Interplanetary Internet (IPN)

Marc Petit-Huguenin writes: "Vinton G. Cerf and others just released an Internet Draft about the Architectural Definition of the Interplanetary Internet (IPN). The first section "Desiderata of Interplanetary Internetworking" is a wonderful text." This is beautiful, both the document itself and the work put into something which, at the present time, has no practical use whatsoever. Bravo... I hope I live to see this deployed.

Comments

*BSD is dying

[Anonymous Coward]

You don't need to be Kreskin to predict *BSD's future. The hand writing is on the wall: *BSD faces a bleak future. In fact there won't beany future at all for *BSD because *BSD is dying. Things are looking very bad for *BSD. As many of us are already aware, *BSD continues to lose market share. Red ink flows like a river of blood.

Let's keep to the facts and look at the numbers.

OpenBSD leader Theo states that there are 7000 users of OpenBSD. How many users of NetBSD are there? Let's see. The number of OpenBSD versus NetBSD posts on Usenet is roughly in ratio of 5 to 1. Therefore there are about 7000/5 = 1400 NetBSD users. BSD/OS posts on Usenet are about half of the volume of NetBSD posts. Therefore there are about 700 users of BSD/OS. A recent article put FreeBSD at about 80 percent of the *BSD market. Therefore there are (7000+1400+700)*4 = 36400 FreeBSD users. This is consistent with the number of FreeBSD Usenet posts.

Due to the troubles of Walnut Creek, abysmal sales and so on, FreeBSD went out of business and was taken over by BSDI who sell another troubled OS. Now BSDI is also dead, its corpse turned over to another charnel house.

All major surveys show that *BSD has steadily declined in market share. *BSD is very sick and its long term survival prospects are very dim. If *BSD is to survive at all it will be among OS hobbyists, dabblers, and dilettantes. *BSD continues to decay. Nothing short of a miracle could save it at this point in time. For all practical purposes, *BSD is dead.

Re:*BSD is dying

[jandrese]

You know, I think everybody who cares saw this the first 1000 times you posted it, you can turn off your autoposting bot now.

Down that path lies madness. On the other hand, the road to hell is paved with melting snowballs.

Re:Insane ping times

[Anonymous Coward]

Even worse if you go like "Excuse me, I have to finger uranus"
heh

Ooh the good ol' Uranus joke never gets old :)


--skurk

Re:uses for IPN

[jandrese]

Actually the ISS is a good canidate for this technology. Just because someone is in orbit doesn't mean communication is easy. First off, the distances involved still add quite a bit of lag to the communction, second the window in which you can transmit before the ISS moves out of range of your antenna is quite small, so you have to really pack as much data as you can into that window, which is what this technology excells at.

Down that path lies madness. On the other hand, the road to hell is paved with melting snowballs.

Actually this is used today

[jandrese]

This is the same technology (SCPS gateways for instance) that is used to talk to some LEO satellites and the Space Shuttle, where you only have window of a few minutes to transmit whatever data you have.

Some of the technologies used in this are also applicable to any low bandwidth high latency connection. IP header compression is a prime example of this. Most people on Slashdot probably havn't considered the consequences of using IP over a link with a bandwidth measured in the low double digit bytes per second where return traffic may take several minutes to reach you, but people working with secure communications or low power long distance wireless links sure have.

One final thing I'm sure will interest many Slashdotters, the SCPS gateway runs on FreeBSD (and many other platforms as well, but it was developed under FreeBSD).

Down that path lies madness. On the other hand, the road to hell is paved with melting snowballs.

Re:Actually this is used today

[Hazzl]

I was actually quite scared when I read the section about billing. Did anybody else see this as the future of the Internet on earth? Or is it just me being a bit paranoid?

"Sorry Mister, but your packet's credit has expired, we will be cancelling all your other packets on the network..."

Literature and Common Sense in the same doc!

[DG]

From the RFC:

Desiderata of Interplanetary Internetworking

Go thoughtfully in the knowledge that all interplanetary communication derives from the modulation of radiated energy, and sometimes a planet will be between the source and the destination. Therefore rely not on end-to-end connectivity at any time, for the universe does not work that way.

Neither rely on ample bandwidth, for power is scarce out there and the bit error rates are high. Know too that signal strength drops off by the square of the distance, and there is a lot of distance.

Consider the preciousness of interplanetary communication links, and restrict access to them with all your heart. Protect also the confidentiality of application data or risk losing your customers.

Remember always that launch mass costs money. Think not, then, that you may require all the universe to adopt at once the newest technologies. Be backward compatible.

Never confuse patience with inaction. By waiting for acknowledgement to one message before sending the next, you squander tracking pass time that will never come to you again in this life. Send as much as you can, as early as you can, and meanwhile confidently await responses for as long as they may take to find their way to you.

