The year 2000 has been 12 months of unfulfilled promises.The death and destruction forecast by the millennium bug fortunately didn't materialise: was it the product of corporate angst or did the heroic efforts of the IT departments' SWAT teams really pay off?
Intel promised delivery of the Pentium 4 and Itanium, but failed to deliver until the eleventh hour - well, month, actually. Oracle and Micro-soft both promised grande-architectures, but got no further than the foundations.
The most positive move of the year saw the Linux penguin move up the food chain and become a hunter and gatherer of rock-solid support from the established suppliers, such as IBM and Hewlett-Packard.
Throughout the year, the Internet revolution continued to dominate everybody's thinking. Java became the skill most in demand in UK IT sites, and more than one in three IT jobs advertised now require e-skills.
The Internet is transforming life in general and ways of doing business in particular. Not only is IT evolving as a result of it, but the Internet is also evolving as it develops from an academic plaything to a business tool.
Events also highlighted the need to treat this tool with respect. The "I Love You" bug forcibly brought home how weak Internet security is, as did "denial of service" attacks on major Internet businesses such as Amazon, eBay and Yahoo, and the successful hack into Microsoft's development source code site.
The global nature of the Internet was underlined as it generated conflicts between national legal systems. China continued to try to control Internet access through state-owned ISPs, one man successfully sued provider Demon Internet for libel, while the ability to distribute items such as Nazi memorabilia and pornography also exercised legal brains.
All the time the spread of the Internet increased steadily. The number of UK households online doubled to reach more than one in four by year end. The Web site naming system showed signs of cracking under the strain. Concerted action was taken against cybersquatters, while a whole new series of top-level domains was introduced late in the year.
Technical developments during the year also helped to sustain the momentum. BT took its first step towards the provision of high speed, always-on access with the gradual introduction of Asynchronous Digital Subscriber Line services both for businesses and consumers. It also introduced unmetered access, allowing connection for a flat fee. The company was followed by AOL, but the commercial risks involved were highlighted by Altavista, which first announced an unmetered service and then withdrew it.
Potentially the biggest increase in Internet access was caused by mobile telephone developments. wireless application protocol phones started to appear at the beginning of the year and hundreds of thousands had been bought by the end of it. The vast majority of owners do not use them for Internet access, largely because of their slow speed.
The first step towards making mobile phone Internet access more viable came with the launch by BT of a General Packet Radio Service offering in June, bringing the GPRS benefits of a packet-switched always-on service, as well as a much needed increase in speed.
GPRS is essentially a tinkering with today's technical infrastructure to make it a bit better: the real breakthrough won't come until the arrival of third-generation services in two or three years.
The Government paved the way for the development of these services in a highly controversial sale of the licences, which netted £22bn and sent phone operator share prices plummeting.
Increasing access to, and use of, the Internet is proving a bonanza for the storage companies. At the beginning of the year, industry and analysts alike expected that migration to storage area networks (Sans) would prove to be the dominant theme because the benefits of removing storage directly attached to servers were already clear.
Instead, it was another server-independent storage paradigm, network-attached storage (NAS), which was to prove the money-spinner of the year. Market leader Network Appliance doubled its sales relative to the previous year in every quarter, while EMC reported NAS sales running at 10 times the previous year's level.
Announcements such as this produced a concerted rush to get a slice of the action. Compaq, Hewlett-Packard, IBM and Sun all introduced NAS systems in the second half of the year, while EMC beefed up its own presence in the market with a string of enhancements to its Celerra product and the introduction of an entirely new lower-level system, the Clariion IP4700.
San growth was, in contrast, rather slower than expected, because of the complexity, the need for a special purpose infrastructure, and the absence of many key standards.
Nonetheless, the industry still expects Sans to become general among large enterprises and significant announcements in this area were made during the year by firms as diverse as Gadzoox, StorageTek and Tivoli.
