vote up 1 vote down
star

I'm doing a lot of research into how certain areas drive customer perception of quality.

One area I'm interested in is compatibility. I'd like to hear any ideas you have on :

  1. Defining compatibiltiy
  2. Testing for compatibility
  3. Measuring your progress

I'll eventually have the same question for other areas - assuming this isn't another tumbleweed question.

flag

2 Answers

vote up 2 vote down
check

Alan, this is an interesting topic, I'd be interested in what your research comes up with.
You are not asking for anything specific... so it is hard to answer :). See if the text below fits with what you've found in your study.

While compatibility can be defined as "capable of existing together in harmony" or "ability to work with another device or system without modification" (here), the ambiguous definitions still leave room for each one and company to come up with their own compatibility tests. Which isn't, completely, a bad thing (it might be efficient for an environment to develop its own dialect and understanding).
One way or another, often compatibility term mingles in docs and practice with integration, interoperability, interface, system, acceptance, etc testing... It is hard to take these apart.

So compatibility may have many facets, which can't be ranked, as each can have different importance at different stages of different projects. You'll have:

  • Forward and Backward Compatibility
  • Compatibility between components of the system
  • Compatibility between different related systems
  • Compatibility between different (less related) systems that may not be from the same company or purpose (The question about "a device driver being compatible with MS Office" may sound ridicule, but I know of a case where that would make sense).
  • Environment compatibility like OS compatibility (or a close counterpart, browser compatibility)
  • Environment compatibility extends to hardware/settings compatibility too.
  • Compatibility with Peripherals, Display resolutions...
  • There are surely more :)


On a practical note, my own little experience confirms what Kaner wrote in the first LAWST report -- OS and HW compatibility testing can benefit from a smart automation almost immediately:

Most of the benefit from automation work that is done during Release N (such as Release 3.0) is realized in Release N+1. There are exceptions to this truism, situations in which you can achieve near-term payback for the automation effort. Examples include smoke tests, some stress tests (some stress tests are impossible unless you automate), and configuration/compatibility tests.


Last, sincerely I don't know how to answer question 3 (about measurement)... But I don't see how the measurement of compatibility testing might be different from other tests.



Postscript:
Of course, I think the most widely known "compatibility test" is the love and dating one, that tries to assert whether a couple is off to a good start :) (often based on silly parameters)... At local newspapers, there are services that would do that in exchange of 50 cent SMS message :). Is this the one you had in mind? ;)

link|flag
excellent answer - thanks. – Alan Dec 4 at 14:26
vote up 0 vote down

Paramter to measure your software compatibility is when you are confident to try your software at an unknown person's machine and also when the other person shows no hestitation in installing/launching your software with a confidence that it wont disturb any of his currently in-progress activities on the machine.

regards Aj

link|flag

Your Answer

Get an OpenID
or

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