Therefore be at peace with physics, and expect not to manage the network in closed control loops -- neither in the limiting of congestion nor in the negotiation of connection parameters nor even in on-demand access to transmission bands. Each node must make its own operating choices in its own understanding, for all the others are too far away to ask. Truly the solar system is a large place and each one of us is on his or her own. Deal with it.

That's beautiful... it conveys all the important information, but yet still manages to be literate, and even a little bit inspiring.

Nice to see there are still visionaries in science.

Re:Literature and Common Sense in the same doc!

[Paradox !-)]

One person's trash is another person's treasure...

I'm just happy to see that folks are thinking about this kind of stuff. Better to plan ahead. Beasides, research on this kind of issue could have immediate benefits to terrestrial networks thanks to the new ideas. I personally like the idea of the "atomic" unit of transport that contains both content and control information in one package. That could be really useful for some terrestrial applications.

Re:Literature and Common Sense in the same doc!

[selectspec]

To bad the satellites dont run in COBOL because you could just compile this excerpt.

Re:Literature and Common Sense in the same doc!

[Lizard_King]

Nice to see there are still visionaries in science.

I don't get it. Are you implying that science currently lacks enough visionaries?

I'll ask before I argue, and you're right, that passage is beautifully written.

Re:Literature and Common Sense in the same doc!

[SlippyToad]

I almost started laughing when I read this. It looks like the "Tao of networking" or "Confucious on IP", or even worse, "Yoda the Network Master." But "literature" -- no way. "Lampoon" is more like it.

Barriers to exploration

[Xenophon Fenderson,]

I finally understand why planet-wide exploration and colonization didn't really happen until Columbus. It must have been as fantastically expensive to cross an ocean then as it is to go to the moon or Mars now. Modern society is facing similar problems to Columbus: Why bother with the expense and danger of exploration & colonization when there is plenty to explore, research, mine, trade, etc. locally? What is really funny is that Columbus was able to get money for his expeditions in the same way NASA has: National ego. Spain had the English and Portugese to one-up, and the American's had the Soviets and the Chinese. I guess it just goes to show that not much changes, even in 500 years.

Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16

Re:Barriers to exploration

[Keysh]

Actually... those are some seriously strange claims! Portuguese colonies in Greenland? Good "secret conspiracy-theory alternate history," not so good "real world" history.

Greenland was settled by Norse explorers in the Middle Ages (and of course earlier by various peoples from North America); the colony was cut off from Europe sometime in the late 1300s or early 1400s, and perished sometime in the 15th or 16th Century. The Portuguese were busy sailing down the coast of Africa and around into the Indian Ocean. They weren't going to Greenland!

(Now, had Columbus somehow ended up in Iceland, and somehow gotten someone to translate a few of the old pagan sagas, he might have deduced the existence of lands to the west. Lands with not much to offer except grapes and hostile natives, according to the sagas, so Columbus probably wouldn't have been too interested anyway.)

The reason Columbus insisted that China was close enough to be reached by contemporary sailing ships from Spain was because he simply got the distance wrong. As you mention, the radius of the earth had been accurately measured over a thousand years earlier, by the Alexandrian Greek Eratosthenes. He gave the earth's circumference in stadia, a common unit in Hellenistic times.

So how long is a "stadium"? Ah, there's the problem -- in late 15th Century Europe, they had recovered Eratosthenes' work, but not enough of the context, so they didn't really know. Columbus argued for a small value, which made the earth small and China not too far away. His academic opponents argued for a larger value, which would mean China was too far away. And, as we all know, Columbus was... well, wrong. His ships would have run out of food and water well before reaching China, but they luckily ran into the Americas on the way.

Re:Barriers to exploration

[Mr. Foogle]

Just some mild criticisms

Columbus wasn't looking for the New World, he was looking for gold, spices, converts to Christianity, in the Far East.

Planet-wide exploration DID happen before Columbus - the Porteguese were busy finding the route to the far east around Africa *years* before Columbus, the Chinese had a huge fleet of ships ready to explore the known world, but internal politics killed the project.

The Porteguese had the only known route to the Orient zealously gaurded - anything else looked like an expensive way to suicide.

The cost to the Spanish crown was, while not cheap, not especially expensive. Three ships, a handfull of men, and King John found a way to cut the cost to himself even further (I think) by forcing a town to pony up the three caravals as a fine.

Re:Barriers to exploration

[vidarh]

Actually, there are some that believe Columbus were looking for a "New World", since he apparently did visit a small Portuguese colony on Greenland during his research for the voyage, and some claim he there must have heard about the viking stories about a land in the west, and may even have had a map giving some indications of distance.

Which might explain why he insisted that well founded claims that China was to far away were wrong (remember: The radius of earth was had been calculated to a reasonable accuracy many centuries before Columbus). It was much easier to fund a trip to China, which after Marco Polo was famed in Europe for it's riches, than fund a trip to somewhere unknown, that most learned people would insist probably didn't even exist.

