Your shout! On how to stop the deluge of spam e-mail

In response to Peter Kidson (Letters, 10 August), who said he has completely eradicated spam with a program that uses a combination of whitelists and blacklists.




Have your say on



On how to stop the deluge of spam e-mail

In response to Peter Kidson (Letters, 10 August), who said he has completely eradicated spam with a program that uses a combination of whitelists and blacklists.

If the product Peter Kidson mentioned gets a 100% strike rate on spam, I am most pleased for him, but it strikes me as unfeasible.

What about e-mail from addresses that are on neither the whitelist nor the blacklist? Presumably some sort of Bayesian algorithm is applied to score them, which will not have a 100% success rate (although admittedly, not far off).

What happens when a spammer spoofs one of the IP/e-mail addresses on the whitelist? A message that is clearly undesirable would skip its way merrily through. Or are addresses on the whitelist tested for sales pitch too? In which case, more false positives seem possible.

I do not doubt Kidson's solution works very well for him, but I doubt his requirements are identical to those of every user in the world. The solution to spam is not better filters, as this encourages spammers to find new ways around them. Stop the spammers, not the spam.

Pete Hall, software developer

On possible rises in Cisco support costs

In response to news that Cisco users could face higher maintenance and systems administrations costs following a shake-up of its support programme (Computer Weekly, 27 July)

Cisco's pricing strategy simply reinforces the view that the resellers who flocked to the once seductive Cisco goldmine are having their fingers burnt owing to already tight margins being squeezed further.

Resellers persistently strive to sell Cisco products in the vain hope users will be seduced by the brand name and that tidy profits will subsequently follow.

The overflowing army of Cisco resellers are battling against each other, slashing prices and making meagre profits, leaving many pondering whether it has been worthwhile.

Cisco's support programme should help its reseller partners make profits and delight its customers, rather than leaving them to pay higher charges that are either absorbed by the resellers or - equally as harmful - transferred to the customer.

Cisco should cut its existing reseller base, focus on supporting those resellers that can offer real value and subsequently banish price cutting.

Resellers tired of flogging best-of-brand products to their customers and suffering tight profit margins should focus on tailoring their customers' needs to best-of-breed products that offer higher margins.

Ian Smith, chief executive, Matrix Communications Group

On managing service level agreements

In response to Jimmy Desai's checklist for getting service level agreements right (Computer Weekly, 10 August)

Where companies outsource their IT, there is generally no mechanism whereby the two organisations work together. Rather, the outsourced operation is managed remotely under the terms of the contract, which have no bearing on the end-user the operation exists to serve.

This can result in an adverse impact on the bottom line, as well as lots of unhappy end-users. The problems lies with the service level agreements. SLAs force the managers of the outsourcer to focus on productivity - and quality goes out of the window.

Although I agree with the need for clarity in terms of the responsibilities of each partner, the hierarchical escalation process as suggested by Jimmy Desai is remote from what is really happening in businesses as well as the needs of the end-users.

Working together to deliver an optimal quality service is the most cost-effective way to run an outsourced service. If the outsourced staff are allowed to take the time to get it right first time, all the time, instead of proving to the client what they are doing, it pays off. Better quality service means lower operating costs, which means lower prices passed on to users.

Alison Widdup, executive director, arete business services

The public sector cannot work the private way

I wish to respond to the recent spate of rather pie-in-the-sky letters supporting public sector IT (Letters, 3 August).

Lee Norris claimed that a public sector organisation, despite lacking a commercial impetus and a profit and loss system, is every bit as efficient as a private sector company. He did not mention what impetus the public sector has, nor did he mention what objective system of measuring success the public sector might employ in place of profit and loss.

Kenny Lang said that public sector IT keeps in touch with economic reality by outsourcing to and benchmarking itself against the private sector. Although this is of course good, it relies on the public sector doing this out of the goodness of its heart.

In the private sector you do not have the choice of whether or not to be competitive - you do it or you lose your job.

Tony Dyer felt that the introduction of market forces into UK public sector industries has proved disastrous. This misperception is largely down to privatisation and the price system exposing problems that are hidden under any system of nationalisation, where prices are skewed by politics, and economic efficiency is virtually impossible.

Common to all these contributors is the idea that management and professionalism are what matters, not economic ideology. Management cannot motivate people unless there are rewards, and in a politicised industry, rewards go to those who advance the vested interests of politicians.

Contrast this with a privatised industry, where rewards go to those who advance the interests of the public. Although management and professionalism are important, under a faulty ideology their contributions can very easily be wasted.

Julia Isaak, human resources manager

Let users go to Longhorn when they are ready

I was interested to read that organisations using Windows 2000 may jump straight to Longhorn as their next operating system. (Computer Weekly, 10 August).

It would be naive to believe that software can continue in one incarnation indefinitely. Correspondingly, it is foolhardy to expect businesses to adopt new versions of operating systems at the frenzied pace currently confronting them.

As a supplier, I am already facing the prospect that within the next few versions of our own software we will be unable to continue supporting Windows 95 or 98 for no other reason than Microsoft's insistence that users move to the latest generation of its operating system.

As a conservative estimate, most analysts believe that 50% to 60% of Windows users are running Windows 95 or 98. We would like to be able to continue supporting these users until such a time as they are ready to move on.

The situation is set to get worse with the introduction of Longhorn - Microsoft has already stated that backward compatibility with previous versions of Windows is not high on the design specification. Combine this with Microsoft's recent changes to its end-user licence, which forces you to agree to upgrade when Microsoft tells you to, and the future begins to look bleak. A precursor to this is the next version of Office, v11, which will only be compatible with Windows 2000 and XP.

As many have found to their cost, you cannot suppress free will, and we can only hope that Microsoft will let current versions of Windows continue to operate fully with Longhorn, or it may find that as well as losing backward compatibility, it also loses users.

Eduardo Loigorri, managing director, Exchequer Software

Outsourcing is not a quick-fix solution

I read with interest your article on outsourcing costs and risks (Computer Weekly, 3 August). Although I agree with the advice, I disagree with the traditional approach to outsourcing.

Although outsourcing does have a role to play in enabling IT innovation, the constant reports that outsourcing contracts are failing to deliver expected business benefits highlights the fact that jumping in at the deep end is a high-risk strategy.

Low-risk, step-by-step outsourcing allows organisations to dip a toe in the water and to take measured, controlled steps into outsourcing and achieve specific business objectives.

Traditional outsourcing is not just handing over the IT department and infrastructure to a third party - it is handing over control of the business.

Organisations opting to abdicate responsibility and ownership should hardly be surprised when outsourcing contracts go sour and undermine business success.

David Blundell, chief operating officer, Netstore

Read more on IT outsourcing