Building a reading diary for my kids
The library was closed. I had a bag of twenty-five books to return, and as I pushed them through the slot, I realized I was forgetting which ones we had read. A few were still in my hands. I thought I could at least save those on Goodreads.
So there I was, hunched over on the library steps, holding Nighty Night, Little Green Monster in one hand and waving my iPhone over its barcode with the other. Eventually I scanned the remaining books and walked home with an empty bag.
Reading had become one of my favorite things to do with my three kids. We went to the library each week. At bedtime, I read a couple of books per child, after an intense negotiation over the order. My daughter usually wanted her book last; my son was happy to go first. Hearing them demand that I “read a book now” made me very happy.
I wanted to remember more of those evenings.
This is a build journal from 2014, when I was learning Swift and developing the app with early testers. Tool references and progress reports belong to that period.
Find out what is worth remembering
I started wondering whether a reading diary could capture more than a book title. Could each child say how much they enjoyed it? Could I notice a favorite emerging? Would sharing those choices help another parent find a book for a child of a similar age? I wanted something we could use every day, that could grow with the children and protect their information.
There were simpler options. Goodreads already existed, though I found the experience awkward for what I wanted to do. A notebook would work, but I worried about losing it. I could leave the moments unrecorded. From my own experience, though, tracking an activity made me more likely to keep doing it.
Later, wedged between my five-year-old and three-year-old, I read Bobo the Sailor Man. One page sent them into fits of laughter. That was the kind of moment I wanted to be able to find again.
I read about family time and reading habits while thinking through the idea. My notes included a report on how little time families spent together, young readers’ preference for print, a discussion of print and ebooks, and a conference paper about recalling a story’s sequence on paper and Kindle. These were leads for exploration, with different sources and scopes. They didn’t establish that anyone would want my app.
Ask, then try a small version
I put a quick poll on Facebook. Twelve people responded. They valued reading with their children, but their routines varied. One person still read with children older than sixteen. In those responses, reading tended to fall to the primary caregiver; some dads read mainly on weekends. This was a small group of people I knew, not a finding about all parents.
The parents I spoke with also tended to prefer physical books at bedtime and limit devices then. That mattered to the design. The app would help us remember the reading, while the book remained the thing we shared.
To think beyond my own routine, I imagined a family. Kara was a working mother of Jackson, three, and Ella, five. She visited the library monthly, met a mothers’ group on Saturdays, and found books through that group, Facebook, Pinterest, and Amazon. Rick, her husband, worked as a startup product manager, traveled, and read with the children on weekends. He found books through friends, Twitter, and airports, and cared about illustration and gender-neutral stories. Al, Kara’s retired father in Irvine, sent books every three months, drawing on recommendations from NPR, Good Morning America, Goodreads, and his local bookstore.
Those were invented people, based loosely on myself, my wife, and my mother. They helped me consider who might choose a book, buy it, read it, or want to know whether a child enjoyed it.
I built a rough prototype in Keynote with Keynotopia. Then, for a week, I tried the underlying routine with my kids. Each night I photographed their faces and the books’ ISBNs. After reading, I asked whether they loved the book and invited them to tell the story back in their own words.
A first reaction was useful, but repeated requests told me more. I began treating a book requested three or more times as a favorite. I changed the design to record both the child’s response to a particular reading and how often we returned to the book.
Prove the technical pieces separately
I learned Swift by building small test apps. The first followed an online QR-code scanning tutorial. I adapted it to read ISBN barcodes, also known as EANs in this context. When it worked, I had evidence that I could build the scanner the diary needed.
Next came a server-side script. It called a book-data API with an ISBN and returned details such as author, illustrator, and title. Keeping that step separate let me work on identifying a book before connecting it to the rest of the diary.
Then I added server methods for “readings”: records that could associate several readers with one book. One tester added himself as a reader. He wanted to track which parent had read which book to his child. I immediately did the same, adding myself or my wife to our earlier readings. I started revising the flows to make room for both adults and children.
Learning Swift alongside new libraries and a backend took longer than I expected. Understanding how delegates, classes, and Storyboard fit together was the hardest part. Familiarity with JavaScript helped me approach the syntax. Automatic memory management and code completion also made the experience easier, though those conveniences weren’t unique inventions of Swift. Parse reduced the backend work I needed to do; relying on a platform was a tradeoff I had to consider along with that benefit.
Let testing change the plan
At the time of this entry, twenty people were testing the app. Every feedback email made me happy. Some who joined when it was rough had stopped using it. I hoped a better version would give them a reason to return.
I wanted to release a build every day. In practice, I was managing about two a week.
The original progress list marked date editing, the revised add-reader flow, reader deletion, and onboarding graphics as completed. Creating a reader had moved from the initial button press to Save, after people found the immediate creation jarring. Deletion also required deciding what happened to related readings. I had considered an adult-oriented question before removal, but a question alone should not be treated as proof that someone is a parent.
The unfinished work was still substantial. I needed loading feedback and protection against duplicate actions while a server request was running; an activity indicator by itself only explains the wait. Favorites needed a more efficient calculation. Settings needed a dynamic table and a scrollable reader list. Social features raised questions about followers, mutual consent, and default visibility. Periodic emails would need frequency and opt-out controls.
Those choices would shape how the app fit into a family’s routine. I started with the feeling of losing track of a book at the return slot. Now I had people showing me which parts of remembering mattered to them, and a working app I could change in response.