Linux optimize

Are we allowed to mention cheaper is not always better?

IT in general becomes a better value with improving technology every single year, but that does not equate to cheaper being best.

The register has published a couple of articles recently that have been gnawing away at me. Brid-Aine Parnell reports that just 5 percent of UK CIOs surveyed by the Corporate IT Forum consider Google a credible supplier to business, citing "missing features" when compared with mainstream offerings from companies like Microsoft.

For the last couple of years it's seemed that talking disrespectfully about any element of cloud computing was corporate suicide. I've sat in meetings where organizations with anywhere from 40 to 100 users have asked us why they can't use Google docs rather than Office with a fileserver, why we recommend laptops and workstations when PC World or Comet have alternatives for sale that may be hundreds of pounds cheaper, and why, on one occasion, we were recommending a phone system that was costly when Skype did everything an office phone did and cost nothing.

On the September 3, Rik Myslewski commented on Net Applications' monthly ‘Net Market Share' survey, which calculates Desktop Operating System Market Share based on internet usage reported by 40,000 websites worldwide.

Windows 7 has just overtaken XP with 42.76% of the market share against 42.52% for XP. That's over 85% of worldwide desktop computers running either Windows 7, the version of Windows currently in the shops and that a number of IT departments have, possibly reluctantly, upgraded their workforce to, or Windows XP, the version that shipped on PC's between 2003 and 2007 and that plenty of large IT departments still choose to deploy.

Of the remaining 14.72 percent, Vista, reviled by users and IT departments everywhere, claims 6.15%. All flavors of OS X account for 7.13% (the largest single contributor was Lion, with 2.45%) leaving Linux with 1.10%.

Most IT pros have used Linux at some point; its appeal is too great not to. It's powerful, stable and incredibly configurable, and costs either nothing or very little. Linux is fun to get working and to work with. So why 1.10%? It most likely comes down to the fact that most PC users are not prepared to put much learning time into using their operating system, and Linux is far enough removed from Windows that, other than for standard tasks, navigating the OS will demand at least some investment. Also, the variety of builds and hardware, support (professional or otherwise) has to be more awkward. It's easy to overlook that one of Apple's greatest strengths is the benefit of designing both the hardware and software and knowing they function well with the other. Support is simplified.

Most people, particularly those making IT purchasing decisions, will have used a PC. Discussing whether Linux is a good fit for them, especially in a corporate environment, would be fairly straightforward. So why do we struggle with convincing people that Skype is not an alternative to a fully-featured phone system, or that Google mail can only be compared to Microsoft Exchange in the most basic of functions?

Part of the answer lies in the hype of the cloud, promoted as all things to all men (at least IT-wise) while saving money too. Has a concept, gadget, or major software release ever generated the same buzz? The triumvirate of new, technically advanced and cheap is an extremely powerful lure. The cloud will change how we work and will have relatively low-cost elements, but is not an entity in its own right. Gartner industry analysts report that cloud computing has passed through the hype stage and is now entrenched in the "trough of disillusionment." This is not a comment on the state of IT in 2012, it's the expected reaction to the hype created by those intent on making a buck in the short term. In the middle of difficult trading conditions, it was easy for decision makers to be seduced by the promise of better and cheaper, without needing to test a product in action.

This last point is pertinent in the small- and medium-sized business (up to perhaps 500 workstations) marketplace. We can tell our clients where we think cloud solutions will suit them and where they won't. Actually giving them a working demonstration is significantly harder; there are numerous difficulties to moving an office, department or team over to Gmail, or certain folders off the file server to Google docs. We've done it and it's awkward and clients don't want awkward from their IT; they want the additional benefits and lower costs advertised.

There's also an issue with IT departments being reluctant to say no. In those meetings where decision makers are pushing for Skype and Gmail, or cheap laptops and workstations, we want to be as helpful as possible and it's absolutely our duty to recognize the benefits of lower cost and to either provide an agreed solution (specification and features) at the lowest price, or to be perfectly clear about the different feature sets at different price points. It's rarely our place to set the budget, but we must be able to deliver the best option at any price point or to demonstrate why a cheaper option may be of lesser value.

That can be a more obvious problem during a recession when companies are receiving dozens of calls every week from IT providers pitching for business offering what they claim is both better and cheaper. We're in no doubt that the cloud will continue to be a huge benefit at a huge number of price points. A $1,600 workstation will be better value to some users than a $320 netbook, while plenty of users will find Gmail does exactly what they need at a fraction of the cost of Exchange 2010. IT in general becomes a better value with improving technology every single year, but that does not equate to cheaper being best.

