CHOOSING SOFTWARE AND HARDWARE

3.1 Introduction

Computers are so common these days that it may seem unnecessary to devote a chapter to software and hardware. However, some readers of this book may have little experience of computers and need a brief introduction, while others may be confused by the technical approach of much computer literature. Although this chapter is partly about choosing a database package, a fair amount of general information is also given. For this latter aspect acknowledgement must be given to Wentk (1996) for his excellent guide to computers; this has been referred to frequently throughout this chapter.

It is essential to bear in mind that a computer system is not just software and hardware: personal expertise is also a vital component. Therefore when choosing software and hardware, it is important at the same time to consider ways of improving your computer skills and to make sure you get the best technical advice and support possible.

3.2 Software first

If you decide to set up a computerised database then you need to find suitable software and hardware. It may be tempting to buy or obtain the hardware first, but this would be a mistake. It is much more important that you spend time choosing software that will do the job you have in mind. Then, once you have decided which software works best, find the hardware to match. This is one of the golden rules of computer buying, as emphasised by Wentk (1996). Software is much more important than hardware in determining how productive your computer will be. Good software should be easy to use and easy to learn to use, while bad software can be a hindrance and may force you to work around it rather than with it. Software dates far more slowly than hardware, partly because when updated versions of software appear they are usually based on earlier versions, whereas changes in hardware are often more fundamental. This is another reason to spend time considering what package to obtain; a good package can serve your needs for a long period of time even though you might change hardware at intervals.

3.3 Database packages: basic choices

The basic choices to be made when obtaining a database package can be shown in a diagram (Figure 3.1). The first choice to be made is whether to look for a general-purpose package or a dedicated botanical package (see Chapter 1). A general-purpose package can be expected to have the advantage of being more flexible and versatile (and possibly cheaper) than a dedicated package, but on the other hand it may be much more complex and therefore take longer to learn to use effectively.

Other choices follow this first decision. If a general-purpose package is required, then the next choice is whether to obtain a flatfile, relational or integrated package. Most, if not all, dedicated botanical packages seem to be of the taxonomic type (see Appendix 1). Some of the characteristics of a few general-purpose database packages are listed in Table 3.1. This is by no means a complete list, but further information can be obtained from computer magazines such as "What PC?" (see section 3.9).

(Figure 3.1 about here)

(Table 3.1 about here)

3.4 Guidelines for choosing a particular package

You may have decided what general type of package you need, but how does one evaluate a particular database package? Some of the chief points to consider are listed below. As it may not be easy to get hold of a demonstration version of a database, you will need to rely to a large extent on personal recommendations and on reviews in computer magazines.

Ease of use

This is of vital importance, especially for relatively inexperienced users. Therefore look for reviews that report favourably on this aspect. Older databases are text-based and these can be difficult to use. The latest types allow you to design data forms (or templates) by moving boxes and other options around on the screen. These databases are easier to use and also usually give better results. It is particularly important to look at the query features of the package -- in other words, how easy is it to get information out of the database?

Help features and manuals

Most modern software comes with built-in help features that you can refer to while working. This means that once you have mastered the basic operations you need not refer to the software manual. Of course mastering the basics may take a fair amount of effort, and it is therefore important that the manual is as clear and user-friendly as possible. For example, try to decide how easy it will be to set up a simple flatfile database from the instructions in the manual.

Costs

Database packages can be expensive, and this is clearly a factor to take into account. You must expect to pay around 100 to 300 pounds Sterling [add equivalent in dollars] for one of the popular general-purpose packages, around 400 pounds for an integrated package, and sometimes considerably more for a dedicated botanical package (see Appendix 1).

These are perhaps the main points to consider, but there are a number of other features of a more or less technical nature that may have to be taken into account, and these are listed in Box 3.1.

------------------

Box 3.1 Some other features to consider when buying a database package

(a) Referential integrity

This means that the software makes sure that each field of every record is filled in and that no entries are left blank by mistake. Not all packages check for this, and without this feature you could experience difficulties with links between tables in a relational database.

(b) Speed

Software varies according to the speed it takes to do the work and speed ratings (or benchmarks) may be given. If you buy a slow package because it is easy to use but want to store a large number of records you may need to obtain hardware that is faster than average.

(c) Menus

Most of today's packages will be menu-driven. It is important to find out whether these suit one's needs, and are understandable at a glance. If many menus are present then these could be a cause of confusion. It would be advisable, therefore, to ask the supplier to set up a small sample database in order to see if the menus are suitable. Some packages offer a "menu-builder" which allows you to set up your own menus.

