I am going to take this column to present my thoughts on what it takes to make an informed decision when buying the right computing hardware for a job.

I do this because I am certain that companies all over the world are spending lots of time trying to figure out what to get. After all, purchasing or replacing mobile hardware is a very expensive proposition, and especially so when the hardware is just part of a larger overall system, and even more so when you consider the overall impact the entire system can have on a company's bottom line. Pick the right stuff and it's a win-win situation where the people using the system are happy, where everything works, and where the results contribute to the bottomline by saving money and possibly opening new markets. Make the wrong decision and you'll be saddled with unreliable equipment that is not up to the job, a frustrated workforce, a potentially disastrous impact on the financial bottom line, and possibly even lawsuits over broken promises, errors, or faulty work.

Unless you start totally from scratch, which few companies do these days, you may already have relationships with vendors that you know and trust. That may or may not be a good thing. The good side is that they know your business. You know them and their performance. And they may be able to save you money by already knowing what you need. And if new or replacement equipment fits into systems already in place, you may also save on training and start-up costs. A good existing vendor will also be able to save you money by making good use of parts that do not need to be replace, and systems conversion costs will likely be much lower.

The flipside of sticking with a current vendor is that they may try to get you to buy whatever they sell, are familiar with, and make most profit on. This may lock you out of considering entirely new solutions that may be better and more cost-effective. Instead you may continue upgrading outdated systems and technologies while contributing to the bottom line of a resistant vendor and not to yours.

By far the best approach is to go through a well-designed RFI (Request for Information) and then RFP (Request for Proposal) process. Developing one can be quite a job and may well require the assistance of a good consultant. That's because inhouse IT is often resistant to change itself and tends to favor current vendors whereas a consultant will be more impartial and up-to-date on emerging technologies.

However, even the most openminded, forward-looking RFI and RFP developer can be hard-pressed to make some of the preliminary decisions needed to narrow down and focus the search.

So let's look at some of the basic issues that can make or break a system.

Rugged versus semi-rugged versus durable. That's one of the decisions you need to make early on. Unless you only use computers in an office, chances are you'll want "rugged." But truly "rugged" machines come at a high price. Do you really need rugged, and if so, how rugged? Don't automatically be turned off by the terms "semi-rugged" or "durable." And educate yourself as to what exactly those terms mean. One vendor's definition may be quite different from another's. There are several standards and testing procedures that can give you a clue as to how much abuse a piece of equipment can take. Read our detailed article on page 62 of this issue to get an idea. What you need to do is figure out, realistically, what degree of ruggedness you need and provide that information to vendors. If you spell out the ruggedness specs, they will either be able to fulfill them or not. And guard against asking for too much. It simply adds bulk, weight, and cost. No need to get a Hummer when a Jeep will do.

Form factor. Computing has come a long way. Today's smallest handhelds have as much or more computing power as Apollo 11 had to get us to the moon. So don't automatically rule out a handheld system because you think it's not powerful enough. The problem with handhelds is not computing power, but whether or not a) your application runs on it or can economically be ported to it, and b) whether the display is large enough to show all the information needed. Handhelds have many advantages over larger systems. Besides their lesser size and weight they cost less, they are usually simpler to operate, they generally have longer battery life, and since wireless communication is available even in small handhelds, the lack of massive internal storage may no longer be an issue since the device can simply download information. On the other hand, the desktop version of Windows has become such a ubiquitous, dominant computing platform that it's increasingly more difficult to find applications running on anything else, or porting a huge Windows application to a smaller system, even if it's one of Microsoft's.

Displays. Picking the proper size and type of display for an application is absolutely crucial. This is especially true when it comes to outdoor-readable displays. Display technology has advanced tremendously and it is all too easy to be seduced by those gorgeous large TFTs we all have in our notebooks. Problem is that those transmissive displays wash out outdoors where a mobile workforce needs them most. That's because they get their vibrant picture from a strong light behind a non-reflective LCD. Outdoors, that light is no match for daylight and the sun, and the display becomes unreadable. There are many attempts at addressing this predicament. Some displays reflect sunlight which makes them good outdoors but almost unreadable indoors. Others are part transmissive and part reflective. Those "transflective" displays are a good compromise, but only you can decide if you can live with a compromise. Yet other vendors use reflective coatings. Some work quite well while others do not. Make absolutely certain you think this through, carefully examine displays under all lighting conditions, and only then make a decision. And include actual users in this decision process.

Notebook or slate. Assuming you have determined you need full Windows computing power, should you go with a conventional notebook or a slate computer? Notebooks have become quite versatile. Not only can you now get them with tremendous performance and very large displays, you can also get versions that are ruggedized to various degrees, and some are even "convertible" into tablets. Many of the new Tablet PCs are actually notebooks whose displays can be rotated and then folded down so that the LCD looks up, allowing you to use the notebook like a pen slate. However, few of them are very sturdy, and you always have to carry around the system unit and the keyboard. So a pure slate may be a better solution. Problem there is that you better make sure you can live without a permanently available keyboard. Many applications work very well with a pen and an on-screen keyboard. Having a small, light slate that you can simply pop out of its cradle and take along for a meter reading or accident report may be just what you need. But, again, make certain to think through all the requirements of your application before you make a decision.

Wireless communication. This is one area where we are nearing a Golden Age. In the past, availability of certain wireless networks often made or broke a deal or application. Today you generally have your choice of a variety of very sophisticated wireless networks with excellent coverage to boot. Many devices come with two or even three wireless systems built in. You may have Bluetooth for an occasional cellphone connection or to print, 802.11b or g WiFi for high speed communication within a building or in the vicinity of a hot spot, and one of the rapidly emerging 3G networks for data transfer when in the field. The trick here, again, is to figure out what you need and then to find a good fit between a network technology and your application.

Performance. With the great advances in processing speed and storage one might think performance is no longer an issue. Unfortunately it still is. While chips are much faster, the software they need to drive has grown even faster yet. A fully loaded Windows system can easily bog down even the fastest processor. It is generally a good idea to run the leanest possible OS/software setup and then find a processor that's fast enough for it without compromising too much on battery life. So trim all the software overhead and fat!

Common sense. You have it, I am sure. But it's often amazing how little of it is applied when making important, costly decisions that can have a huge impact. So take your time, think things through, read Pen Computing Magazine, and then decide. -

Conrad H. Blickenstorfer, a former corporate CIO, is editor-in-chief of Pen Computing Magazine and Publication Director of Digital Camera Magazine. He can be reached via e-mail at cb@pencomputing.com.