Took a quick look through the sources (I'm aware they're not officially intended for local builds) but haven't found anything interesting (yet). First impressions:
- The API keys are obtained using flutter_dotenv, and omitted from the Git source but trivially extractable from the Play Store APK (using a rooted phone).
- After installing the .env file and building t4t locally, I was unable to log into the Android app using a Gmail account; every attempt left the app in a not-logged-in state. The same thing happened after replacing my local app with the Play Store version? (EDIT: After restarting the app I was able to proceed?)
- (nitpick/observation) When I tap a post, the app opens the user's post history, and I try scrolling the list, the view lags behind my finger. Can't tell if it's latency, dropped frames, or both together.
You can check the review history on either app, the project has existed for a full year. The timing is a coincidence, but I do understand the paranoia.
I am taking security very seriously. All sensitive rows are protected by RLS, and I have gone even further by adding random noise to all location data. Locations are locked down, but in the unfortunate event of a hack what would be leaked is location within a 5 or so mile range, not exact location.
If it is possible on Supabase I would like to eventually obscure emails and oauth info.
5 mile range is pretty serious, it narrows someones locations down to a single town in the entire world. Just dont store it at all and you're all good.
I'm wondering if it's more or less safe (from doxxers attacking individual users, or the entire service being compromised) to share location with this app vs. Lex. In any case I chose not to share location yet, and don't know if Lex is any safer (they require location).
I got an offer of a "code review" like that once for an authentication system and never heard back; it was open source anyway so anybody could have downloaded it and found my rookie mistakes like
signed_token = content + MD5(secret_key+content)
which doesn't stop anyone from appending to the content (might not have really been exploitable, but any honest review from somebody who knew more than me would have turned it up)
I'm sorry, but considering your rapid pace of development and team size, and the extremely sensitive content and vulnerable user base, I have no faith that security and privacy are well-maintained against malicious users and attackers.
This is no disrespect to the quality of the prototype you have built. This is a real-world pragmatic observation.
If I were in the US, and particularly in one of the seriously intolerant red states, there's no way I'd want and of my details in their database.
I don't think it's catastrophising to consider there may well be very real risks to being openly trans in the US in the near future. Risks of not only blackhat 4chan hackers, but also from government and legal system attacks on sites like this.
There's nothing new about this, sad though it is. People, particularly insecure people, have tried to suppress minorities since the dawn of time, and they're constantly inventing new ways to do it.
A decent percent of people are voluntarily paying for it right now, enough that it is already profitable at about a 10:1 ratio of profit vs cost of operation. The platform can grow quite a lot without expenses going up since my main expense is just server power, and I am not using much of that at all right now.
I think I have engineered it all to scale well, and I don't anticipate costs increasing. The plan right now is to not run ads, and to make almost all features free for all users in perpetuity.
Thank you for the feedback. As a design decision I did want to completely forego landing pages and upsells, both the sites and the app just "drop you in". I could probably make it a bit more clear without losing the minimalism of the page though that the real experience is in the app, not on the site.
The site serves mostly to allow people to share public posts and profiles from the app.
I think the main complaints overall with apps like Grindr, Lex et al comes in a few parts:
1) Mainstream dating apps are full of very profit driven patterns, and feel manipulative and corporate. A fair followup would be why would t4t not eventually become like this? I think probably the main differentiator is that t4t is written to run extremely cheap, and my goals for it are mostly just for it to support me financially. I have gotten quite far as a single person team, and currently I am using something like 2% of the processing power within my payment tier on Supabase. There is huge growth potential with little cost increase.
2) Probably the bigger issue is that there is a generational divide between "old queers" and "young queers", and the divide does seem to fall largely on trans comfort. Many older gays feel like their space is invaded, and many younger ones feel discriminated against. I think this pattern plays out regardless of gender. It is helpful to set intentions from the start, and so the intentions of this app from the start are to be most friendly to the "new school" way of seeing things.
3) Finally, these apps are all pretty explicitly sexualized. I know that might sound funny given how sexual the app I created can get, but I am not really pushing that. This is just a free and open community space nothing more. You can use it for whatever you like. You can date on it but you can also just share hot takes and thoughts.
OP, I love your passion, but as someone who's been in these trenches for a long time as a visibily queer person, you are a Titanic heading for an iceberg of legal issues in the US, UK,and a few other places. You are putting yourself and your users at an extremely high risk of outing, doxing, and potential legal issues. The fact that you haven't addressed these issues scares me and it should scare you.