(d) Precautions against loss of data

The ability of the package to prevent loss of data needs to be taken into account. For instance, does the package allow automatic saving periods and can the user specify how often this should be done? A good test would be to set up a trial database, the loss of which would not affect one's work, and then switch off the computer during various operations. If no damage to the database occurs then one can feel reasonably confident about the safety features of the package. (Other ways of protecting your data are outlined in Chapter 5).

(e) Sub-strings

This refers to the ability (in some packages by the use of "Wild Cards") to find data items in composite fields -- that is where data have not been stored in atomic form (see Figure 1.3). Composite fields should be avoided, but sometimes they are present in data files from another source and then it may be necessary to have a package that can isolate specific data items.

(f) Structural changes

How easy will it be to alter the structure of the database? For example, how easy will it be to remove superfluous fields or to add new ones? More specifically, can this be done once the database contains data, or can it only be done when the database is being set up?

(g) Deleting duplicated data

It sometimes happens that when two or more databases are joined together some records in the combined database are exact duplicates of one another. A useful feature in a package is the automatic elimination of such duplicates, thus reducing the size of the file and making access faster.

(h) Indexing

Most packages will have this feature incorporated. However, it is important that the index is easy to use, and that additional indexes can be made if required. If an index is not made, a search for specific items is performed by the computer on a sequential basis, where each record from first to last has to be looked at in turn. This can take some time. But once the records are indexed, access to these records can be almost immediate. For example, if a set of names is indexed in alphabetical order then the computer will know where to look for a specific name. Indexes take time to build and take up storage space; hence the more efficiently the package handles these aspects, the greater the advantage to the user. Small files of records can be handled quickly by such methods, but as files grow in size the indexing takes proportionally longer. For example, a file twice the size may take four times as long to index. As index-building can lead to storage space problems (and therefore slows the database down) indexing should be handled with discretion and only applied to fields that really need it.

(i) Sorting

It is an advantage to have key fields in a database arranged in order. For instance, if a database of plant species is sorted alphabetically according to species name within genera then it will be relatively easy to look at all species falling within a certain genus. The efficiency and ease of application of the sorting mechanism is critical, especially when large numbers of records are involved. For instance, a database of South African plants called PRECIS (the Pretoria National Herbarium Computerised Information System) contains over 30,000 entries.

(j) Simple calculations

Some ability for calculation may be useful (e.g. obtaining totals) and the ease with which the package handles such operations needs to be taken into account. A separate package should be used for more advanced calculations (see Chapter 1).

(k) Data transfer

This subject was mentioned in Chapter 1. If you need to import or export files of data it is important to discover how easy this will be.

(l) Dimensions of the database

This could be a factor to consider if a large number of fields or a large number of characters per field is needed, although the dimensions of most packages would normally be more than adequate. For instance, dBase allows up 255 fields and up to 1024 characters per field to be incorporated. It also allows 47 fields to be indexed, far more than one would be ever likely to need. There is virtually no limit to the actual number of records that can be stored using modern packages (see Table 3.1).

(m) Degree of programming required

Some packages need no extra programming, while others require the user to learn, to various extents, a programming language or program statements. For many users the former category would be more suitable, but the latter category could be preferable to those with experience of programming, as these packages permit a greater range of potential operations to be carried out.

(n) Language generators

The programming languages and Structured Query Languages needed to manipulate a database are often complicated. Some packages allow the user to generate equivalent languages based on their own needs.

(o) Demonstration databases

One of the features of more recent database packages is the ability to convert a database into a form which can be run on any compatible computer. Such a database can be viewed (but not altered) on another computer without the need for the original software.

(p) Pictorial fields

Various packages are available which can handle the inclusion of photographs, drawings and graphics. Examples of such packages are Visual FoxPro 3.0, Visual dBase 5.5, Microsoft Access 2.0 and Paradox 5. These packages need an appropriate operating system (e.g. Windows 95/Windows 98) and modern hardware.

------------------(End of Box 3.1)

The final choice of database package may depend on factors other than those listed in Box 3.1. Perhaps, for example, the organisation to which you belong already owns a package which you can use and which is reasonably suitable for your purposes. And in other cases it may make good sense to obtain a package that, although perhaps not ideal in all respects, is currently used by other people in your area. This means that if you get into difficulties someone nearby should be able to help you.

Once you have decided which software package to use you can then choose matching hardware. The choice of hardware will depend on a number of questions, such as what operating system is required? How much hard disk space does the package take? How much memory does it need? How fast is it? These questions raise queries about the nature of hardware itself and how to select the right model for a particular task.