Re:Barriers to exploration

[vidarh]

Actually, first of all, the radius had been reasonably accurately calculated by several people in Europe as well, much closer to Columbus time, and in fact this was one of the objections that was raised at Salamanca: That he chose to ignore well founded recent calculations. Actually, the more common theory of why he used the wrong number is based on indications that he studied Ptolemy's works on geography, not Erastothenes.

Secondly, the Portuguese were active in parts of the Northern Atlantic in the 1400's, and increasing rapidly thereafter, particularly due to several expeditions searching for a northwest passage. Now, why they would search for that in the shores around Greenland, if they at the time did not have knowledge that America likely stretched far north, is an interesting question.

There are claims concerning Columbus visiting both Greenland and Iceland.

Some claims about Portuguese activity in the North Atlantic can be found here: a message referring to claims about Portuguese slave traders, an article (in Norwegian, unfortunately) referring to theories about Columbus reaching Labrador in 1477, with subsequent Portuguese activity in the Northern regions as a result, a claim that Vatican records tells of a slaving raid in 1418, and information about a possible Portuguese expedition to Greenland around 1479, an article about possible contact between Columbus and Vikings on Iceland, based on memoirs written down by his son.

Much of this is of dubious quality, though, and I'm certainly not judging their quality, but it is an interesting theory whether correct or not.

While brining up more or less weird theories, though, there's a few people that have presented a theory that Columbus was originally Scandinavian, member of an important family with roots in royalty throughout Europe.

Decide for yourself whether to laugh at a funny story, or believe there's something in it. But either way, history from that far back isn't always as straightforward as people tend to think - there are very few parts of history from that time period that is comprehensively documented in trustworthy sources.

Re:I could have used this decades ago

[Xenophon Fenderson,]

Ever heard of UUCP? As far as I know, it works over IP as well as dialup.

Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16

Re:Slashdottet?

[Rob Wilderspin]

Spooky, given the Slashdot quote at the bottom of the page...

"It is your destiny. - Darth Vader"

How do you feel, knowing that your future holds such destruction and terror?

Re:Come on, Give them a lottle credit...

[Luyseyal]

That reminds me of that project to transmit IP over SMTP. It actually worked. Here's an RFC draft to create a standardized MIME for IP: http://www.imc.org/draft-eastlake-ip-mime

-l

The Larger Mesh...

[dew]

IPNSIG (the InterPlanetary Internet Special Interest Group) submitted this document to the IETF. It's interesting to note that IPNSIG is looking at very long-term solutions, but (to me at least) it's equally fascinating to read about current space communications standards in development that already take into account many, indeed nearly all, of the "far reaching" recommendations made in the post.

Readers may be interested in the CCSDS (Consultative Committee for Space Data Systems) homepage which has many protocols, proposals, and drafts available for public review. Take for example their file transfer protocol (PDF - start reading on page 20) that already "bundles" data and looks to be somewhat comprehensively thought out.

Food for thought; these principles have not only been conceived before, but reduced to standards (and implementation).

David E. Weekly

Re:uses for IPN

[WhiteDragon]

Does this remind anyone else of A Deepness in the Sky or Fire Upon the Deep by Vernor Vinge

Back to the future

[bbrelin]

So, we're back to a store and forward network, eh? Can anyone say "uucp?" Perhaps I should dig out my O'Reilly books on Managing UUCP and USENET. "Paging Peter Honeyman, Dr. Peter Honeyman, please pick up the white courtesy internet phone in the lobby of the Space Hilton Hotel" Braun Brelin bbrelin@yahoo.com

Re:Back to the future

[castoridae]

Yes, I agree. I think the underlying concepts aren't applicable to just interplanetary communication, or to uucp; they describe a method of communication in a system where the time delay in receiving an acknowledgement is potentially greater than the time until another message needs to be sent. I believe old modem/BBS networks such as FidoNet worked this way.

Re:interconnection with mirror universe

[mandolin]

say, evil google or - shudder - evil slashdot?

If this is the good slashdot, I'd hate to see the evil one.

Slashdottet?

[Accumulator]

Wonderful seeing the progressbar going slower and slower and slower...

Re:Slashdottet?

[HiQ]

What's nicer is that maybe in the future we will be able to slashdot whole planets :-)

Re:Slashdottet?

[HongPong]

I bet it would look like an awesome dark green and white beam, demolishing all in its path...

--

Expires November 2001

[Accumulator]

Does that mean Interplanetary Internet will be old by then? Well, why not, I always hear people telling about how fast the internet grows.

Re:Lag vs. Speed of Light

[Delphis]

Of course I have to wonder, reading that comment, what sort of computer would be running AT THE SUN. .. can you say HEATSINK REQUIRED?!? :D

Even putting machines closer to the sun than earth is presently seems like a bad idea.

As has been mentioned before though, a supercomputing 'farm' on the dark side of the moon or mars would be kinda cool (literally).

