What would it take for me to be able to connect my computer directly to the internet without having to use an ISP ? Why don’t people do it ? Is it even possible?
In a word, it money. Of course you could do it. It would involve leasing a line from a telco, purchasing a range of IP addresses, etc. The telcos essentially own the backbones that make up the internet, then they lease connections to these backbones, typically for ISPs. So you would technically be your own ISP. You would also need networking equipment (routers, switches, etc). For instance, you could lease a true T1 line from the telco, but it could easily range $1000 per month. There are different classes of lines, so you could buy whatever you could afford.
Of course there is more involved, but yes, you could do it if you had lots of money.
IF it is possible to connect to another computer (when you know all the information) then why can’t you do that with a web-site?
You can connect to another computer (on your network) because you have a physical connection with wired/wireless routers, etc. between you and the other computer.
If you want to connect to someone outside of your physical network, you need to have an interface to the backbone of the internet, which is what your ISP provides.
Can you rephrase this question? I don’t quite understand what you are asking. Why can’t one do what with a web site?
Because the website doesn’t talk to computers, it talks to the Internet via a protocol involving networking, routers, and IP addresses. Unless you are part of that network, the Web Server supporting the website you are connecting to has no way of addressing you.
When you type in a URL you are connecting to a web server that responds to requests from the Internet to serve pages. In order to serve those pages it needs to know what address to send them to. There’s no way to tell the web server “Hey, over here, just send it to my machine, forget about all this IP address stuff!” It can’t, all it knows to do is to respond to requests that come in over the net. There’s no phone line on the webserver to call, it is connected to the rest of the world through an Internet line (greatly simplified) and can only respond along that line.
Well, yeah, it can be done. It’s expensive, though, both in terms of hardware and fees. Here is an overview of the process. For more info, float around the ICANN site for a bit of light reading.
A bit of terminology clarification first, just to help with the answer:
An internet (note the lowercase “i”) is a network of networks. Basically, it’s groups of computers connected to other groups of computers. The Internet (note the uppercase “i”) is the specific massive worldwide internet through which we’re using this message board.
The Web is a software protocol that allows a client computer (like the one you’re using as you read this) to get hypertext documents from a server computer (like the Straight Dope Web server), using the Internet.
The Internet is a large, amorphous blob of equipment at this point. There isn’t a single connection place you can point to and say, “That’s the Internet. Pop a jack in there and you’re connencted to it.” There are, indeed, connection points (MAE West and East, for example), and there are “backbones” (not just one, as Fat Bald Guy said, but lots of them) running across continents and countries. Some are privately-owned, others are public or semi-public.
As beltbuckle pointed out, connecting to a MAE or a backbone requires a substantial investment in equipment and serious monthly fees. Unless you require massive bandwidth, it’s far cheaper to find someone else that has a connection, and plug in to them. That someone else may be a school, a local business, or a company that exists solely for the purpose of selling Internet connections–an ISP.
Does that straighten things out?
Ill jump in here with some highly-suspect rumors:
-
What is now the “web” (nee “internet”) was established in the early 1970’s by the US DoD as a means to inter-connect command & control facilies so that a hit on just about everything would still leave them with c&c. (The computer networks (using the term very loosely by current standards) of the time were all centered around mainframes/mainframe sysplexes, and a hit on them would take down everyone running on it - the idea being it was a lot preferable to trust the AT&T dialing/routing facilities than to blast-proof all the processors and their networks).
-
I (many years ago) heard a passing reference to ‘insiders’ snickering about the suckers who didn’t know the modem commands to connect, and had to pay an ISP for access. I have never met such a person, nor have I ever heard any other reference to them, but: since the DoD was presumably running the ‘internet’ before anyone had heard of conditioned lines, let alone T1, T3, etc. and ISP’s were not even on the horizon, how did those remote sites connect, and would that technology still work?
Anyone else heard of hackers getting into strategic DoD machines? Are they still on the web? (I would hope that they have exceedingly efficient firewalls these days)
Highly suspect indeed. In fact, almost completely false.
The Internet, which is and always has been an entirely different creature from the Web, was not designed to withstand nuclear attacks, as is commonly thought. It was funded by the DoD (specifically the Advanced Research Projects Agency, precursor to DARPA) in order to develop packet-switch technology for computer communications. The primary users were the universities who developed it. The web is simply a particular service which uses the internet for communications. Email, Usenet, IRC, FTP, and other services are entirely independant of the web, but use the Internet.
