Just see, what they do to skype, if you don't know what I am talking about.
I hope you agree, that skype, that was a great software, has been systematically wrecked by microsoft. It has not only become unusable, it gathers more and more bugs (don't start to talk about the MS classics).
So MS did not get it in 40 years. You are still in good mood, that they have something valuable to tell about software development? Certainly, there are always new developers. But the culture did not change. Good luck with your hope to still learn something valuable from MS. The chance is just not too high, I guess. Yes, the author is not the company. But he takes his experience from MS as an example.
And, btw., my viewpoint is neither uninformed nor shortsighted. It is pure anger about 40 years of bad software and especially about ruining skype.
You one the other hand, assume things of me from 2 lines of comment. How can you believe, this is less shortsighted than you think my comment is?
Also, testing user acceptance/ ease of use of use of the interface is also important on a major product with millions of users. Let's not forget that.....
To be honest — I don't see that mistake much any more. Indeed I'm generally much more upbeat about the current state of testing that the author of this article seems to be.
The thing that strikes me most about testing is how much the profession and practice has changed over the last twenty years. All for the better.
In the 90's and early 00's it was very much as the article says. Seen as something for larger organisations only. Generally something done by a separate person or team alone. Almost never something integrated with ongoing development work (those ghastly integration sessions that took weeks and months… shudder…). Almost never something that developers were involved with themselves — which led to some god awful code coming out the end of the sausage machine. Testers were generally seen as folk who couldn't hack programming. And there were an awful lot of folk that I encountered with a testing job title in those years who were bloody awful at their job — who were basically human automated tests running through checklists and providing relatively little value to the organisation.
These days every language has half a dozen test frameworks, supporting different styles of automated testing. Knowledge of testing at some level is expected of developers.. Anybody but the rankest newbie should be able to discuss it — even then I'd be pretty shocked if even the freshest of faces didn't at least have an opinion and knowledge on the topic. Almost every codebase that I interact with will have some kind of automated test suite. Often really excellent ones. Testing is integrated into the development process rather than being a separate phase. Because of this a huge number of the dumb errors that the checklist-style tester had to catch are now caught by machines. Which has freed up the professional tester to spend their efforts on the hard problems. You get so much more value out of testers these days because they're spending their time coming up with evil edge cases, isolating those bastard intermittent issues, or feeding back a metric shed load of issues from sessions of exploratory testing. Stuff they never had the time to do when they were dealing with the errors that the devs should have caught themselves.
Don't get me wrong. I still encounter teams that really, really need a specialist tester but don't realise it. I still encounter teams that could be way, way better at automated testing than they could be and their shipping problems show it. The testing profession is currently having to deal with the idiocy that is ISO (see http://commonsensetesting.org/stop29119/OpenLetter/).
Life's not perfect.
But compared to 20 years ago — the state of testing is good!
