--------------- FIDO MESSAGE AREA==> TOPIC: 193 LAN Ref: DEY00001 Date: 10/28/96 From: MIKE BILOW Time: 08:46pm \/To: RASMUS BECK (Read 8 times) Subj: ETHERNET LAN NEWBEE Rasmus Beck wrote in a message to All: RB> I am about to set up an Ethernet LAN with about 3 - 4 RB> computers. Are you planning to use any software? RB> I would like to know if there are any traps I should look RB> out for? This echo doesn't have that kind of bandwidth! RB> Which would be preferable Ethernet or Fast Ethernet, RB> prisewise. Fast Ethernet cards are not a lot more expensive than for standard Ethernet, but Fast Ethernet concentrators are about 10 times more costly. For a network with only 4 nodes, Fast Ethernet would likely be overkill. -- Mike --- * Origin: N1BEE BBS +1 401 944 8498 V.34/V.FC/V.32bis/HST16.8 (1:323/107) --------------- FIDO MESSAGE AREA==> TOPIC: 193 LAN Ref: DEY00002 Date: 10/28/96 From: MIKE BILOW Time: 08:49pm \/To: MARK WRIGHT (Read 8 times) Subj: Knowlidge needed Re: NetWare for OS/2 Mark Wright wrote in a message to Mike Bilow: MW> Thanks for the overview, it really sounds like a workable MW> product to me, I wonder if it had have been given away with a MW> purchase of netware... Savy marketing is what makes or breaks. What killed NetWare for OS/2 was that it was a solution looking for a problem. If you want an OS/2 file server, it makes a lot more sense to use LAN Server (now Warp Server) instead of wasting time with NetWare. For that matter, ripping out a NetWare server and replacing it with an OS/2 server was a reasonable action under certain circumstances. NetWare for OS/2 was for the very narrow segment of people who had a need to run NetWare, and specifically NetWare, in an OS/2 shop. An example of such a case would be where a particular piece of software had to be used, and it was only available as an NLM. -- Mike --- * Origin: N1BEE BBS +1 401 944 8498 V.34/V.FC/V.32bis/HST16.8 (1:323/107) --------------- FIDO MESSAGE AREA==> TOPIC: 193 LAN Ref: DEY00003 Date: 10/25/96 From: ERIC SMITH Time: 03:50pm \/To: MIKE BILOW (Read 8 times) Subj: coax headaches 1/2 MB>Eric Smith wrote in a message to Mike Bilow: MB> > ES> If you have one with a long enough extension on it, unplug it > > ES> from your TV and slowly just push it back in without screwing > > ES> or touching the outside parts. You will see you TV signal come > > ES> to life. It works this way because of the Skin effect. MB>MB>This is dead wrong. MB> ES> No it's not. Try it. MB>No, I realize that the experiment will succeed. What is wrong is your >assertion that this is somehow connected with the skin effect; it is not. Then how do you explain it then. Does it have a secret transmitter and transmits the signal through the air???? How does it complete the circuit? MB>MB> ES> The electrical signal is sent through it's core. Each channel > > ES> has it's own frequency, which moves out slightly from the core > > ES> towards the surface (Note I could have the core and surface > > ES> backwards, but I believe the charge is in the core). MB>MB>This is dead wrong, too. MB> ES> No it's not. If you don't believe me. Talk to a CATV engineer > ES> at your cable company. A RCDD may know that as well. MB>What makes you think that anyone at the cable company would have better >engineering credentials than I do? :-) Tell me what your credentials are. MB>MB>If you had a way to look at an infinitesimally tiny spot along > >the cable electrically, you would see a single instantaneous > >voltage. At any particular instant, that single voltage would > >be a component of a single waveform that is determined by the > >algebraic sum of all of the signals at different frequencies > >which go into that cable. By analyzing the series of voltage > >measurements over time, you could deduce (filter) any > >particular frequency waveform. MB> ES> The signal is there yes, but it travels along the core at a > ES> certain location. MB>The important point is that the physical location in the conductor has almos >nothing to do with the frequency of the signal. It's not meaningful to sk >where channel 3 is in the cable with respect to channel 4, for example. The >frequency and time domains are interdependent and