all time view for a given page (boo!) but it's been like this for a while.

https://img.skitch.com/20120222-rrw4rqrbxg755khx3xfj1i9nqk.png

2) Showing the latest blog posts might be good, but we don't have any news beyond that, really.

3) Our current "fancy graphs" suck in my opinion. In fact, I've been debating pulling them *all* out of the app. They suck to maintain and quite frankly, they're ugly. We can't afford to keep stats around either for > 90 days now, and Evan and I are in the process of backing that data up and getting it out of the system. I'd rather provide that kind of data via the API (and backed up via S3) and let other, more creative folks have a ball with it.

4) Search is a big deal. Way bigger than I thought it would ever be. I'd love for a new design to focus around that.

[Kurtis Rainbolt-Greene]

Agreed on the graphs, they can be chucked out the door. If they're absolutely needed (for whatever reason) I'd write some simple sinatra app that consumes the API and acts as a satellite application on heroku or whatever.

Dropping charts, restyling for a search focus, and other small bits could also bring performance improvements for the website (which would lower costs).

--

Kurtis Rainbolt-Greene:

 title: "LIMist, Hacker, & Father"

 address: "1631 8th Street, New Orleans, LA 70115"

 phone: "(202) 643-2263"

[Nick Zadrozny]

On Wed, Feb 22, 2012 at 1:24 PM, Nick Quaranto <ni...@quaran.to> wrote:

4) Search is a big deal. Way bigger than I thought it would ever be. I'd love for a new design to focus around that.

I'm still down to help make better search happen, but unfortunately the current master got way ahead of my branch. So perhaps a fresh start is just what we need on that front.

--
Nick Zadrozny

http://websolr.com — hassle-free hosted search, powered by Apache Solr

[Kurtis Rainbolt-Greene]

I've currently got a fully working (well, the tests are still wonky) branch based off rails-3.1, merged with the latest master, named rails-3.2.

I'm cleaning the views, helpers, etc up real quick. I could share the branch, help you pull in master?

--

Kurtis Rainbolt-Greene:

 title: "LIMist, Hacker, & Father"

 address: "1631 8th Street, New Orleans, LA 70115"

 phone: "(202) 643-2263"