--
Delphis

take it all seriously

[jonbrewer]

If you look at the author list, they're all from MITRE or JPL. (read Defense Department sponsored.)

This is likely as important and visionary as (Rand Corporation) Paul Baran's 1964 "On Distributed Communications" series, and I would take it as a look at the near future.

Comparison

[rde]

Now:foo.com protested to the whole world that its servers were blackholed. /. listed the story three times, and after two years and $1.2M, foo.com was once again off the list.
Then:foo.com.cygnus.alpha was blackholed. A memorial will be held during the Month of the Space Lizard.

Not so far off

[apsmith]

I've been spending quite a bit of spare time working with various grass-roots groups on lunar mission planning for the relatively near term, and one of the significant issues (out of a couple dozen) is the communications infrastructure. If you don't want to keep hogging the "Deep Space Network" of radio antennas, there is a real need for a simple and flexible standard that allows you to route information through one of a collection of geosynchronous satellites - a packet-switched protocol has some big advantages over traditional fixed circuits, and if you're going to go with packets, why not IP? So this has some real practical consequences for current planning, and I'm glad to see progress is being made.

IPN huh?

[Luxury P. Yacht]

Does this mean I will one day be able to finger Uranus? Doesn't sound too appealing...

Avian Carriers?

[NoseyNick]

Why does this remind me ever-so-slightly of "IP Over Avian Carriers"? Low bandwidth, high latency...

Re:Avian Carriers?

[interiot]

Except that "farmer with shotgun causing network outages" is replaced with "Ferengi threatening IPN backbone if they're not paid soon".
--

Re:Avian Carriers?

[plover]

We were discussing it over lunch. CP/IP (Carrier Pigeon/Internet Protocol) would probably benefit greatly by transitioning to IPN. It parallels space-borne communications almost exactly:

• Messages will be scheduled. The pigeon fanciers may not want to ship messages at oh-dark-thirty.
• Resources are very scarce. No-one owns an infinite number of birds.
• Bandwidth is limited to the paper you can tie to a pigeon's leg. Of course, technological advances might have future sysadmins sending lots of microfiche tied to its leg.
• Bundle transmission may be interrupted by a shotgun, hawk, or a pigeonne fatale.
• You need encryption to protect against pigeon impostors sending bogus messages. Or at least a child-proof cap on the message capsule.
• Every hacker on the net will want to get a pigeon packet ping going.

The only thing we didn't discuss was money. Perhaps a roll of Starbucks gift certificates tied to the other leg would work?

John

Why no UFO technology?

[wfberg]

This work was performed under DOD Contract DAA-B07-00-CC201, DARPA AO H912; JPL Task Plan No. 80-5045, DARPA AO H870; and NASA Contract NAS7-1407.

Surely the Area 51 people would have figured out some better way by now?
--

Everything old is new again

[mwood]

As I began reading, I was reminded of three things:

• uucp
• BITNET
• A story about interplanetary communication by Isaac Asimov, I think it is titled "My Son, the Physicist".

(Come to think of it, there's a lot of good SF inspired by communication. Clarke's "Meeting with Medusa", Smith's Venus Equilateral stories, Stasheff's _Escape Velocity_ (more social / political than technical).)

InConsistent acroNym

[GlobalEcho]

Am I the only one who is bugged by the fact that the acronym IPN is formed like "InterPlanetary interNet"? Is the word "inter" important enough to form an acronym boundary, but occasionally too trivial to get a letter into the acronym?

How about Solar System Network (SSN) instead? Then a node on an asteroid still counts, too!

Re:InConsistent acroNym

[cuyler]

I just figured that IPN would stand for InterPlantery Network. In that case it'd make sense.

Re:Sounds like FidoNet

[E-prospero]

that's a lot like FidoNet (and uucp). The asymmetrical bandwidth is the only real difference.

Not really. Didn't you ever use a 1200/75 modem back in the day? 1200 baud down, 75 baud up.

Alternatively, ever connect to a BBS that had a sponge filter? This doesn't affect physical bandwidth, but it sure put a limit on my effective bandwidth on a number of BBSes...

Russ %-)

Re:uses for IPN

[addaon]

Why is the ISS any better a testbed than, say, asia? It's a lot closer most of the time, you know (from my POV here in the states).

SSN

[wiredog]

SSN is already in use. An SSN is a nuclear powered submarine. I don't think those would work well for interplanetary networking.

Re:SSN

[AddressException]

Hmmmm...I thought everyone had an SSN

Everyone in your small part of the world, perhaps....

Words to live by

[wiredog]

Therefore rely not on end-to-end connectivity at any time, for the universe does not work that way.

Re:uses for IPN

[interiot]

Talking to ISS would have half the ping time that satelite traffic has today. Satellite internet traffic doesn't require IPN.

A planets, its moons, and its immediate satelites can essentially be linked into one internet. IPN is only used between planets or further.
--

I could have used this decades ago

[interiot]

IMHO, it would have been nice to have a standard store-and-forward protocol years ago.

