I tried it out for a bit and I think it's generally pleasant.
Some thoughts:
- The cleanliness and whitespace make it feel like information's been lost compared to the previous version. Probably just psychological and I'll get used to it.
- I wish the header search button was a search bar instead. While it looks cleaner, I actually had a moment where I thought they removed search completely. My preference would be to have a large, highly visible search bar in the header. Increasing the click/tap area would help as well, as I misclicked on the dark mode toggle.
For anybody who comes into this thread in the future, I've closed off public signups for the iOS beta. I've hit a critical mass of testers and want to make sure the app isn't overrun by trolls and predators. Reach out to me if you really want in [0]!
To know what's on the roadmap, checkout my public trello board [1] for more info.
My impression is that my use of Flutter has been successful (especially in an iOS build) because of my heavy use of custom-designed components. There are packages built by the Flutter team to mimic native behavior, but they aren't good enough in my opinion. I'd go with react native if you want something that feels (at least on iOS) native.
The biggest plus for Flutter in my opinion is its great animation system. It's difficult to learn, but it makes things that would have been near-impossible in react native relatively straightforward.
I'd use Flutter again for sure. I'd rate it like an 8/10?
My path to using flutter was that I wanted to use SwiftUI because of its native-ness, but found it was too buggy. I know react native well, but I realized that a lot of the interactions I wanted (e.g., animations) would be much more difficult. Flutter was a compromise that's overall been great to work with.
I do feel very at-the-whim of the flutter team's progress (mainly native behavior I can't hide behind custom UI like text editing), but that's not unique to flutter.
If you don't mind a link to reddit, I also wrote a comment about the decision [0].
On your first point: scroll down on your mailbox screen to see a history of what you've opened up! Sorry about that. It's a poor design decision on my part to make it so undiscoverable. Will fix it soon.
Reactions of some sort are one of the high priority features I plan to build.
No strong opinion from me on horizontal cards and swiping. It is intentional to add some friction to skipping a message (so it feels a bit more weighty), but I'll play with it!
Hey! For now I have it locked to portrait orientation and no official support for iPad. Very sorry for the trouble with this — it does sound annoying. I'll look into enabling iPad-specific rotation.
Not at all arguing that this is novel or innovative — rather my own take on the genre. I'd argue that some aspects of the system incentivize better behavior:
- It'll be a purchased app (not free), so the stakes of being banned are higher. Right now in beta, it is more open for sure.
- The 1:1 message sending doesn't give any feedback to bad actors, so attention-seeking behavior is more limited.
- I'm also hopeful to continue to design an experience that's calm and cheerful, and provides guidance about good behavior (in a way that's more in-your-face than most social media).
Of course, there will always be malicious people, and neither I nor this project are perfect. As I mentioned elsewhere in this post, I'd be quick to shut things down if it got out of hand.
Thanks for the info! I guess it also only happens if you have some VR device setup. Sorry for the trouble with that. I'll swap out the vimeo embed soon since that sounds like annoying and unnerving behavior.
That's... unusual and concerning. You're talking about the gentle.app site, not the testflight link, right?
The only thing I could think of to cause that is the embedded vimeo iframe. But why? If you're open to share, I'd love to know what browser you're using.
Hey! It's a Flutter application, so I'm using their concept of Hero animations [1]. It renders a new component (the flight shuttle?) that does the visual animation.
I think there's better scaffolding/explaining to do with content guidelines and expectations, but the general idea is that content that causes concern like this should be reported. A lot of explicit content would probably already be caught by a filter for manual review.
It would be taken down and I would send a message to the writer with mental health or counseling resources (ex: crisis text line). I take all messages seriously by default, even if it appears to be trolling.
If that person continued to write content like that, I'd do a bit more digging. This is where it'd be subjectively up to me as the moderator to take action. Banning or shadow-banning would be a last resort, generally.
There's no strong accountability on my end (beyond people not using the app), but I'm interested in creating a system people can trust. If the app continues to gain traction, I'm thinking about writing formal moderation rules and expectations and seek input from the community.
Thank you! I'm a bit wary of giving big hints to how the filters work, but I was inspired by:
- npm packages like redact-pii, bad-words (and numerous other "bad word" lists)
- askismet's open source spam code
- regex unicode categories
I also wrote a variety of checks for word counts, combinations of words, known "bad patterns", etc. You may not realize it, but the text input forms block a lot of potential inputs.
It's not perfect, but that's why I'm manually handling things that surface. Happy to answer any follow-ups.
Some thoughts:
- The cleanliness and whitespace make it feel like information's been lost compared to the previous version. Probably just psychological and I'll get used to it.
- I wish the header search button was a search bar instead. While it looks cleaner, I actually had a moment where I thought they removed search completely. My preference would be to have a large, highly visible search bar in the header. Increasing the click/tap area would help as well, as I misclicked on the dark mode toggle.