Joel – You should write the core of your application
Jeff – Joel is right – I’m making a web application & therefore i can rewrite stuff

wow !! How much more wasteful can this debate get?

REWRITE – What really comes to mind ?

“How much effort are we talking about here”. But the sneaky thought that might sway my decision is often the challenge involved (provided of-course all other factors, money and time included are willing.)

Human Factor – Play / Brag
I would never try to rewrite notepad coz the main effort is sheer human time but there is not much programming challenge compared to say writing a compiler or a core Database module or something similarly difficult (for me).

I would though, jump at the first chance to write a compiler, if offered even half a chance to do so. This, by the way, is the number one reason why enthusiastic programmers make bad enteprenaurs.

Tim berners Lee has the www covered.

But going back to my notepad / XML parser / web engine discussion these would never get the rewrite treatment from me simply because they are all well solved and solved many times over that too i might add.

I would not see any reason to write new code unless i feel i can do a better job out of any of these tools vis via

  1. ForMaintainability (think Mozilla vs Netscape though that was a bad move for Netscape)
  2. Features (think Notepad  vs Notepad++)
  3. or even new shiny code (Paint .NET) that grows into something big (bad though it is as a reason compared to others above)

Scientific Standards !!

Had the original debate been about, rewriting HTML input validators and was followed up with a list of valid libraries which are evaluated and proved unfit for use / exemplary for general consumption, the discussion would have been of a much better value IMHO.


But of-course all that real work is for the dumb review sites – we talk only about insights here in the blogosphere.