Trying to predict 2008

Trying to predict 2008

Summary: With the exception of the Linux opportunity presented by the OLPC project, the right word for IT in 2008 will, I think, be "continuation" as existing trends continue to reveal themselves and little new enters the market.

SHARE:
TOPICS: Linux, Open Source
13

The IT word for 2008 will, I think, be either "continuation" or "consolidation" as existing trends become more obvious to more people and little new enters the market. Herewith, however, some predictions I hope not to be apologizing for next December.

At the top of the list of continuations is SCO. No matter how the legal action pans out, it will continue to dominate direction setting in the Linux community - and until or unless IBM gets its collective head straight on the issue and cleans house, the polarization this case has led to will continue to undermine Linux legitimacy.

Oddly, however, the number one Linux influencer for the year may turn out not be SCO, Torvalds, IBM or any of the currently more visible Linux distributions. Instead, the Linux community may find its second chance at the kind of public tipping point that would give it widespread credibility and mass media supported desktop appeal coming from Negroponte's OLPC (one laptop per child) project.

The reason for this has nothing to do with any of the OLPC project's stated goals - in fact, the project is intended to give under-privileged third world children a better chance at education and economic participation, but will actually reduce their opportunities because they'll mostly just go to people committed to preserving existing local power structures.

Basically what's important about the OLPC has nothing to do with its nominal purpose and everything to do with the intended interface. Ultimately traceable to David Gelernter's "Lifestreams" model, this is not just a remake of Apple's evolution of the original work done at Palo Alto, but something new - and probably as appropriate to home computer usage as Sunview and CDE are to systems administration. As such it's both a threat to Microsoft and a critical opportunity for Linux: get the OLPC accepted for use among the least privileged and most ignorant in the United States (almost all of whom are orders of magnitude better off than their third world counter-parts) and their success could easily trigger enormous mass media support for Lifestreams on Linux for the rest of us.

It could happen in 2008, or it could not - money is largely rational and thus roughly predictable, but politics, media enthusiasms, and do gooder fads are not.

And on that same basis I think Sun will get "Victoria Falls" out on schedule and is somewhat more likely than not to get its "Rock" chip sets out by late summer, early fall - because I think the changes now expected to delay "Rock" until early to mid 2009 are ultimately more tied to commercial/strategic considerations than technical ones.

Meanwhile, Microsoft's inability to transcend simple minded x86 programming will continue to drive wintel back to the gigahertz race while delaying the move to PPC - with HP taking a bigger hit from dollar devaluation than Dell and the acceleration of existing trends to an iPhone/Sun Ray style of computing becoming more and more obvious as Microsoft's inertia grinds down its wintel partners.

Among other things Microsoft's organizational limitations will mean that we'll probably see lots of press hype about advanced Intel products like a 32 core wafer running on micro-watts, but no significant products because Microsoft can't get the software working well enough to matter and a half dozen demo sites running Linux do not amount to an industry tidal wave no matter how heavily hyped.

Key management trends from last year will continue - although I think we're starting to see some push-back against the absurdities of PC/mainframe style virtualization and no-one will figure out what SOA means in 2008 either. What will happen is that the emphasis on "security" will get worse as more auditors train up on 1920s data processing control models and company legal departments get heavily into the act.

The outsourcing reversals and repatriations going on now will, I think, continue; largely because they're driven from oil related dollar devaluation. As a result I expect to see a minor gold rush in IT hiring -and that spells good news for American IT workers as increasing numbers of legal visa holders choose to go home and more and more illegals either get deported or take pre-emptive flight.

As a side-effect, IT worker shortages could lead to an upturn in the market for less labor intensive technologies - i.e. all major flavors of Unix: Solaris, MacOS X, and Linux.

However, the U.S. can absorb an economic adjustment like this a lot better than most of what used to be western Europe can. There the combination of immigration rules simultaneously working against retention of the more educated and in favor of illegals with little beyond unmet expectations, children, and religious fervor will (I think and barring a recession), cause severe problems for IT hiring managers. Oddly France, because of its commitment to nuclear power and political response to the riots two years ago, may become the hot spot for IT investment in the EU - not something anyone would have imagined possible in 2005!

One odd side effect, particularly in Britain, may be to further limit IT innovation as demonstrated allegiance to technology choices made for high cost (and failure prone) government IT projects plays an increasing role in determining hire-ability among career IT people.

