On Fri, 22 May 2009 04:34:44 -0400, [email protected] said:
> 
> As I've been reading the forums, and exploring d-star and it's uses it
> seems as though there is some serious "infighting" between those using
> dplus, and those using other methods such as dextra, rtpdir, and other
> software implementations of d-star.  The debate has gotten so heated,
> that it is causing some bitter factions.  As a "user" of d-star I
> understand what it does, and why it does what it does, and I know how to
> make it work from a users perspective.  

Wow... you're reading too much emotion into it.  Ed and I have been
bickering back and forth, but as I've said before... "I use both." 

> From reading the forums it seems that this debate is over security of the
> currently working d-star network.  Repeater owners, gateway admins, and
> the trust server admins are concerned that the expermental open source
> software can crash their gateways when they have spent lots of time and
> effort in making sure that this system that we know and love is d-star 
> in it's current form,works affectively.  

Never seen anyone worried about D-Plus in this regard, not even the
skeptics.  D-Plus is fine.  I'd be far more worried about owners that
don't keep their Linux patches up to date, and the vast number of
unnecessary software packages the Icom instructions have you load... but
I'm paranoid... I run public Linux servers (web, mail, DNS, whatever)
for years, and the "rules of the road" there are... "Only install
packages necessary for the machine to operate."  And... "Compilers don't
belong on publicly accessible systems."

Neither the extra packages nor the presence of the compiler is actually
all that "insecure", it's just not "best practices".  Again, more
emotion being read in than I think there really is about the topic... 

> The open source software developers are concerned that this approach flys
> in the face of the experimentation aspect of ham radio and d-star's
> development.  What if an "x" module were reated on the repeater stacks? 
> This could be virtual, or could have the ability to be placed on a
> repeater if desired.  However, this "x" module could be done in such a
> way that if it crashed would not affect the  gateway, but still give
> communication access to the current d-star network.  

In my world... I don't "experiment" with infrastructure.  If I were
adding functionality to D-STAR, I would have a SEPARATE Gateway and/or
Gateway network to test on before things went anywhere near the
"production" network.  Robin and the Trust Server folks DO have such
setups. 

Much of the dislike of some of the newcomer zealout developers with NO
experience operating/maintaining production-quality networks is that
they're NOT testing elsewhere... they test right on other people's
systems, without communication, or a back-out plan.  That's just
inappropriate.  TEST ON YOUR OWN GEAR.  One particular developer doesn't
even have a repeater or a gateway.  That's just ridiculous of someone
who refuses or is unable to pay to play with their own infrastructure to
point their untested basement-coder quality code at a system another
admin is trying to get and keep running properly without even talking to
them. 

> Also what atempt has been made to develope a standard framework for
> d-star software develop, and the use of d-star:  when to use
> linking/unlinking, or callsign routing etc?  

Gateways aren't part of the D-STAR standard.  Icom's implementation is
proprietary and not documented.  Those with the resources, talent, and
time to build some additional Gateway machines (or a bunch of virtual
machines in their home lab) can certainly reverse-engineer it.  But in
Icom's view of the world, the Gateway system is NOT open for
experimentation.  It's a commercial software product you buy, plug your
"standard" D-STAR repeaters into it, and it handles linking.  The fact
that we also have some pretty neat add-ons today, is a testament to
Robin and other's work ethic and their drive to come up with new things.
 But they do it "right" and test their stuff OFF NETWORK unlike some
other folks, and then carefully release changes.

> Developing a "d-star compliant standard would give everybody some
> guidelines in which to develop alternatives to dplus if desired, while
> still being interoperable.  f I'm not on the right page, pleas feel free
> to correct me, but I don't see how calling people frauds dictators, due
> to dplus and the dvdongle and attacking people's natonalism is going to
> get us anywhere.  We're all pioneering a method of communication that is
> too new for this stuff.  What can I the d-star proponent/ user do to help
> bring  both sides to the table?  If we don't resolve this d-star will
> never take off.    

I think you're wrong here, but not a in a bad way. D-STAR *has* taken
off.  It works.  Icom did a "pretty good job" building the world's ONLY
international proprietary digital-radio to IP linking system.  Everyone
screaming and waving hands saying "I want more access" only needs to
grow up, buy a couple of Gateway servers, set up a home lab, and do
their experimentation as they see fit... OFF NETWORK... then work with
the rest of the world to integrate their new code/features/whatever in a
way that recognizes that they're ASKING PERMISSION of an already up and
working system, if they can MAYBE break it... for "X" gain.  Show the
world a new way to do things that's WORTH breaking something... the ham
world will do it.

That's my thoughts anyway... experimentation isn't stifled or dead at
all in D-STAR.  All you gotta do is learn how to treat it like
infrastructure and do a bit of a "sales pitch" to get buy-in from people
who ALREADY HAVE A WORKING SOLUTION to their IP linking needs.

Does that help make it clearer as to why there's really not a problem
until someone "experiments" all over all the other Gateway operators
toes?  One of the developers out there is screaming and hollering that
people should "contact the FCC!" about Robin not allowing access to
ROBIN'S ADD-ON CODE.  It really shows how stupid this person is...
Robin's code is Robin's code.  Want to have a port of your own, write a
Gateway module and get some Gateway operators to install it.  Tell them
why they should, etc.  IT'S NOT THAT HARD.  Why is this developer
frustrated?  He's BROKE and doesn't have a Gateway to even work on,
unless he convinces someone to give him access to THEIRS.  Now he says,
"Contact the FCC"?  Since when is the FCC in charge of who gets access
to PROPRIETARY PRIVATE OWNED INTERNET SERVERS?  

He's lost his mind.  He has as much "right" to access my computer here,
as I have to go into his house and steal his computer.  Seriously...
it's to the point where most of us just ignore him.  He's off the deep
end.

There are examples of OTHER developers BESIDES Robin who've "done it
right"... the folks that coded the D-STAR HotSpot for example... never
"blocked" from accessing anything, and haven't caused ANY trouble.

The whole thing is a non-issue.  Want to develop code for D-STAR... go
for it.  Learn to play nicely with others... is the only requirement.

Nate WY0X
--
  Nate Duehr
  [email protected]

Reply via email to