Okay, so we lost some time last week. First things first.
Elodie, Xin and Glen got to work immediately on correctly implementing the caching system, which they fixed by Monday afternoon. Once that was done, Elodie and Glen continued server and cache optimization.
Jeremy continued polishing the data entry screen and the photography system. The photo system exists in two places with slightly different functionality; the kids can use it to take pictures of their projects for submission to the database and the staff can use it when entering new items to take their picture. Jeremy has the latter up and running and is working on implementing the former.
Xin spent another busy week on the front end. Working with the others, he has added a text search function with an auto-complete option. This allows kids to search for items by name and makes back end data entry so much smoother. He also worked with Allison this week implementing new UI textures, and continues to iterate on interactions in the workspace and on how we look up and display connections.
Allison, aside from her work with Xin and the textures, has begun building the pieces of our menu. On Friday, she and Chris went to Makeshop and took the majority of the pictures we will use in the app. We took the pictures using her iPad since all other pictures in the database will be taken the same way.
Chris, aside from helping Allison with photography, finalized the menu layout, which Allison usd to design the look of the menu. He also began outlining the four documents we will need for our final product: a user guide for the front and back end, a script for the staff to follow when using the app, a style guide for taking pictures in the app, and a paper addendum to the inline documentation for the app.
Also, we returned on Wednesday to conduct out missed playtest. There were very few children but we were able to spend time with several of the staff getting their impressions and suggestions.
But then, it happened again.
After last week we decided we had been too lax in our build pipeline. We had decided to change code the day of the test without a backup version that was worth testing, and we paid the price. This week we decided to a) constantly test systems each and every day, and b) build the app and test it the night before.
We have been building the different systems separately, which is to say we had one app for the front end, one app for data entry, and one for photography. During the week they were testing fine separately. Because of the apparent success with these builds we continued developing on Saturday instead of calling feature lock on the playtest build, deciding instead to meet early on Sunday and integrate all systems.
On Sunday, once we integrated all the different systems into the app, we began having seriously app-breaking memory issues. Again, we realised the cause of the issues later that evening, and again we cancelled a playtest. We rescheduled for Wednesday next week.