Is it bling or innovation?

Is it bling or innovation?

Summary: I read a great piece by my colleague, Chris Dawson, this morning entitled I love Linux, but it’s not going to save the world and after reading a few of his readers comments, I felt compelled to get in my two cents.


Marc WagnerI read a great piece by my colleague, Chris Dawson, this morning entitled I love Linux, but it’s not going to save the world and after reading a few of his readers comments, I felt compelled to get in my two cents.  His point (at least what I took away from the piece) is that just because Linux can run on old hardware does not mean that you should keep that old hardware lying around -- despite the somewhat specious argument of a UK study that pointed out the old hardware which was still in use was not in a landfill somewhere.  (More likely, it will end up in a storage room somewhere, waiting for someone like Chris to get the time to fix it!)

To be sure, Linux is a remarkably flexible operating system and a suitable replacement for UNIX in settings where UNIX was once king.  Further, it can be configured with a robust GUI and perform as a remarkably flexible desktop operating system. 

In other words, both UNIX and Linux can be configured to run on very modest hardware, as long as your needs are modest.  And, when your needs are not so modest, it's up to the job when deployed on robust hardware.

Some of Chris's readers scoffed at the 3-5 year life-cycles to which Chris referred -- arguing that with Linux, much older hardware would do just as well.  Others took pride in being able to turn old Windows 95 systems into Linux thin-clients.  Well that's fine and dandy -- if you ignore the human costs associated with maintaining old hardware. 

It's not about cost, it's about productivity. 

Chris, with his depth of knowledge of IT, can spend his time replacing broken floppy drives or he can be answering his students computing questions.  He can be hunting down an intermittent memory or hard drive problem or he can be teaching computer science courses.  Or, he can be writing a proposal for a new service which will help his educators teach -- and thus his students learn more effectively, using the latest tools.  Which do you suppose is the best use of his time?

In his own school, Chris has an extensive mixed environment designed to meet a variety of needs for his students and his faculty.  In a university setting, the goals are exactly the same, except that the range of needs is far greater -- as is the diversity of skill levels of the user population -- faculty, staff, and students.  Further, the extent and reliability of services provided by the university can have a far-reaching impact -- not just on faculty and students on your campus but on researchers throughout the world who have become dependent upon those services for collaboration with your faculty. 

Cost pressures, plus lack to expertise (in both accounting and IT) on the part of Administrators and Ed Tech staff often lead to a consumer's approach to IT upgrades:

  • If it ain't broke, don't fix it. 
  • Buy it and use it until it breaks.
  • It must be less expensive to fix than to replace.
  • Buy no more than what I need now.
  • Buy the least expensive, the vendor doesn't matter.

The enterprise (be it in a business or university setting) must ask larger questions:

  • What do each of my users need?  How long will they have this need?
  • Will this product still meet those needs for three to five years?
  • Will the vendor fix it if it breaks during its life-cycle?
  • What on-going expenses will be incurred by choosing this product?

Many readers assume that three-to-fives years is an arbitrary period of time for hardware life-cycles.  It is not: 

Accountants like five years because hardware is a capital expense.  Tax law dictates how capital expenses are treated and whether you are in a not-for-profit environment (in which educational institutions find themselves) or in business, these rules of accounting still apply. 

IT professionals like three years because of Moore's Law.  Moore's law states that processor power doubles every 18 months.  Though merely a supposition, it has proved to be a reliable measure for nearly thirty years.  This means that the machine you buy today will be, at best, 25% as powerful as the machine you will buy three years from now. 

A number of our readers like to criticize the apparent trend toward bloated (others might say feature-rich) software.  Their argument is that adding all that bling to software makes it no more useful to the average user yet makes us dependent upon more and more robust hardware -- thus shorter hardware life-cycles.  (I find it ironic that other readers argue that an apparent lack of competition thwarts innovation.  So which is it? Bling or innovation?

Looking to the second point above, it becomes clear that software which is upgraded often will need more robust hardware sooner rather than later.  Considering that software publishers always lag the capabilities of the hardware, three years comes out about right.

Case in point:  Vista will run (albeit poorly) on six-year-old hardware and it runs easily on properly configured three-year-old hardware -- though it runs best only on the newest hardware (at any price-point). 

When the dynamics of the environment dictate the need -- not the average user --  the needs of the few today should dictate life-cycle purchases for the many who must still be able to use that hardware in three to five years.  Try to get more than five years out of your hardware and you will leave your most robust users (be they educators or students) out in the cold with no way to meet their needs.

