Linus Torvalds (torvalds@osdl.org) on 3/3/08 wrote:
---------------------------
(snip)
>
>Sometimes low-level languages are simply better in the
>long run, and any "prototyping" in a higher-level one
>should literally be considered a throw-away stage.
>
>But you seldom know up-front whether it's worth it (or even
>what the problems will be), and from a practical perspective
>you may well end up being stuck with your initial choice
>(even if you intended for it to be a throw-away, you
>might not be able to afford to rewrite it all).
>

This actually happened to me. I wrote a 'proof of concept' application for internal use (essentially source/change management to replace a really crappy in-house written one). It was written in something called CLIST and ISPF dialogs. Fast to develop, but not very efficient in terms of resource usage (and a simple language with limited functions/features). After the demo, the manager told us to go forward... but we weren't allowed to rewrite it in assembler (as we had anticipated). So, now we have a mixed-mode application - some in assembler, some in CLIST and some in REXX. It performs well enough that nobody really gives a damn (a few pieces have been rewritten when they were identified as being a bottleneck).

Regards,
Dean

>
>Linus