What the hell?
I just had an argument with another developer, even though we were both
agreeing.
It has to do with error messages. Currently there is no way of
displaying error messages (because we don't want the user to see error
messages-- it might give the wrong idea). We do have the ability to
internally log error messages, but this particular one is not propogated
from the GPS decoder. It has to do with the validity of the GPS signal
not getting propogated into the position validity test. I agree that it
should be, and I have argued that for over a month but I have been shot
down because it would mean changing things too much.
So I agree that there should be an error message. The problem, is that I
don't have the information to generate an error message. And it took
about 15 minutes for me to get that idea across.
It's a very stressful time right now.
I just had an argument with another developer, even though we were both
agreeing.
It has to do with error messages. Currently there is no way of
displaying error messages (because we don't want the user to see error
messages-- it might give the wrong idea). We do have the ability to
internally log error messages, but this particular one is not propogated
from the GPS decoder. It has to do with the validity of the GPS signal
not getting propogated into the position validity test. I agree that it
should be, and I have argued that for over a month but I have been shot
down because it would mean changing things too much.
So I agree that there should be an error message. The problem, is that I
don't have the information to generate an error message. And it took
about 15 minutes for me to get that idea across.
It's a very stressful time right now.
Profile
magicmarmot- Tom Ramcigam
- Magic Marmot Studios
Comments
*sends a few REALLY deep breaths*
This will be OVER soon..... keep repeating that.
We had a "bug" written up that an untrapped error was being passed from the back end systems through our software, when it was formerly trapped.
Since our testers are the least technically knowledgeable testers I've ever worked with, they write bug reports that say, "When I do X, I get this error." (Instead of something useful like, "I expect Z, but I get Y when I do X")
Well, the real deal is that the back end system changed how it responds to our call, so the conditions we used to trap the error they saw no longer exist. Any conditions we don't know how to respond to, we pass the error through to the user. Not pretty, but at least there's an error.
So, technically the system's doing the right thing in passing through the error. The tester doesn't know enough to understand this, so they write up unhelpful bug reports that cause us all to waste time divining what they intended to request.
IlovemyjobIlovemyjobIlovemyjob.
I'm doing some integration testing. Currently with EXACTLY the same code, each time I run, something different happens. Some part or other fails, or doesn't load, or something.