2

I am about to release a piece of software that will be licensed with the GPLv3 (the project itself is "for fun" and has no commercial dimension in it).

The software is intended to help designing digital clocks and I know the core functionality works as - beside having unit tests - I used it for designing a couple of clocks myself. Yet, before getting the software advertised on hackers and makers sites, I would like to get a few voluntary testers... trying to use it and reporting any problems.

The only tests I did so far in my life were with fellow programmers or with the customers. So I would appreciate input from the community on:

  1. Where to find voluntary testers (is there a well-known website that escaped my googling perhaps?)
  2. How to brief them (fellow programmers and clients know a great deal on the software they are going to test... should I try to brief in depth these "unknown" testers or should I contrarily try to let them "on their own"?)
  3. How to structure the tests (I'm not asking what to ask them to do here, but if I should structure my test in some specific way: for example in more phases, like having one group to make a test, then fix bugs, then having a second group test again... or collecting data in a particular way [e.g. with a questionnaire with scores rather than with free-form e-mails])
  4. How to give maximum visibility to their contribution (Again: I can't afford to invest also money on this project, but I would be happy to give them as much glory as I can to them!

The focus of the test would be primarily usability (Does the GUI make sense? Is the workflow easy to follow? Are there unexpected behaviours?). But of course any bug that should emerge in the core functionality would definitively receive all my attention too!

Thanks in advance for your advice and time.

1
  • 1
    Besides a very light briefing to tell the potential testers what they're getting into, I would only provide them with the official documentation. That way, you'll test both your documentation and the software itself at the same time.
    – Maxpm
    Jul 4, 2011 at 4:23

3 Answers 3

3

Friends, Family, and co-workers is probably the easiest. You could also try some social media (ie: facebook, twitter). These work very well for User Acceptance type tests. If you want more thorough testing performed, you could always try one of the many testing forums on linkedin or even see if someone like Weekend Testers would be willing to take a look at it. Although many testers may spend their day testing, they're often willing to take a look at something off the clock just for curiousity or practice. Even many programmers forums would have people willing to test it. One thing about many of us in technology have in common, if it's useful or interesting, most people are willing to check it out.

As for structure and briefing, if they work in the technology sector, release notes would probably be good enough. Otherwise, a list of what it should and (possibly) shouldn't do.

1
  • Couldn't edit the previous comment: Weekendtesting seems a good lead that I was not fully aware of.... I surely will investigate more.
    – mac
    Jul 3, 2011 at 23:56
1

If the project is "for fun" and have no commercial aspects to it, then what are you concerning getting a group of beta testers for? Just release it already. That is my tongue-in-cheek answer. Just look at Linux when it got initially released into the wild?

For me, I don't really mind having bugs in your code, but how you go about resolving the reported problems. If it is quick and to the point, then I would be more than happy.

However, sourcing a group of beta testers and then managing them is a complex objective. All those question that you mentioned are valid. One thing that I found useful is to select areas for the group to concentrate on. eg, testing buttons, and then messages in the dialog boxes and so on..

2
  • 1
    Thanks for this! (+1) - On why I am concerned about testing the software: first is because I have a couple of leads with major hacker/maker online editors to publish news about the project and the software I developed for it. When you release free software "in the wild" the adoption curve is very gentle, so that major bugs (hopefully discovered early) affect few people. In this case I could easily get thousands of people downloading the software as soon as the article[s] would be published. Secondly, I simply don't see GPL as an excuse to be a sloppier-than-commercial-level work! :)
    – mac
    Jul 3, 2011 at 23:53
  • I don't think releasing it into the wild will get as much feedback as formal testing. How often do you think the average computer user thinks "Gee, I'd better report this to the developers?" The answer is rarely, if ever. They just curse the software for its glitchiness.
    – Maxpm
    Jul 4, 2011 at 4:29
0

I think letting your software free in the wilderness will get the testers you need.

If your software is worth using, the people that start using it are becoming also testers of it. I don't know of any professional testers that are doing it as a hobby too :)

1
  • I am not looking for necessarily professional testers. Average computer users will do, I think. Also: my question is not about if I need testing prior to release. It's about how I can get the testers I need. Articles about this project (and its software) will be published at least on one site that gets various million unique visitors per month, probably two of them. So I definitively would like to get the software tested with a small group of people before getting millon of users frustrated for some stupid bug I could have sqashed before release...
    – mac
    Jul 3, 2011 at 13:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.