It’s out of the question that software testing is a discipline which requires well structured working – no matter which type of test has to be done!
Often your test (managmenet) tools can help you to organize your testcases (create, edit, delete), and manage your testdata, test iterations, logs on executions of testcases, information about what testcases have been run and not run on SUT (system under test) for testing release x. Also it’s possible with test tools to create and manage your found bugs. That’s all nice.
But is it for example possible for your test (management) tools to figure out at a glance which testdata has been used for your test executions for a special software version/release/test iteration you’ve tested? Can you save the outputs of your SUT with your test (managment) tool so that you use them by way of comparsion for regression testing in further software releases or test iterations? Do you know directly the right logfile and the right testscenario document with screenshots you have created for a bug in software release x when you only use your test (managment) tools? This information is very relevant and usefull if your current testcase will result in the same or similar bug. Can your test (managment) tool handle testing technics like ‚exploratory testing‘ or the methodologies of ‚rapid software testing‘? In relation to this: Is it possible to create, save and manage your models for your SUT only by the help of your test (management) tools? Maybee it’s better to use a mind mapping software for this kind of task and save your mind maps in a certain folder. Is it also possible with your test (managment) tool to add and manage some other data like requirement documents, release notes, use case diagrams or other stuff which belongs to your SUT? Have you something to do with customer missions (for example retesting of bugs found by cusomers)? If yes, how do you organize your work for this?
Maybe that there are pretty good test (managment) tools which will alow to manage all of these tasks.
But I think that every test (managment) tool has it’s week points and some day there will occure an issue which you have to check with your test (managment) tool and the tool doesn’t allow that. And with the questions above I only named examples…
In my practical experience as a software tester I found out, that I unfortunately have to answer all these questions above with ’no‘ or at least not completely with ‚yes‘, when I only wan’t to use my test (manamgement) tool for my daily work.
And that’s the point why I’ve early decided to organize all my testing data not only in my test (managment) tool but also in a well defined folder structure on my operation system I work with.
As you know software testing is something you often have to do in a loop – again and again in different iterations and for different software versions and releases for your system(s) under test – keyword regression.
So I created a folder structure which concidered this. That means that the folders/directories are allways the same and should also be named the same for every time they will be created again but for different test iteration, software version or release I have to test.
The screenshot on the left shows this folder structure:
The root folder is called ‚Test_Ralf‘ and it contains different folders which I’ve created for testing different releases, software versions, spcial deliveries and so on.
Under every ‚release/test_iteration/sw_version-folder‘ you can find the folders for the different systems under test. Often it’s more then one, so I sperated it in different subfolders called ‚SUT1‘, ‚SUT2‘ and so on. In real practise I won’t name them ‚SUT1‘ or ‚SUT2‘. I name them for example by it’s normal project name. Every ‚SUT‘ folder has the same subfolders in which I can save and edit different data which is relevant for my daily testing work.
In the folder ‚Defects‘ for example I save screenshots, logfiles, videos, descriptions of test scenarios of bugs in SUT I’ve found. So I don’t need to ask my test managment tool, which data I’ve assigned to bug number 4711 I found on 2012-07-04 for example. I can check it right now in the right ‚Defect folder‘ and don’t need to search in the test (management) tool about that. Often this will cost more time than checking my test folders for that.
Further example: In the subfolder ‚Test_Documentation‘ all my test scenarios I’ve done can save as documents. Especially those testing work which I’ve done during ‚exploratory testing sessions‘. My test management tool doesn’t support ‚exploratory testing‘ in the way I need it. So this is the only option I have to save my working results.
Another subfolder is called ‚TestData‘. This is a very important folder I think. Every testcase needs its testdata. And testdata isn’t fix at the most cases. It’s depending on time periods, settings, user rights, master data as precondition on SUT and so on. Some testcases requires that the SUT makes outputs in form of XML files or as .xls files. And often also special input files are necessary for example to test special import behaviour of the SUT. However, all these data can safed in this subfolder. My test (managment) tool doesn’t allow that in that way. The only thing it does allow, is the option to describe the required testdata I need in every single test case (scribt based testing). So I’ve to save my testdata in textfiles or something else in my ‚TestData‘ folders directly.
There are a couple of other subfolders which purpose I haven’t described yet. But I created them because my test (management) tool can’t manage to save that corresponding data in the way I want, so also here this is the only way to save and manage it.
And my experience shows me, that working with these folders is more faster than useing my test (manamgement) tool for some isues.
The only thing is, that it’s realy exhausting to create these folder structure in a manual way – especially when you have to do this again and again for every test iteration, software release under test and so on.
Thankfully I’m able to program :-)! I’m working on a Windows OS and so I’ve created a small ‚.bat file‘ which can do this. I only need to type in the name for the testfolder for the release or software version or test iteration. I also have the option to name all my folders and subfolders I want and to create more or less subfolders. The script creates all empty subfolders in the known structure (see screenshot above) for a new test iteration. With the script I don’t need to do this work by hand.
If you’re more intrested in this ‚.bat script‘, then please don’t hestitate to send me an email to firstname.lastname@example.org and I can send you the ‚.bat script‘ and a description of how it works and how to use it.
kind regards, Ralf