Tests not being up to current specification is quite common. But if it takes a day to find this out likely either the script or the software isn’t really written in a maintainable manner.
Lelorgen
1 month ago
And that’s why every testing proj should start with a test along the lines of Assert.True(True), to see if the test are even running
igorski81
1 month ago
Maybe an unpopular opinion, but I would argue that while frustrating, this isn’t a “wasted day”. If a change in implementation affects the tests, the test should be updated to reflect the new requirements, that’s part of a software lifecycle.
Now, if a full day was spent before finding out that the test was wrong, I would argue that the person that made the change should have been aware of the test and its implementation, even when the test isn’t maintained by themselves but provided through their supervisor afterwards.
ThiccStorms
1 month ago
this graphic is so cute, whats the source? love the art
nujuat
1 month ago
Once I did this physics assignment; in classical physics to first order a system can either have a natural frequency of oscillation or a natural decay rate, or both.
The question was to find the frequencies of oscillation of a particular system. But i found that it had one frequency of oscillation and one natural decay rate. So I spent like a whole day of undergrad trying to figure out where I went wrong. Eventually I gave up and just wrote what I found.
It turns out that the system did only have one natural frequency and one natural decay rate, but the professor didn’t check to see whether or not they were both frequencies because it’s theoretical physics and details like that don’t matter lmao
CrashGaming12
1 month ago
Now you need to become your supervisor’s supervisor
No-Con-2790
1 month ago
You just learned a valuable lesson. A few more of those and you will be a senior.
Piotrek9t
1 month ago
Consider it a lesson learned, most of us have fallen for this at some point. For exactly this reason, when receiving a bug report, I defaulted to checking for any obvious errors on my side first but seconldy to check the specifications and test methodology before actually jumping into debugging.
-Redstoneboi-
1 month ago
my classmates wondering why their binary adder machine isnt working (the given circuit diagram had two switched wires)
DisastrousCrow11
1 month ago
If only I got a penny every time this happened.
I was the one who wrote the test script.
DevAlaska
1 month ago
In which industry are programmers and testers still divided?
Unusual_Flounder2073
1 month ago
I didn’t get a job because my code worked per the requirements but failed the test case. Even though I told them the test case was wrong.
basvas4
1 month ago
+ 8 paid hours, sounds great!
PastaRunner
1 month ago
To be honest, unless you’re an intern or something, it’s on you to verify the tests as well.
Tests are generally written to verify some specific behavior. Your code not meeting that exact behavior but doing something else which is also correct is a very common state to be in.
Da_Di_Dum
1 month ago
Been there
DoNotFeedTheSnakes
1 month ago
I’m sorry did you say “the test script provided tonus”?!?!
Bro if the test script is provided, I’m making the MVP (minimum viable program) that clears it.
Monkey patching and class mocking all the way through!
TimeSuck5000
1 month ago
If you haven’t wasted an entire day yet, are you even a developer?
inotparanoid
1 month ago
Who is testing the tester? Who is professoring the professor?
Works out if your codes being run in tests and passes it
sebnukem
1 month ago
We have all been there.
rdmille
1 month ago
My manager, himself, set up the tests to simulate a part for a product we were developing. The customer told him it wasn’t working, and manager argued that the scripts said it was, end of subject.
Customer knew me from other projects and called me. I ran the scripts, and yes, they said it worked. I ran the sims myself, looked at the results, and of course, it wasn’t working. The sims were set up incorrectly, and his results were all “don’t care”. Which he saved over the reference results. So, his comparison script was right, since “don’t care” matches anything…
Took me a week and a half to fix his shit.
CryonautX
1 month ago
I don’t understand why you would take the whole day on this? Just find out why the test script is failing your code and you’ll quickly get to the source of the problem which in this case is that the script is wrong. How exactly were you going about this problem? Just shooting shit in the dark?
Precious_Owl7733
1 month ago
Spent 8 hours debugging, only to realize it was a semicolon missing.
PhogeySquatch
1 month ago
This reminds me of when we were wiring a two way switch in shop class. We had this little wooden section of a wall with a power cord, switches, and a lightbulb. Student after student tried wiring it up, but the light would just not come on. Turns out the bulb was shot.
