Frequently Asked Questions

(Home)

• What does the BCS in BCS Technology Limited represent?

BCS is an acronym for Bill's Custom Software.  Originally, what BCS Technology Limited's founder did was develop custom software.  (Back to Top)

• Why should I purchase from BCS Technology Limited instead from one of those retailers who advertise on TV?

Because we add value that retailers can't provide and we're there when you need us most.

The prices quoted by retailers are mighty tempting, especially if you are just starting up your business and are trying to minimize expense.  However, you only get what you pay for.  What you don't get for the low price is quality equipment or expert technical service when you need assistance in setting up your system or solving unforeseen problems.  When was the last time a retailer came to your business to help you out of a jam?  And how many times have you talked to telephone technical support without actually resolving anything?

When you purchase from BCS Technology Limited, not only do you get a top-quality package tailored to your specifications, you get the help you need when you need it—from someone located right in your area, not on the other side of the world.  And if you are not able to set up your system we can do it for you.  When was the last time a retailer offered that kind of service?  (Back to Top)

• Why are computers so complicated?

Because users continue to demand more power, more flexibility and greater ease-of-use.

Ease of use accounts for much of the complexity, as any computer technology that is easy to use is inherently difficult to design, build and configure.  Also, there are certain vendors who don't want to adhere to long-accepted industry standards, which grossly—and unnecessarily—increases the complexity.  (Back to Top)

• Upgrading is such a hassle!  Do I really need the latest technology?

Not unless your system is really out of date.

If what you have is performing adequately and has all the features that you need, why replace it?  Unfortunately, there are computer industry vendors who depend on businesses constantly upgrading or replacing their systems to generate continuing revenue.  These vendors will often spread misleading information about system requirements so as to plant FUD (Fear, Uncertainty and Doubt) into the minds of computer users.  The fact is that any reasonably good system purchased in the last several years will run most of the currently available software, although not necessarily at an optimum level of performance.  If your technology can't do what you want then you should consider upgrading.  (Back to Top)

• Which processor is better: AMD's or Intel's?

It depends on how you define "better."  Intel's design has been optimized to support high clock speeds, which sounds good in sales literature and TV ads.  In some cases, the processor's high clock speed will accelerate the performance of certain types of software where moving data around in memory is the main task to be executed.

In contrast, AMD processors may run at a lower clock speed, yet get more work done per clock tick.  In fact, at any given clock speed AMD processors generally will be faster at number manipulation and will tend to outperform Intel products in many graphics applications.  If you factor in purchase cost, an AMD-powered system is usually "better" because you get more performance for your money.  (Back to Top)
 
• So you're saying clock speed isn't an accurate measure of performance?

That's right!  Comparing the performance of one system to another using clock speed is analogous to comparing the performance of two different automobiles by measuring how high the engines can rev.

There are numerous and often conflicting factors that determine overall system performance.  On average, Intel's processors may run at higher clock speeds than the AMD equivalents.  Independent performance testing has generally shown little difference between the two processors in the software typically run by most small businesses.  Where the differences generally appear is in number-intensive applications like spreadsheets and graphics applications (e.g., CAD).  AMD's more powerful floating point numerical engine generally gives their processors an edge in such cases.

It's important to understand that the processor is only part of the picture.  Perceived performance is affected by the rate at which data can be read from or written to the hard drive, the speed at which the video graphics controller can drive the display and a host of other subtle factors.  For example, Intel's advertising claims that used to assert that the Pentium 4 processor would "enhance your Internet experience" were nonsense, as the limiting factor in Internet performance is the speed of the connection to your Internet Service Provider (ISP).  It is highly unlikely that you will see any difference in Internet performance between the latest computers and older machines with slower processors.  In any case, purchasing a system on the basis of processor clock speed is not a good idea!  (Back to Top)

• What makes a server so special?  Isn't it just a PC running a server operating system?

Using an ordinary PC as a server is analogous to using the family sedan as a delivery truck.  A true server is designed and built for the job, and often bears only a superficial resemblance to a PC.  Here are some of the attributes you should demand in a server:

• High performance microprocessor and mainboard.  The resources of the server will be shared by everyone in your organization, and possibly by others in remote locations.  For this reason you should invest in as much performance as your budget will allow.  The microprocessor and mainboard will often dictate the upper limits of the machine's capabilities and thus must be carefully chosen.  In some cases, multiple processors may be required to handle the expected load.

