Evolution vs. Revolution

Chris Jackson s Semantic Consonance Evolution vs Revolution | pool toys

Chris Jackson s Semantic Consonance Evolution vs Revolution | pool toys — Thu, 18 Jun 2009 11:58:50 GMT

PingBack from http://pooltoysite.info/story.php?id=6768

Evolving make « LSD::RELOAD

Evolving make « LSD::RELOAD — Sun, 11 Nov 2007 18:52:54 GMT

PingBack from http://lsimons.wordpress.com/2006/04/14/evolving-make/

re: Evolution vs. Revolution

Travis Owens — Tue, 26 Jul 2005 21:40:06 GMT

By the way, that article was a very enlightnenig read, while I haven't done it yet, I will admit to being in the "rewrite it from scratch" camp but thes article really opened my eyes.

I won't be so quick to rewrite code next time, I'll give porting code a more serious attempt.

re: Evolution vs. Revolution

Travis Owens — Tue, 26 Jul 2005 21:32:42 GMT

Well on the issue of throwing away old code and start anew or fixing old code....

It actually differs, overall I'd say fixing old code is the best thing, even if it's possible more work than starting anew. At least with a OO language you should be able to rewrite the code and keep the app working as you refactor. If it's an older scripting language, it may not be that easy.

Despite, my point is, if you were the original author and you are going to throw away the code, you should slap yourself on the wrist for writing something so bad in the first place. Of course chances are you are repairing code you wrote as you were learning (but then aren't we always learning).

If it's somebody else's code, well the code could be so bad that you might not have a choice. A coworker had to repair a project that was VB.Net code gen from some custom tool. We didn't have the tool or the original specs that generate the code, we just had the generated VB.Net code. In the long run the app was better off being rewritten as the project went horrible and only 1/5th of the features they needed were added as the code gen architecture is in such a serious need of refactoring, it hurts.