The core developers of the internet probably have very little trouble getting full-time connections to the internet. Until the advent of DSL and cable though, many of us were stuck with dialup connections. I've longed for an SMTP-like protocol that my programs could use to talk to each other when their connection was down.

Even for "permanent" connections, this protocol is useful. Connections go down from time to time, so apps that absolutely need 100% reliability need to build in code to wait and retransmit later when the link is back up. Instead of having each app implement this in a different way, a standard store-and-forward tcp protocol could have been designed.
--

Re:Wouldn't it be nice..

[Paradox !-)]

I suspect that's why they're working on it now. It's a helluva lot harder for a vendor to, ah, "embrace and extend" a protocol that was established long before the vendor realized it would be significant.

I complement the team for building this now. Solving problems way in the future is what true planning is all about.

Now, about Social Security... ;-)

NASA's DSN Link

[Paradox !-)]

Here's a PR PDF on NASA's Deep Space Network, which is likely what this IPN Bundle protocol would first run over.

http://www.cdscc.nasa.gov/pdf/dsn.pdf

Nice idea but...

[oPless]

Okay, at first I thought this was going to be a half baked idea, indeed the domain hierachy is a quite sensible recommendation (personally I hate seeing domain b*st*rdisation like *.uk.com uk.*.com etc. anyhow thats a different rant altogether :-)

Reading on, store and forwarding in nodes? Sounds like another fido/uucp implementation to me... not that I disagree that high latency links shouldn't have a store or forwarding system they *should*. But if one already has a workable system, why invent another?

On a lighter note, its a good thing(tm) that people are thinking about this now rather than implementing some half-assed propriatry system.

Re:Imagine: Microsoft Outlook Interplanetary Editi

[oPless]

OMG! :-)
Noooooooo! I'd rather die :-)

no, seriously :-)

Re:interconnection with mirror universe

[PurpleBob]

Evil Google would be a page full of banner ads where you click one that seems to relate to what you're looking for.

Evil Slashdot would be where the users of Evilnux, the operating system which you can be shot for not using, mock the surviving users of Good Windows. Occasionally Jon Anti-Katz posts a brilliantly-written column, loved by all its readers, in which he protests that the same people who claim to be defending immorality on the Internet actually aren't taking away enough rights, or that high school kids are having way too much of a good time.

On the Evil Internet, sending personal or informative communication using e-mail is grounds for being kicked off of your Evil ISP. Several Evil ISPs have started to install Evil Spam Filters to ensure that nothing but spam is served to their customers.

But the most prevalent thing on the Evil Internet would be the vast number of decent, moral sites showing pictures of cute puppies, hiding underground so that self-declared defenders of evil don't shut them down in the name of "harming the children".

--

Ping!

[Jarvo]

Depending upon which planets are connected to this network, the timeouts on IP packets will have to be increased. You'll have to wait for several hours to find out if the remote host is up.

Or you might get:
$ ping pluto
....
Destination planet unreachable.

Re:Tha battle rages on

[rtaylor]

I'd imagine that sex.uranus would be more popular.

applications to LEOs etc.

[Swordfish]

There are two obvious applications of this sort of protocol. One is to a network consisting of a very small set of LEOs. Suppose you are a really poor country with just one LEO in your military comms network. Then to communicate between your command and control centre and your troops, you have to be able to cope with large delays, waiting for the LEO to come overhead etc.

A second application is to a network of submarines which rarely surface, or to network nodes which are often out of communication, e.g. military back-pack radios which are used only once a day for 1 or 10 seconds, to minimize probability of detection.

They should have phrased the protocol more generally so that it takes into account all of those sorts of nets too.

Re:Come on, Give them a lottle credit...

[kevinank]

I don't even think that it is that far out. Wireless, disconnected devices, and packet radio connected subnets are all examples of systems that need a similar solution. I for one would quite like to see an HTTP implementation that is hosted on a message oriented transport protocol instead of TCP. E-mail would also translate well of course, as would usenet.

HTTP would have to be modified slightly to bundle subrequests automatically (a wget protocol?) and I haven't yet read far enough to see how the draft RFC proposes establishing a route between the source and destination. If those routes change very rapidly (as for example might be true over packet radio) I can imagine some difficulty in finding a path to the destination.

Re:Come on, Give them a lottle credit...

[kevinank]

Yeah, I remember reading about IP/SMTP back when it came out. The ping times are supposed to be 30s or so, but with the IT infrastructure at my company I think they would end up being closer to five minutes or more.

In any case TCP isn't going to run well over that kind of latency, you'll just fill up the pipe with retransmit requests until the TCP connection times out.

To really make use of this kind of application you need some protocols that aren't oriented toward interactive latency. IP datagrams over avian carrier are all well and good, but at an hour per packet they had damned well better be carrying complete messages.

