Join 34,000+ subscribers and receive articles from our blog about software quality, testing, QA and security.
 

Configure required fields when adding test results


#1

I don’t know if it’s already possible, but I would like to force some entry fields in the “add test result” dialog to be mandatory.

For example, we really want the users to enter an “elapsed time” for test execution as well as the “version” tested.

Can we change the fields configuration so they would be required ?

Thanks !


#2

Hi Louis,

It’s not currently possible to make the built-in fields required. You can mark custom fields as required and we are also considering this for the built-in fields. In general we recommend making not too many fields required, as it makes it more time consuming and difficult to enter test results. For example, if you are mainly worried about accurate data for the time forecasting when users forget to enter the Elapsed Time field sometimes, TestRail will automatically calculate average times from past results. So having 100% accurate data for this is not needed.

I hope this helps.

Thanks,
Dennis


#3

Thank you for your response. Good to know that it calculates the forecasting using the average instead of using only the elapsed time values of the executed test cases.

Unfortunately, the version is really important for us because we use the same test run with different builds. For example, a tester may start testing using the version “6.0 b456” and will finish testing with version “6.0 b487”. Then, if needed, we can do a query in the database to verify which test was executed with which version. If we don’t enforce that field, we cannot measure it after. :frowning:

Thanks,


#4

Hi Louis,

Thanks for the additional feedback, that makes sense. We are looking into making this customizable in the future. You could simply add another “Build” custom field or similar, possibly even as a dropdown field so users cannot make any mistakes when entering the build number. You can make custom fields required and we will also consider this option for built-in fields.

Thanks,
Dennis


#5

How often are the testers moving to a new build? Every day? Every week?

I guess it depends a bit on the project, the number of testers available, etc.

We have nightly builds but only designate weekly (sometimes every two week) builds as our evaluation or milestone builds. We use the “cloning” feature of test plans to effectively rename our test plans per milestone build. Typically a tester will only be marking results in one milestone plan at a time – although they certainly could mark results in more than one.

In a sense we simply ignore the version field in our workflow since every plan is uniquely labelled as corresponding to a build.

This can and does result in some early milestone builds only being partially completed since we’ll move on to a more viable build at times… but it’s a conscious decision w/o much consequence and we have accurate records implicitly.

Just sharing one alternative… your mileage may vary.


#6

Hi Donald,

Thanks for the feedback, that’s very useful. We also have other customers that use test plans for their builds (or iterations) and it’s not uncommon that they don’t finish the entire test plan until the new iteration starts. As you mentioned this isn’t a problem though and it builds a nice history of results for each build/iteration as the project progresses.

Regards,
Dennis


#7

[quote=dgurock]Hi Louis,

Thanks for the additional feedback, that makes sense. We are looking into making this customizable in the future. You could simply add another “Build” custom field or similar, possibly even as a dropdown field so users cannot make any mistakes when entering the build number. You can make custom fields required and we will also consider this option for built-in fields.

Thanks,
Dennis[/quote]

Hi Dennis, I was wondering if you were planning to implement this soon, because it is important for us. Since we mainly have 1 project with different versions, having the possibility to enforce the version field gives us better results when we query the DB directly and want to extract the test executions done for a particular “Version”. If not, the field can be “null” and we cannot determine the version used to execute the test, etc.

Thanks in advance,

Louis


#8

Hi Louis,

Thanks for your feedback. This change is not currently scheduled for a specific update, but this doesn’t necessarily mean that it won’t be added in the coming months, we don’t have a timeframe for this at the moment though. This is a request we’ve received multiple times now and it’s on our list of things we want to add eventually. You could create a required custom field for this in the meantime if this is critical for you, or ask your project members to enter version numbers for every test.

Regards,
Dennis