Slashdot Mirror
U.S. Military To Create Its Own Internet
An anonymous reader writes "The New York Times today reports 'The Pentagon is building its own Internet, the military's world wide web for the wars of the future. ... The Pentagon calls the secure network the Global Information Grid, or GIG. Conceived six years ago, its first connections were laid six weeks ago. It may take two decades and hundreds of billions of dollars to build ...' Members of a consortium formed 9/28 include Boeing; Cisco Systems; Factiva (Dow Jones and Reuters); General Dynamics; Hewlett-Packard; Honeywell; I.B.M.; Lockheed Martin; Microsoft; Northrop Grumman; Oracle; Raytheon; and Sun Microsystems."
DISA (Defense Information Systems Agency) issued RFI's on this in 2002. In Decembver of 2003, DISA confirmed they'd contracted Juniper, Cisco, Sycamore and Ciena to provide equipment for this network. Total business is about $100 to each of the 4 through 2005. Now wouldn't it have been nice for some oversight on this 2 years ago?
o/~ Join us now and share the software
Link to Skynet5.
I spent ages trying to think of sig, but never did
No, actually it is "hooah" in the Army. It was originally short for "Heard, Understood and Acknowledged."
If life is a waste of time and time is a waste of life, let's all get wasted and have the time of our lives.
On a side note, the DISA site I work at has seen no long term planning schedules for this GIG network. My guess is it will lose steam long before it becomes its own 'Internet'.
actually i thought arpanet.
Well, ya, but ARPA is now DARPA. Shouldn't it be the DARPANET?
Has the AdaOS actually ever gotten off the ground? If it has, and the OS is working properly, then many of the following comments may not apply, but the last time I saw it, it was still early in the design phase.
Ada has lots of nice features, but if you don't need them (and you usually don't) then there are better choices. I include both Eiffel and D (Digital Mars D). There are probably others.
The problem with Eiffel is that the community is unfriendly to library developers, but the DoD could handle THAT. It's small enough to just be taken over. (And if you base the code with a fork off of SmartEiffel, it's FOSS, so you can adapt it as you choose.)
The problem with D is that it's too new. It's still short of version 1.0, and it's therefore missing many needed libraries (but work is proceeding rapidly).
Both include garbage collectors, which even Ada2005 doesn't appear to include. (And you can turn off both garbage collectors when you need to.)
D is more similar to C/C++, being in a sense (the designer's sense [Walter Bright]) the properly designed successor to C. Unlike C++ it doesn't attempt to maintain compatibility with C, and this allows it to avoid many of C++'s shortcomings.
Personally I would consider D a superior choice, but either would be superior to Ada. The code to accomplish identical functions is much shorter, e.g.
OTOH, Ada is clearly superior to either C or C++ from an error detection and prevention standpoint. (So are both D and Eiffel.)
If, however, you need the specialized features of Ada, it is incomparable. But almost nobody does. And if you don't need them, the overhead is about a factor of 2 in the LOC count. (D, OTOH, is nearly as compact as Python, partially because it borrows many of the control constructs from Python. (But NOT the space sensitive indentation! Thankfully.)
I think we've pushed this "anyone can grow up to be president" thing too far.
The only real news here is that someone has finally forced a single standard. You'd be suprised how many different communication systems DoD uses and all the work that has to go into making them all talk to each other. Now everyone's going to be TCP/IP (with a mandated change to V6 looming) and VOIP. Hopefully much easier to talk to everyone.
As far as the other things like giving the grunts more toys to break and throwing metric buttloads of money into commercial sattelite time, these things have all been in the works, now we have a nice easy to pronouce TLA to call all of it.
Luckilly the powers that be have cut out a large portion of the red tape. Some of these systems were slated to be on a 25 year timeline. Now they're all supposed to be done in five. This was the problem with most of the systems: you'd get a 1980's system fully fielded in 2000. That's only about 13 Moore's Law doublings or 8192 times crappeir than what you could just go out and purchase commercially. Even now they plan to make improvements to systems _as they field them_. So the first units look nothing like the last ones. Of course, they'll have to go back and upgrade the older ones, but it's a nice feedback loop.