I could see the Palm Pilot application synchronization being done over non-interactive protocols. Store and forward messaging would work fine, and of course the web would be okay if a bit slow (over SMTP.)

So there are plenty of potential uses, the question is really whether those uses will become prevalent, or the latencies of handheld wireless devices will drop to where they are no longer needed.

Banner for Linuxgram?

[ahde]

This morning I see a banner for Linuxgram "News that means business" or something like that. I click it and discover a slash-knockoff site that looks a little out of date. I think to myself, how does someone's homepage afford banners on Slashdot? Then I see a sidebar that says "beginning June 1st, LinuxGram will be available only by subscription - $100 per year (£70 in Europe)." I notice a VA banner at the top, and at the bottom; sourceforge on the side. Back at Slashdot I now see a VA banner.

*Click*

I will leave the whois paper trail as an exercise for the reader.

Lag vs. Speed of Light

[Sirius25]

so now my bandwidth won't matter?!? ping theSun... request timed out.. request timed out.. 8 min. HPB..

What it needs...

[Bitsy Boffin]

Ok, traditional store-and-forward protocols will translate well to interplanetary network applications. So email takes several minutes to mars, big whoop. Streaming applications, and downloading files will work fine to, so it takes 5 mins to get going, again, big whoop, once it is it can be fast as it is now.

But, the web, and anything that requires such interactivity (instant messaging, IRC, ...) is a big problem. I don't think the instant messaging problem could be solved.

But for the web all we need is a bigger, more powerful google on each planet/moon which uses it's cache as default (it would also cache images) location for the page, and does link translation in the document again into it's cache.

If user wants to get the latest version of the file they enter thier email address (james@martians.mars) and google.co.mars will notify them with a link when the page comes in off the wire, err, make that laser beam.

Of course dynamic page content is a different story (can you imagine requesting a page on earth from mars that calls a page on venus to do something, submit, 2 hours come back).

---
James Sleeman

Wouldn't it be nice..

[meckardt]

if just once we actually had a plan to implement a system before the system was needed? I expect that it will be another 10-15 years before there will actaully be much demand for this, but once we do start operating out there, its going to turn into an explosion as big as the Internet of the mid-90's.

Re:Wouldn't it be nice..

[basking2]

Ahh, but even if we have the protocol early, how long will it take vendors to support it all? :)

Sam
--
"The Son of God became a man to enable man to become sons of God."

Re:Wouldn't it be nice..

[a.tomaka]

Amen to that. Seriously, what is wrong with being prepared. People need to learn from past mistakes and realize getting stuff done before it is needed is definitely the way to go!
-------------

Re:uses for IPN -- Actually

[kjoyce]

The ISS orbits around 300 miles high - considerably lower than the 22,500 geosynchronus satellites orbit at. Meaning - ping times between New York and Cleveland are slower. But- there is no reason you can't test this using servers in the same room.

Re:Insane ping times

[Digital Anvil]

How about

touch uranus

Not with a Bang....

[Wintermancer]

Great, just what we needed.

Some pissed off alien getting dropped from a Quake death-match and starting an interstellar war.

"But...but...Alpha Centauri got /.'ed! It's not our fault!"
"We told you to upgrade that node years ago. Die, Earthling!"

And, so goeth the Earth, not with a bang, but with a lag.

I would hate to be an Onsite Tech for this.....

[os2mac]

"Bones, The Uranus subnet is down again could you get someone to go reset the switch? Dammit Jim I am a network technician not a proctologist".

Links

[LazyGun]

InterPlaNet

Mars Network

Unfortunately...

[clickety6]

... ICANN have already allocated all interplanetary domain names. All Mars based domain names belong to a certain candy manufacturer. Pluto domain names have been handed to Disney Corp. Saturn to the car manufacturer, Sun to the computer manufacturer. Venus was of course given to our Nathalie.

ICANN have reserved Uranus for themselves to keep up with their corporate policy of:

All Uranus are belong to us!

IP Addresses..

[heytal]

Have you used the IP address alotted to earth ? or are you using jupiters range ?

This could be useful now

[ibirman]

This scheme would also work in linking together multiple terrestrial networks. There are lots of internal Intranets that are not connected to the Internet. This protocol could make it easier to connect Intranets to the Internet and to each other while still maintaining security. There are many reasons why some people don't want to be on the Internet but still need its resources. The people that work at the CIA for example have to have two computers on their desks - one connected to the internal network, and one connected to the Internet. Maybe this scheme coule be used to securely bridge the two?

Tha battle rages on

[saider]

I wonder if ICANN will release new top level domains for the major bodies. Imagine the rush for sex.moon.

Marketing the interplanetary network

[isomeme]

Proposed slogans for interplanetary networking:

• "Putting the 'Distributed' in 'Distributed Denial of Service."
• "In space, everyone can hear you ping."
• "ISDN meets DSN."
• "...For those times you're 2.3 AU from the telco CO."
• "Sunspots...an exciting new excuse for network outages."