• Lots of memory.  In any server, memory is a lot like money: more is good!  Data access from memory can be up to ten thousand times faster than from the hard drive.  You don't need to be a computer expert to understand that the more data that can be kept in memory the more responsive the server will be to user requests.  Incidentally, only error correcting memory should be installed into a server.  Non-error correcting memory (the type typically found in a PC) may trigger a system crash if a so-called "soft" error (occasionally caused by background radiation) occurs and goes undetected.

• Capacious mass storage.  The volume of data that many servers are expected to handle can readily overwhelm the type of mass storage installed in a typical PC.  This is to be expected, as the server's storage will be shared by everyone who has access to the system.  Your server should be equipped with one or more SCSI (Small Computer System Interface) hard disk drives, which are designed to efficiently process the multiple simultaneous data requests encountered in a shared environment.  In some cases, it may be wise to specify RAID (Redundant Array of Independent Disks) mass storage, where the data is spread out over several hard drives to minimize the likelihood of loss in the event of a system crash or hardware failure.

Don't let someone try to convince you that less expensive SATA storage is just as good.  If this were true no one would be using the more expensive SCSI hardware.  As always, you only get what you pay for.

• High Capacity Data Backup.  Modern hard drives are very reliable.  However, they are not infallible.  A hard drive failure will usually result in total loss of data—including the operating system itself—which for many businesses would be disastrous.  The only means by which protection from data loss can be assured is by performing frequent data backups and storing the backup media in a safe place.

Here again, the volume of data that many servers must manage demands a backup solution that is up to the task.  This means either a four millimeter DDS or LTO cartridge tape drive should be used, along with redundant sets of cartridges.  If the data volume exceeds the capacity of any one cartridge an automatic tape loader should be specified.  An external hard drive is not a suitable backup device, as it can fail for the same reasons that the server's hard drive(s) would fail.  Also, achieving redundancy would require the use of multiple external drives, which would simply not make sense from a cost perspective. 

Along with a suitable tape drive, reliable archiving software is a must. We recommend the use of Microlite's Backup Edge® archiving software for Linux and UNIX environments.  Backup Edge is designed to work with a wide variety of tape drives, as well as other writable media (e.g., DVD-RAM), and has numerous features to verify that backups are reliable.

• Ample power supply and cooling.  Most servers are expected to run around the clock.  This so-called "24 by 7" operation isn't possible if the server's power supply and cooling system are marginal, as electrical and thermal issues—the silent killers of computer hardware—may trigger system instability or outright failure.  Features like RAID and multiple processors greatly add to the electrical and cooling load, demanding that adequate capacity be available at all times.

• Field reparability and upgradability.  Any server you purchase should be built from industry standard components that can be readily replaced if failure occurs.  Also, it should be possible to field upgrade the server in the event your computing needs change and more processing capacity is required.  Unfortunately, many name brand servers are built from non-standard components that are not available from any source other than the original manufacturer.  In an emergency, you may find your operation shut down by a server that cannot be quickly repaired due to parts unavailability.

All of our Uni-FI servers possess these attributes, assuring you that your technology investment will be a sound one.  (Back to Top)

• How can a network help my small business?

Setting up a local area network (LAN) will allow your employees to share expensive computing resources, such as disk storage, printers, Internet connections and so forth.  Over the long haul, such sharing will save you money, as less needs to be invested in hardware to assure that everyone can work uninterrupted.  For example, instead of placing a printer at each person's desk, you could set up one or two shared printers.  In the former case, each printer would probably spend a substantial amount of time not being used—which is wasteful.  With shared printers, you will get more service in exchange for your technology investment.  Plus with fewer printers on your system, you will have fewer printers to maintain and repair.

A significant benefit of a LAN is the ability to share an Internet connection through broadband or DSL.  With such an arrangement, you get more value from your Internet service expense, as multiple users are able to simultaneously go on-line at any time.  Also, productivity can increase due to the much quicker response of broadband or DSL Internet access.

