Is anyone trying to back this up with a proper scientific study of whether users find it easier or harder to use? Seems just like opinion and debates about aesthetics otherwise?
If I wanted to claim that JavaScriptCore was getting slower and I didn't post any data to back that up people would laugh me out of the conversation.
As long as we miss SI unit for the "hard to use" attribute we will have to fall back to user opinions and debates - and potential randomly constructed half cooked polls.
The opinions and debates are not looking well for Apple though.
We already have usability testing, but there are more dimensions to usability than can be encapsulated by one SI unit. As an example one measure might be the number of seconds required to perform a set sequence of actions such as copying text from a web page to a Notes entry. That's a completely different task to copying a pre-written comment from Notes into a web page, even though they both involve the same apps and are measured in seconds.
Nice isolated example that tells nothing about the "hard to use" factor. We can repeat such for days without yielding reliable empirical results. The issue is too compound.
Opinion of the users, especially repeated ones ought to be enough to justify a concern. There are too many to doubt its validity even without exact but half cooked partial numbers.
You still could set up a study where you have tasks that users must complete and you get countable results. But this would quickly get expensive, so you need a sponsor. This could either be Apple, Google, or some other company in the market interested in making their opponent look bad. And I wouldn't take such a study for granted.
Not only expensive but making a comprehensive one that people actually finish is practically impossible.
And there are many many further aspects that make the measurement non-reproducible - and so unreliable - or biased.
Right. Like I said, the only parties interested in taking all that effort and that have enough money are probably Apple, Google or maybe Samsung. Independent companies won't go down that hole.
> we will have to fall back to user opinions and debates
So what's the point? If it's all just random subjective opinion why are we indulging in it? If you can't even demonstrate any issue in the first place then why bring it up?
But that's a great example of what I mean - are HN users grumbling a statistically significant result? Do they represent average users? Should Apple listen to it or ignore it? I don't know - do you?
As far as I know, the last time anybody used science to help make a good UI was for Windows 95. If you can find "The Windows 95 User Interface: A Case Study in Usability Engineering" you can see they were thorough. I would love to be wrong, but I don't think anything like this was posted for any more recent Windows, or Mac OS X or any of the many mobile operating systems.
If I wanted to claim that JavaScriptCore was getting slower and I didn't post any data to back that up people would laugh me out of the conversation.