September 5, 2002
Porting to Oracle
I'm doing this contract work, mostly porting exsisting perl code from MySQL to Oracle, and the situation reminds me of when I was writting C apps on Linux and IRIX. Here's why:
Under Linux, I used gcc (natch), but ran into problems when moving my code to IRIX, where I used the MIPS Pro compilers. You see, gcc will allow you to get away with a lot. It will happily compile things and never warn you of anything that might be suspect with regards to the quality of the code, unless you tell it to. So I developed some sloppy coding habits, since gcc never talked back. Then I had to move the code to IRIX and was shocked by the number of errors and warnings the MIPS Pro compiler reported. So I had to change the code and be very strict about it, no sloppiness! This was a good thing, and I have been reather 'anal' about my code ever since, turning on all warnings and changing code until it compiled cleanly. This is good because it means that my code is generally very clean and managable, not to mention readable.
Porting from MySQL to Oracle reminds me of those days, because MySQL allows you to get away with a lot, while Oracle is as picky as can be. So much so that it can be a real pain in the ass. In the long run, however, these programs will be better off for it.
The more effort you put in to being correct the first time round saves headaches in the long run. Of course, if these people had done that I wouldn't be working right now... Maybe I should just shut my trap, eh?
Posted by ed at September 5, 2002 9:14 PM