Only a decade ago the sight of an electronic box on a diver's wrist was guaranteed to draw comment ranging from envy to scorn. Though dive computer development can be traced back to the 1950s, the machines we associate with diving today only had their genesis in the early Eighties. Pioneered by Divetronics' Decobrain, advanced by Orca's Edge and popularised by the original Uwatec Aladin, computers have overcome initial scepticism to become commonplace. They have had a profound effect on how we dive, making table diving all but redundant for many recreational divers. And they promise even greater advances for the future.

Dive computers offer many safety benefits. They can formulate decompression schedules in an instant, should a diver unintentionally overstay his no-decompression limits. This relieves the diver of attempting to locate, read and remember how to use possibly unfamiliar dive tables, perhaps in darkness, at depth, while suffering nitrogen narcosis and probably under stress.

They can also help a diver to maintain a slow and more easily controlled ascent rate without recourse to following the unreliable guide of the 'smallest bubbles'.

Repetitive dives can be calculated without having to have a PhD in maths. And computers remember to treat the late, late night dive and the 7 a.m. pre-breakfast descent as part of the same series, instead of assuming 'new day' equals 'new first dive'. Some machines use a 'smart' algorithm that reacts to provocative dive profiles to try to reduce risk; some can even handle using different gases for diving and decompression stops.

No longer are we bound by conventional square-profile tables that force a diver to stay shallow to gain time, or accept dramatically shortened dive times to gain depth. The multi-level computer diver can descend to 40 m or so and begin a leisurely ascent that could turn a 10 minute no-decompression dive on tables into a dive restricted more by air, stamina and warmth than no-decompression limits. The computer diver can speculatively drop into deeper water for a look, without the fear that not only will there be nothing to view, but that he must now surface immediately to avoid decompression, having largely wasted a dive. Repetitive dives can also be made with far shorter surface intervals and greatly extended no-decompression times compared to following table rules for the same profile.

Still, nothing is for nothing. The user-friendliness and freedom that computers bring must be tempered with responsibility and as full an understanding of the known risks and assumed risks as possible. There is no such thing as no-risk diving, but an educated diver can at least take an informed decision and more clearly assess and minimise the dangers.

The dive computer's brain is its algorithm. Like a dive table, it calculates the diver's gas loading. However, it accomplishes this much more accurately since it does not have to round up figures or use worst case penalties, as tables often have to do to take account of th almost infinite number of combinations of first-dive time, depth, surface interval and subsequent repetitive dive profiles, and still fit all of the information on to an easily read slate! Instead, the time and pressure information is checked every few seconds and compared directly to the precise saturation limits permitted by the computer algorithm. Being a number cruncher, a dive computer can relate gas loading directly, pretty much regardless of profile, to maximise dive time.

Just as different groups of divers choose different tables depending on their likes, purpose or type of diving and agency rules, so there is a choice of computer algorithm. Some computer algorithms provide very generous dive times compared with others. Bottom times without decompression stops for a 30 m dive can run anywhere from nine to twenty-five minutes. Unfortunately, there is no statistical evidence that we are aware of that proves and particular algorithm is demonstrably safer than another. Currently, the sheer costs involved prohibit thorough testing of tables or computers to the point where all possible combinations of dive profile can be said to have been tested to yield, let's say, a one per cent rate of decompression sickness (DCS). In fact it is quite likely that ultimately research will only prove that a no-DCS rate is unachievable and that the best we can hope for is a lower number of hits or a more thorough understanding of DCS, which will enable divers to predict which profiles are most dangerous. Estimates of an algorithm bends rate are based on an approximate number of unit sold, best guesses as to how many are in actual use and assumptions about how divers use them (whether they do even follow instructions and recommendations). This makes for a very grey area for buyers.

The ability of modern computers to track many hours of diving for recall at a later date may help to provide guidance in the future as to which algorithms, if any, do have a clear advantage. It should be possible to monitor the actual dive profiles being followed for proper statistical analysis. The 'best' algorithm may very well depend on the type of diving being done. Certainly such a study will be very beneficial but is probably years if not decades away.

Many computer manufacturers have also tweaked their algorithm in the light of experience, in much the same way as recommendations on how to use a table may change over time.

Ultimately an algorithm, like a table, is just an opinion. Be guided by it, but not enslaved. At times err deliberately on the side of caution.

Computers can be wrist, hose, console or decompression ring mounted. As long as it is easily read, can be turned to show a buddy, and is unlikely to snag in a wreck or drag through the coral, the choice is yours. Consoles can be draped over your forearm during ascents to make it easier to read them and maintain ascent rates, while leaving your hands free to control buoyancy or signal to your buddy. Information may be badly presented because it is crammed into too little space, digits and graphics can be too small to be easily read, information is shown that is irrelevant at the time or several information screens are flashed up sequentially, requiring the user to study them closely to work out what he is actually being told.

Further confusion can be caused when having to read an unfamiliar display, such as the decompression information after accidentally exceeding the no-decompression limits. A confusing display adds to stress and can increase the chance of an accident.

Some displays can be difficult to see under strong or poor light. On a night dive the low light means you can't see the information at all, and a powerful torch simply wipes it out. The answer is a back-lit or luminous display or a low-power back-up light.

Some computers have user-changeable batteries, others must be returned to the store or factory for replacement. Changing the battery requires that seals be disturbed. Factory-sealed batteries need no special care, but user-changeable battery doors should be examined regularly for o-ring wear, debris or leakage. Batteries should be checked prior to a trip. Practice changing batteries if your computer only gives you limited time to make the switch before losing all your information, Air-integrated computers display your remaining tank pressure and usually estimate how long your air or gas will last. This is not just convenient, but can help with decompression management.

Air computers can be connected via a conventional high pressure hose like an ordinary pressure gauge or information can be transmitted via a radio link from your regulator first stage.

Disadvantages to air integration include a possible computer crash for whatever reason, which will knock out your air information as well as your decompression status; problems with transmitters failing (we have reports of this happening around camera flashguns), and higher costs.

Audible alarms are provided on many computers and typically warn a diver that he or she has entered decompression, has risen above the stop ceiling or is ascending too fast. Oxygen toxicity alarms are provided on nitrox computers.

Computers are (excepting Orca's talking Pilot Audio which speaks to the user) passive instruments. They can tell you nothing unless you look at them. Audible alarms are useful, but they should not be relied upon in place of good diving practice. You must check your computer. Remember, alarms cannot always be heard over the noise of heavy breathing or through a hood.

Some computers are always 'on'. When they are in sleep mode the still continue to monitor surface pressure and are aware of ascents to altitude to dive in a lake, for example. They automatically adjust for the lower barometric pressure and the increased nitrogen arriving in the diver's body and adjust the times accordingly. Obviously, for the computer to accurately track the diver, computer and diver should travel together!

Some other computers require the user to set the altitude zone manually. Some divers set a higher level than they need to thereby provide them with a more conservative dive.

Most dive computers assume that you will dive on normal air. Many divers who use nitrox still follow the schedules laid down by the air computer, arguing that this results in a more conservative dive profile. However, the use of nitrox by recreational training agencies has encouraged several manufacturers to offer computers that can be programmed by the user for different nitrox blends. Some machines are aimed at divers who dive on one mix and decompress on another.

Machines designed to handle nitrox have an additional algorithm that monitors oxygen exposure levels and warns the diver if he or she approaches or oversteps the algorithm limits.