I sort of had this as an IT guy. Spent 2 hours trying to figure out what was wrong, had to tell my boss I didnt know, he spent an hour, turns out they were using the program wrong. The program they had been using daily for 5 years. Good times.
VoidZero25
1 month ago
Is it not common to test your code first before the test? At least my experience thus faris that, it’s always the test that is wrong.
Now we need test for the tests
Tests not being up to current specification is quite common. But if it takes a day to find this out likely either the script or the software isn’t really written in a maintainable manner.
And that’s why every testing proj should start with a test along the lines of Assert.True(True), to see if the test are even running
Maybe an unpopular opinion, but I would argue that while frustrating, this isn’t a “wasted day”. If a change in implementation affects the tests, the test should be updated to reflect the new requirements, that’s part of a software lifecycle.
Now, if a full day was spent before finding out that the test was wrong, I would argue that the person that made the change should have been aware of the test and its implementation, even when the test isn’t maintained by themselves but provided through their supervisor afterwards.
this graphic is so cute, whats the source? love the art
Once I did this physics assignment; in classical physics to first order a system can either have a natural frequency of oscillation or a natural decay rate, or both.
The question was to find the frequencies of oscillation of a particular system. But i found that it had one frequency of oscillation and one natural decay rate. So I spent like a whole day of undergrad trying to figure out where I went wrong. Eventually I gave up and just wrote what I found.
It turns out that the system did only have one natural frequency and one natural decay rate, but the professor didn’t check to see whether or not they were both frequencies because it’s theoretical physics and details like that don’t matter lmao
Now you need to become your supervisor’s supervisor
You just learned a valuable lesson. A few more of those and you will be a senior.
Consider it a lesson learned, most of us have fallen for this at some point. For exactly this reason, when receiving a bug report, I defaulted to checking for any obvious errors on my side first but seconldy to check the specifications and test methodology before actually jumping into debugging.
my classmates wondering why their binary adder machine isnt working (the given circuit diagram had two switched wires)
If only I got a penny every time this happened.
I was the one who wrote the test script.
In which industry are programmers and testers still divided?
I didn’t get a job because my code worked per the requirements but failed the test case. Even though I told them the test case was wrong.
+ 8 paid hours, sounds great!
To be honest, unless you’re an intern or something, it’s on you to verify the tests as well.
Tests are generally written to verify some specific behavior. Your code not meeting that exact behavior but doing something else which is also correct is a very common state to be in.
Been there
I’m sorry did you say “the test script provided tonus”?!?!
Bro if the test script is provided, I’m making the MVP (minimum viable program) that clears it.
Monkey patching and class mocking all the way through!
If you haven’t wasted an entire day yet, are you even a developer?
Who is testing the tester? Who is professoring the professor?
Just use https://github.com/auchenberg/volkswagen
Works out if your codes being run in tests and passes it
We have all been there.
My manager, himself, set up the tests to simulate a part for a product we were developing. The customer told him it wasn’t working, and manager argued that the scripts said it was, end of subject.
Customer knew me from other projects and called me. I ran the scripts, and yes, they said it worked. I ran the sims myself, looked at the results, and of course, it wasn’t working. The sims were set up incorrectly, and his results were all “don’t care”. Which he saved over the reference results. So, his comparison script was right, since “don’t care” matches anything…
Took me a week and a half to fix his shit.
I don’t understand why you would take the whole day on this? Just find out why the test script is failing your code and you’ll quickly get to the source of the problem which in this case is that the script is wrong. How exactly were you going about this problem? Just shooting shit in the dark?
Spent 8 hours debugging, only to realize it was a semicolon missing.
This reminds me of when we were wiring a two way switch in shop class. We had this little wooden section of a wall with a power cord, switches, and a lightbulb. Student after student tried wiring it up, but the light would just not come on. Turns out the bulb was shot.
https://youtu.be/v_0iuus8qjQ?si=6xLd1rnJybisZSHU
I sort of had this as an IT guy. Spent 2 hours trying to figure out what was wrong, had to tell my boss I didnt know, he spent an hour, turns out they were using the program wrong. The program they had been using daily for 5 years. Good times.
Is it not common to test your code first before the test? At least my experience thus faris that, it’s always the test that is wrong.