On Sat, 2010-02-27 at 00:54 -0700, Nate Duehr wrote:
> 
> 
> On 2/26/2010 9:47 PM, Woodrick, Ed wrote: 
> >   
> > You mean in an environment where every ham could do it better than
> > every other ham?
> > 
> > 
> 
> Fully in agreement here.  Every new programmer (with less experience
> than his peers) thinks he can write a better version of [insert
> software here] and ends up writing a new one with completely
> different, but just as silly bugs, as the last guy.  Code-reuse is a
> farce that never happens, because programmers also write new languages
> to program in constantly, thus creating their own Tower of Babel.
> 

Well,  I think that an open system would be better in this case than the
closed one we live in with D-Star.

The Linux kernel works great with contributors offering up patches and
those patches being approved by those with commit privileges.  What is
better?  Having a handful of eyes on the gateway software and the d-plus
software or having potentially 100s of eyes on those projects?  

I'm thinking that some of the concern with keeping these things closed
is the same concern that keeps IRLP closed.  Possibly a bug in one fork
could kill the whole network.  That is true but policy could prevent
damage from that.

Also note that the term 'open' as it is used with UNIX and other systems
as "Open Systems" does not mean you get the source.  What open means is
that their APIs and protocols are open (specified) to allow
interoperability.  Unlike Microsoft's "embrace and extend" concept.
Embrace the protocol and extend the spec in a closed manor.

There is a lot I don't like about D-Star.  The fact that the 8 character
is used for commands in my opinion is a cancer that we are going to be
living with for years to come.  It is a hack to allow us to do things
not intended on the closed system of the gateway.


> > Remember, Linux is a copy of Unix and then attempts to copy
> > everything that Windows does.
> > 
> > 
> > 
> 

Linux was written to provide features not present in MINIX (Educational
UNIX based OS).  Linus relied heavily on the SunOS manuals as he was
creating a kernel.  No UN*X kernel is a copy of another UN*X kernel.
UNIX is a specification of openness where Windows is totally closed.
Linux is still not fully POSIX compliant.
 
> And Windows GUI was a copy of the original MacOS GUI... which was a
> copy of a user interface prototype at Xerox PARC... for copy machines.
> 
True.

> And ... the MacOS GUI was also copied by Amiga, Tandy, and others, but
> not by TI, Atari, and others who died by not copying it and keeping
> up.  Those that copied lasted a little longer, but lost to the
> marketing budgets of those that survived.
> 

Correct.  A good marketing campaign promotes inferior products.  We see
this many times over and Windows is a prime example.

> And... to top it all off, just about every TCP/IP stack in existence
> is a copy of the original code from Berkeley... in fact known-bugs in
> BSD's TCP/IP stack were still used as "fingerprints" to show that
> Microsoft used the BSD stack, years after those bugs were fixed in
> BSD's own code and copy of Unix.  Microsoft eventually patched their
> IP stack...

Correct.  BSD license does not require recognition or even disclosure.
MS did copy the BSD stack for its own TCP/IP implementation.  Fully
legit too.  It is copying though :)

> 
> So... I think you're trying to state that Microsoft isn't a copy-cat,
> but an innovator -- and that's simply not true. In reality everyone
> copies everyone in this industry... and it matters little over the
> long-haul.  True innovation is rare.

Correct.  Reinventing the wheel costs time and money.  It is best to
copy ideas and improve upon them.

> 
> So I'm not sure what your point is, other than Linux as a desktop OS
> is not as polished as Microsoft Windows, which is a bit of a
> non-sequitur, since most Linux folk don't care... they can go back and
> forth between command-line and GUI for whatever they need to
> accomplish, with ease.

I'm in X now but routinely run up to 10 or more xterms per day accessing
the various machines I work on.  I prefer the CLI.  

Chris

Reply via email to