substantially quivalent, >something shown mathematically by Fourier hundreds of years ago. The >bandwidth >of a cable is mostly a result of its attenuation and the fidelity with which >instantaneous changes in voltage can be propagated, which are in turn >determined in large part by the capacitance of the cable. It's just plain >wrong to think that you can look at a cross-section of a conductor and see >different places occupied by different signals. I am afraid you mis-understood me then. It uses the skin affect. But that is just a basic explanation for those not that familiar with it. You have to keep the original thread in mind here. We are already way to technical for that discussion. MB>MB>If all of those frequencies travel in different parts of a coax cable, it > >would follow that thicker cables have higher bandwidth. It > >would also be kind of hard to explain why fiber optic cable has > >bandwidth orders of magnitude higher than any coax. MB> ES> Apples and Oranges. Fiber uses LIGHT and is transmitted at > ES> different frequencies. Other cable is metallic, and uses an > ES> electrical signal. The thickness has nothing to do with it > ES> except for the amount of electrical current it can handle, > ES> its shielding, casing... MB>Visible light is not qualitatively different from Ethernet: both are >transverse >electromagnetic radiation, and are subject to the same general rules of >refraction, reflection, and attenuation. All of the techniques commonly use >with light, such as waveguide bending, would work with radio frequency >Ethernet >signals equally well except that that the waveguide would be impractically >large because of the increased wavelength. Fiber typically uses non-visible light. Fiber equipment have some sort of photo sensor at the connection points, which regular electrical equipment dosen't. Both light and radio are at different frequencies. But that is just about where the similarities end. Light is immune from electromagnetic interference. And going over fiber, it doesn't have the db loss like the electrical currents do. MB>MB>I think you are confusing braiding and twisting. The two conductors whic > >form MB> ES> You say Tomato, I say Tomotto. (Continued to next message) --- * QMPro 1.53 * Ross Perot - America Needs Him. --- WILDMAIL!/WC v4.12 * Origin: Aspencade BBS (1:226/0.0) --------------- FIDO MESSAGE AREA==> TOPIC: 193 LAN Ref: DEY00004 Date: 10/25/96 From: ERIC SMITH Time: 03:50pm \/To: MIKE BILOW (Read 8 times) Subj: coax headaches 2/2 (Continued from previous message) MB>No, braiding is not twisting. These are very different things. If you were to take apart a CAT-5 cable and look at the twisting carefully, you will see that it is also braiding. Like a girl would braid her hair. I know what your talking about though. I was just saying that in general terms, they are the same. MB>MB> ES> CATV cables have the same problem, except there is only one > > ES> wire in it, so it must be shielded. MB>MB>This is also dead wrong. The shield in coax is an integral part of the > >system, MB> ES> No it's not. It must be shielded or there is nothing to > ES> prevent electromagnetic interference. MB>Shielding in coax does not prevent electromagnetic interference, either. On >of the great challenges in antenna design is getting the transmission line t >operate as a pure transmission line and not as part of the antenna. This is >known as "decoupling" the antenna and transmission line, and a number of ver >sophisticated techniques have been developed to do it. If the signal >encounters an impedance discontinuity, either along the line or at the >feedpoint of the antenna, then the shielded feedline itself will happily >function as an antenna. Furthermore, there are antenna designs actually >based >on this property, such as slotted radiators often used in broadcasting. >There >are also transmission lines, such as unshielded and untwisted "ladder line," >which are among the lowest loss and best quality you can get. I never challenged any of that. BUT, I do challenge that use where networks are concerned. What you just described, almost ALWAYS has Interference in it. Way to much to be viable for High speed and accurate computer communications. If the shielding in not important, then why do they have so much interference that way? I have installed whole networks where specially shielded cables was needed, and in one case an entirely fiber LAN because they were right next to a radio tower, and their PCs needed heavier shielded cases also. Now, before you go and challenge that. I used the fiber, and replaced the cases, and now it works. The way you are talking, one would be lead to believe that interference is no problem at all. --- * QMPro 1.53 * Ross Perot - America Needs Him. --- WILDMAIL!