Wednesday, January 11, 2017

A Quick Note on Performance and Usability Improvements

As I mentioned in the release notes for the Anniversary Release of the Manager, I made the decision back in November 2015 to slow down on feature development and focus on usability and "quality of life" (QoL) improvements.

What follows is a bit of an extended meditation--a sort of "State of the Union"--on how the Manager will improve during 2017, with respect to general usability, performance and so on.

I haven't written out anything as fancy/formal as a road map, but I have taken the time to set down some goals and a rough agenda that I thought regular users might like to read.

If that sort of thing interests you, hit the jump for more. Otherwise, I've got a release coming out later tonight/tomorrow, so stay tuned for that.

And thanks for using the Manager!

Prior to the Anniversary Release (AR), most controls on the Settlement Sheet required a page refresh: you would select an option from a drop-down and then the page would reload/refresh completely. After the release, Locations, Innovations and Settlement Storage are the only controls on the Settlement Sheet view that require a refresh.

To quantify this, prior to the AR, there were 12 things you could click on the Settlement Sheet (including Timeline Controls) that would trigger a refresh. After the AR, there are three elements you can click on the Settlement Sheet that trigger a refresh: that's a 75% reduction in control operations that cause a page refresh.

And though the AR did a lot to solve the problem of page reloads/refreshes, recent feedback confirms that these are still a major usability issue for users:

Bracketing the comments about design/layout, the point that comes across to me in a comment such as this is mainly that any page refreshes are going to be an annoyance/impediment to users.

Which is why the current development milestone has been re-designed to implement fewer new features, and to focus more on QoL and usability improvements, such as reducing the number of reload/refresh operations on the Survivor and Settlement sheets.

Another, perhaps more subtle, QoL/usability concern (that is implicit in the comment above) is general performance.

Think With Google's new-ish Test My Site tool shows that, while the AR moved the needle on overall performance, there are still some things I need to do to optimize the experience.

Here's the site's score prior to the AR (2016-12-13):

And here's the score afterwards (I ran this just now, i.e. 2017-01-11):

The point of the AR was to start to move the legacy webapp to the API and to "refresh" and refactor the UI/UX of the Manager. Which, to put that another way, is to say that the point was not to optimize the site for performance, and I didn't really expect major gains.

But the incremental gains were nice to see. And this also is something that I'm going to focus on in the near-term.

And, as I mentioned up top, all of this will be happening while API development and work on The Watcher ramps up. The net effect, if things go as planned, will be that gradually becomes better/faster/easier-to-use on the front-end, while the back-end technology (i.e. the API) becomes more sophisticated and bears more of the "weight" of providing the service.

Finally, by this time next year, I would really like to write a post about how the Manager has really awesome Google performance scores, requires minimal page reloads and is being phased out in favor of The Watcher.

Ideally, I will be talking about I spend most of my time optimizing the API and supporting third-party applications that use the API, etc.

At any rate, that's the plan.

As usual, if you've got ideas/comments/feedback, hit me up. Otherwise, thanks again for using the Manager!

No comments:

Post a Comment