Expensive contractors, take responsibility for your work!

[minestrone]

so you admit you did not do your job and you ask who is at fault.

this is a tought one

Gordon Brown of course!

Of course!

I changed the code and had the BA test it on my local machine which he was happy with, I assume that his test was then written into the overall test document. I left the team and I assume that I never checked in a file as I knew the fix straight away.

[Drewster]

If someone like you thinks I am "You pointless *****" I'll take that as a compliment.

No No No. Seriously you would have to be seriously lacking in judgement if you considered that a compliment.

It really makes no sense at all.....
You may think he is a total winker.... but the "worst" that you can think is that his opinion is not worth anything... there is no circumstance that his poor opinion (whatever its merits) of you is a compliment.

You really ought to think before you type....

[original PM]

So you seriously think that by abdicating responsibility for something to one of your team, in the eyes of your stakeholders you will no longer be held responsible for that piece of work??

erm no that is not what I said.

I am saying it is not my direct responsiblilty to complete a task.

I am directly responsible for the success of the project (yes I am a PM)

However some people I have come across who do think they are PM's obtain lots of signatures on pointless bits of paper who do think they are abdicating responsiblity by obtaining a signature.

on a different note the original post was talking about tulip coders who deliver untested code - to clarify I was talking mainly about upgrades and code changes where they are only small changes and when put into test throw up an error.

Obviously if we are talking a large software development then of course we would expect bugs and it would be put through a team of dedicated testers

I guess to drive home the point that team of dedictaed testers would be RESPONSIBLE for ensuring the code had been properly tested.

Have a great weekend!

[threaded]

So, it falls into two camps.

1) Threaded etc. who do what the client asks for, even if we know that not only it isn't what the user wants, but it probably won't work anyway.

2) Other who try and keep the client sweet, all the while doing what they deem the end user requires.

I really used to try and do #2, but got where I am today doing #1.

[oracleslave]

S

I really used to try and do #2, but got where I am today doing #1.

Sounds like a recipe for becoming an overweight chippy manager. Err hang on

[Ardesco]

on a different note the original post was talking about tulip coders who deliver untested code - to clarify I was talking mainly about upgrades and code changes where they are only small changes and when put into test throw up an error.

You do realise that something as simple as a missing semi-colon or a single misspelt variable name could do anything from break the entire product, to change the way the product works under the hood without anything obviously visibly wrong for a long period of time.....

All code changes to a system require a version update (even if it is minor) which then means a full regression should be performed before it is released to live. Anything less is a cockup waiting to happen.....

[suityou01]

You do realise that something as simple as a missing semi-colon or a single misspelt variable name could do anything from break the entire product, to change the way the product works under the hood without anything obviously visibly wrong for a long period of time.....

All code changes to a system require a version update (even if it is minor) which then means a full regression should be performed before it is released to live. Anything less is a cockup waiting to happen.....

Worse still the OP is coding in C++, so there could be stray pointers and time bombs lurking. It's a trading app IIRC.

[shoes]

I can't be bothered to read all this thread, but here is my industry experience based observation : Testers prefer developers to give them code that works. This is because if the code already does what it is supposed to they have no pressure to do their jobs correctly or at all and can sit there talking about big brother.

Coders on the other hand prefer the testers to actually do their job so that they don't have to do their own job and that of the testers as well. They also prefer not to overhear conversations about big brother whilst performing the only demanding part of the life cycle.

[Menelaus]

I can't be bothered to read all this thread, but here is my industry experience based observation : Testers prefer developers to give them code that works. This is because if the code already does what it is supposed to they have no pressure to do their jobs correctly or at all and can sit there talking about big brother.

Coders on the other hand prefer the testers to actually do their job so that they don't have to do their own job and that of the testers as well. They also prefer not to overhear conversations about big brother whilst performing the only demanding part of the life cycle.

Or those of us who are model builders in SAS, R, S-Plus, MATLAB, Synapse (etc) - we're coders and testers

[The Lone Gunman]

So, it falls into two camps.

1) Threaded etc. who do what the client asks for, even if we know that not only it isn't what the user wants, but it probably won't work anyway.

2) Other who try and keep the client sweet, all the while doing what they deem the end user requires.

I really used to try and do #2, but got where I am today doing #1.

I am customer driven, I try to provide the best service I can, but there are limits to how far I can or will go.
I write and test to spec and advise if I think it wont work, but buggered if I will perform tests which I have been explicitly told not to (as the OP states) as it is not in the rate.
He seems to think that a taxi driver should get him to Heathrow when he has only paid to Victoria, even if the cab is an expensive limo.