Topics: Hardware, Emerging Tech, 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.


Log in or register to join the discussion
  • Conflating issues

    You've served up a rather rich soup of issues here. It's rather difficult to discuss them intelligently as long as they're stirred together that way.

    For instance, hardware lifecycle. Are you talking about "broken" or about "inadequate?" The two don't happen together, please note. For instance, a bleeding-edge machine won't last as long as a lower-performance one at the same technology node because it's under less stress (voltage and thermal mostly.)
    Yagotta B. Kidding
    • I purposefully chose not to bring in issues ...

      ... about which choices are made for that very reason.

      You can purchase "middle-of-the-road" hardware today and expect it to meet "middle-of-the-road" needs three years from now. If you have extraordinary needs now, then you might have to accept shorter life-cycles to get the performance that you need. (But probably not -- "bleeding edge" choices need only be made for "bleeding edge" needs, which are truly rare.)

      That doesn't mean that ALL of your machines have to be the same. A few machines with extraordinary capabilities and a lot of middle-of-the-road machines may be what you choose.

      Time is the 'enemy' -- choosing only robust, high-end machines will not necessarily allow you to extend your life-cycles but choosing low-end hardware will most certainly shorten them!
      M Wagner
      • But don't forget warrantees and support...

        If I'm talking about a server with something important running on it then down time is death. I recently sent 3 old Dell Servers running Windows off as surplus because they were 4 years old and the service agreements were expired. Those 4 year old machines were still running Windows Server 2003 just fine - performance of the old hardware - even under Windows just wasn't the issue.
        • Yes, warranty costs need to be figured in ...

          ... when asking those questions. In general, buying extra years of warranty from the vendor is less expensive than fixing it yourself if it fails after the warranty expires. These costs are best considered at the beginning of your life-cycle though -- not three years into it.

          By selling at surplus, there is still an opportunity to re-purpose the hardware for less demanding needs.

          In other scenarios, the need to upgrade to Windows server 2008 (due out soon) might be reason enough to retire old hardware.
          M Wagner
      • Needs vs Fanboys

        The needs of the target users determine the software and hardware choices, from home to super computing. Plus, the software needs will drive the hardware choices.

        After that is determined, everything must be fit into a budget, which will refine the hardware choices.

        Gaming Fanboys always say 'go for top of the line!' Price just went way up.
        • Yes but Education IT cannot determine ...

          ... those needs in a vacuum. Faculty need to be part of that determination. Once the need has been identified, then educators and EdTech and the budget folks can determine what mix of solutions will meet the greatest need for the least amount of money -- and time! There is no single solution.
          M Wagner
          • re: Yes but

            Faculty would be part of the Target Users group. My comments were general in nature, because they cover the entire spectrum from individuals like myself, to super computing and clustering, and not just education.
  • Different strokes for different folks

    I can make a solid argument for some of our staff to get new hardware every two years. I can make just as solid an argument that others on our staff only upgrade once every six years. One of our light file servers is eight years old and still has another two years worth of use in it. We have another server which hosts databases which gets upgraded religiously every two years.

    All hardware and software decisions get made based on the application in which it is used. I base my decisions based on what course of action will minimize our total company's man-hours. Both hardware and software are cheap next to personnel expenses.
    Michael Kelly
  • I would like to see you quantitize...

    all of the bling you just spewed.

    Learning more effectively for instance.

    IT has been saying upgrading is good since day one. But ask them to tell you how these upgrades help the bottom line in a business and they have no answer.

    Show me the numbers and then I'll believe your opinion.
    • The "numbers" are different in every ...

      ... situation and that's why no one in IT can answer the question directly.

      There is always a point of diminishing returns and to know where that point is, a thorough cost/benefit analysis is required. By not establishing a life-cycle replacement funding plan based on a thorough understanding of your needs, you become vulnerable and find yourself desperately needing something which is not planned for and which you cannot afford. It is far better to have the funding available and NOT need it than to make no plans for the funding and not have it when your users do need it!

      In the case of Education IT, "need" is based upon what your educators determine the need to educate their students. IT prioritizes need when funding is tight. By giving educators a certain amount of discretionary budget for IT, they become part of the decision-making process and they determine whether they need it enough to put in some money!
      M Wagner
  • LTSP

    In your article and the article from which this one has been derived the premise remains the same.

