5 ms·

Robot With Long Finger Wants to Touch Your iPhone Apps

17 points by hugs 13y ago

vsComputer 13y ago
"Catch’s flagship product, Enterprise Tester, actually automates this part. It takes specs created by analysts and software architects and automatically generates test plans. These can then be handed over to manual testers to run." <--- this is how they test software in Hell.
- hugs 13y ago
  I agree. That is hell. There's good automation and bad automation. Bad automation is just automating the creation of more busy work for people. Good automation does the boring, tedious things for people -- so they can relax and go back to focusing on the things that make them happy.
philk10 13y ago
If only the robot could do all the thinking and observing that a skilled manual tester can do whilst testing an app... I'm all in favour of automating the boring repetitive parts and using automation for things humans can't do but AI ain't here yet
- hugs 13y ago
  I totally agree. The way I see it, there are two types of manual testing that often get conflated in conversation: 1) Make sure everything that worked yesterday is still working today. 2) Check out this new feature. Let us know what makes sense, and what doesn't. Automation is great for scenario one and (still) terrible for scenario two. Some expert testers prefer to call scenario one "checking". They don't want their high-skill work (which it is) to be confused with low-skill button clicking.
vsComputer 13y ago
There's an implied dichotomy in this article between "manual testers" and "testers who know how to program" which is unnecessary, in my opinion. There's a whole spectrum of technical skill that's being ignored. All software testers are using computers to do their work, and they are all using tools to make them more effective. I think this tool sounds really interesting and would allow a tester to generate lots of interesting tests. It could conceivably also speed them up, which is great, and it could allow for automated regression testing of touch features, which is also great. That doesn't really mean anything with regard to who is using the tool. You don't have to become a full-time developer to learn to use Selenium for tasks where it makes sense, and I assume that this tool is the same way. If it's not, then a collaboration between somebody who has those automation skills and the "manual testers" could still create a wrapper that allows the testers to use the tool effectively.
perze 13y ago
My non robot hand has 10 fingers and can execute up to 10 point tactile dependent tests, but that's not really the point of this article. First, there seems to be a very uninformed notion that being able to "program" makes you a better tester. Maybe, but that really depends on what your context is. What's important is the mindset of the tester and how you can bake that into the engineering culture that it's in. Second, speed. Agile isn't about speed. Agile is about people and how the people interact with each other to produce better "things". I think one of the places where bugs are introduced in any development cycle and that's when the developers write code. The faster your "speed" is, the faster you introduce "bugs". The ability to test "fast" is nice but you have to test intelligently to get the most bang for your buck and you need a human for that. Third, Manual testing was referred to as the "lowest form of life on the dev cycle" is not a reflection on the tester but how bad the organization's hiring methods, and engineering processes are. How you employ a tool or a person does not define and limit what testing is for. cum hoc ergo propter hoc. Not to take away from Jason's accomplishments in the test automation tool space and what he's done to revolutionize and simplify certain tasks that would have been really tedious and even annoying to perform over and over again. I thank him for that. But, if people who doesn't seem to understand what testing is for and make summary judgements and/or conclusions on the future of people with brains in software testing, I agree wholeheartedly with Keith Klain as to where they can put that Robot's finger (https://twitter.com/KeithKlain/status/362941136207233025 https://twitter.com/KeithKlain/status/362941136207233025). Peace.

17 points by hugs 13y ago

"Catch’s flagship product, Enterprise Tester, actually automates this part. It takes specs created by analysts and software architects and automatically generates test plans. These can then be handed over to manual testers to run." <--- this is how they test software in Hell.

hugs 13y ago
I agree. That is hell. There's good automation and bad automation. Bad automation is just automating the creation of more busy work for people. Good automation does the boring, tedious things for people -- so they can relax and go back to focusing on the things that make them happy.

philk10 13y ago

If only the robot could do all the thinking and observing that a skilled manual tester can do whilst testing an app... I'm all in favour of automating the boring repetitive parts and using automation for things humans can't do but AI ain't here yet

hugs 13y ago
I totally agree. The way I see it, there are two types of manual testing that often get conflated in conversation: 1) Make sure everything that worked yesterday is still working today. 2) Check out this new feature. Let us know what makes sense, and what doesn't. Automation is great for scenario one and (still) terrible for scenario two. Some expert testers prefer to call scenario one "checking". They don't want their high-skill work (which it is) to be confused with low-skill button clicking.

vsComputer 13y ago

There's an implied dichotomy in this article between "manual testers" and "testers who know how to program" which is unnecessary, in my opinion. There's a whole spectrum of technical skill that's being ignored. All software testers are using computers to do their work, and they are all using tools to make them more effective. I think this tool sounds really interesting and would allow a tester to generate lots of interesting tests. It could conceivably also speed them up, which is great, and it could allow for automated regression testing of touch features, which is also great. That doesn't really mean anything with regard to who is using the tool. You don't have to become a full-time developer to learn to use Selenium for tasks where it makes sense, and I assume that this tool is the same way. If it's not, then a collaboration between somebody who has those automation skills and the "manual testers" could still create a wrapper that allows the testers to use the tool effectively.

perze 13y ago

My non robot hand has 10 fingers and can execute up to 10 point tactile dependent tests, but that's not really the point of this article. First, there seems to be a very uninformed notion that being able to "program" makes you a better tester. Maybe, but that really depends on what your context is. What's important is the mindset of the tester and how you can bake that into the engineering culture that it's in. Second, speed. Agile isn't about speed. Agile is about people and how the people interact with each other to produce better "things". I think one of the places where bugs are introduced in any development cycle and that's when the developers write code. The faster your "speed" is, the faster you introduce "bugs". The ability to test "fast" is nice but you have to test intelligently to get the most bang for your buck and you need a human for that. Third, Manual testing was referred to as the "lowest form of life on the dev cycle" is not a reflection on the tester but how bad the organization's hiring methods, and engineering processes are. How you employ a tool or a person does not define and limit what testing is for. cum hoc ergo propter hoc. Not to take away from Jason's accomplishments in the test automation tool space and what he's done to revolutionize and simplify certain tasks that would have been really tedious and even annoying to perform over and over again. I thank him for that. But, if people who doesn't seem to understand what testing is for and make summary judgements and/or conclusions on the future of people with brains in software testing, I agree wholeheartedly with Keith Klain as to where they can put that Robot's finger (https://twitter.com/KeithKlain/status/362941136207233025 https://twitter.com/KeithKlain/status/362941136207233025). Peace.