I once knew a guy that was into handwriting analysis. You see or hear about it sometimes, where how you write is suppose to say something about your personality. So he asked me for a sample of my handwriting to see what he could determine about my personality from it. I wrote out a page of my thoughts (I never write briefly) and gave it to him. He said he only needed a sentence or two, so I told him just to take his pick. He replied Never Mind, and said he would get back to me with the results in a day or two.
I got back my original paper with these words added all over and arrows pointing to portions that apparently supported what he claimed were indicators reflecting my personality traits. These are a reflection of about 90 percent of the words he used.
Stubbron, bull-headed, iron-willed, obstinate, difficult, insistent, demanding, assertive, overriding, inflexible, opinionated, argumentive, mule-headed, and something about acting like a jackass, but I forgot the precise reference made.
Some of these words appeared repeatedly, and some words had arrows connecting them to different parts of where I had written something. meaning it was not one indication but many involved. I looked at him, and he seemed dead serious. So you know what I thought? Must be something to this analysis thing, because he really had it right. I was all those things and more. Realizing it may have soften me a bit over time, but I am not sure.
So what is the point of saying this? Because I often do not start off with the simplest problem or simplest configuration, but decide to see how far the matter can be taken, and that is from the very start. What I have here is a D-Link router and wiring for up to four PCs to connect with the internet, and that works fine. So do the PCs that connect to it. But I was taken by a chance to get a notebook PC at a great price with many features, like a large screen and keyboard, so I went ahead and bought it. It comes with both wire and wireless internet capabilities, and I just used wire initially. That was fine.
But my nephew across the road and a bit north of here has little money, and a deal for cable access that he shared with his brother-in-law next door just went to pot, so he got his own access from the cable company. But he is so busy with helping others that he has little time or interest in using it, and the expense mounts up. Having a PC is only half the picture, because the internet brings you the world, just like TV does, but in more useful ways.
So I got a Geeks' Wireless-N router with four ports at a bargain price, unplugged my notebook from the D-Link, and just attached the Geeks' router there, and it works fine in that configuration. I just had a dicken's of a time getting the wireless adapter in my notebook to connect to it, but that has all been solved. So 4-port D-Link, and port 4 runs to the Geeks' as its WAN, and the notebook is wireless from there.
Some will tell you that this is a bad idea, and others say the Geeks' should just replace the D-Link, but I am sort of stubbron, you know? Oh, of course you do. So I am not going that way, because it is just the way that everybody else goes. My nephew's involvement is that I want him to work on wirelessly connecting to my setup so that he can get away from that heavy expense each month, when he gets so little use out of it. He's a stubbron cuss though, and not inclined to follow my lead on this. Not sure why.
I got a USB to Wireless adapter at a great price, and he can pick up about six AccessPoints in the neighborhood, but the signals are kind of weak (actually, pretty dang low). I'm trying to convence him to look at some of the many online postings about making your own directional antenna or reflector, some boosting signals by 12 dB to 24 dB, and that is a tremendous amount of gain, so if you get any signal at all, you should be able to really pull it in with the right antenna.
But my nephew don't believe it because... Hey, I don't know why he doesn't believe things sometimes, but it kind of makes you mad when you know you are giving him the real lowdown, and he won't take your word for it and he won't bother to check it out for himself. People can be like that, you know? Just frustrating and inferiorating, but I don't like to give up on things because... well, that's just the way I am.
So what now? I want the PCs connected by these two routers to act as they are part of one domain or one workgroup. I already know this might be super hard because of my chosen configuration, and made even harder by the fact that the PCs might be running either Ubuntu or Windows 2000. From what I've found so far, this is a task that is just past the scope of what Windows 2000 can do, and you have to move on to XP if you want this to happen. With Windows, nothing seems beyond its inability at times, but I won't be pushed in that direction. There has to be another way, and I just have to find it.
Ubuntu has a Network button under Places. and under it, you are suppose to connect to a server. What server? What does this server have to be set up as, and how? Don't know yet. I installed both Samba and smbfs as some reading has suggested, but nothing significant from either yet. I may actually have to read something about both. There is a samba configuraton file, and some suggestions for settings in it in a couple of posts, but some of these posts go way back, and the contents of the config file have changed over time.
What is a domain, and what is a workgroup? With an IP mask, every binary bit set to 1 reflects a part of a license code that is used in the public domain to direct network traffic from one place to another, It is like a business address, which brings mail and packages in he door and into the mail room. Every bit that is left a zero is for internal use only, and can designate specific maildrops in the whole of the business itself. A 201 code for internal use might indicate the first maildrop on the second floor. Another business might use 201 to indicate a certain office on the second floor. It makes no difference, because it is just an internal reference.