Nonsense. Everyone must connect through somewhere. Even your ISP must have an upstream provider or a peering agreement to connect to other networks.
T1s and T3s are just bundles of copper phone wires, and tons of people had them long before the internet was around. The only difference was that they were connected to telephone switches instead of network routers. In the olden days, most people who wanted to connect to the internet used a modem to dial-in to a mainframe (probably at their university) which was connected. This is pretty different from dial-up connections these days, where personal computers have an actual IP address and the software to do all the routing and whatnot.
Important DoD machines are not now and most likely have never been accessable on the public Internet. There have been cases of dial-in numbers for certain important machines accidentally becoming public information, though.
Of course, you can log in free if you “steal” service, as you can with telephone and cable.
T1, T3, et. al.:
Around 1982, I was working at a small office 35 miles from the mainframe. Somehow, I ended up being the software person designated to contact the phone co. when the line went down, and there was no hardware person around (which was most of the time - this was the infamous CFIS project (Bechtel)). There was a specific phone number to call, and the expression “data-conditioned line” was key to getting their support person to check the status/condition of the line. I have no idea what distinguished that line from the voice lines, but if I did not mention “conditioned”, the guy would end up spinning his wheels. Thus, the data communication lines, even then, were a different breed than the voice lines. I spent a few months working for the billing adjustments MIS/IS/IT group at Pac Bell in the 90’s, which was just enough to give me a hint of all the different kinds of wires they ran. If you know the term ‘NPA’, maybe you can answer this Q: What is/was a ‘metal line’?
hijack (I hope the TSA doesn’t come barging in, guns blazing…)
Any others scarred for life by CFIS?
This is debatable. The RAND Corporation was commissioned by the USAF in the 60’s to study computer interconnect technologies. From this work Paul Baran (and others) developed and endorsed a packet-switched network as the best solution. What was the first benefit listed in the Summary Overview of this technology?
Not long after the RAND reports were published, DARPA hired Lawrence Roberts who was furthering network technology research. Roberts was familiar with Baran’s work and also that of Kleinrock who had focused on packet-switching network theory. Roberts then became the driving force behind what would become ARPANET in the early 70’s. ARPANET was the first viable implementation of a packet-switched network.
While Roberts and other researchers probably did not deliberately modify their research to satisfy DoD demands, it is probable that the DoD chose to fund this particular project and to hire specific project leaders based off of the RAND assessment.
The universities that participated in developing ARPANET were being funded by grants and contracts from DARPA. As they were the ones creating this technology, they were also the primary users.
Making a statement like that is the quickest possible way to destroy your credibility in a discussion like this one. “Nee” means “previously known as.” The World Wide Web and the Internet are two different things. The Web is not a new name for the Internet.
The Internet is a set of interconnected networks communicating through IP protocols. One of those protocols is TCP/IP. Many different services can be offered using TCP/IP (e.g. ftp, streaming video/audio, email, instant messaging, VOIP…). One of those services uses a higher-level protocol called HTTP (hypertext transfer protocol). Computers using HTTP to distribute hypertext (and other) documents are called Web servers. The set of all Web servers on the Internet is the World Wide Web.
I know it can seem very fuzzy to a non-techie. Email is an Internet service, not a Web service, but you can create an interface program between a Web server and an Email server and come up with a product like Yahoo Mail that makes it look like email runs over the Web. It doesn’t. In that case, the Web is being used to view your mail, and to create new emails. The email is not delivered using the Web. It’s delivered using the Internet.
Does that help?
Here is a very well answered previous thread on the subject.
In other words, it would be like asking someone to call you when they don’t know your phone number.
Back in the early days of the internet, there wasn’t a whole lot in the way of computer security. I can remember back in the early 80’s that companies started putting their computers onto the internet, some with security, some without. A couple of banks got in trouble because of the lack of security on their computers allowed people to move money that didn’t belong to them around (which prompted a lot of other people to start taking computer security seriously).
Back in those days, if you knew the modem number of a computer you could dial it up, and if it didn’t have much in the way of security you could pretty much use that computer to jump around to anywhere on the net.
Computer security has come a long way in 20 years. The idea that you could still find a local computer with free modem access is downright silly. However, there are areas now that have free wi-fi access. If you are within reach of their coverage area, you can have free interent access without paying an ISP.