I have developed a way to a virus free internet...

Quote:

---

Originally posted here by mark_boyle2002
[calling nokias bluff]
come to think of it nokia. Could you name a senior member with full ap who is a 15 year old kid who knows nothing. I am struggling.
[/calling nokias bluff]

---

MemorY

Quote:

---

Originally posted here by gbrowne40
I have successfully --on paper--designed a technology to outright eliminate the existence of viruses

---

You're getting rid of Microsoft? I have Virus free internet right now, 2 Slackware boxes and a BSD workstation.

Basicallly to do what your doing, your going to need your own programming language. If your planning on using languages that already exist, then it won't work, someone would have already seen the possibility...

"If it seems too good to be true, it probably is."
--probably_tons_of_people

I'm waxing verbose today, so let's begin. Examine the criteria. Viruses. First, I believe we can all agree their most important aspect is that they are algorithms, or sets of instruction. Secondly, these sets of instructions are able to replicate, increasing (through their influence) the number of concurrent and/or sequential instances of their execution.

To make a virus-free internet, you must either mean eliminating the internet as a vector, or having the participating nodes of the net all be both virus-immune and incapable of transmitting viruses, or else the viruses would still 'exist' on the internet.

That latter meaning is so obviously reaching that I will bypass it, since it would mean virtual perfection of control over all hardware and software systems, at least those with IP addresses. And assigning all availible IP addresses to chunks of inert rock isn't too useful.

Now for the former interpretation. The internet as a medium is virus free, so it cannot transmit viruses. Since you mention the internet specifically and presumably the current version of it, this will be packet data, the scope being anywhere between the ports of end-node computers.

To prevent viruses, you must split them from useful traffic. This means identifying either useful or virus traffic. I do not believe with present technology that useful traffic can feasibly be 'signed' cryptographically, at least not to an extent that the speed of transmission and processing is both useful and that viruses cannot forge it.

So we have to default-allow everything and detect viruses. These viruses will often be split up over more than one packet, requiring reassembly, scanning, and purging. Given the structure of the internet, each router would be hard-pressed to verify the number of hops that a packet has taken to get to it, so each router would need to scan to be safe unless specific DMZ routers were set up.

So we need to reassemble every packet stream, scan it for *known* viruses, scan it in a perfect heuristical sense (or try a sandbox) and then either pass or reject the packet stream. The fastest way to do this may be to drop packets only once you are sure it is a virus, hoping the fragmentary first part is missing enough code to be inert or disabled. Caching the whole stream would be impractical in terms of space and processing, but also in terms of latency---packet data would surge and flow in waves and floods and other water-imagery, pretty much destroying the whole point of packet-based switching.

So in the end you have every single router snooping every single stream of data and assembling it for scanning, meaning you've just more than more-than-linearly increased the necessary processing power to keep the internet moving depending on the number of nodes. (As for distributed virus scanners that cooperate, see my earlier post.)

In light of that particular set up assumptions and chain of reasoning, I prefer the inert rock plan. Cheaper with just as many end-user features.

And no, quantum computing principles are not kosher as a defense. :)

:) Well , if i wasn't confused at the start of this thread i bloody well
am now. :confused:
In saying that, i will also add that it is one of the best debates on this site
in a long time. No neggs and insults (well not many) and a good all round
discussion is an improvement that is welcomed.

Does that mean i can throw away my crayons and scribbly pad?

Torr...I like what you are saying but I could forsee "smart-viruses" in the event of that technique. Viruses broken up from packets and would have to have a way to be reassembled. I would develop a virus to work in multiple stages ultimately reassembling to execute. That is an approach I would take to battle that...I like the way you think. Things could be so much easier if we start from the original square one. And there is no reason to try to identify which is malicious and which is legit. I am sure. I notice you are from Seattle. What do you do? And for who? Your quote says alot about what I am going for...

I don't want to sound too pessimistic, but are you sure there would be any takers? Looking back at the old VHS vs Betamax scenario, or CD music vs those nasty MP3s :cry: the best technology doesn't always win. The internet is already widespread (no kidding!) and getting ingrained in its ways... IPv6?

Big companies like Billy's would rather do things their own way (Trustworthy Computing, anyone?), and that would be a tough battle regardless of how good your product could be.

Then again, if you could find a way of pushing this to the porn industry... :D

Don't get me wrong, I'd like to see this succeed, but it might be a hard slog.

As a community, we should be open-minded and skeptical at the same time. If this idea has merits and has been discussed by some of our more experienced community members, then the rest of us should try to help.

I offer my computing resources if needed (though I doubt they would be).

Gl with it

you've got my attention and id be interested in seeing the idea even if it meant signing the NDA.

i2c

I've signed NDA's for weirder stuff..

so if ever.. count me in !!

Although it is "realy" dangerous for a GPL minded person to ever sign a NDA...
I have signed them for my current job too ;)