Skip to content

Engineering Note on 10 31 2016

eric_hochwald edited this page Nov 1, 2016 · 13 revisions

Engineering Note

Test of the main feature of our app in preparation for Leg 4.

Lint

The initial lint scan result is in the folder named tests. We fixed all of the relevant errors, although there are some left over due to a bug we have yet to deal with and some incomplete features. After editing the errors from this lint report, the final lint report is stored in a folder named [lintreport] (https://github.com/JoshuaW1990/comp150mob/tree/experimental/lintreport).

Fabric

We have installed fabric in android studio and start to utilize crashlytics to record the crash events when testing the app in the emulator. Here is the screenshot of the fabric dashboard of the crash report of our app:

Static Analysis

In order to test our app statically, we ran our compiled .apk file through Veracode. With a low policy evaluation our app scored a 90 as a security rating. A screenshot of the above report can be seen below: alt text After looking at the assessment report, it seems that that majority of the most critical security issues stem from careless authorization information management and loose variable handling. The steps to remedy this involve cleaning up our module structures and consolidating information within our modules before passing data around. Luckily, our app handles very little sensitive information right now (barring password information that would be bad if leaked if our users use the same passwords across all their accounts), so these security issues are not incredibly detrimental if exploited. However, with this being said, the security of our app will be looked at with more scrutiny as we move from a MVP to a production ready program.

In Addition / To Do.

We have also figured out a way to store a list of a user's liked and disliked artists into Firebase using an instance of the interface as part of the User object. Next we must implement the album rating feature, and the user like and dislike artist feature. We must also clean up the front end displays of the app to make them prettier. Finally we have to move beyond our test database, and use the web crawler to set up our complete database of artists and albums from sputnikmusic.com .