Not sure what was more amazing, the story or the reactions I've read. Some of you seem to get the joke, but most of you are clueless. The DoD has had it's own isolated networks (yes, several, and they are actually isolated and independent from the W3) for a long time. The GIG is old news. DoD is refining updating it and will go on refining it and may even call it something different in the future. The new consortia (Net-Centric Operations Industry Consortium - NCOIC) is still trying to figure out it's own charter and mandate. It's all based on big money and it costs lots to join. There's a foundation (Net-Centric Operations Industry Foundation - NCOIF) that predates it and it has within it the Association For Enterprise Integration (AFEI - www.afei.org). This one is trying to be all inclusive (low cost of membership and all sizes of companies welcome. More the open model even if some of the same bigger players are involved in both. There's lots of this sort of stuff going on and it's been going on for a long time. I will conceed that many of the important DoD web sites that used to be visible are now protected and restricted access, but there's still lot of information in the public domain... if you're looking. Something you might be more concerned about is the waste of time and effort as different parts of the DoD try to protect their rice bowls. They are not all on the same page, and it's going to continue to cost more than it should for the functionality that gets deployed. I guess that's not a new story either... ;-)
Has the AdaOS actually ever gotten off the ground? If it has, and the OS is working properly, then many of the following comments may not apply, but the last time I saw it, it was still early in the design phase.
:-)
Dunno. You're probably right about it, but that's how Ada programs are supposed to work anyway: design, design, design, design, write the perfect code.
Ada has lots of nice features, but if you don't need them (and you usually don't) then there are better choices. I include both Eiffel and D (Digital Mars D). There are probably others.
Ada allows lower level manipulation than either Eiffel or D. They're really designed for different purposes, but Ada is a lot more flexible (at the cost of being more complicated, but not by much).
D is more similar to C/C++, being in a sense (the designer's sense [Walter Bright]) the properly designed successor to C. Unlike C++ it doesn't attempt to maintain compatibility with C, and this allows it to avoid many of C++'s shortcomings.
Any language with a garbage collector won't succeed C. Doesn't matter whether you turn it off or not, it's still too bloated for some of the places C needs to run. Ada compares to C in these places, it's just far more unpopular (and has syntax that I hate).
Personally I would consider D a superior choice, but either would be superior to Ada.
It depends on what your purpose is. If you just need to scratch out a cheap, quick program to accomplish some purpose, then ksh or perl might be the right tool for the job. If you're going to write a program that you intend to be in use for the next 40 years, and hundreds of people will examine the code over its lifetime, then code clarity is more important and Ada is your tool. D and Eiffel fall somewhere in the middle.
If, however, you need the specialized features of Ada, it is incomparable. But almost nobody does. And if you don't need them, the overhead is about a factor of 2 in the LOC count.
A side effect of Ada is often claimed to be ``code correctness.'' The idea is that it's cheaper in the long run to write the code correctly the first time and have to debug it less later on. Would Microsoft not benefit from taking a little longer to write Windows and then needing half as many coders to maintain it for the next 10 years?
But NOT the space sensitive indentation! Thankfully.
Yes, this is a horrible feature. I personally believe that scripting languages would benefit from a more verbose (and thus more exact) syntax, but spacing isn't any part of that.
I can guarantee you, though, that if the 1944-era US military had to take Falluja, the city would be rubble, and all of the civillians would be dead or refugees.
As opposed to this?
Or this?
Or this?
Or this?
Or this?
Or this?
Or this?
Or this?
The problem with ANY government project is that it's done by government contractors -- who don't really have a stake in the project's ultimate success or efficiency. Has anyone here ever eaten at a civilian-run military "dining facility"? They are horrible. The civilians have no stake in providing a good product, people get "fed", they get paid. Go to a military-run DFAC and you'll have decent food because they have an interest in their fellow servicemembers' food not sucking.
Government contracting needs serious reform, but no one wants to do it because the government and big business are having one giant sex party.
$5 billion for setting up encryption? The military's done some good things with adapting to an environment where off-the-shelf merchandise often can fill certain military needs. What's wrong with IPsec using AES? And is a worldwide private network REALLY neccessary or is it just another way for the Republicans to pay off their wealthy contractor friends? The project itself is ludicrous and unneccessary. It could be done a lot cheaper on private, secure VPNs on the Internet.
We have enough issues as it is. During Operation Iraqi Freedom, troops resorted to using their embedded journalist's satellite phones because their communication systems failed. How about we work on something like better strategic planning for theater-wide communication systems THAT WORK.