3.5 Hardware: basic components and functions

Many people are familiar with computer equipment, but it may be helpful to give a brief outline of components and functions as a prelude to the following sections. A typical desktop computer consists of a screen (or monitor), a keyboard, a mouse and a base unit (Figure 3.2).

(Figure 3.2 about here)

Some information on screens, keyboards, mouse pointing devices and other equipment is given below. All of these components are dependent on the base unit; this contains the hard disk drive, the diskette drive (or drives in older machines) and the main electronic circuitry -- known as the motherboard.

The motherboard represents the heart of the computer: it contains the processor, also known as the Central Processing Unit, a silicone chip that does all the calculations and another chip (or chips) -- known as Random Access Memory or RAM -- that holds the information that the processor uses. Both the processor and RAM work together and determine the power and speed of the computer. Various models of processor are available, and are known by a numbering system that relates to their power (e.g. 8088, 80286, 80386, 80486, 80586 -- the last also called the Pentium); these numbers are often used to refer to different types of computers, as will be described later. RAM, on the other hand, is referred to according to its size in bytes. A byte is the basic unit of information stored by computers, equivalent to a single digit or letter. Modern computers have RAMS measured in thousands or millions of bytes -- kilobytes (Kb) and megabytes (Mb) respectively. The larger the RAM, the more capable the computer will be to handle complex software.

Hard disk drives (often shortened to "hard disk" or "hard drive") and diskettes are used to store data and software. Although the computer memory, or RAM, holds information when the computer is working, all this is lost once the computer is switched off. In other words it is only a temporary memory and any information that needs to be kept must be transferred to the hard disk before switching off. Hence the hard disk can be thought of as kind of filing cabinet: it provides longer-term storage for all data and software and only that needed for a particular task is loaded into the computer memory for processing. Modern hard disks can store huge amounts of information, often measured in gigabytes (Gb), one gigabyte being equivalent to a thousand megabytes. However, it is possible to an extra hard disk if required -- for example in order to upgrade an older machine.

The management of files of information on the hard disk can be quite complicated and you will have to consult user-guides and computer books (e.g. on DOS or Windows) to find out more about this. Procedures can be carried out, for example, to delete unwanted information and to rearrange the remaining information in a more compact manner. This will allow faster access to the information on the hard disk. When this is done it is essential to make copies of important files onto diskettes (or other storage medium; see below) in case any information is lost by mistake.

The diskette drive is used to get information into and out of the computer: a diskette is placed in the drive and by the use of suitable computer commands information is either loaded onto the hard drive or off-loaded from it. It should be noted that new diskettes have to be prepared for the operating system being used, a process called "formatting". Software is commonly supplied on a diskette, or diskettes if the package is large, although Compact Disks (CDs) are increasingly being used for supplying software.

Two types of diskette drive and diskette exist: the older 5.25 inch size and the more recent 3.5 inch size. The storage capacities of the diskettes are as follows:

5.25 inch diskette - 360 Kb capacity

5.25 inch diskette (high density) - 1.2 Mb capacity

3.5 inch diskette - 720 Kb capacity

3.5 inch diskette (high density) - 1.44 Mb capacity (currently the standard type of diskette)

There may be some confusion about the names of diskettes: the 5.25 inch type may be called a "floppy" and the 3.5 inch type a "stiffie" - while, more commonly, the latter type is also called a "floppy". Because of this possible confusion the term "diskette" has been used in this book to cover both types.

Other means of data storage have appeared in the last few years, namely tape-based systems, Magneto-Optical disks and CDs. CDs are fast becoming the standard medium both for holding information and software and for making backups. In order to use CDs you need a CD-ROM (Read Only Memory) drive on your computer and, in the case of making backups, a CD-R unit. These last units are expensive at present but prices are likely to fall in the next few years.

Data storage and protection against accidental loss is a vital issue for database users and is covered in Chapter 5.

3.6 Choosing a suitable computer

Choice of computer depends on what software you have selected. It is therefore important to follow any recommendations regarding specifications for the computer system needed to run the software. The following points need to be checked in order to find out what model to obtain.

(a) Brand

The main choice is between IBM (International Business Machines) or Apple Macintosh, the former making the PC (Personal Computer), the latter the "Mac". Although they use different operating systems, the same kinds of software are widely available for both machines so it may not matter which one you choose. The most commonly available type is the IBM - or the IBM-compatible standard as made by a huge range of manufacturers. Recent models of the "Mac" can also handle IBM operating systems and thus offer a dual standard.