There are two basic types of networks suitable for small businesses installations: peer-to-peer, which is a network with no server or other controlling entity; and client/server, in which workstations (the clients) are under the control of a server computer, which makes resources (files, programs, printers, etc.) available to the clients on a demand basis.  A client/server system can be described as homogeneous, meaning all machines on the network run the same basic operating system software (e.g., Windows); or heterogeneous, in which the clients and server run different operating system software (e.g., Windows or MacIntosh clients communicating with a UNIX or Linux server).  (Back to Top)

• Which is better: peer-to-peer or client/server?

As with the "which microprocessor is better" question, your perception of "better" is dependent on how well you understand the differences between a Windows peer-to-peer network and the client/server version.

For a small office with two or three machines and no link to the outside world (i.e., the Internet) a peer-to-peer network may be entirely satisfactory.  A peer-to-peer network is readily configured with a minimal amount of additional hardware.  All of the required software is built into all current versions of Windows, eliminating the need to purchase additional software.  With proper configuration, users will find the resulting environment quite friendly, as they can "point and click" to get to resources.

Unfortunately, in a Windows peer-to-peer network, security is enforced by individual workstations and anyone with access to a workstation can create a log-in account for themselves.  In simple terms, there are no unauthorized users.  Obviously, if the system is attached to a shared Internet link (e.g., broadband or DSL) any reasonably skilled intruder could gain access to resources.  You should also know that the crash of any one workstation can temporarily disrupt the entire network, especially if the crashed machine has resources needed by other workstations.  And because there is no server, performance will rapidly deteriorate as more machines are added, which means the network will not scale well.

Most installations are best served by a client/server network so as to avoid the peer-to-peer weaknesses.  Although the initial cost is higher, and configuration and administration are more technical in nature, the gains in performance, reliability and scalability are generally worth the extra expense and trouble.  (Back to Top)

• Why would I want to run a heterogeneous client/server network?  What's wrong with using Windows on everything?

It may surprise you to know that you are already using a heterogeneous network (the Internet) as you read this page.  Of course, that hasn't answered your question, so perhaps an analogy might be useful.

We all know that the Serengeti in Africa has a richly diverse animal and plant population.  There are carnivores, herbivores, insectivores, predators, prey, vegetation of all kinds, and so on.  Scientists know that such diversity is essential to the long-term survival of all species.  In fact, this type of ecosystem—a heterogeneous mixture of animals and plants—tends to be self regulating and self preserving.  Disregarding the effects of human beings, only a catastrophic event, such as a meteorite impact into the region, would threaten the long term viability of the system.

Now imagine how the Serengeti would be with only one species of animal—wildebeest, for example—and only one kind of vegetation—a homogeneous environment.  There's no question that the bovine-like creatures would have an easier time searching for food, as there would be no other herbivores competing for the same vegetation.  Since there would be no lions or other predators to avoid, the population could multiply much more rapidly than it does now.  However, consider what would happen if one of the animals contracted a contagious fatal disease.  The disease would quickly spread to and decimate the entire population, as no one wildebeest would be able to resist the disease's effects.  Or suppose something fatal to the vegetation were to wipe it out.  The herd would expire from starvation.

A homogenous Windows network, either peer-to-peer or client/server, is like our theoretical "one species" ecosystem.  "Propagating" the network is relatively easy because of the homogeneity—everything works the same.  However, the resulting environment will be such that viruses or other malicious software will be able to readily attack both the clients and the server with impunity, simply by exploiting the "all Windows" aspects of the network.  Simply stated, if your network were homogeneous, it would be possible for a single virus to knock out your entire system.

In contrast, if your network were heterogeneous—supported by a UNIX or Linux server—malicious software might be able to clobber the PC's.  However, due to the way in which UNIX and Linux operate, it would be highly unlikely that the software would be able to successfully attack the server.  The fact is the overwhelming majority of viruses in circulation today attack Windows, primarily because the Windows environment is easily attacked.  In contrast, the number of successful attacks that have been mounted against the UNIX/Linux environment is so small no one has ever seen a need to develop any kind of anti-virus software to deal with the "problem."  (Back to Top)

• What is required to set up a heterogeneous client/server network?