So in an IP address mask, what does 255.255.255.0 represent? Translating each of the four goups into binary format, each 255 represents a group of 8 bits all set to 1s, and the 0 represents 8 bits all set to 0s. In other words, this is what a binary representation of 255.255.255.0 looks like: 11111111.11111111.11111111.00000000. Now this is lined up with an actual IP address, such as 126.96.36.199 when seen in decimal form, but let's convert that to binary as well: 11011000.11101111.00110011.01100111
So let's stack these two numbers together:
11011000.11101111.00110011.01100111 <-- IP Address in binary form
11111111.11111111.11111111.00000000 <-- IP Address Mask in binary form
11011000.11101111.00110011 <-- the part of the IP Address for public routing
Now the logic rule that is applied here is the AND rule. This states that when binary numbers are aligned like this, the only time you get a "1" below the line is when there are two "1"'s aligned, one below the other, above the line. The leftmost side has a 1 over a 1, so we put a 1 there below the line. Anywhere else, we would have a 0 instead.
Now I did not show the rightmost eight 0's for a reason, which is that based on the 1's in the mask. the ones shown below the line make up the public address when used together. The rightmost group, not brought down as zeros, would be the designator for the mailbox, or the device, that is only a concern of the business itself. This is referred to as a Class C license.
Note that all the bits in either group total up to 32 bits. If all 32 bits were flagged with 1s in the address mask, that would be a Class D license, the lowest class, because it then deals only with the public address and leaves nothing left for directing internal routing. Working backwards, from right to left, if the first 16 bits in the mask are set to 1's, that would be a Class B license, and that would leave you with 16 bits in the IP address for internal routing purpose, which is quite a few. Coming over to the left one more group, if the first 8 bits in the address mask is set to 1's, then that would leave the remaining 24 bits for internal routing purposes, which is a tremendous amount.
It is a good scheme, but the license administrators did not allow for the extensive growth of the world of networking and gave out way too many Class A and Class B licenses at the beginning. Many large organizations have tried to get multiple Class C and/or Class D licenses for their needs as a result.
The real result is that the world has been running out of IP addresses for quite awhile, and several solutions have presented themselves. One is that instead of assigning static IP addresses to each machine, the machines have to rely on requesting an available IP address to be loaned to them for awhile, and that would be a dynamic IP address. It would not be the same numbers in the IP address each time.
A second approach is to determine a method of associating a specific port number for each machine to both idenfity the machine and to send traffic back to it. An IP address with added port designation would look something like this: 188.8.131.52:12345 <-- the port address is tagged on the back As it happens, though, some port designations are already reserved and used for specific services, such as port 80 is used by web browsers to browse the Internet, and you can employ only one port designation per address.
We may ultimately have to move on the third method, which is to find a way to add more bits for use in the IP address and possibly in the address mask. They have already come up with a method for doing this, but does not fit in with what we are doing now.
Notice that there are just four groups in the present IP address structure, so that is referred to as IPv4, and that is what is being used today. The new form adds two more groups, taking us from 32 bits to 48 bits, and that is called IPv6. As more and more networking devices are made available that have the option to work with IPv6, we may eventually get to a point where we either switch over, or subdivide the world network where part of it continues to use just IPv4, the rest uses IPv6, and we divise methods to interchange traffic between them IPv6 offers such a vast array of addresses that there is some confidence that it will never be depleted. But of course they thought that about IPv4 once too, It isn't that we can't switch over. It is just going to be expensive to do and disruptive, and how much it will cost and who will be hurt or damaged by mistakes made is impossible to determine beforehand.
Anyway, that is the good news. By good, I mean we aren't there yet. But what I want do do now is get a home network set up behind an interactive pair of routers, one serving as an access point, and be able to interconnect and talk between them and use shared things like printers.
Keep in mind, like I said earlier, that this is to be a combo of Ubuntu and Windows 2000 working together, if I can find a way to make it happen.
So ideas are welcomed, tips even more so, and maybe some progress will get made. I'd like to think so. There seem to be so many settings involved, from Samba to the routers themselves, and it is hard to find information on what all those settings are for and what they ought to be set to.
I paid my dues by telling you some of what I know, but there is much I haven't had a hand at yet. I guess my first real questions are:
How to get all the PCs involved to see each other? They actually only see themselves at this point.
What should the local server be, Windows or Ubuntu? Can I make it so that it can be either? What needs to be done to make this happen?
And that is for starters.