Notebook - the thorn in the thin client's side

Notebook - the thorn in the thin client's side

Summary: Not a year goes by where Sun and others don't extol the virtues of thin clients -- devices of limited intelligence that provide access to enterprise applications.  [Sun Ray 1g Ultra-Thin Client, at right.

TOPICS: Laptops

Not a year goes by where Sun and others don't extol the virtues of thin clients -- devices of limited intelligence that provide access to enterprise applications.  [Sun Ray 1g Ultra-Thin Client, at right.]  To be a really good thin client these days, a device must support the primary protocols of terminal-based access to certain applications.  Such protocols invariably include the Web and Windows Terminal protocols, and Java Virtual Machine capabilities, as well as some older ones like X Windows and 3270 access.  Supposedly, not only are the terminals cheaper (actually they've had a tough time beating the cheapest PCs), but their inability to run executable code or to save information to a local storage device (eg: a USB key) also means that that they can't be the source of major security headaches.  Another big benefit is that you can log on anywhere, and your "configuration" follows you everywhere you go. 

But one cat that the thin client approach has never quite skinned is the mobility cat.  Until we have 24/7 access to a network (regardless of where we are... the beach, a plane, etc.) there are millions of people who won't be able to get work done when they need to get it done if all they have is a thin client device.  And that number is on the rise as more and more employers offer or require that their employees work from locations other than their offices or cubicles.   For this reason, notebook computers and increasingly, PDAs -- both of which are intelligent devices with local storage -- continue to gain ground as the preferences of end-users.   I was reminded of this fact by a recent news story that marked an important milestone (one that thin-clients will probably never reach).  

According to a recent study, May 2005 marks the first month in history during which more notebooks were sold than desktops.   That trend was obviously helped along by another trend -- that of home office users establishing a preference for notebooks over desktops.  Uh, doh.  Given how just about everybody has a WiFi access point in their homes these days, what would you want?  A desktop that you have to drag around the house from your office to the living room (if you want to catch some TV) to your bed?  Or, a notebook? For more than three years,  I've been recommending against the purchase of desktop computers.  In fact, outside of my son's tricked-out Alienware gaming system (which, he laments, he can't take very easily take it to tournaments), I can't remember the last time I touched a desktop system.

Topic: Laptops

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


Log in or register to join the discussion
  • TCO versus TBO

    How many times did we hear about TCO when it came to thin clients? My argument against that was that if it were all about TCO, then we wouldn't have computers as the cost of zero computers is zero dollars. But if we look at Total Benefit of Ownership (TBO), we need computers, and the ability to work away from the office is a huge item on the benefit side of the ledger. Trying to mix thin clients (for personnel not expected to work at home) with thick has its own set of problems. Thin clients might make sense in a setting such as a call center, but as the article states, it tough to beat the price of a PC (though the management cost of a thin client could be lower). The thin client idea has been around for 10 years, its time has not yet come, and in fact seems to be farther away than ever.
    Rodney Davis
    • The cost is NOT the hardware

      Thin client has a completely different meaning these days. It's no great trick buying more processor, memory and hard drive than you need. As a matter of fact, it would be hard not to exceed requirements.

      Seat costs today are driven by the cost of support. If you'll take a look at almost any enterprise IT budget you'll find you could buy every user several desktops per year for the cost of the support staff to keep them running.

      The reason for the corporate push to thin clients - browser plus office suite, and I haven't seen a design review exception approved in the past two years - is to reduce labor costs, not hardware.
      • I agree that...

        ...the cost is not the hardware today, but a big part of the thin client idea 10 years ago was lower hardware cost (that was when a top-notch PC cost $4000).

        Unlike you, I don't see the "corporate push to thin clients", and I think the market stats will support me on that, though we are using more and more web-based apps on our Windows PCs. The ultimate goal of the thin client proponents was to replace MS Windows on the desktop. They have failed. Part of the reason they failed was because PCs got cheaper, the bandwidth to run thin clients was late in materializing (for smaller, satellite offices this is still an issue), and, though you may disagree, PC got easier to manage (compare XP with Win 95).
        Rodney Davis
        • Replace Windows?

          Just a wee bit of a flame here, but at no time when we were developing thin clients were we ever considering removing Windows, or in the case of Linux based terminals a windowing manager, from the desktop.