Aside from a local area network (LAN) and some PC or MacIntosh® workstations, you will need a UNIX® or Linux® server (such as one of our Uni-FI Ultra models) running the Samba® Windows-to-UNIX filesharing software suite.  Once the UNIX server is up and running, and Samba has been installed and configured, the server appears on your LAN as the Windows primary domain controller.  All required services, including domain login, file sharing, centralized printing and roaming profiles, are made available to client workstations in a totally transparent manner.  Samba is so good at doing this your users will have absolutely no idea that they are not "talking" to an actual Windows server!

Unlike Microsoft's Windows 2000/2003 server software (or other Windows file sharing solutions for Linux and UNIX), there are no per seat licensing costs associated with Samba.  In fact, the number of simultaneous users that a Samba installation can support is limited only by the server's ability to handle the processing load.  If the load becomes too much for any one server, several machines, all running Samba, can work together to share the load and maintain performance.

Speaking of performance, tests by several independent organizations have shown that Samba running on UNIX or Linux typically responds faster to service requests from client workstations than Windows 2000/2003 does running on the same hardware.  Also, because Samba utilizes the underlying UNIX/Linux operating environment to manage files, printers and network connections, the resulting system is much more reliable than Windows 2000/2003.  This reliability, especially under heavy load, is legendary in the computer industry and has led us to dub heterogeneous client/server networks running on Linux or UNIX crash-proof.  (Back to Top)

• What's the difference between broadband and DSL?  Which should I use for my business?

Firstly, the term "broadband" actually refers to the electrical characteristics of the data link, not the service itself.  In itself, this is of no concern.  The real differences are found in the medium by which services are delivered.

The service typically referred to as "broadband" is delivered via CATV (Community Access TeleVision), using a band of frequencies not allocated to television reception.  DSL is delivered through the PSTN (Public Switched Telephone Network), usually over existing telephone wiring, and employs bandwidth or data capacity not used to carry telephone conversations.

CATV is a shared medium, in which performance will tend to degrade as more subscribers are attached to the system.  Think in terms of everyone in the neighborhood tapping into a local water tank through a single pipe.  Since the pipe's capacity is finite, as more faucets are opened less water becomes available to each home.  Also, CATV is easily disrupted by many external influences, resulting in less than sterling reliability. Again using the water analogy, if a break occurs in the shared pipe no one gets any water.  Needless to say, this is a very undesirable characteristic if your company is dependent on reliable Internet service to conduct business.

In contrast, much of the part of the PSTN used to deliver DSL is not shared media, so performance is much less affected by the number of subscribers attached to the system.  Using the water analogy once more, each DSL subscriber's faucet is directly piped to the water tank, which means the flow isn't reduced as more faucets are opened by others in the neighborhood.  Also, the PSTN in the USA and Canada has become highly developed, resulting in consistent performance and excellent reliability.  For this reason (as well as other more technical ones) we generally recommend the use of DSL for business Internet access.  If possible, you should arrange for SDSL (Symmetric DSL), whose performance is much better than ADSL (Asymmetric DSL), especially if your system is to be remotely accessible.

A properly configured DSL link can provide excellent performance at an attractive price.  However, keep in mind that such an arrangement provides a fixed bandwidth that must be shared by all the users in your business.  Most business DSL providers can offer higher bandwidth if needed, generally up to a maximum of about 1.5 megabits per second (Mbs, about 180 kilobytes per second usable data transfer speed).  1.5 Mbs can be expected to support about 20 simultaneous users and still produce reasonable performance.  If you plan to have more than  20 users accessing the Internet we recommend that you consider a T1 connection to the Internet.  (Back to Top)

• Can I get access to my office files from another location?

Yes, if certain conditions can be met:

• Your system should be client/server, not peer-to-peer.  While it is theoretically possible to allow remote access to files on a single PC, such an arrangement would be akin to leaving the front door open to your business and posting a flashing sign saying:

Welcome, burglars and vandals!

