Hacker Newsnew | past | comments | ask | show | jobs | submit | bwoodruff's commentslogin

I was an advocate for putting out the blog post early, despite the fact that we're currently only testing this with our employees. As you say, we knew it would be something the community would have questions about, rightfully so, and wanted to be as transparent as possible.

-Ben, 1Password


It tells the extension which item you've selected to fill. It isn't possible to use the Open & Fill feature without it. If you navigate to the website in your web browser and then fill from 1Password's inline menu, instead of using Open & Fill, you can avoid it. Hope that helps. Please drop us an email if you have further questions: support at 1password dot com

-Ben, 1Password


I love that idea. We'd have to be super careful with the de-identification of associated data (which we're doing anyway), but having automation behind figuring out filling failures could be a huge boon. I'll share the thought with the team.

-Ben, 1Password


I wrote more about the consent aspect here: https://news.ycombinator.com/item?id=35706897

tl;dr If we roll this out to customers, we'll be asking for consent, and won't be collecting telemetry data unless we have it.

-Ben, 1Password


I wrote more about the consent aspect here: https://news.ycombinator.com/item?id=35706897

tl;dr If we roll this out to customers, we'll be asking for consent, and won't be collecting telemetry data unless we have it.

-Ben, 1Password


That's much more reasonable than the wording on the linked page. Thanks for your response.


Happy to help. In addition, while we're in the early stages and this design is likely to change, it may help to visualize how we're thinking about this process:

https://bucket.agilebits.com/ben/telemetry-consent-draft.png


I wrote more about the consent aspect here: https://news.ycombinator.com/item?id=35706897

tl;dr If we roll this out to customers, we'll be asking for consent, and won't be collecting telemetry data unless we have it.

-Ben, 1Password


I wrote more about the consent aspect here: https://news.ycombinator.com/item?id=35706897

tl;dr If we roll this out to customers, we'll be asking for consent, and won't be collecting telemetry data unless we have it.

-Ben, 1Password


I wrote more about the consent aspect here: https://news.ycombinator.com/item?id=35706897

tl;dr If we roll this out to customers, we'll be asking for consent, and won't be collecting telemetry data unless we have it.

-Ben, 1Password


I wrote more about the consent aspect here: https://news.ycombinator.com/item?id=35706897

tl;dr If we roll this out to customers, we'll be asking for consent, and won't be collecting telemetry data unless we have it.

-Ben, 1Password


One anecdote as to why it has become clear telemetry is needed:

When prioritizing what we needed in order to launch 1Password 8 we did not prioritize an Apple Watch app. We rarely heard from customers about Apple Watch, and so the assumption was that very few people were using it. When we launched without it, it quickly became apparent that was a poor assumption. People came out of the woodwork to ask where our Apple Watch app went. If we'd had telemetry, we could've known that lots of people were using the Watch app, and just didn't have a reason to write to us about it.

-Ben, 1Password


This is a top class example of organizational blame shifting at it's extreme, manifesting as user hostile behavior.


If better data had been available to us, we would've been able to make a more informed decision. We've decided that privacy-preserving telemetry is one of the ways worth exploring to improve that data for the next time. If it is something we roll out to customers, customers will have the choice to participate. We will not collect any telemetry data unless we've obtained consent.


That could also have been avoided by not rewriting in Electron, which I truly think 0 users wanted, whereas the old app warranted nominations for the Apple Design Award.

It’s still not too late to reverse that decision.

Incidentally, how would telemetry tell you that people miss a removed feature if said telemetry was not in place before removal?


1Password for iOS does not use Electron. It uses Apple's SwiftUI. Our iOS app is what contains the Apple Watch app.

The point was if we'd had telemetry while prioritizing this work, we would've known it was something many people were actually using. The little data we had showed that was not likely the case. Had we known, we would've prioritized differently.


Fair enough, I stand corrected, and agree that telemetry would be useful there.

Makes me wonder though, if there is a SwiftUI codebase, why not maintain the macOS app as a native app too?


Or you know you could just go to Wikipedia and see how many Apple Watches have been sold, then make a good solution for that platform (wikipedia estimates 100M units as of 2020): https://en.wikipedia.org/wiki/Apple_Watch

This is absolutely no justification for telemetry.


Nobody was questioning how popular the Apple Watch is. The question was how many people were using it to access their 1Password data. The evidence that we had at the time suggested very few. Better data would've helped us make a better judgement.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: