Building Type.
May 31, 2017A friend asked how hard it was and how long it took to build Type. While I thought of, meant to, and wished I had recorded the time for each aspect, I didn’t. That’s too bad, but I did come up with some estimates on how long each part took me. I’m ordering it in this way because that was the order I approached it, though the backend and web implementation went mostly hand-in-hand. These are all approximate estimates and are in no way gospel.
- UX/Design: 8 hours. This is hard to say for sure. I had thought about it for a while, starting with while I was at Viddy thinking I could teach myself something new. While I might’ve spent more time generally thinking about it, this was probably the amount of time I actually sat down and thought things through and sketched them out.
- Database Design: 4 hours. It’s pretty straightforward with just three tables.
- Backend Implementation: 30 hours. Making sure all of the corner cases for social login and expiring tokens were handled took most of it. Other than that, setting it up on Heroku was simple other than trying to cheap out on my SSL setup.
- Web Implementation: 20 hours. I hadn’t defined much of the smaller UX bits until I got here, so that slowed me down. I also wasted a significant amount of time trying to perfect the mobile web experience. I’m still not happy with it, but I realized I was getting bogged down.
- Type iOS Implementation: 16 hours. This was over a couple weekends while I stumbled my way through Swift.
- Type Android Implementation: 8 hours. This was over one weekend, mostly on one day.
- Facebook Assets/Approval: 8 hours. I got analysis paralysis thinking about how to best handle the demo video. I also had to add a privacy policy at this point so that took some time to track down.
- App Store Assets/Approval: 4 hours. I blew it and managed to do all the app icons the wrong size initially and didn’t understand that that was the problem when I tried to create the app to submit.
- Google Play Assets/Approval: 2 hours. At this point, most of the assets had been created and Google Play is easy to push things through.
I’m happy with how it turned out overall. Maybe there should be some onboarding or explanation, but I purposefully avoided that.
Some potential upcoming versions and features as I see them are:
- All: Tumblr support. I think the different post types here will be interesting.
- All: Google+ support. What do you mean it’s dead?
- iOS and Android: Offline warning and saving of posts. Let a user know that they are offline and save a post for later when they come back online. This could also handle any outage of the API.
As for long term features, these are ones I know about, but don’t plan on implementing any time soon.
- All: Multiple accounts. Right now, I don’t have it set up to switch accounts or use different accounts. That would be nice to change.
- All: User and hashtag autocompletion. Pulling the hashtags and users from different platforms might be tough, but it would help with tagging people across platforms.
- All: Uploads. Ideally, Type would detect image versus video versus pdf, etc, and then know where to share it and show the appropriate account types. This would probably mean adding YouTube, etc, support where appropriate as well.
I’ve got one of my 2017 resolutions taken care of by releasing Type. I’ll have to get on the next one by getting to work on an update. Luckily or unluckily, I haven’t gotten an app crash in Crashlytics, so I can’t use that as my easy update.
