Re: Ada The Best Language?

["Marin David Condic" <marin.condic.auntie.spam@pacemicro.com>]

There are no accepted standard measures of productivity in the software
business. If we were talking concrete sidewalks, we could stipulate
dimensions and industry accepted construction materials and measures of
quality, then start laying sidewalk and measuring time. (Or dollars. Same
thing) The problem is that software almost *never* attacks an identical
problem twice and even among similar projects, you almost *never* do it with
an identical set of tools. Hence you can't control all the variables going
in, so you can't attribute any perceived changes as being due to the
language (or some other single variable of study.)

Its a really intractable problem. If we at least had some quantifiable
measure of units of work and some quantifiable measure of quality, we might
stand a chance of counting hours or dollars in. But I defy anybody to
stipulate some units of work and some units of quality that isn't going to
start a firestorm here with arguments all over the place about why the
metrics picked miss some critical factors that change the whole game. I
know - I've been there and done that. :-)

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/


"Leif Roar Moldskred" <rmoldskr@online.no> wrote in message
news:5VB57.119$%1.8915@news3.oke.nextra.no...
> codesavvy <codesavvy@aol.com> wrote:
>
> > Questions regarding productiviy are bogus?
>
> Umm, yes they are actually - unless you clearly define what metric
> "productivity" is to be measured by. Lines of code per hour? Average
> time to release? Total life-cycle cost? Man-hours needed to reach a
> certain level of quality? What level? Cost per line of code? What kind
> of projects should be measured? What weighting should be given to the
> various kinds of projects?
>
> If project A finishes in 2/3 of the time it takes project B to finish
> (with identical scores of the results), but the code of A is so shoddy
> that any major changes to it means a complete rewrite, whereas with
> code B you can just add a couple of modules and write some interface
> code - which project was the "most productive"?
>
> Questions about productivity that are as general as yours were, _are_
> bogus. (And don't get me started about "synergy").
>
>
> Leif Roar Moldskred
> professional lurker