/WC v4.12 * Origin: Aspencade BBS (1:226/0.0) --------------- FIDO MESSAGE AREA==> TOPIC: 193 LAN Ref: DEY00005 Date: 10/25/96 From: ERIC SMITH Time: 03:54pm \/To: MIKE BILOW (Read 8 times) Subj: coax headaches 2/2 MB>Eric Smith wrote in a message to Mike Bilow: MB>MB> ES> The cheep stuff you get in stores has a aluminium foil type of > > ES> material inside the black rubber, then a white hard rubber, > > ES> then the single wire. The better cables have a raid of wires in > > ES> place of the aluminum (Which is probably what is being referred > > ES> to) and often has a type of jelly in with it so that the cable > > ES> can bend and move without breaking those brades. MB>MB>Jelly?! You need to starting buying your cable elsewhere. MB> ES> You need to study the field more carefully. I have been > ES> working in the communications field for a decade now. I have > ES> been a Network Engineer for 20 years. I have see hundreds of > ES> different types of cables from a multitude of manufacturers. MB>Oh really? So, you've been a network engineer since four years before the >IEEE >Ethernet standard was adopted, and one year before ARCnet was invented? t >must have been pretty tough back in those days when you had to whittle our >protocols out of wood. :-) You really are out of touch. Networks have been around since the 60's. Not all networks use the Arcnet and Ethernet protocols. By the way. Both Arcnet and Ethernet were around in 1986. I installed my first computer network in 1987, Arcnet. When we needed more speed, we upgraded it to Ethernet. MB> ES> You can get some to be direct burried, direct burred with > ES> jelly for "Mechanical" reasons, some with heavy tin > ES> shielding for protection against rodents...The list goes on > ES> and on. If you doubt this, I would be more than happy to > ES> email a list of manufacturers and some of their cable specs. MB>I've bought a lot of coax cable in my day, too. Meaning? --- * QMPro 1.53 * Ross Perot - America Needs Him. --- WILDMAIL!/WC v4.12 * Origin: Aspencade BBS (1:226/0.0) --------------- FIDO MESSAGE AREA==> TOPIC: 193 LAN Ref: DEY00006 Date: 10/28/96 From: AHMAD KHIARY Time: 01:29pm \/To: ALL (Read 8 times) Subj: Volume Conflict Hi all, I have a Novell Netware server running Netware 3.11 20 user. I had a 1GB drive SCSI-2 on an Adaptec 2940. I also had a rewritable magnito optical SONY drive connected to the card externally. The server drive is F: and this optical drive is H:. The optical drive was working fine until . . . I decided to install a new 2GB scsi drive internal. The 2GB drive is mapped to G: and woks find. But, no more H: !?! I can detect the drive on the server console using Monitor. But, when I try to map the volume into a drive letter (namely H:) I get an error message saying that it can not detect the drive path. The command I use is : MAP H: = EEE\RMO: where EEE is the server name, and RMO is the volume name. Please, help. I need to use the optical drive asap. Regards. Ahmad Khiary (Jeddah, Saudi Arabia) Internet ID : khiary@netbox.com --- RoseReader 2.52 P007485 --- QScan/PCB v1.19b / 01-0045 * Origin: Saudi BBS. Jeddah. Saudi Arabia +9662 667-2293 (2:542/100) --------------- FIDO MESSAGE AREA==> TOPIC: 193 LAN Ref: DEY00007 Date: 10/28/96 From: AHMAD KHIARY Time: 01:29pm \/To: KURT HILL (Read 8 times) Subj: RJ-45 and 100Mbps? Reference your message, KH>> 1) RJ-45 (Twisted Pair) is rated at 10 Mbps. Shielded RJ-45 is, KH>> I assume, OK for 100Mbps. But regular coax, the stuff we all KH>> use for thinnet at 10Mbps is *already* OK for 100Mbps as it is KH>> already shielded. KH>> KH>> 2) Hubs are better from a reliability standpoint. OK, we all an KH>> agree on this, yes? But not a strictly necessary item. KH>> KH>> Then *why* aren't hubs made with coax connections? Moving to ast KH>>ethernet (100Mbps) requires the use of *shielded* RJ-45, so we can't use our KH>>coax which is already rated for 100Mbps, and additionally we all have to find KH>>$$ for the 100Mbps hubs... There must be a good reason why coax was thrown KH>>out the window! 