The Bloor Perspective: The death of free email, Linux databases and P2P prize-giving

For much of its history the internet has been intimately associated with freedom, and even more with the idea that the majority of services available through it should be free. The philosophy of the early web entrepreneurs was founded on the belief that the internet would deliver commercial benefits as an indirect consequence of the volumes of traffic generated, rather than via the more usual model of 'pay for what you use'.

The three most commonly utilised internet services accessed by members of the public at large are the access service itself, information sources, and email. During the internet boom, these services were all available for free, funded through indirect channels such as advertising and sponsorship.

Even at the height of the dot-com explosion few services managed to generate revenue through these channels in volumes that exceeded their basic running costs. Gradually businesses went bust and free services began to become subject to increasingly restrictive usage constraints.

There has been much debate about the future provision of content without charge. Obviously the vast majority of internet users are extremely reluctant to pay, however, several well-respected publications have already started to differentiate between their free-to-access services and those only available at a price. The ISPs themselves are taking a similar approach in their efforts to sign people up for annual contracts, offering various additional services to make the paid service more attractive. Of the big three functions highlighted earlier, free email is the latest to feel the hot breath of economic reality.

Two of the largest email sites have now announced charges for previously free portions of their services.

Yahoo! this month starts to charge users for checking email from other sites, and Microsoft now charges its Hotmail users for additional email storage space to top up that available for free. It has been reported that MSN users are now receiving exhortations to upgrade to the chargeable service or risk losing email.

With the majority of email sites putting in place increasingly severe restrictions on free email usage, it is clear that the suppliers will encourage the use of premium services.

However, in return for cash, the overall quality of these services must improve dramatically. It looks as if the free internet is on the way out but it will not disappear without a struggle.

Sun scoops P2P prize

The IT world is full of conferences, the content of which ranges from that designed to appeal only to dedicated propeller-heads to those that are nearly technical-content free. The JavaOne conference, recently held in San Francisco, attempted to combine these two extremes.

As with any show of this nature there were a number of awards up for grabs. The winners of major prizes included JBoss for Best Application Server, and Sun's Forte For Java winning the JavaWorld Editor's choice award for Best Java IDE.

Sun has made a number of announcements around Java lately, including products addressing the mobile device space and web services. All companies active in these areas need to ensure that the marketing and business reality do not get too out of step if we are to avoid another spectacular let down.

On an altogether different level, the Most Innovative Java Product or Technology award went to the comparatively unknown Project JXTA from Jxta.org. The JXTA technology is a set of open, generalised peer-to-peer (P2P) protocols, defined as XML messages, allowing any connected device to communicate and collaborate in a P2P manner. Supported devices include mobiles, PDAs, PCs and servers. JXTA peers operate by forming a virtual network where interaction can take place between peer devices and resources directly-even when some of the components are behind firewalls or utilise different network transports.

Despite being formed as a Sun Microsystems research project, JXTA has had community involvement from its outset, encouraging open development and adherence to standards. Everything from the design through to the resulting code is made public. Of all the news from JavaOne, the promotion of Project JXTA to the limelight may prove to be one of the most important.

Can Linux crack the database challenge?

The tracking of Linux is difficult, not least because anecdotal information indicates a much deeper market penetration than press stories sometimes suggest. The simple problem is that nobody really knows how much Linux is in use out there - particularly in the server space.

Measuring vendor sales doesn't always tell the whole story. For instance, last year Linux-loaded servers only accounted for only 8.6 per cent of the total US market for new shipments according to Gartner. IDC puts the figure at a much higher 27 per cent. The actual figure for servers running Linux is likely to be much way higher taking into account older machines and the fact that many companies are not even aware of the extent of their Linux usage.

There is less confusion in the embedded space. Linux is third behind Wind River's VxWorks and Microsoft's MS-DOS, according to Evans Data Corporation, but it is expected to move into the number one position quickly, possibly this year.

It is clear that Linux on the PC is now a serious proposition, with some of the impulse for this coming from government sites across Europe, where standard procurement procedures make it difficult to defend paying over the odds for PCs. There is also a great deal of enthusiasm for Linux in the less advanced economies, primarily because of cost.

At the high-end server market you also see Linux doing well. It has found a life for itself on IBM mainframes, and in the building of super-clusters, arrangements of computers that cluster together hundreds and even thousands of CPUs.

Indeed Linux stretches much further in its range than any other OS. It goes lower and higher and stops at all points in between.

But as yet Linux has not managed to infiltrate large database servers for big transaction systems or big data warehouses. The question is whether there is a problem with Linux or whether the IT departments simply do not contemplate using Linux in such performance critical and scalable roles.

It is probably the latter. Either way if anyone knows of Linux being used on heavy duty database applications, we'd be interested to know.

Bloor Research is a leading independent analyst organisation in Europe. You can find out more at http://www.bloor-research.com or by emailing mail@bloor-research.com.