There was a recent discussion on the value of rephrasing statements and then a post appeared on my radar that almost dismissed the value of “it works on my machine” and even to some extent attracted meme’s and comments like “we’ll send your machine to the customer then”.
What’s your view on it, is it worth rephrasing the stance on how it is viewed?
My own comment on that discussion was along the lines of.
“I find it incredibly useful.”
Firstly it means someone has actually run the application and maybe even done some testing which is a plus particularly when its a developer stating this.
When it’s a tester stating this, its also a clarity statement where they are not broadly saying that it works but they are adding more specific information as where it works.
When we explore and find scenario’s where something does not work, knowing where it does potentially work allows us to narrow down our investigation. Its not a point blank it does not work, it empowers a comparison where the differences can be narrowed down and investigated.
So I find “works on my machine” is incredibly useful to testers especially if we understand what makes it “their” machine compared to “our” machine.
Usually I hear “works on my machine” used to dismiss someone else’s concern, so I think that’s where the negative connotation and meme come from.
I consider it in the same way as(not sure which one I heard more often): I can’t reproduce it, I followed the scenario and nothing happens, I did this and got another problem but can’t see the one you described, I tried to see an issue there but I haven’t, looks to be as designed/expected/specs.
It happens once every couple of hundred bugs reported.
I find it is such a good way of learning:
have I missed to add a detail in the report for the issue;
is there some system setup in which this can actually work?
what is the impact overall, have I thought about that? how often would the problem actually occur and to whom and what impact might it have(least/worst);
I always provide a gif/recording of the issue and not just screenshots. I’m always going through the whole flow, so they can SEE that it doesn’t work. And they’ll often replay the video over and over to reproduce the steps.
Hence I never get “it works on my machine” because they literally see something is wrong.
Its a really interesting one this. Historically I used to get this a lot from devs, especially with desktop apps in the day. The dynamics between dev and test were very different as well.
But nowadays, I agree with @kristof, I never get that response anymore. Our bugs are better, maybe we’ve learnt over the years to put a bit of extra work in to give the best evidence to dev because in the end it saves time. The tech is different, its deployed in the cloud so as long as we’re testing against the same build as dev are looking it at, it won’t happen. Also, these days we’re all one team so we’re all working to achieve the same goal. Far more positive relationships.
So if ever I got “works on my machine” as a response, my response would be, lets get on a call and look at it together. I don’t think rewording the response is going to add much value. Take it on the chin as its a fact to whoever says it, its how you respond thats the key.
I think that’s very valid, I never really get it in a dismissive way these days which is why I see it as a positive.
I do find myself asking does it work in your environment quite a bit, some of that is because a lot of the developer testing is in their own environment, maybe using docker, stubs, mocks or a debug run that is maybe more forgiving than the environments I often test in.
When it does it usually means we can narrow things down fairly quickly.