    That, it takes new hardware to provide access to new software.

    Having actually used LTSP, I have to argue strongly with that.

    Of course technically, the new software is being run on a reasonably powerful machine, but the access clients do not require that level of sophistication to provide 'better' than adequate performance to cutting edge software, with local sound.

    Even with the additional cost of suitably robust hardware for the server, the performance of a machine suitable for 20 users in a room is relatively small compared to the price of 20 fully specified machines to carry out the same tasks.

    As I explained in a response to the first article, machines are made redundant based on time scales - which if your premise is followed is a reasonable thing to do - but if you step back and look at what is achievable with thin client technology then that so called redundant hardware is actually much more valuable.

    Your premise states that 20 machines go into a school, 3 years on the warranty ends, the drives start showing signs of wear, then it is reasonable to throw them out. Replacing them with bright shiney new machines. Hardware numbers stay the same, no additional benefit to users, additional drain on funds.

    Your argument that the extra cost of maintaining these devices seems robust on the surface, until you recognise that as thin clients the mechanical drives that are most likely to fail are not actually required. Pressed into service as an LTSP client, the user sees a machine that has all the performance and software required, but does not need to be encumbered with mechanical devices. Stripped, and as a solid state machine with a decent keyboard and mouse it can last for years longer.

    In fact, with the stability of the OS, and the solid state nature of the clients such a system requires less maintenance rather than more.

    From experience in schools, the vulnerable parts from students are the floppy drives and cd/dvd components. Harddrives take a significant bashing of course, so why have one per machine, where two or three on a RAID equipped server can reduce maintenance.

    In this way the 20 old machines become a fully equipped computer suite, with only a small additional outlay for the server, the new hardware gets brought in on its replacement cycle and the technicians have it serviced under warranty for those first few months as the hardware is burnt in and failures are most likely to occur, and the school equips an additional room.

    Hardware costs are written off after 3 years for replacement, and any catastrophic failures get thrown out rather than repaired. My experience of thin clients is that they are as close to plug in and use as you can get on an existing server, so where is the maintenance? Keep a few machines in the cupboard for replacement, and get on with teaching.

    OSS allows all the servers to be identically equipped with software (licensing) so building an identically suitable configuration for all of the servers is straightforward. So each classroom has its own server. If you want more than 20 in a room then perhaps two servers would be better? Still cheaper than 20 fully specified machines. Off course, given the savings on hardware, having a couple of pre-configured servers on hand to swap out a faulty device until it can be fixed should be straight forward.

    Strangely, what seems to concern the people in charge of the cash is - how can we justify spending ??40 a machine each year for the M$ tax for licenses, when these machines and software are costing us nothing, and are too old to run M$ software! But then off course the next response is, how can we justify requesting more funding if we don't need to spend that much after all? Is it because big budgets, require big salaries for the responsibility?

    Perhaps the hardware and the software are not the problem. Perhaps its the consultants justifying their significant salaries for the extortionate budgets involved in equiping schools with hardware and software....?
    • LTSP is a great solution when it can meet ...

      ... the complete spectrum of your students' and edcuators' needs. Further, if re-purposing machines which were originally overkill allows you to re-balance your hardware environment then fine. A sensible approach to life-cycle funding is to buy a few pieces of hardware every year so that, no hardware is too old and some is newer (and therefore more robust) than others. In a LTSP environment, the next time around, you will buy thin clients to begin with -- which themselves will be on a reasonable life-cycle.

      No matter what life-cycle you adopt, this needs to be built into your funding model or "the powers that be" will find a way to justify cutting your budget on a whim because they will perceive that everything can be re-purposed indefinitely.

      Even in a mature thin-client environment, a certain percentage of your needs cannot be met with a thin-client solution. Those needs are addressed with more robust solutions.

      Education IT cannot adequately determine "need" in a vacuum. IT needs to engage educators to find out the tools they have deemed appropriate for their students. If their needs can be met with thin-clients and OSS, fine. If more robust solutions, (or even proprietary solutions) are needed to address the instructional needs of faculty and their students, then that's another story and a balanced introduction of more robust machines can (and should) be introduced into your environment.
      M Wagner
      • The world is flat, not round, don't waste your time"

        In his article Christopher Dawson made his case against Linux saving the world, which you supported - as it would take up too much of his valuable time.

