Hexawise – more coverage, fewer tests

Hexawise is an interesting concept. It’s worth checking out their 3-minute introduction to see what they’re talking about, but basically the idea is that with every application, there are many variables that lead to massive amounts of possible test cases. So you just enter all the different variables into Hexawise and it whittles that number down until you have a small set of tests that will result in the best feature coverage.

Sounds pretty handy, so I might give it a go after I’ve found a decent test management tool (still looking for one – any suggestions?)

4 thoughts on “Hexawise – more coverage, fewer tests

  1. Justin Hunter

    Trisherino,

    I was checking out http://www.softwaretestingclub.com and came across a link to your blog posting. Thank you for the positive mention! We’re starting to see our user base grow pretty rapidly now almost entirely due to word of mouth, tweets on twitter, and blog posts like these. It is really satisfying to see users sharing positive experiences with our tool after the months of late nights we put into building the tool and making lots of small iterative improvements to its features and user interface.

    Our goals in launching the Hexawise test design tool recently have been to help testers: (1) design test cases faster, (2) leverage powerful, proven, applied statistics-based algorithms “behind the curtains” to generate test cases that will consistently find more defects per tester hour than manually-created test cases, and (3) make the test design process fun and collaborative.

    We believe that great software testing comes from great testers; test design tools help them improve their aim and increase their efficiency and effectiveness.

    For any of your readers interested in checking out Hexawise, we have a permanently free version available which is suitable for most test design tasks as well as a fully-featured commercial version which has free trials. Free no obligation sign-ups are available at http://www.hexawise.com/users/new

    Thanks again and don’t hesitate to let me know if you have any questions or feedback!

    – Justin
    ________________
    Justin Hunter
    Founder and CEO
    Hexawise
    http://www.hexawise.com/users/new
    “More coverage. Fewer tests.”

  2. Ashwin

    Also, please explore http://www.TestersDesk.com – The Pairwise Test Case Generator does a similar thing. That is, it takes the test parameters and produces a test set that covers all pairs of values.

    You can also check http://www.pairwise.org that lists some other tools which base on the same technique, with slight variations from each other.

  3. Michael Bolton

    Hi, Trish…

    I’d recommend that you check out two online resources. The first is one of mine.

    http://www.developsense.com/pairwiseTesting.html

    The second is a paper by James Bach and Pat Schroeder:

    http://www.testingeducation.org/wtst5/PairwisePNSQC2004.pdf

    Those who read carelessly may believe that the second article negates the first, but it doesn’t. The second paper tempers the first. That is, for any test technique, or for any test strategy, or for any test tool, it’s important to know when you should or should not use it; when it’s a great idea, a bad idea, or worth a try.

    I haven’t evaluated Justin’s tool either. I’d like to–I had a very stimulating set of conversations with Justin in March, and he sent me a very important book, for which I thank him–but a good occasion for using Hexawise hasn’t yet presented itself, so I’d prefer to wait until I encounter a productive environment for it. In any case, a good understanding of combinatorics, orthogonal arrays, and pairwise approaches will help us to make the best possible use of it.

    Cheers,

    —Michael B.

  4. Trisherino Post author

    Hi Michael,

    Thanks for providing those resources. I agree with the concepts – that pairwise testing can be useful, but should not be treated as a silver bullet solution to all situations.

    As a clever developer said to me today, it is accepted that there are always multiple ways to approach a programming problem, so it stands to reason that there will be multiple ways to approach a testing problem.

    That said, I can certainly see the temptation to subscribe to one or two techniques as “best practice”. After all, if I have followed “best practice” then it’s harder to blame me or my methods when my strategy fails. So am I more scared of the risk to the product or the risk to my career?

    Getting back to pairwise testing, I had heard of the concept but didn’t know much about it. So thanks very much for the links, they explain the concepts very well and I’ve found them to be very educational.

    Trish

Leave a Reply