 asardar says: Joel hits it again about why one needs to be careful if presented with the idea of restarting from scratch. Most times it will be a losing proposition. In general, i agree with what he's saying, but i'm not sure i can fully support this statement: "when you start from scratch there is absolutely no reason to believe that you are going to do a better job than you did the first time." Whether writing code or doing anything else, you certainly can learn things from the first try and then apply what you have learned when you try again. For the individual developer I couldn't agree more with you. The value of the learning experience would be great. From the company point of view, there is a high cost of starting over. And there is risk that you won't build anything better, just something different. I could be wrong though, there are example where people flipped a .NET or PHP website with Ruby on rails, in less time and the results were much better. Good article. Old code, like old flattop guitars, can have more resonance and depth and reliability. Great clip, but the problem I have with his article (written almost 6 years ago) is that his prime example is Netscape 6. Netscape 6 proved to be infintely better than Netscape 4.x -- the demise of Netscape was not that it scrapped 4.x and went with a new engine, it was that it didn't do it soon enough. During the 4.x - 6.0 lag, IE caught up and stole the show/marketshare. Oh well hindsight is always perfect. But at the end of the day, I believe, it's who has the better product. And sometimes, a rewrite is necessary. I had been a Netscape user until 6.0 came out. I found it strange and discontinuous from the earlier versions. I gave it up fairly precipitously, so I can't describe it with much authority, but I remember having some kind of trouble with bookmarks, and wasn't there something forced into the toolbar section that defied removal? I can't completely agree with what he is saying in that article, but I'm sure there is some truth to it. I think if you are re-writing to just get rid of bad code, then that's probably is a bad idea no matter what. If you are re-writing because what you have isn't going to cut it any more or you are going in a different direction, then I say you start from scratch and steal what you can from the old code. But to try to mold something old into something new is generally not a good idea. As for Netscape, that is a really bad example. Sure they still have major problems with the 6+ version, but the important thing was that they scrapped the old rendering engine. I'm sure that was 90% of t... 10 GOSUB 300 20 'As an amateur coder, with years' worth of reliable backups, I'm sometimes beffuddled by my own code, even though I learned almost from the start to code clearly and economically, using meaningful (to me) variable and subroutine names, sectioning off code, and trying to modularize functions for central use and easier decyphering. (I gave up "spaghetti code" around 1986, when I found I couldn't understand what I'd written maybe a week ago). 30 'But sometimes I'm simply amazed at the beauty and balance and ingenuity (heh) of some old code, after I've thoroughly figured it out! 40 'No doubt I'd lifted code from one of those fat $35 books I bought when I was learning, or as... Rewrite start over if ok way do it i know microsoft did it with Vista & not as good as XP time will tell if it will work ok not look so good now. Old code been well tested new like a new car it may good or bad you test & hope it OK Vista all new & its got some bad code like drives & WGA playing up say not genuine |
Embed This Clip In Your Site...
|
||
|
|
|||
