Hacker News Reader

Comment by stavros | Hacker News Reader

> Companies have existed for centuries without needing constant telemetry

And if you're fine with software from the 1700s, you're welcome to use that, but if you want UX improvements to Audacity, there's only so much that can be done without telemetry.

View on HN

Quantitative (ie telemetry) data gives us the what, not the why.

Qualitative (usability testing and other observation methods up to and including ethnography) can give us the why (users seem to do what we didn't expect).

Modern UX design can make use of both modes of user research such that they support each other. I.e discover the what via quant and understand it better via qual, to land at a solution.

Or discover both why and what via qual, and then determine severity (how many users get stuck in a given way) via quant.

But of course, there needs to be explicit consent such that trust is not breached, no matter how inconsequential the data collected is considered to be. The decision belongs to users.

Apparently there wasn't consent.

Edit: Ideally, should anonymised usage data in OSS projects be public? Such that we can see the data and design decisions made based on it? Perhaps this would generate more of the trust that was desired here.

Companies like Apple and digital research were producing great UI/UX in the 80s, well before widespread internet usage made telemetry a viable option for companies.

(I mean the 1980s, just to stave off any attempts to send me back in time to observe computing in the 18th century).

Apple OS pre OS-X, and Windows 3.1/95, used to crash all the time. A big part of the improvements in crashes (talking to people in the companies) is automatic submission of errors.

Also, Apple was selling Apple IIs for the modern equivalent of $5,000. I'm sure Audacity could achieve some great in-person UI studies is people were willing to pay that kind of money.

doesn't take $5000 to open a survey monkey account.

I don't understand this argument. "Apple did good UX without telemetry in the 80s, so you should be able to do good UX without telemetry now, Muse".

Like, what do people think they want the telemetry for? Are they planning to sell the treasure trove that is "how many people clicked on the 'select audio sink' button" to the highest bidder?

You can tell them "do it without telemetry" all you want, but in the end it's just not going to be as good as if they had feedback.

not any more, they've blown all their good will

I'm pretty certain Ardour and Reaper don't have any telemetry, and they have UX a million times better. Audacity can be muuuuch better with having to stoop to telemetry.

So is the UX unusable right now?

Edit: >but if you want UX improvements to Audacity

You insinuated that the only way to improve the UX (your comments insinuates that it needs improving) is to use telemetry

Who said that?

> Who said that?

Whoever wrote: > > > but if you want UX improvements to Audacity, there's only so much that can be done without telemetry.

Oh look, seems to have been some guy called "stavros". Anyone you know?

I feel like UX improvements would be best driven by user feedback, not inferred from telemetry.

Durex get by without knowing exactly how their products are used. Product improvement is not one to one with privacy abuse.

Reply on news.ycombinator.com