Comments on: We don’t need no steenkin’ code reviews!

By: omegamom

omegamom — Sun, 09 Mar 2008 18:07:21 +0000

Jozet–Har! Don’t worry, and come back out of your corner sometimes!

Spacemom–I believe you. I also believe that the code reviews on the spacecraft side are pretty stringent–when I was at Large Aerospace Company, every step of the software side for new aircraft was scrutinized with a microscope.

Johnny–Yeah, I could see that after a while code reviews would get…um…mind-numbing. 1–All the variables were variants in ASP, so he didn’t have to worry (I checked after the fact to be sure we didn’t overrun anything), and 2) They weren’t hard-coded; it was done on the database server side.

By: Johnny

Johnny — Fri, 07 Mar 2008 16:36:01 +0000

Actually, I was once the Nice Young Intern at a Mega-lo-Defense-Contractor. The secret with code reviews? After a while, they are all a rubber stamp anyway. When looking at locking code for 2 hours, the eyes glaze over and you just say, “Yeah, looks good.”

The thing I would have asked or noticed is:
1) why did you pick this number? Why not 99999 or 999999? Does it overrun your variables? If so, maybe you need to think the size of your variables.
2) please, please, please don’t tell me the numbers are hardcoded into the raw lines of code. Please say he used global variables!

By: Spacemom

Spacemom — Fri, 07 Mar 2008 15:48:07 +0000

Actually, if I could tell you what passes for code reviews in the actual data system here for a NASA project, (the on-ground side, not the spacecraft side) you would shudder too!

By: Jozet at Halushki

Jozet at Halushki — Fri, 07 Mar 2008 05:44:03 +0000

Okay, I understood about half of that after the bit where you shuffled your feet, lol.

But in spite of not knowing, I do have a suggestion: could you just switch to Roman Numerals after 9999?

Okay, I’ll go sit back in my corner.