LOCAL area networking (LAN) is a fashionable topic in the world of computers today but be warned - satisfying either the whims of your computer salesman or your own ego may prove to be an expensive and pointless exercise.
You do not need to spend thousands of US dollars to share a laser printer between a few personal computers.
You do not need to have your carpets torn up or your ceilings pulled out to install cables to connect a handful of PCs together just to share data once or twice a month.
Although I am a strong advocate of electronic mail, there are other ways to achieve that across an organisation without great expense or inconvenience, often with better effect than with a private LAN.
Understandably, most small company operators balk at the costs of installing a LAN and hence do nothing - and achieve nothing.
It is an unfortunate problem exacerbated by some networking vendors who induce a feeling of shame if one admits to not being able to afford their offerings.
I have never understood why, when a prospect is presented with a quote to connect his PCs that is greater than the entire original cost of the PCs, the vendor is surprised that his quote is rejected.
What is worse is that these insensitive vendors seem never to learn from the experience. The inevitable solution, when they discover they are not selling enough, is to raise their prices.
Companies of all sizes should examine their requirements critically before acquiescing to vendor or management information systems manager pressure and implementing LANs throughout the organisation.
For example, if only one person really needs to use information from overseas headquarters on an on-line real-time basis, then why waste fortunes providing everyone with access to it? Look carefully at who actually needs regular access to local information and when you do, you may find that a complex and expensive network is not what you want.
High speed local area networks are only of any justifiable use when there is a true demand for a multiple of users to transfer large quantities of data across the network.
I believe this is the exception and not the rule, though many over-enthusiastic vendors would not agree with me.
Do not be misled by claims that high speed networks are essential to transfer graphical data.
It is true that high definition graphics take copious amounts of disk space, but do you really need to transfer them around a network? Very rarely in my experience.
What is needed is the ability to transfer simple text data, not graphics, and I have yet to meet the executive who finds speeds of 9,600 bits per second too slow.
At that speed the system is still working faster than the user can absorb the information it is displaying.
But the graphical user interface (GUI) raises its ugly head in this equation also.
It is simply not essential or even desirable to have magical windows being dotted around the network creating overheads that you would not believe.
Take care, because those GUI overheads do very little other than increase the cost of the network.
A factor which recently seems to have been lost is that most business people using PCs would like to have access to lots of information, but do not necessarily want to manage that information or transfer it to their local machine.
I have not met many executives or secretaries who would be happy to accept direct responsibility for the physical maintenance, security and integrity of the main company information files.
Here we are back at a very old concept - a centralised database managed by the people with the expertise to do so, which still demands a combination of working knowledge of the data and the system.
At the top end of this concept on PCs today the PICK operating system stands out as an extremely satisfactory solution for small to medium in-house systems.
It functions on the basis of one central system being shared by many users who may have cheap dumb terminals or PCs.
Do not be fooled by the fact that Microsoft chose not to build a multi-user capability into its MS-DOS.
The average office PC today can handle a multiple of terminals with ease and a quick demonstration of PICK will soon prove it.
There are many products which can provide an MS-DOS multi-user platform, not the least of which is PCMOS.
This operating system emulates MS-DOS but splits the main memory of the central PC, which can amount to many megabytes these days, into a multiple of user-defined partitions.
Don't listen to the so-called experts who claim that today's PCs are not powerful enough for these age-old ideas.
The PC on your desktop more than likely is far faster and has larger and speedier disks than the mainframes which drove 200 terminal airlines reservations systems a couple of decades ago.
Apply normal common sense to your theoretical networking needs and you may find that you could save thousands of dollars and end up with a better result.