          The systems we initially derived to deliver a windows desktop, be it CE, PocketPC, Windows NT/XP Embedded, SmartPhone, Linux, PalmOS, whatever, whereby we could deliver applications that made sense to be on the local level. Those applications, Citrix ICA, RDP, 3270, a web browser, whatever, were and are server-based.

          Remember, all that a PC is in a properly networked environment with network services and the such, is a really, really fat thin client.

          As one of the developers who were around in the beginning and who is still around in the market, I'll be the first to tell you that Windows belongs on the desktop, it should just be fast, secure, highly manageable, and it shouldn't allow users to make changes that are detrimental to their computing health.
          • How was Sun going to make money?

            The thin client pitch, circa 1996, was about replacing Windows. The antitrust suit was about Netscape as a basis of a competing PLATFORM (NOT a competing application) being crushed by MS. I'm not arguing with your view of thin client, in fact, it was a smart view. I'm just saying that was not the sales pitch of Sun, et al.
            Rodney Davis
        • Choice of OS?

          Although you may have been embroiled in Windows vs thin client arguments in the past, this is not today's ground. Today's thin client might better be termed "simple client," or some other term that conveys the opposite of complex.

          In the client-server era most mission critical apps, as well as most commercial software, required loading a client on the desktop. The desktop grew successively "fatter" and more complex - and more expensive to maintain. Now as we (here I mean the corporate "we," as reported by META, Burton, Gartner, etc) replace our legacy client-server apps with web-based apps and the commercial packages have made their tranisition, the client is on a rapid reduction program.

          But it matters not what operating system, you could have a thin client with any OS, and I don't know where one would find "stats" on the thinning of the client - that's just not the kind of thing that is easily quantified or collected.
          • Evolution

            I agree that the idea of the "thin client" has evolved into something else today, but I don't know that the client is on a "rapid reduction program." Despite the rise of web-based apps, MS Office, the classic fat client set of apps, is still the dominant piece of software that you'll find in corporate America.
            Rodney Davis
          • Not complexifying

            If you have an entire enterprise of OS, browser and office suite, you tend to have a pretty manageable configuration. From an O&M perspective, we certainly consider that to be in the spirit of "thin." Interactions among the software can usually be throughly checked out ahead of deployment and, in the case of trouble calls, the tech arriving on the scene can usually deal with the problem in a single trip - maintenance headaches go way down. The fact of Microsoft OS and office suite implies ongoing cost that might or might not be addressed in future years, but what enables all of this is the devolution of special-purpose client software at the edge. We still have pockets of special-purpose clients and standalone software, machines that do demand more care and feeding. I expect these to continue to diminish over time, but never to vanish entirely. So even in the extreme case of OS swapout, there would be lingering pockets of Windows, same as there have generally been lingering pockets of Macs.

            The thin client arrived on the scene with relatively little fanfare, and I think it's about as thin as need be. I understand from DISA colleagues some of the newer and larger DOD installations make use of a suite that does not retain office suite at the client, but I haven't looked into the product set and don't see off the top of my head what the benefit would be. The interesting side benefit of the thin client is that it gives management options it has not had had for many years, whether the choice is made to avail itself of the options.
          • Interesting observation

            Whatever it may have been in the past, thin client is really a user movement, rather than a push by any vendor. I note that sales teams present their pitch in terms of how their products fit in the thin client environment, rather than trying to sell a product as "the thin client solution."
    • I see what you mean but...

      In fact, a vast number of people use their PCs as thin clients.
      Students and so on with messenger programs and webmail,
      ebay, amazon etc.
      Also look at netsuite and
      (there are also MRP solutions, and a million more web apps out
      there, being used right now).

      Many corporations will be slow at moving to this kind of
      application, since people are used to their old apps and old way
      of working. In my view the webapp approach has so many
      benefits, it's amazing.
      It took me a while to get used to not downloading data (eg from
      mail server), but now I leave the mail on the server, it's visible
      *anywhere*. I'm not chained to a peice of equipment that might
      get damaged. I can go to an internet cafe, borrow a wifi laptop,
      borrow a desktop, use a kiosk, and so on. The data is visible to
      me everywhere.

      In places without wifi, as long as your webapp isn't dressed up
      with lots of pointless images, it'll run quickly over a mobile
      phone connection.

