Authors: Chetan Giridhar, Rahul Verma
image: http://freedigitalphotos.net/
Performance tests are instrumental in keeping a check on product quality in terms of reliability and scalability. One essential area of performance measurements is application launch time which is the time elapsed from the operation which triggers the launch (double-click/command-line etc.) and the moment at which the application becomes responsive. This measurement would be useful in atleast two contexts:
– Measurement of launch time of the application under test
– Measurement of launch time of applications on a system with and without your software installed (e.g. for AV products)
The scenarios could fall under the category of performance testing or benchmarking based on whether you are conducting the test to:
– Measure the time to launch (and compare against the performance criterion in requirements, if you are lucky!)
– Measure and compare against the previous version of the product
– Measure and compare against a competitor product
The above could be achieved with a stopwatch, which wouldn’t be a very reliable method most of the times as:
1. start-up time would be in a couple of seconds
2. knowing exact point of “responsiveness” is a little tricky in manual testing
3. what falls after the decimal point is interesting in such tests
4. such tests require multiple readings which would be good enough to force someone to resign in a few days
AppTimer is a free tool for the purpose and this article discusses how you can use the same for the performance testing and benchmarking in terms of application launch time.