--

Store and forward...

[YKnot]

Our approach, which we refer to as bundling, builds a store-and-forward overlay network above the transport layers of underlying networks. Bundling uses many of the techniques of electronic mail, but is directed toward interprocess communication, and is designed to operate in environments that have very long speed-of-light delays.

Just when you think someone has figured out how to make interplanetary Quake matches possible, they tell me about store and forward...

Leadership?

[SubtleNuance]

Vinton G. Cerf and others just released an Internet Draft about the Architectural Definition of the Interplanetary Internet (IPN).

Is this the same Vint Cerf that is a member of the ICANN? The very same Vint Cerf that is discussed in this article at The Reg. Is this the same Vint Cerf who is aptly portrayed in these (part 1 and part 2) cartoons describing the actions of ICANN and their mis-handling of the Root-DNS?

If it is I suggest we steer clear of his current intentions - I wonder if he'll still be interested building a community owned/run, democratic, system with the proper 'goals'.

Vint certainly deserves accolades for his TCP/IP work, technically he's top-notch, but based on his reported actions within ICANN Im wondering if he hasnt been compromised by personal/corporate ambition.

True leaders are usually humble people - Vint sounds to be neither. I would feel a little better about this proposal if he were a more honest person.

rfc1149

[bartjan]

I wonder when we'll see someone implement rfc1149 using Space Shuttle technology.

Re:rfc1149

[RexxFiend]

You would need space suits and jetpacks for the pigeons!


A crash reduces
Your expensive computer

Insane ping times

[KarmaPolice]

Kinda interesting since our servers at the dorm are called Mars and Venus so ping mars and ping venus actually makes sense to write.

Just a quick thought:
Excuse me, miss. Can I borrow your computer? I need to ping Uranus?
**SMOCK**

Pockets of activity

[gailwynand]

I think this is really neat. Another aspect of this, is that I have always been interested in the thought that on each planet we colonize there could be a completely new and unique internet, only accesible from other planets in highly latent, or mirrored (also latently) form.

This is just one way in which different human societies will develop as we wrest life from our solar system.

Sounds like FidoNet

[b1t r0t]

Intermittent connections, store-and-forward bundles, scheduled connection times, that's a lot like FidoNet (and uucp). The asymmetrical bandwidth is the only real difference.

Ah yes, back to the good old days of e-mail, usenet, and ftp-mail. Just imagine trying to browse the web (or even just trying to use Slashdot!) over e-mail.

I can remember back in the early '90s how my only access to a very few select newsgroups was through a gateway to Fidonet, with my BBS (which wasn't really intended for users--they would just hog the connection) dialing long distance at midnight to get the feed.

Needs a new name

[Gaijinator]

We should call it the 'Sub-Etha Net' in reference to the HHGTTG and immortalize Douglas Adams.
----------
"Remember, your friends will stab you in the back for the price of an Extra Value Meal."

Come on, Give them a lottle credit...

[hillct]

It is a well written document, and really, the internet as a whole was once just a bunch of hopes and dreams.

The RFC process just puts some structure around publishing those hopes and dreams

--CTH

--

Wait till the scriptkiddies come...

[not-quite-rite]

1 0wn j00pit3r.

hehe

How about some this applied on earth?

[ishrat]

Could these guys keep a limited range and first test it out on the planet surface first? As both the wireless and the broadband industry don't seem to be coping up.

IPN Bundle Received

[FrankDrebin]

[spock@ncc1701a]$ mail
No mail for spock
[spock@ncc1701a]$
IPN bundle received.
You have new mail
[spock@ncc1701a]$ mail
"/home/spock/mbox": 147731 messages, 147712 new^C
[spock@ncc1701a]$ grep -c -i 'work from home' /home/spock/mbox
255412
[spock@ncc1701a]$ grep -c -i 'horny housewives' /home/spock/mbox
291337
[spock@ncc1701a]$ rm -f /home/spock/mbox
[spock@ncc1701a]$

Upgrades/Repair?

[Shoten]

Remember always that launch mass costs money. Think not, then, that you may require all the universe to adopt at once the newest technologies. Be backward compatible.

This might be a hell of a good place to use a really fault-tolerant and flexible architecture, like a cell matrix. That way, upgrades need not be so significantly hardware-dependent, and repairs can often be a matter of routing around damaged processors (not to mention the lesser specificity of devices means it's easier to keep spares on hand).

The Cost of It All

[OverlordQ]

Not many of the broadband companies are able to keep from going bankrupt, so just think about how quickly something like this would start costing.
-------------------------------

Violating principles

[jthill]

This notion:

When the central authority determines that a source application is out of funds, it reports the source application's bankruptcy to all bundle agents; from that time on, all service requests and packets received from that source application are rejected.

violates so many different principles it's hard to know where to start.

Denial of service is a last resort.