      Obviously this doesn't work so much for things like Powerpoint.
      But only a minute proportion of people use powerpoint, and
      frankly far less should in my opinion.
      (ever sat through a morning of "presentations"??)
      Hey do something different, make a small video, write some
      jokes in a regular paper notepad, instead of using all your
      "creativity" to find an 'amusing' animated gif that fails to run on
      the day.

      What is a "thin" client these days anyway. You can put Knoppix
      on a 1 gig compact flash, and run it on a 500MHz machine with
      no problem. And in the compact flash, you have thousands of
      desktop apps including a raft of word/excel/powerpoint like
      programs. I just reckon web apps are going to be way more
      productive than old office apps.

      LAMP (Linux Apache MySQL PHP) server man, you'd be AMAZED.
      I'm writing a certificate generation system at the moment, and
      it's so fast to develop in,
      it's truly awesome. (pdf-php is really class at making excellent
      pdfs very very easily - put a picture here, this text there).
      I've written 1400 lines of code so far, can refactor to maybe lose
      30% of that, and the functionality is just jaw dropping. Try it.
      • But you still need...

        ...a compelling reason to switch from what you're used to AND there are some things you can't do. Even if it's only 10 percent of your user base that needs to run PowerPoint or edit 30MB Word documents, that group's needs may dictate your whole computing strategy.
        Rodney Davis
        • 10% dictating strategy?

          Why can't you run 90% on thin client, and the other 10% fat
          client (if they feel they need to).
          I really can't see why you HAVE to have one universal desktop. It
          doesn't make sense. The 90% will generate virtually no work
          compared to the current situation, allowing you to better tailor
          the fat system to the 10%, or reduce cost.
          If your competitor's IT people can handle this minor
          complication, and you can't, then your corp will suffer in the
          market. (higher overheads).
          Why would you automatically discount this technology?
          • Division of Labor

            Sorry, my post wasn't very clear (at all!) on my point. You may have 10% that are need to run anything from Photoshop to Access and require a thick client. But I think it's naive to assume, in the vast majority of companies and government agencies that I deal with, that we can draw a line to separate this 10% from the other 90%. In a call center situation, yes. In a consulting organization, much more difficult. These days, we are expecting that every new hire has computer skills. Though an employee may not normally create a PowerPoint presentation, we like to have workforce flexibility when it's crunch time.
            Rodney Davis
          • not naive, use something like VNC

            or windows terminal services perhaps for those that *might*
            need access occaisionally to an app server where they can run
            powerpoint etc.
    • In reality, yes. In the eyes of those who think 'short tem $ savings',

      they're as mindwarped as ever, delusional that things have to cost money at all.
  • False Dichotomy

    You drank the hardware-vendor Kool-Aid. They have an interest in pushing hardware that is strictly limited to use as a thin client, but that's their agenda. It doesn't mean that the compromise that they ignore isn't viable.

    "Thin client" is a mode of interaction, not a religious obligation.

    Example: many of the local engineering staff use notebooks both for the usual mobile applications (word processing, e-mail, etc.) However, when the time comes to do core engineering work they use them as X servers and run the serious work on compute clusters.
    Yagotta B. Kidding
    • very true

      "zen and the art fo motorcycle maintenance" again. A peice of
      beer-can 'CAN'T' function as a handlebar shim for an expensive
      motorcycle, it isn't labelled for that purpose.

      What is a "thin client" these days anyway (like I say above). You
      can put Knoppix on a 1 gig flash or MMC card (thousands of

      Personally I prefer webmail - mail is visible anywhere, the
      machine loses it's importance, I can borrow others, or use public
      terminals. I think wordprocessing is so 90s.
      If you're writing a communication, use email.
      If you're writing to a friend, use real paper, and a pen (or crayon
      ;) ).
      If you're writing corporate documents, use XML editors and an
      XML repository designed for object re-use. It will save a
      FORTUNE for your corporation.
  • The WiMAX steamroller

    should take care of the connectivity problem - in about a year or two. Until then, its a hodgepodge of Cellular and WiFi (and some lesser known protocols).
    Roger Ramjet
  • Did it ever occur to you

    that vendors WANT you to buy laptops instead of desksides - BECAUSE its a proprietary platform? There are few to none interchangable parts between laptops - including such things as memory, video, sound, etc. Vendors can charge you more for initial sale and later support with laptops. Not to mention the risk of exploding batteries! ;)
    Roger Ramjet
    • Good Grief!

      You have an excellent point! LOL how funny!
      Linux User 147560