...
- We lost a lot of crucial time dealing with development issues in the beginning of the implementation process. Computers that were running on Windows could not be used to develop and eventually , so we first tried to use the athena cluster to develop. It turns our that any software you install on an athena computer is wiped on log out so we had to re-install everything all over again. Eventually we had to start booting Ubuntu from flash drives connected to our computers.
- Scripts was not working, so we had to host on Heroku and in order for our back-end to work, we had to pay $20
- While developing, our ".gitignore" file was being buggy and would not ignore certain files making pushing and pulling a hassle. Eventually we got it to work
...
- Keeping in mind the feasibility of any given feature during brainstorming. A voice interface might be awesome idea, but implementing it might be too much to ask for, given the time we had.
- Everybody is not going to agree on everything, but disagreement can sometimes be a good thing. Discussion has to be a delicate balance between not disagreeing too much that nothing gets done and agreeing too much that an environment of group think fosters which leads to the implementation of bad ideas.
- Things that might seem simple because you see them everywhere in your daily life might not be so simple. For example, a simple shopping cart can be very difficult to implement depending on your website's back-end.
- Don't bite off more than you can chew. We severly underestimated the amount of work that a fully working shopping site requires.
If we could do it again
Looking back, we might have chosen a better user population than the elderly. In the city, it is difficult to user test the elderly. We would have also started using ruby-on-rails since the computer prototype. We ended up implementing everything for the computer prototype using javascript, css and html, and eventually had to scrap all of that code and start over for our ruby-on-rails implementation.
...