100Mbps hubs cost $1,500 and up, and I can put together a PC KH>>with two 100Mbps NICs in it to act as a router for the same or less, so the KH>>"hub/reliability" issue is not a convincing one... Or perhaps I just do not KH>>understand -- this is most likely the case, as I am, in fact, a moron. Any KH>>help and good, in-depth technical discussions would be most preciated... If you want to use UTP with RJ-45 on highspeed hubs (100MBPS), then use UTPs that are category 5. They are certified for highspeed. Regarding the Coax cables, they are capable of pubping data at 100Mbps. But, the cards in the market are not. I don't see a problem in using Coax cables (55Ohm or better) for highspeed with proper cards. The best configuration for a highspeed LAN (in my himble view) is multiple stars on a bus. That is interconnecting the servers and the hubs using a Coax (could be thick or thin net) and from the hubs spreads the star using UTP to the stations and peripherals. Regards. Ahmad Khiary (Jeddah, Saudi Arabia) Internet ID : khiary@netbox.com --- RoseReader 2.52 P007485 --- QScan/PCB v1.19b / 01-0045 * Origin: Saudi BBS. Jeddah. Saudi Arabia +9662 667-2293 (2:542/100) --------------- FIDO MESSAGE AREA==> TOPIC: 193 LAN Ref: DEY00008 Date: 10/28/96 From: AHMAD KHIARY Time: 01:29pm \/To: BRIAN LEO (Read 8 times) Subj: NETWORK HELP Reference your message, BL>>If you have good hands on experience with: Novell lite-personal, BL>>Novell, Windows 95, Windows 3.11. I'd really like to hear back from you. BL>>Thanks for any replies. Brian, Hi. State your problem or question and who ever feels confident to handle the message will. Shoot . . . Regards. Ahmad Khiary (Jeddah, Saudi Arabia) Internet ID : khiary@netbox.com --- RoseReader 2.52 P007485 --- QScan/PCB v1.19b / 01-0045 * Origin: Saudi BBS. Jeddah. Saudi Arabia +9662 667-2293 (2:542/100) --------------- FIDO MESSAGE AREA==> TOPIC: 193 LAN Ref: DEY00009 Date: 10/28/96 From: AHMAD KHIARY Time: 01:29pm \/To: DIETER VISSER (Read 8 times) Subj: Time to backup Reference your message, DV>> I'm maintaining a 25 user Novell 3.12 network. I want to make backups at DV>>evening when everybody, including me, went home. We are using the program DV>>Arcsolo from Cheyenne and a SCSI tapedrive to make the backups. It runs from a DV>>workstation. DV>> Is there a way to have the backup made automatically at a time DV>>I can choose? I think Arcsolo it self has a scheduler that allows you to run the program at any time with any frequency. I use a program called Sitback which does exactly what you are asking. Regards. Ahmad Khiary (Jeddah, Saudi Arabia) Internet ID : khiary@netbox.com --- RoseReader 2.52 P007485 --- QScan/PCB v1.19b / 01-0045 * Origin: Saudi BBS. Jeddah. Saudi Arabia +9662 667-2293 (2:542/100) --------------- FIDO MESSAGE AREA==> TOPIC: 193 LAN Ref: DEY00010 Date: 10/28/96 From: AHMAD KHIARY Time: 01:29pm \/To: GABRIELE DE GENNARO (Read 8 times) Subj: low level programming Reference your message, GD>>I just wonder if anyone can help me find out a bit more about low level GD>>programming, i would like to control my security system with my computer. GD>>i will appreciate anyone advice. If your security system has a PC interface, then you only need to find how to interact with the port it connects to (serial or parallel). This is the simpler approach. If your system does not have a PC interface, then you need to buy an Analog/Digital converter to handle the signals coming from and going to the security system. What is the system you have ? Don't worry I am too far to do any thing picky to your house . Regards. Ahmad Khiary (Jeddah, Saudi Arabia) Internet ID : khiary@netbox.com --- RoseReader 2.52 P007485 --- QScan/PCB v1.19b / 01-0045 * Origin: Saudi BBS. Jeddah. Saudi Arabia +9662 667-2293 (2:542/100)