Upgradability

Commercial pressures coupled with constant improvement in hardware performance and capability lead software manufactures to regularly upgrade software. The upgrade may provide functionality identified as necessary by users or which is present in a rival version or it may take advantage of improved processor speeds or increased memory or new hardware features like a larger screen size in a new generation of hardware. A new operating system for a particular hardware platform may result in upgrades of applications software to take advantage of the new operating system's features.

Since the purpose of providing software upgrades is both to gain new customers while retaining existing ones, upgrades are usually upwards compatible. That is to say that the new version of the program will be able to process files created by previous versions. The reverse will not always be true.

Files created using an earlier version of the software may need to be converted if the upgrade uses a different file format. This will normally happen automatically the first time the file is accessed by the upgrade but it may be a time consuming process. Problems can arise if the conversion algorithm fails to work properly in all situations.

Other problems can be experienced when upgrading software. The existing hardware platform may either be insufficient for the upgrade version or else run it very slowly. The upgrade may require more disc space than the older version. If the user interface has changed then experienced users may take time to adjust so that initially the upgrade is used less efficiently than the previous version. In addition, if production of the upgrade has been rushed for commercial reasons, it may actually contain more bugs than the version it replaces.

Software firms often encourage customer loyalty by offering upgrades of existing packages on favourable terms. Another marketing strategy is to offer older version of the software at reduced prices and aim to make money when the customer upgrades to the latest version.

Reliability

All but the most trivial items of software can be expected to contain bugs. The programs are so complex and produced under such time constraints that this is inevitable. On the whole, users are generally satisfied if the software broadly behaves as expected. They accept the occasional failure philosophically. Indeed in some cases, where a known sequence of operations causes the software to fail, they will simply avoid that sequence of operations.

Software firms attempt to remove the bugs from a program before it is launched. This is done by testing the operation of the program or computer system before putting it on sale or commissioning the system. In the case of software testing is carried out. Here a pre-launch version of the software is given to a number of potential users to test in the field. This process has become so common place that it has it's own terminology Where an alpha version is tested in the software house itself, then after correction a beta version is given to selected "partner" companies. These companies have the benefit of seeing the software first and my get cheaper final copies. In return they test the package and report faults. When a package is finally released it makes sense to wait six months before purchase and let other people iron out the glitches!

© LEV