Similarly, what's going to happen with respect to time sharing and social networks won't be anything new - existing trends will merely continue to strengthen the former while weakening the latter, at least until (or while) current structures get replaced.

Beyond that? its an election year in the United States and you'd think that would lead to some focus on E-voting - but it won't, at least this year. What's going on is that existing e-voting technologies are absurdly vulnerable and everybody knows it, but the democrats prosecute nearly all election related lawsuits in the United States and they liked the results last time, so nothing's changed since - and nothing will unless the Republicans win big in November - and then the changes will come in 2009, not 2008.

So, bottom line? I expect lots of continuations and consolidations from last year - making it a bit boring from a bleeding edge perspective, but overall promising a good year for people focused on improving IT productivity.

Topics: Linux, Open Source

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

13 comments
Log in or register to join the discussion
  • More specificity please

    You say that IBM needs to get its collective head together re the SCO case. Do you have more specific recommendations on how they should proceed? After all, they appear to be winning.
    John L. Ries
    • "they appear to be winning"

      Actually, they appear to have wrestled defeat from victory. The problem is that if they succeed with the Novell ploy, they won't have addressed the issue - the breach of contract.

      Ever watch a TV crime program in which the putative villain escapes on a technicality? That's great if you're rooting for the bad guy, but leaves the majority of the audience wanting a replay - and, in most cases, it's the replay that drives the program's emotional content.
      murph_z
      • You didn't answer the question

        Which was, what should IBM be doing in regards to the SCO case? "Getting their collective head together" is hopelessly vague.
        John L. Ries
        • Ahh - that's a hard question

          As i've said before, the basis for settlement depend on one's perception of the original fault. If, as i think, the problems mostly originated with the mainframe linux project (with people from Dallas explicitly warning IBM management that AIX expertise was being applied to the Linux kernel port there) and then got into the GPL tree initially through SuSe then..

          1 - IBM should kill that code base
          2 - IBM should compensate SCO for losses arising from IBM's behavior
          3 - all parties involved should release all remaining rights in AT&T and derived code under some open source license like the CDDL
          or GPL2.

          In practice none of this would be easy, but the result could be to end this nonsense of Linux not being Unix and, therefore, the increasing bitterness with which people like Torvalds attack both Solaris and the BSDs. That conflict didn't exist prior to the lawsuit and is fundamentally wrong: we're all Unix supporters and the enemy is ignorance (often proxied by MS!), not the other guy's Unix.
          murph_z
      • SCO's Contract Claims

        "The problem is that if they succeed with the Novell ploy, they won't have addressed the issue - the breach of contract."

        Here is SCO record to date with claims involving breach of
        contract.

        SCO alleged 1 claim for breach of contract against Daimler
        Chrysler. It was dismissed (by Judge Chabot).

        SCO alleged 2 claims for breach of contract against Novell.
        One was dismissed and most of the second one dismissed as well
        (by Judge Kimball).

        Novell also alledged 2 claims for breach of contract against
        SCO. SCO motion to dismiss one of them has been denied (by
        Judge Kimball).

        SCO seems to have a serious deficiency in its ability to
        read contracts.
        sk43999
  • delay the move to PPC? what?

    The comment about SCO is baffling enough, but the comment about PPC is just bizarre. We all know PPC is more elegant, but if the chips aren't there, not much one can do with software. Microsoft *used* to have a PPC version of Windows and gave up on it... and of course, finally Apple had to as well. It may be superior -- but just like OS/2, that's not all it takes.

    This goes double for Itanium, which you didn't even mention.

    Plus, x86_64 works around the key deficiency in the x86 architecture, which is lack of registers. We'll be seeing it or years.
    mattdm
    • Oh?

      1) the chips exist, they're just being used in embedded apps and are therefore invisible to PC users.

      2) MS Office now runs on the X360 - but the intended network OS does not so the product isn't being released. Ask why not, and no one will say - but the truth is that MS has created huge internal barriers to change and now can't change..

      I think MS will eventually be forced to switch to Linux or BSD for this.

      3) yes the Mac now uses x86 and a 2007 dual Intel 2.7Ghz runs faster than a 2002 G4 single core at 1.2Ghz - but the mac now costs more than the PC (it didn't before because you got more for the buck) - and head to head PPC vs x86 comparisons using current chips from both groups show ppc significantly ahead. Remember: cell and Xenon are both PPC, and check out the embedded industry benchmark site.

