Wonders of team work...

Kornfeld Eliyahu Peter

So she added some new features, fixed it at the system level as needed, and changed the tests... Those features passed the tests but broke down plugins were installed on live systems... So a bug report filled and she expressed her will to fix it immediately (but set the severity of the bug from Critical to Low) - it took 5 days and need to be applied manually, as it will be part of the next release at the end of the month... Her mistake was to change the tests and not to add to them a new rule, so I expressed my opinion, that let the same person write the code and the tests is not a good idea as it can be the source of the infamous 'it works on my machine' bug. I also expressed my opinion, that the person who has assigned (by boss or by chance) to a bug report, should not be the one who set the severity level, especially if it is the very same person who ruined the code in the first place... So I got a lot of buzz and she got the glory (not clear of what)... I feel a bit like a bad dream...

Skipper: We'll fix it. Alex: Fix it? How you gonna fix this? Skipper: Grit, spit and a whole lotta duct tape.

CodeWraith

Kornfeld Eliyahu Peter wrote:

So I got a lot of buzz and she got the glory (not clear of what)...

You just earned some Code Monkey Points. Congratulations. When you have done something, it works and someone who had nothing to do with it steps in and gets all the glory: Lots of CMPs. When someone else has done something, it fails spectacularly, you had nothing to do with it and still get the blame: Lots of CMPs. When those who actually did it still get some glory for nothing: Platinum CMPs for CM who got blamed. Seriously: I hate office politics, and that usually is what is actually going on when such strange things happen. Be a little careful.

I need a perfect, to the point answer as I am not aware of this. Please don't reply explaining what method overloading is

Pete OHanlon

Sounds like you guys need something like a pre-commit code review cycle.

This space for rent

Matt L

I don't fully agree that the person who writes the code shouldn't also write the tests. The only benefit I see is the person writing the test will be in a different mindset to the coder, which can lead to some tests that the coder wouldn't have thought of. The bad side is that it would take so long to have them bounce code and tests back and for until both were happy, I'd think it would cost more money than it would save. Also, you assume the person writing the tests have a complete understanding of all units of logic that go towards fulfilling the requirement. Even if the code was documented and we'll written, there would still be induction period for the person writing the test to get acquainted with the new code that the coder wouldn't face. It's all a cost/benefit ratio that I think favours the coder to do both.

Kornfeld Eliyahu Peter

To write a test all you need is the method to test (the signature)... No need of knowing what is inside, but what is expected by requirements... Also - a test oriented people will not drop an existing test of an existing method, just because there is a new requirement (especially if the method actually an API for externals too). The one who wrote the code, and made sure all parts of the system play nicely with the new code may think that all is good and drop (change) the existing test, neglecting the fact that there are callers from the outside too...

Skipper: We'll fix it. Alex: Fix it? How you gonna fix this? Skipper: Grit, spit and a whole lotta duct tape.

Kornfeld Eliyahu Peter

We have. Very ineffective... I did a statistics last year - 83% of the bugs customers reported were there in our code and approved by QA to commit to the latest stable...

Skipper: We'll fix it. Alex: Fix it? How you gonna fix this? Skipper: Grit, spit and a whole lotta duct tape.

Pete OHanlon

So, the takeaway from this is that your existing processes aren't fit for purpose, and they need to be tightened up.

This space for rent

Kornfeld Eliyahu Peter

Pete O'Hanlon wrote:

tightened up

Blow to pieces and rebuild from the ground you mean... Tell me about that... :(

Skipper: We'll fix it. Alex: Fix it? How you gonna fix this? Skipper: Grit, spit and a whole lotta duct tape.

CodeWraith

Matt.L wrote:

The only benefit I see is the person writing the test will be in a different mindset to the coder, which can lead to some tests that the coder wouldn't have thought of.

So you never overlooked something when proofreading something you wrote? Did you never fall into the trap of repeating errors based on some misunderstanding over and over again? Any tester can tell you that the author usually is the last person who should conduct or write any tests. He usually is not objective enough.

I need a perfect, to the point answer as I am not aware of this. Please don't reply explaining what method overloading is

Kornfeld Eliyahu Peter

CodeWraith wrote:

not objective enough

Not objective at all, you should say! After all, it is his making! He loves it!!!

Skipper: We'll fix it. Alex: Fix it? How you gonna fix this? Skipper: Grit, spit and a whole lotta duct tape.

CodeWraith

Kornfeld Eliyahu Peter wrote:

He loves it!!!

As if that were the only problem. I also have already seen what happens when the tests become some kind of ritual which everyone must perform, but nobody really gives a moist fu.. about it. The only thing that counts are andless rows of green tests that actually test only very little.

I need a perfect, to the point answer as I am not aware of this. Please don't reply explaining what method overloading is

Lost User

Any tester Anybody can tell you that the author usually is the last person who should conduct or write any tests. Furthermore any auditor or certifying authority will tell you the author must never write nor perform the testing and in some cases should not even have access to the testing suite.

Sin tack the any key okay

Lost User

Once again, the Simpsons has the answer.. :laugh: [Homers Enemy-Nuclear contest scene+Frank goes crazy - YouTube](https://www.youtube.com/watch?v=rOgS8gTATv8)

Now is it bad enough that you let somebody else kick your butts without you trying to do it to each other? Now if we're all talking about the same man, and I think we are... it appears he's got a rather growing collection of our bikes.