• Your server must be accessible via the Internet using a static IP address.  Your Internet service provider must be willing to allocate at least one static IP address (Internet Protocol address, a number that uniquely identifies a computer on the Internet) for assignment to your server.  "Static" means the IP address is guaranteed to not change from one session to another.  Without a static IP address, it will not be possible to remotely contact your server, just as it would be impossible to call your office if the phone number constantly changed.

T1 and business DSL providers will assign a block of static IP addresses upon request.  The policy with broadband operators varies—some small local or regional operators may not be willing to assign static addresses.  In any case, an interconnecting device called a router will be required to link your server to the Internet (think of a router as a kind of traffic cop that keeps traffic moving in the right direction).  It may be possible for the server itself to act as the router.  Otherwise, we can provide and configure a suitable router.

• The server must be configured to authenticate remote log-ins.  This requirement is inviolate: unless the server authenticates the remote user no access will be granted.  While on the subject of authentication, all users should be assigned secure passwords.  Otherwise, anyone who manages to guess a valid username could gain unauthorized access to the server and all that it has to offer.

If the above requirements can be met, we can give you remote access to your office files and other server resources from almost anywhere in the world.  Please contact us if you are interested in having remote access to your system configured.  (Back to Top)

• Why should I use Mozilla's Open Source Firefox web browser instead of Microsoft's Internet Explorer?

For several reasons: we think Firefox is technically better software, both in terms of ease-of-use and features, including features that are not present in the browser.  Firefox doesn't get into your face all the time, dragging you over to the Microsoft website or surreptitiously downloading and installing "enhancements" without checking first to see if it's okay to do so.  We dislike any software that automatically does things to a system without being given the opportunity to say "no" to proposed changes.

Most importantly, Internet Explorer and its inseparable companion, Microsoft Outlook, can expose your system to attack by defective or malicious code embedded in web pages or E-mail files.  This is because these programs are entwined with the guts of Windows (a fact that was behind a lot of the wrangling in the antitrust suit brought against Microsoft a number of years ago by the U.S. Department of Justice).  If something attacks Internet Explorer or Outlook it will attack Windows itself, possibly resulting in system fatality (yes, that's as serious as it sounds).  With Firefox, the attack will most likely only kill Firefox, leaving the rest of the system in a functional state.  (Back to Top)

• What is transaction processing software and why would my business need it?

Transaction processing software, also refered to as "vertical" software, is software that is designed to automate your business.  Transaction processing refers to repetitive tasks that most businesses perform on a day-to-day basis, such as order entry, accounts payable, customer database management, and so forth.

Many aspects of doing business are similar in all fields of commerce (e.g., accounts payable and accounts receivable are nearly universal).  Other aspects are peculiar to a particular area.  For example, an automotive repair facility would need software designed to manage workflow through the shop, as well as to write work orders, estimate jobs, and maintain a history on each vehicle so customers can be notified when it's time for the vehicle to be serviced.  On the other hand, an insurance agency wouldn't need any software related to running a shop, but would require software that could maintain detailed records about clients, the types of insurance they have purchased, covered assets, claim history, and so forth.  It is these specialized aspects of the software that makes it "vertical."

Almost all businesses who sell a product or service for which detailed records need to be maintained are candidates for vertical software.  It is possible to run a business using stand-alone software, such as a word processor, spreadsheet and database manager.  Many small business owners, often under pressure to quickly get something going computer-wise—as well as limit expense—have done so.

However, they have also discovered that it can be very time-consuming to set up everything to support what is needed and that it can be difficult or impossible to properly integrate the data from standalone packages into a cohesive processing environment.  This last problem occurs because few industry-wide data interchange standards exist in the Windows software industry (even Microsoft's products don't always interchange well), and those that do exist are often ignored by vendors who don't want you using their competitors' products.  Adding to the woes, some software doesn't always behave well in a networked environment, which can lead to corrupted files and data loss.

For some fields of commerce there is a large body of vertical software from which to choose.  For example, the operation of an insurance agency or real estate office is such that off-the-shelf or "canned" vertical software will often fulfill all needs.  The same is true for warehousing and distribution operations where conventional methods are employed to manage inventory and process shipments.  However, many businesses require special features to manage their operation and data.  For these companies, custom transaction processing software, while initially more expensive, is ultimately the best and most cost-effective solution.  (Back to Top)