      4) I did not mention the itanium because its time has passed - again because MS couldn't deliver the software for a non x86 architecture.
      murph_z
      • Itanium

        Can't blame MS for Itanium's failure. Linux has been available on Itanium for years, as has HP/UX. Itanium could have succeeded in the UNIX world and hasn't because Intel hasn't made the case for it. The fault lies with Intel alone.
        John L. Ries
        • Hey, since it's just you and I talking..

          I can defend Intel, right? Ok - maybe not.. :) but:

          1 - itanium is actually a pretty good chip but Intel's agenda was to get a patentable instruction set to put AMD out of business. HP's agenda was to become the services leg in a Wintel trio - and tried very hard to sell HP-UX customers moving to itanium on Windows instead - and MS's agenda was just to sell licenses while protecting its installed base.

          2 - Sun responded by helping AMD get x64 into the market. That put intel between rocks and hard places and let MS off the 64bit hook - because had Itanium been the only 64bit route for wintel users (as intel intended) MS would have had to adapt.

          3 - now Linux on x86 is faster than it is on itanium - and for a lot less money. HP-UX is functionally dead, MS's itantium software is so piggy (because of the x86 optimization built into both code and compilers) on Itanium that it actually runs faster in x86 emulation mode..

          So who's fault is it? Well - McNealy's - his AMD bet made the world what it is now - but
          that's like blaming Wellington for Napolean's loss. The reality is that MS could have made Itanium (and thus HP's strategy) a success, but didn't.
          murph_z
          • I think I'll buy that

            But I'll add that Intel played the "aura of inevitability" game (what MS has done for years with Windows and Office) with Itanium and AMD called their bluff (kind of like what's happening with Windows Vista). You might recall that we were all supposed to be running Itanium 5 years ago.

            I notice that SGI's Itanium efforts don't appear to be going well either (at least, the company that employs me has yet to see a request to port our software to Linux/Itanium; we've had one for HP/UX-Itanium).
            John L. Ries
          • I'll also add...

            ...that delays in getting the chip out after years of hype didn't help matters. Proper management of expectations is a good thing.
            John L. Ries
  • 2008

    Microsoft will continue its top of the line programming and Vista will reach around 40% adoption by year end with the rest of the world using XP.

    Rudy will continue to believe that MS software doesn't work and third party Windows developers, both commercial and open source, will continue to churn out useful and non useful applications.

    Someone will try the thin client scam again in a vain attempt to control those troublesome users.

    Linux will once again fail to reach single figure global usage, but be mentioned in the majority of posts to ZDNet.

    Apple will continue to release security patches in an attempt to catch up.

    The same less than 1% of the global population will know what an OS is and select what they want. The rest of the world will just keep using Windows, because in the end, it's the applications, standard GUI and compatibility that matter.

    Third world goverments will sue Nicholas Negroponte under equal opportunity legislation as he prevented them from accessing to the global OS.

    Nerds who find they still can't play World of Warcraft properly on Linux, buy a Mac or PC and are shocked by the features and speed.

    BitTorrent traffic increases by 300% as consumers get sick of their TV shows being out of date and adulterated with ads.

    Apple's iPhone attains 75% of the features of other smart phones.

    Paul Murphy has to take a Windows job to pay the bills and is shocked by the features and sophistication of the platform, leading to his own conversion on the road to Damascus.

    OK I was kidding about the last one, but if it worked for St Paul then who knows ;-)
    tonymcs@...
  • RE: Trying to predict 2008

    Well Mr. Murphy, you speak alot of truth I don't often comment but what you say about the OLPH project is quite true. I have worked for do gooding organisations in developing country's and as you say any allocated resources will indeed been benifical to the local administration only. These people get plenty of finance to improve their infrastucture/education and trade already. However they are smart enough to sit and wait until which point the money stops arriving before they retire to a nice beach credit of western tax payers who can't even raise the money to pay their own rent let alone start feeling sorry for the little brown people, who can more than take care for themselves.

    Also If Vista will be adopted or not? Sure it will be! the real question is? is it any good? Unfortunately the large percentage of people who are likely to answer that question are not exactly qualified to. I probably missed a few point's there as i'm done, bored.
    blueglue