Disc capacities rose to 73Gbytes per drive, while Seagate set new performance standards with a drive that spins at 15,000 rpm. Enterprise level tape also moved forwards with the launch of products based on the competing LTO Ultrium and Super DLT technologies, both offer capacities of 100Gbytes per cartridge before compression.
The rest of the industry was also looking to take advantage of the spending spree generated by the rise in Internet usage. Many companies repositioned their product development in line with major strategy rethinks, led by Microsoft with its .net announcement in June.
Apart from setting its seal of approval on XML as the basis for future development, Microsoft said it was converted to the view that software should be delivered as a service rather than a product. This was the most significant recognition yet that existing IT pricing models are inadequate in the Internet world.
The irony is that Microsoft had until then been resisting the notion of IT as a service provided from the centre. Its historic antagonism to the thin-client concept had put the company at odds with the likes of Oracle and Sun, which have been heavily promoting it.
The rise of the application service provider (ASP) during the year was another reminder that shrink-wrapping products, slapping on a price tag and putting them in the shop window is not necessarily the best way of keeping customers up to date with technical advances.
Conspicuous among the companies that did not announce a new Internet-led direction during the year was IBM. The company came under fire from the financial analysts for a series of less than thrilling trading returns in which the holy grail of double-digit growth remained a long way over the horizon. IBM candidly admits that it cannot foresee how the Internet will evolve and argues that therefore a detailed strategy is rather pointless.
Onestrategic move IBM did make was to rename its whole range of computers as eServers and condemn its established brand names, such as RS/6000, AS/400, and S/390 to the dustbin of history.
Big Blue also made the bold decision to commit to open source as the best means of keeping pace with future developments. This resulted inasteady streamof Linuxannouncements throughout the year.
As well as building Linux features into its 64-bit version of AIX, the company demonstrated the open source operating system on everything from a wristwatch to its top-end mainframes - but failed to convince the analysts or anyone else this was the route to Eldorado.
Linux was, according to all the market research reports, the fastest growing server operating system throughout the year. A closer look reveals it is very strong in two areas: single processor servers and the fast-growing ISP and ASP community. It has yet to make much headway on the desktop - mainly because the user interface compares very unfavourably with Windows or, in high-end servers, because multiprocessor and clustering support is still embryonic.
The challenge of Linux in the enterprise market will strengthen next year and it will be in there fighting with Microsoft, which really moved into this market sector with the launch of Windows 2000 (n‚e Windows NT version 5) at the beginning of the year. The real test will be how well the Datacenter version is accepted in the heartlands of Unix dominance.
During 2000, the top-end server market remained the domain of Unix and mainframes. There were significant changes nonetheless. Technically, the major change was the arrival of 64-bit computing in the mainstream, with launches from HP, IBM and Sun in the Unix arena. Compaq appeared to keep quiet about its 64-bit heritage of Alpha servers and software running back to 1992, pre-dating Compaq's acquisition of original owner Digital.
Commercially, Sun was the big winner, capturing the hearts and minds of the dotcom companies, while IBM was the big loser, dropping to third place in the sector.
Intel lagged behind and general availability of its 64-bit Itanium processor has been postponed until 2001.
IBM introduced its first 64-bitmainframes,the zSeries, in October. But it was a poor year for mainframe buffs. Hitachi announced its withdrawal from the market in March after sales of its Trinium line collapsed, and was followed by Amdahl's abandonment of any future mainframe development in October. IBM itself saw sales fall substantially in every quarter of the year. Intel "mainframes" with up to 32 processors were launched by Compaq and Unisys.
IBM has been arguing the mainframe is the best platform for Web serving, and the technical argument is hard to gainsay. The failure to make this tell in the market is one of the biggest surprises of the year, but a recent deal with Nordic ISP Telia may be a sign of change. As Java rose to the top of the skills league table, Cobol dropped out of the top 20.
The mainframe lost ground during the client/server movement, but it comfortably survived. During 2000, it has fallen from favour faster and further than ever before. The big question for next year is whether this is a terminal decline or an unusually prolonged downturn.
What they said in 2000