(b) Operating system

The software you select will be designed to work with a particular kind of operating system (e.g. DOS, Windows 3.1, Windows 95, Windows 98). The operating system software is normally sold as part of the hardware as both have to work together to make an effective system. Most modern machines will include a version of DOS as well as one of the Windows operating systems.

(b) The processor chip

The processor chip does most of the work in the computer. The chips are known by their model numbers and this gives an indication of the speed and power of the computer as a whole. IBM machines are in fact known by their chip numbers (e.g. 8088, 80286, 80386, 80486, 80586) and by their sub-class codes (SX, DX or SLC). The sub-classes vary slightly in the features they offer and in their efficiency. All processor chips are also rated according to their speed as measured in MegaHertz (e.g. 25 MHz, 40 MHz, 150 MHz, 233 Mhz). Therefore the model number, sub-class and speed rating of the chip gives a very good idea of the power of the computer and hence whether it is suitable for the software that you have chosen.

(c) Memory

Computer memory (or RAM) has to be taken into account because software is designed to run at particular RAM sizes and will not work at all if the computer's RAM is too small. An unnecessary excess of memory, on the other hand, will add to your computer costs. Memory size is measured in Megabytes and can vary from 2 Mb or less in older machines up to as much as 256 Mb in some models. Most modern software needs at least 4 Mb of memory, but will probably function better at higher values (e.g. 8 or 16 Mb).

(d) Diskette drive

All modern machines have a single 3.5 inch 1.44 Mb diskette drive (this will also take the 720 Kb diskettes). Much older machines have drives for the 5.25 inch diskettes which are now becoming obsolete. As most software is only available on the 3.5 inch diskettes (or on CDs) it is generally necessary to obtain a computer with this kind of drive. It is possible, however, to upgrade some older machines from the 5.25 to the 3.5 inch diskette drive.

(e) Hard disk size

Hard disk size is measured in megabytes or gigabytes. Disk sizes of at least 750 Mb are common in modern machines and offer a huge amount of room for normal database applications. It should be borne in mind that your software will be stored on the hard disk drive and that some modern software can occupy a lot of space. For example, an integrated package can take up over 60 Mb of storage space. On the other hand, some software is much less demanding and a smaller hard drive is quite suitable (e.g. the Demonstration Diskette takes up no more than 2 Mb and a 20 Mb hard drive will suffice).

These, then, are the chief points to consider when choosing hardware for the software you have selected. Although it has nothing to do with software as such it is also worth taking care over obtaining a suitable screen and keyboard. Both of these units are used so frequently that you need to feel comfortable with them. As regards screens (or monitors) various factors have to be considered -- such as dot size, scan rate, screen size and radiation output -- but the main point is to choose a screen that is reasonably safe and unlikely to cause eye-strain. This means that you should check the screen type before buying the computer. The same principle applies to keyboards: several types are available and it is a matter of choosing one that feels right for you.

Other factors which may influence your choice of computer are the support available for maintenance and problem-solving and for basic training. There are clearly a number of decisions to make and it may well be necessary to seek help from a computer-literate friend or colleague -- or perhaps from a computer consultant. The last option can be the easiest and least time-consuming way of buying a computer but is also likely to be expensive. As costs are usually a critical factor these will be looked at next.

3.7 Costs

Costs of hardware -- base unit, screen, keyboard and operating system -- fall into a number of price bands, as described by Wentk (1996). The hardware and their price bands are outlined below, and the main options can be shown in a diagram (Figure 3.3). Please note that although these prices are realistic at the time of going to press, they can only offer a rough guide.

(Figure 3.3 about here)

(a) Vintage (50 - 100 pounds)

These machines are only available second-hand and include IBM models such as the 8088 and 8086. They cannot be upgraded and are unlikely to work at all with modern software. But they are still useful for wordprocessing and for other less demanding packages. The Demonstration Diskette can be run on a machine of this type provided the hard disk size is at least 20 Mb and that a mouse can be fitted.

(b) Dated (100 - 250 pounds)

This represents a step up from the vintage category and include models such as the 80286. As with vintage models, they are only available second-hand. They are difficult to upgrade and unable to cope with modern software. However, they can handle wordprocessing and spreadsheet calculations.

(c) Recent (250 - 500 pounds)

These machines are probably only available second-hand. They will work with most of the latest software but only very slowly (although not suitable for Windows 95 or later). They can be upgraded within certain limits and include models such as the 80386. Machines in this category would be very suitable for operating the Demonstration Diskette.