        You seem to have just agreed with all of my points, including the difficulty of making the decision makers aware of the alternative ways that computing technology can be used for the benefit of its users and reduce the detrimental environmental effects on the planet.

        Of course as you say it needs the decision makers to be educated as to how that process can be achieved.

        It is unfortunate that the premise that was supported by yourself and Christopher is that it is a waste of time to get tangled up in this process of changing the thinking and attitude of those decision makers, and blog in a manner calculated to suggest that they like yourselves have better things to spend their time on - like checking out the specs of all that new hardware for Vista perhaps?

        So, of course it is more constructive and informative to those looking for answers to declare - don't evaluate Linux. Don't waste your time evaluating it to see if it can reduce upto 80% of your standard computing needs. Don't find out how it can be used to reduce your energy bills, and those of - How many schools world wide (with how much of an energy cost- and environmental impact)? Don't find out how it can provide more of your students with greater computing ratio's than they can dream of at present.

        I, personally, do not have my own Blog pages, supported by a major international company that is respected and read across the globe. But if it did - I would want to use it to educate those I could reach - rather than tell them not to bother.

        "The world is flat, not round, don't waste your time"
        • Neither Chris nor I said ...

          ... 'don't waste your time'. The point is that there comes a time where the cost (in time and money) of extending the life of old hardware exceeds the value of the effort.

          For instance, if you can meet your needs with a thin client model, then it is probably less expensive, in the long run, to buy new thin clients than it is to nurse along some old hardware designed for Windows 95 which has been re-purposed to run as thin clients.
          M Wagner
          • It's do able, it's not brain surgery....

            Christopher's summation:-
            "Most importantly, it isn?t an excuse for administrators to stretch my hardware lifecycles to 8 years, just because some British hardware company says it?s doable. We still need reasonable lifecycles of 3-4 years, regardless of OS. I would still rather spend my time teaching kids to use computers as powerful tools, rather then spend time tinkering with hardware to keep it alive beyond its normal lifespan."

            and from yourself:-

            Some of Chris?s readers scoffed at the 3-5 year life-cycles to which Chris referred ? arguing that with Linux, much older hardware would do just as well. Others took pride in being able to turn old Windows 95 systems into Linux thin-clients. Well that?s fine and dandy ? if you ignore the human costs associated with maintaining old hardware.

            Both of these statements scream 'Don't bother'...

            Both declare that using extending the life of hardware is going to add to the administrators troubles and not really worth the effort. Leaving the students to limp along on pathetically inadequate hardware and software.

            Both of these articles 'scoff' at the suggestion of going to this trouble... and to any of the Linux advocates that are suggesting taking such hardware and seriously trying to limp it along with some 'cut down' distribution I have to agree with the premise that you both advocate.

            However, extending the life of hardware through the use of LTSP, and ditching the mechanical hardware as previously explained seriously extends the useful life of the device without adding to the administrators workload. Particularly if it is understood that subsequent failure means the device is ditched. There is, after all, little that can be done when solid state devices finally go to "silicon heaven" (not impossible, but it is generally extreme).

            But can schools get better value from there investment? - Yes.
            Can they save the environment by extending the useful life of the hardware they have purchased? - Yes
            Can they even reduce some (not all) of the complications associated with running networks in school? - Yes
            Can they reduce the power consumption drain from these devices? - Yes
            Are these tools powerful? - Yes

            Do I think it is important, given these answers, to advocate this technology rather than dismiss it to the people that are likely to read your Blogs? - Yes

            Can you still teach kids computing? - Yes - ( even more of them and mathematics, english, language studies, sciences, etc, etc)

            It's do able, it's not brain surgery, and it can make a difference. But way too many people aren't even aware it can be done. There are, after all, a lot of schools out there. That is a lot of resources, and a lot of energy.
      • 'soap' salesman.

        One question.

        "LTSP is a great solution when it can meet ...
        ... the complete spectrum of your students' and edcuators' needs.

        Why does LTSP or any other technology have to meet "the complete spectrum of your students and educators needs." ?

        If you had an 80% market share in your line of business would you not be happy with it? Perhaps ecstatic?

        So why then would it not be worthwhile to utilize a tool like LTSP, combined with OSS to make significant budget and energy savings, across education?

