Moderation-wise, I have a dashboard where I can see all drawings and quickly delete anything bad. Users can also report stuff, and anything with enough reports gets auto-hidden. Looking into automating it more down the line as well.
You should see a toolbar at the bottom with a pen/marker icon, tap either to start drawing. No payment needed, premium is really just to support the project.
You may be right, I may need auth down the line. I wanted to reduce friction and let people draw right away though. I think the bridges thing is an artifact of Mapillary, the API I use for panoramas. Dashcams tend to overrepresent highways. Working on better curation.
Thanks! There's a report button if you hold on any drawing, plus I have admin tools to delete flagged stuff. Working on automated moderation to catch things faster as it scales
Right now it's limited to ~2,000 curated panorama locations. You can only draw at existing bubbles on the map. I'm working on expanding coverage, but if you have a spot or location you'd love to see added, let me know :)
Yeah, I see that does seem confusing. The heatmap currently shows all available locations, not just ones with strokes. I need to think through the UX more like trying to balance showing coverage with making it obvious where actual drawings are. Thanks for the feedback!
Haha yeah, can't have nice things I guess. I'm currently working on automated moderation to catch that stuff faster. In the meantime, Report button exists if you stumble into anything awful.
About 2,000 locations so far. Mapillary has a lot of imagery but finding the 360 panoramas is trickier since they are a lot rarer. I'm actively expanding coverage
I hear you on the friction. I’m experimenting with the 'ink' limits right now, also the ink actually regenerates. I think I need to make it more obvious in the UI
Currently, the JS, HTML, and CSS are rendered client-side, which was relatively straightforward to implement. However, I’m in the process of moving these files to the server and serving them from there, which will enable me to execute languages that aren’t natively supported by the browser
Good question — before diving into the project, I did have the chance to choose. I looked into both OT and CRDTs and found that OT is widely used in collaborative editors like Google Docs, so I leaned that way. I wouldn’t say it was a deeply informed decision — I just had more reference points for OT at the time. I think with more hands-on experience with CRDTs, I’d be better equipped to weigh the trade-offs more clearly
Yeah, you're very correct actually. I chose Java because I'm most comfortable with it, but in retrospect, Node might have been a better choice. There’s some shared logic for the collaboration part, and it was a bit of a hassle to ensure the client and backend had the same logic. Having both in JavaScript would have made things smoother
Thanks for the feedback! I clear the session URL after joining because I thought it looked cleaner, but I might start leaving it in if it helps with reload confidence.