Tuesday, March 06, 2007
Working Hard... Not Smart?
Ah, the joys of a release. My companies Windows product that I'm working on was scheduled to go out Monday. Everything is testing fine until Friday morning. When we put everything together, the installer explodes, and the Bank Reconciliation programs decide that they prefer a hard crash for every little strcpy. And for the last weekend and Monday and today, we've been working overtime just to figure out what went wrong and how to fix it. Ah, the joys of salaried pay.
Subscribe to:
Post Comments (Atom)
3 comments:
Crash on strcpy??? That's past bizarre! I hope you've solved that. It seems fairly easy to track down though. The "strings" handed to the function aren't truly strings (maybe char arrays?), or the place that strcpy is being used doesn't know about strcpy. It's been a long time since I've worked with strcpy, or any C++ code.
The installer blowing up sounds like the tougher fix.
So what are the solutions you've found? Since my own coding is largely Bash scripting these days, I have to live vicariously through your programming! :^D
The tester was using different data, that reached into previously untested code. That any of the code remained untested was a surprise. After our tester gets through with it, we rarely have any problems show up. The original program was made using a different standard and I need to change them: setmem->memset, movmem->memmove, sprintf->sprintf.
The problem this time was a parameter in sprintf that was treated slightly differently. %*r should be %*c.
You live vicariously through my programming? This is a sad, sad day. :-) Say hi to the Mrs. for me!
Sigh, what wonderful memories it brings back dealing with pointers.... NOT!
Post a Comment