        Providing computing resources in schools is one of the most difficult areas to resource. The shear volume and variation of software and intention of use is staggering when compared to other business users. But I would be confident in guestimating about 80% of resources in modern schools could be provided with off the shelf OSS packages, under the 20% 80% law, (i.e. 20 % of the software covers 80% of the requirements.) Even at 80% we are talking big savings.

        So does it have to be all or nothing?

        Will students really get a second rate education if there software is not purchased from a large Monopolistic corporation?

        How much will it cost us to BUY a 'complete spectrum' package for education?

        Are we teaching them to be independant thinkers, to be more concerned with the content of their efforts.

        Or are we promoting a particular brand of software because its the 'right one to use'?

        I didn't get into education to be a 'soap' salesman.
        • The 80-20 rule applies in most instances ...

          ... but education is unique because meeting the needs of MOST and NOT ALL means that gifted students go unchallenged and less-gifted students fall behind without the added resources.

          Educators cannot afford to be hamstrung by Ed Tech's unwillingness to consider those special needs. It's Education IT's job to advocate for their educators -- not just tell their educators they have to "get by". This means accepting mixed solutions -- not just Linux, not just OSS, not just Macintosh or Windows, and not just LTSP. The answer lies in a variety of solutions available to a variety of students and educators.

          Administrators are all too often ignorant of the importance of life-cycle funding based upon a sound cost-benefit analysis not just a seat of the pants approach to making do. It's our job to show them.
          M Wagner
          • Better capital investment

            Sure, proprietary software does provide 'some' additional advantages at this moment in time. But is it really true that the needs of the few outweigh the needs of the many.

            Where all hardware software combinations should be based solely on these proprietarty packages at the cost of creating a digiatl divide so wide that it becomes devisive.

            Do you really think students cannot be challenged (educationally not BLING) by making use of the wealth of OSS and extended life of hardware through using LTSP ( or something similar on a propriatery system)?

            However, when capital is freed up to provide those "essentail" additional resources that may be propriatary, and make them available to an even wider audience is that not GOOD for education?

            I repeat a previous question, as educators am I concerned with the content of what my pupils write, or whether they write it on propriatary software?

            Does a mathematically formula loose its validity because it has been produced in LaTeX?

            Does a graphic from GiMP or Blender have poorer composition?

            As for those small areas whereproprietary software has the edge, that is changing. Linux has the ability to move on and change, propriatary software has by design to build 'Lock in' into its software. Should the educationalist lead the software or the software lead the educationalist?
          • OSS solutons can very often meet the ...

            ... personal productivity needs of students and educators but a growing number of textbooks include software which requires Windows or Macintosh. Further, many UNIX and Linux applications have been ported to Windows and Mac OS X and more and more faculty are asking for Windows and Mac ports for the applications they once ran under UNIX/Linux. Further, some educators want to to teach their students to use the tools they will encounter in the enterprise -- which very often are commercial products.
            M Wagner
          • Recursive argument

            Computers become landfill when they are no longer capable of running proprietary software adequately. The mind set has to be changed as to what Linux can do with those systems to clarify why amd how they are much more cost effective and energy efficent than the proprietary systems.

            The savings and advantages for the environment are real.

            If educationalists are aware of these advantages, and of the impact they can have on saving the environment and on the administrative advantages, then they would be asking for the software to work on Linux.

            Strangely, a large percentage of the software that schools make use of are information based. Displaying html, pdf, or Flash files in an interactive sequence. All of this information can be read on Linux, but invariably they come with some form of serving utiity that runs as an executable under windows. A small change in the structure of these packages would make them work on either system.

            But educationalists although computer literate are seldom capable of making such requests of developers when they identify off the shelf software. Instead we need to buy it off a different shelf. Invest some of that educational cash in devloping these packages.

            As long as the marketing engine of the proprietary systems leaves the user believing there is no alternative then there will be none.

            In some cultures they believe that if they are cursed, they will die... and they walk away and die...

            Not so long ago people believed the world was flat, and they would fall of the edge.

            Belief is a very powerful force to influence the minds of the less informed.

            Even though 90% of the planet, get their software from one source because they belive it's the best choice (convenient choice when bundled with a machine - such a great marketing tactic). It doesn't make that belief absolute.

            The problems we are having with the environment and energy use are real, education is a big market of influence, ( which is why so much proprietary software is fed to schools under dubious contracts ).

            But, Linux users will not walk away and die just because others believe that it is dead. It's growth is increasing. More and more users are seeing the reality for themselves.