[time-nuts] Releasing sources (was Re: Brooks Shera)

Jim Lux jimlux at earthlink.net
Sun Mar 31 10:51:11 EDT 2013


On 3/30/13 2:58 PM, Lizeth Norman wrote:
> What a bunch of hooey. Another so called expert wasted hours of my time
> because he can't be bothered to either note that code is buggy or just
> can't be bothered..
> If you don't want to release it, then don't. If you do and it's a POS,
> Expect emails.
>


Let's talk about that "wasted hours"..

You had a need.  You had two alternatives:
1) write the needed code yourself
2) use something someone else has written.

Presumably, you figured that #1 has large cost (if it were trivial, you 
wouldn't even start considering #2)..

The value of satisfying the need is Value(#1)


So you make a *speculative investment* in trying #2.  It pays off and 
you are ahead of the overall game by Cost(#1)-Cost(#2).  You've just got 
a substantial return on your small investment (you spent Cost(#2) and 
you got Value(#1) in return)  If it doesn't pay off and you've invested 
Cost(#2) without any return.

This is not "wasted".. this is "a speculative investment that didn't pay 
off".

A smart investor might look at the quality of documentation, or at the 
source code, or look for support groups.  Such things sometimes exist 
and make the probability of usefulness go up (In modern terms,the 
"Software Reuse Readiness Level" is higher).

Sure, sometimes you invest blind, and find that the program doesn't work 
well, etc., but that's not wasted.  You've basically paid for information.


Finally, what is a POS for you may not be a POS for other people.  A lot 
of freely released software was written to satisfy a tiny niche need, 
with NO intention that it be used for anything else.  If you want to use 
it as a starting point, fine, but don't come whining when it doesn't 
happen to do what YOU need.

  This is especially true of software written to provide an interface to 
a piece of test gear or equipment, for which the writer has exactly one 
instance.  All they care about is that they can get their counter, 
timer, antenna tuner, or whatever to work.  They have neither the time, 
money, nor inclination, to make the software work for ANY model of that 
piece of test equipment, nor to accommodate all the manufacturing 
variations.

Or maybe someone wrote software to extract data from a published source 
for some need, and then the published source changes its format.  The 
extraction software is now broken, but it met the original need, it 
might provide a framework for a future user to modify.

I don't have a problem with this.


More information about the time-nuts mailing list