0 reply
0 recast
0 reaction
Now that we've launched @pinetree, I want to start "building in public" more and sharing our experiences, both good and bad (don't worry, no walk and talks from this pseudoanon yet)
First off, it was really great launching at FarCon. Wanted to echo our earlier statements about how supportive this community has been, it's been a shot in the arm for us 馃挏
We had a hard deadline to ship the MVP by May 1, as a forcing function for us to move fast
The good news: we shipped and y'all liked the product!
The bad news: there were several issues with auth, mobile views, and other bugs across the app
As @deodad wisely reassured us, we picked the lesser evil by launching early and starting to onboard users. But I care about building a great product, and tbqh it was painful shipping something that felt incomplete in many ways (and had bugs)
Now, after getting back from FarCon (and taking a week to regrow our brain cells), @giu and I are focused on the following work, in order of priority: 3 replies
5 recasts
29 reactions
1) Fixing the mini app, particularly authentication
In building the MVP, we spent 60% of our time working on the webapp, 30% on mobile web, and 10% on the mini app. Our thinking was that video creators would need a desktop experience to upload and manage videos, as mobile wouldn't cut it
I think this was correct, but we made an error in assuming that viewers would also use desktop. This was totally wrong -- most viewers have used the mini app or the mobile browser (if I had to guess, 65 / 35 split). Turns out, if you ship a mini app, people are going to use it even if you have a full-fledged app
This was particularly bad on the authentication front -- users expect to be automatically authenticated into the mini app with their FC account, and not to have to enter their emails or connect a signer key. Furthermore, the signer key request didn't minimize the mini app and sessions don't persist 馃槵
We are fixing this now and will start treating the mini app as a first-class citizen rather than an afterthought 1 reply
0 recast
8 reactions
2) Improving observability
Another learning from launching quickly -- before you have users, your surface area for bugs is much smaller. Once real people begin using your product, you get a much wider set of devices, browsers, settings, etc. This surfaces bugs that you would have never found testing (especially as a small team)
This is what happened, and when we were helping customers with bugs and trying to figure out what went wrong, we realized we were flying blind without observability
In addition to improving the product and the UX, we are going to invest time into adding logging (probably Datadog?) and making sure that if you ever have an issue with Pinetree, that we can debug it effectively without making you record your screen and sending us your console logs (sincere thanks to everyone who has done this so far!)
If y'all have any advice / thoughts on how best to do this, taking cost into account, let me know! 2 replies
0 recast
6 reactions
