3 ms·
Sure, elegance is attainable in any language and I definitely agree that large rewrites (especially involving a language change) are at best perilous, however i
by knieveltech 16y ago
Sure, elegance is attainable in any language and I definitely agree that large rewrites (especially involving a language change) are at best perilous, however it seemed to me that the OP's point was largely about common attitudes.
Having seen a few large projects detonate under similar circumstances as described in the movie I believe he has a point. In all instances the timeline went something like this:
- several year old codebase (now bloated with so many change requests the original design is completely buried)
- new project manager enters stage right with a gleam in his and and a plan to "modernize" the codebase
- Ruby selected as the language of choice for the rewrite
- 8 months to a year of code churn later project is abandoned
Now given language hopping large scale rewrites are typically a bad idea, especially if you're converting an existing shop with a different language of choice and established workflows, one has to ask what would lead any otherwise rational individual to decide engaging in a project of this scope is viable.
In all of the instances of this kind of project flameout I've observed a PM (or higher) was so enamored of some "sexy"* new technology that all other concerns where cast aside and predictably carnage ensued.
* For values of sexy that include: Ruby, NoSQL and just about any other arguably (im)mature tech that's getting a lot of column inches in the tech blog arena.
- InclinedPlane 16y agoThe problem with rewrites is that most people don't appreciate how difficult they are or know that there are special skills required to execute them. When done correctly rewrites can be wonderful things, but more often than not they are failures. Most software projects, rewrite or no, are mis-budgeted and mis-scheduled, rewrites are harder, more expensive, and take longer than the average green-field project, and so are proportionately less successful.