% fortune -ae paul murphy

Six things the techpress isn't talking about

"Everybody knows" - all kinds of weird and wonderful things. Unfortunately most of them are wrong, we mostly don't know they're wrong, and a lot of us base daily decisions at least partially on some of these certainties.

As a public service, therefore, I thought I'd list five important technology changes now underway that the main stream technology press is either largely ignoring, or mainly mis-representing.

  1. Everybody knows storage is getting cheaper, but how many articles have you read pointing out that ZFS is getting ported to most OSes and not only implements the first real RAID (redundant Array of Inexpensive Disks) solution but makes all the pretenders, and their associated software, obsolete?

  2. Everybody knows convergence will drive handhelds -but how many articles have you read pointing out that Apple's iPhone is a pocket Mac (and apparently PPC based at that!); the first product genuinely capable of combining with network based services to wipe out the volume market for notebooks?

    It's also a great opportunity for people who want to provide the missing bits: a web service, for example, to which the iphone user downloads a voice file and gets back a text transcript.

  3. Everybody knows that the race to multi-core has displaced the race to high megahertz. Intel, for example, now offers two dual core CPUs in one package and is promising to eventually integrate larger numbers of cores.

    Great, but how many articles have you read pointing out that the ability to use these devices to make most applications run faster depends almost entirely on compilers - and Microsoft's compilers are pretty much limited to four way instruction issuance?

    Intel bought Kuck and Associates back in the nineties to get its hands on the company's C and C++ compilers for parallel processors, but one reason Itanium hasn't met its targets is that the key people left and Intel hasn't advanced the art since.

    IBM has recently made some breakthroughs on this front (not that you've seen that widely reported either) but they're focused on the Altivec extensions within the PowerPC architecture; not x86.

    So what's the bottom line you're not getting told about? the value of additional cores for personal computers slopes down dramatically after two and falls off a cliff after four - meaning that Microsoft can either fix the compilers (not so easy), face a dead end on performance gains (unacceptable), hope Intel gets back into the gigahertz game (bound to happen) or accelerate its switch to PPC.

  4. in the United States Congress some democrats have taken time off from planning show trials to draft and present a lobbying control bill which, among other things, sets out the conditions under which anyone trying to influence public opinion via internet means such as blogging has to register as a paid lobbyist - and subsequently abide by the regulations affecting that profession.

    The draft apparently isn't particularly onerous because people whose rights to free speech will be curtailed under this law must meet at least three conditions: someone's paying for their work, the total of their applicable costs and/or revenues must exceed $25K in a quarter, and the audience must exceed 500 members.

    Personally I'd love to see this become law just long enough to find out more about Groklaw, but consider almost any technology industry CEO - most are using blogging directly or indirectly to reach audiences of more than 500, most of them make enough that even a few minutes of focus per day would put them over the compensation limit, and, of course, all of them are paid to propagandize for their firms.

    So the law takes away Sun's freedom to have its CEO write a blog - but it's not applicable to you, right? Wrong, It's not just lawsuits alleging that putting a bumper sticker on a car amounts to an improper campaign contribution; it's a general attack on use of the internet in politics.

    So why haven't you read about this in the techpress? Because the focus on bits and bytes, the need to be upbeat, and the risk of offending democrats in the audience enforces a see no evil, hear no evil, and report no evil policy under which the press pretends that politics, the force most directly driving national economic behaviour, has no place in the data centers of the nation.

    If only it were so - but it isn't; and rights that aren't defended, are usually lost.

  5. There's a quiet little tsunami sweeping through e-communications that hardly anybody seems to be talking about. Here's the first part of the problem: you send off an email, and don't get an an immediate response - now: did somebody's ISP send it to bit heaven? did your intended recipient's email server trash it as spam or otherwise fail to deliver it? did the intended's own spam filter get it? or is the intended recipient simply not answering because he doesn't want to? isn't there? has a corrupted file system and can't read it? or because police have seized his PC in respect of a criminal investigation or civil seizure order?

    This is getting to be very common - in fact, most people in business have had this happen to them with increasing regularity and I'm getting calls, as I'm sure you are, that start out with a variation on "did you get my email?"

    Right now text messaging is still marginally better - but that's starting to suffer from many of the same problems while voice mail remains trustworthy -except that if it goes VoIP at some point you lose control over where and how often it's delivered.

    The second part of the problem is an exploitation of this. With e-discovery legislation in the United States now in force, and similar laws either in force or on their way elsewhere, you can be forced to produce emails your spam filters declined to show you - and any shyster worthy of the name can send you emails that will get treated as spam when they arrive, but hang you when they're "discovered."

    So why aren't the gentlemen of the technology press all over this story? The cynical explanation is that they don't get paid for sending billion dollar industries into bankruptcy by pointing up what we all know - that the post hoc solutions the industry sells not only don't drain the swamp, but actually feed the alligators.

And if those five haven't made your day - consider number six, ( from Saturday's TimesOnline)

A mass study of the long-term impact of mobile phones is to be undertaken amid fears that people who have used them for more than ten years are at greater risk from brain cancer.

More than 200,000 volunteers, including long-term users, are to be monitored for at least five years to plot mobile phone use against any serious diseases they develop, including cancer and Parkinson's and Alzheimer's diseases.

Professor Lawrie Challis, who is in the final stages of negotiation with the Department of Health and the mobile phone industry for the 3 million that he needs to fund the study, told The Times that research has shown that mobiles are very safe in the short term but that there is a 'hint of something' for people using them longer.

In an interview, Professor Challis, a world expert on mobile phone radiation, and chairman of the government-funded mobile telecommunications health research programme, emphasised that the 'hint' was just that. One European study has found a slight association and using a mobile for more than ten years. The few long-term users developed more acoustic neuroma brain tumours which were found close to the ear used for phoning.

And that, I think, is a failure to report along lines we've seen before.

Paul Murphy wrote and published The Unix Guide to Defenestration. Murphy is a 25-year veteran of the I.T. consulting industry, specializing in Unix and Unix-related management issues.