(d) Entry level (500 - 750 pounds)

This represents the cheapest of the modern machines and includes models like the 80486SX and SX2. They can cope with the latest software (including Windows 95/Windows 98) although they may appear slow for more demanding work.

(e) Established (750 - 1000 pounds)

These machines are considered by Wentk (1996) to offer the best price to performance ratio of any computer: they are very suitable for modern software and can easily be upgraded. Models such as the slow Pentiums (75 - 100 MHz) fall into this price band.

(f) Advanced (1000 - 2,000 pounds)

Machines in this category are intended for demanding software where speed is important. Applications include advanced desktop publishing and graphic design.

(g) Leading edge (1500 - 10,000 pounds)

These are the latest machines available. However, they are far too powerful for personal use and are designed for very demanding software -- for instance of the sort needed for sound and video editing.

The prices given above are for the main hardware only. You also need to allow for insurance against theft or damage, for possible training, for maintenance, for diskettes (and perhaps CDs) and for other equipment -- such as a printer and a Modem (see below).

3.8 Other equipment and options

(a) Mouse pointing units

A mouse -- or hand-operated pointing device -- is an essential piece of equipment for using most modern software. It is usually included with the main hardware of modern computers but may not be present with older machines. They are relatively inexpensive (10 - 80 pounds) and as they are used such a lot it is worth buying a reliable model. When purchased a mouse comes with a diskette of software that must first be loaded into the computer in order to allow the mouse to function (see Chapter 7). Note, however, that Windows operating systems include their own mouse software.

(b) Printers

A printer is an important extra piece of equipment, not only for producing documents but for making copies of database records as a safeguard against accidental loss in the computer. There are three main types of printer: dot matrix, ink-jet and laser.

Dot matrix printers do not produce a high-quality output but are cheap to run; they are therefore particularly useful for making backup copies of database records and reports. Dot matrix printers come in two basic models: 9-pin and 24-pin, the latter giving better print quality. Costs of basic machines lie in the region of 150 to 400 pounds.

For certain documents and for graphics an ink-jet or a laser printer will probably be required, especially if a high quality of presentation is important. Ink-jet (or bubble-jet as they are also known) printers give reasonable results at moderate cost (e.g. 200 - 400 pounds), while laser printers offer high quality print but at higher purchase cost (300 - several thousand pounds) and increased running costs.

It is important to obtain a printer that is suitable for your software and hardware. There is no point, for example, in buying a printer which has printing options far in excess of the capabilities of the software being used. Alternatively, make sure that the software you obtain has a "driver" for the printer of your choice.

(c) Modems

You will only need a Modem if you intend to use the Internet or otherwise exchange information via the telephone line. A Modem -- a contraction of the words modulator-demodulator -- is able to convert digital information from a computer into audio signals that can be transmitted by telephone and received by another Modem and computer elsewhere. Extra software is needed to operate a Modem and, apart from Internet use, the help of a specialist will probably be needed in order to transmit and receive information by this means. Modems work at different speeds. Very slow Modems, although cheap (about 25 pounds) may turn out to be a false economy because of higher telephone bills; it may therefore be better to spend more on a faster model (e.g. 14,400 bits per second or faster). In any event, it is important to obtain good technical advice before buying a Modem.

(d) Voltage regulators and lightning protectors

Where the electricity supply is erratic -- for example when a small generator is being used -- it may be necessary to install a voltage regulator to protect your computer equipment. And if you use a Modem it is a good idea to fit a lightning protector to help counter any lightning strikes entering by the telephone line. However, power surges coming down the electricity supply during a lightning strike can also cause severe damage to computer components, even if the computer is switched off. The safest thing to do, therefore, is to unplug your computer completely and (if necessary) the telephone line whenever there is any chance of electrical storms occurring.

(e) Portable computers

Portable computers (also known as "laptops") are an option to consider when working away from the home or office. The main disadvantages are that portables tend to be much more expensive than a desktop with similar specifications and that battery life is limited (no more than a few hours of use at present). However, battery technology is improving and it is possible that portables will eventually replace desktops to some extent. Such portables could be of great advantage to botanists working in the field.

3.9 Training courses, books and magazines

Although it is possible to teach yourself how to use software and hardware many people find training courses very useful, if not essential. Computer training can either be directed at general computer literacy or at learning to use specific software. The former kind of broad training is probably more useful, especially for beginners. Information about computer training courses can often be found in local newspapers and in the "Yellow Pages" of telephone books.

For those wishing to teach themselves there is no lack of computer books and magazines; the problem is to know what to choose. Wentk (1996) gives some useful guidelines and provides a survey of the magazines currently available.