Pretty, but is it "literature?"

[Artifice_Eternity]

Umm...yeah, I like that they phrased it this way, rather than in dry technical language. But it ain't exactly great poetry. It's an intentional parody of this thing, which has been floating around on inspirational posters for years. It was apparently written and copyrighted by Max Ehrmann (1872-1945) in 1927.

Imagine: Microsoft Outlook Interplanetary Edition

[Artifice_Eternity]

I can't come up with any funny jokes about M$ embracing and extending their reach to the Oort Cloud, but I'm sure someone can...

A whole solar system of new opportunities

[WIAKywbfatw]

Great. Soon we will be getting spammed, cybersquatted, defaced and DoS attacked from all over the solar system.

Ain't progress grand?

Linux knows :)

[beaviz]

Reminds me of some old fortune snippet:

/*
* [...] Note that 120 sec is defined in the protocol as the maximum
* possible RTT. I guess we'll have to use something other than TCP
* to talk to the University of Mars.
* PAWS allows us longer timeouts and large windows, so once implemented
* ftp to mars will work nicely.
*/

-- from /usr/src/linux/net/inet/tcp.c, concerning RTT [round trip time]

this document will self destruct in..

[waspleg]

Cerf, et al. Expires November 2001 [Page 58]

about 6 months it looks like to me

so much for that 10 to 15 years heheh

i just skimmed through it, but it seems to me like having relay nodes (their idea) along the way to and from where ever using lasers (my idea) would make the most sense

Interplanetary internet? This sounds like...

[A Dapper M]

...that one Isaac Asimov story. You know the one.

interconnection with mirror universe

[cosmo7]

this is all well and good, but what about internet connections to the evil mirror universe, the one where Steve Jobs has a moustache?

they presumably, already have an evil internet up and running, but running on evil protocols. how long before we can connect with, say, evil google or - shudder - evil slashdot?

Re:interconnection with mirror universe

[guuyuk]

Since Microsoft runs on the current Good Internet, then that would mean a Good Microsoft exists on the Evil(tm) network...

Connectivity check

[lunaman]

$ ping bradbury.mars.nasa.gov
PING bradbury.mars.nasa.gov (139.169.196.201): 56 data bytes
64 bytes from 139.168.196.201: icmp_seq=0 ttl=4096 time=1665294.5 ms
64 bytes from 139.168.196.201: icmp_seq=1 ttl=4096 time=1665297.6 ms
64 bytes from 139.168.196.201: icmp_seq=2 ttl=4096 time=1665294.1 ms
64 bytes from 139.168.196.201: icmp_seq=3 ttl=4096 time=1665297.6 ms
64 bytes from 139.168.196.201: icmp_seq=4 ttl=4096 time=1665294.0 ms
^C
--- bradbury.mars.nasa.gov ping statistics ---
5 packets transmitted, 5 packets received, 0% packet loss
round-trip min/avg/max = 1665294.0/1665295.5/1665297.6 ms

what's important in life

[spacefem]

Therefore be at peace with physics...

There are too many applied computer scientists in the world, that's why I admire documents like these. So it's far fetched. Not terribly applicable to today's needs. I thought it spoke to me, in the way pure science speaks to me. When Marie Curie was discovering radium, her goals were entirely lofty, it's science, don't apply it, don't make money with it, don't win wars with it, don't launch Unreal Tournament with martians, just think about it, it's a beautiful thing. Know your limits, appreciate what you've got, admire how the universe can work for you.

Be backward compatible.

uses for IPN

[mrericn]

Does this have no application for the ISS. The ISS seems like it could serve as a wonderful testbed for extra-planetary networking. Sure it's still in orbit, and the distance is almost trivial (in Inter-planetary standards), but Alexander Graham Bell first made a call to another room.

Here's some interesting info about the use of laptops on the ISS but AFAIK no point of presence on the Internet for the floating condo yet...

I'm waiting to fork over my 20 million until I can get my /.!

For pie are squared..... (unless pie are round)

[RalphTWaP]

Alright, first off... let's point the masses at some good reading

Some Basics Of Radio Astronomy

... and some really specific reading.

Chapter 2. The Properties of Electromagnetic Radiation (PDF)

Now that that's done, let's revisit your comment again. Please see the section regarding the inverse square law and EM propagation.

Thank you,
-Goodnight


Nietzsche on Diku:
sn; at god ba g
:Backstab >KILLS< god.

What a great concept...

[Em Emalb]

Too bad our internet here on earth is suffering from the strongest known force in the universe.... the slashdot affect. I will be writing an rfc on this in the near future.... btw, anyone got a mirror yet? OMG, you slashdotted the IETF! You bastards!

basis for a frivolous law-suit

[rassie]

from the draft:
3.4. Bandwidth Allocation via Market Mechanisms: "Starbucks"
I didn't realise that coffee was actually used for the protocols as well, I thought only us sysadmins needed the fix.