11 comments
michael.burgess
michael.burgess

Does anyone remember the quote from Armageddon where Rockhound mentions that they are sitting on over on millions of pounds of fuel, close to 300 thousands of parts and all built by the lowest bidder... In one should apply a Gaussian curve involving lowest price bidding, purchases etc.. Essentially, the lowest and highest prices are rejected where all other fit into a best fit scenario.

georgeal
georgeal

And, of course, more expensive is not always better either. 1.1% Linux market share ? First, please show the percentages for Servers as well not just Desktops. Then consider the market share of Desktop Linux in India and China. The figures for Linux are totally inaccurate as it is a hidden quantity, with Microsoft and other figures being based on sales - but there are no sales of Linux to compare. Then there are so many Linux flavours that a comparison becomes impossible. As a further example, what about the people and companies who are following Linux From Scratch and building their own Linux flavours. Then consider all the devices that have embedded Linux like D-Link NAS devices. Linux with KDE is easy for any Windows User to use, and many applications now in use on Windows started life on Linux, BSD, and Unix. If you doubt that Linux is marching forward and think Microsoft will continue to dominate, remember Digital Research and CP/M, most people thought they would never be displaced, but where are they now ? Also consider that Apple OS/X is really a Linux "type" system and Apple has been outselling Microsoft.

sysdev
sysdev

Everyone wants to save money. Saving money is good. That is what everyone learns growing up. But there are so many ways that saving money is not even close to being the best way that it is hard to convince people. Saving money in the short term can and frequently does do the opposite in the long term. A case in point. Most people are at least aware of situations where short term 'savings' become long term disasters. The Royal Bank of Scotland (RBS) made a decision to save money by outing a large portion of their technical staff. They thought that they were saving money. Those positions were replaced with cheaper offshore people' They were totally and absolutely wrong as they discovered on June 9 this year. Some of the people who they had not discharged attempted a weekend update to CA7 (a commonly used scheduling system worldwide). For some reason, the upgrade failed and needed to be backed out. Only the people who installed it did not have the proper authority to do that, so it was done by an outsourced resource. The backout was successful, but the person in India also wound up deleting their entire production schedule. They couldn't run any of their systems as a large organization must do things in a certain order or things will not work properly. With no production schedule, RBS was essentially turned into a 19th century bank. They had no access to their computer based data RBS is (or at least was) the largest bank in the world. They could open their offices to accept and disperse small amounts of money, but major systems were (and as of mid October I have not heard anything about them recovering) unavailable. Some time ago I saw an estimate that their direct costs in this fiasco were in excess of 100 million Pounds (Great Britain). I have seen nothing about the cost to recover or what it is costing their customers. So they saved a minuscule amount of money and did major damage to the business. The problem was clearly caused by an inadequately trained person with the wrong level of system authority. That person was inexpensive compared to the person he replaced and the damage he did was enormous. This type of problem is happening with outsourcing all over the world. Short term cheap translates to long term expensive (or catastophic).

john-paul.sivori
john-paul.sivori

yes - think back a decade to the irresistable prospect of being able to virtualise ALL your production servers, reduce power and hosting footprints, maximise use of server resources, provision servers in minutes and deliver resilience beyond imagining. Then, the counter arguments were customers were putting all their eggs in one basket, that they needed phenomenal amounts of hardware to deliver performance+resilience which in many cases wiped out the advantages of virtualisation. At the time I counselled using virtual hosts for develpment environments where performance and resilience weren't so key. This time I say the same thing. Use it for development servers first and when you've uncovered all the gotchas, move your prod environments on there. But will they listen? Will they heck!

Deadly Ernest
Deadly Ernest

don't think you can win, just point out how all that cloud stuff is supposed to be platform independent and that they can save a hell of a lot more by not buying Windows at all but putting Linux on to run the browsers and thus extend the life of their current hardware by many years. Go with Zorin OS with the Win 7 or Win XP desktop GUI, and it should be an easy sale. I've found the best argument being the prison sentence they face for violating various federal and state laws about the management of privacy related data.

Tony Hopkinson
Tony Hopkinson

Are we allowed to mention that cheaper isn't always good enough and will never remain good enough? The answer to both questions is of course, no. Well not unless you wish to suffer the traditional fate of a heretic...

dave
dave

good, fast, cheap. Pick 2, you can't have all 3. You have to remember that when you use external be it the cloud or VoIP (Skype, MagicJack, Primus, etc) you are putting your business in the hands of someone else. What if you cloud provider goes down? How long are they down for? Can you run your business without that data and how much do you lose per hour down? What happens if Skype is down or not available? Maybe employees can use their cell phones to call out but what about customers calling in? You may be able to give sales a tablet but not a design engineer or programmer. One has to do a risk analysis and decide how long can you be down and how much does it cost per hour when you are down. CIOs and management need to do this rather than jumping on the last ad on the back page of that tech magazine.

Deadly Ernest
Deadly Ernest

a very expensive high dive. A few more bucks spent on making sure everyone was on the same page and using the same specs would have saved that mission from being written off.

David.Izen
David.Izen

The figures for desktop market share were compiled by interrogating data from 40,000 websites and the OS's that visited those sites; the data could never be interpreted to show worldwide adoption of Linux, but rather is indicative, I feel, of the breakdown of OS's in use by the average Western user. Point taken about West vs. East - I suspect a study of desktop OS's worldwide along with the amount of uptime or fettling time required to keep them working would make interesting reading. I feel in the longer term you're right that OS's will be cheaper, but would bet recent trends of Apple and MS charging way less for the OS will see them maintain popularity and market share. Ease of support will be critical and the single-flavour of OS, along with the greater tie-in between hardware and OS (Apple today, but MS soon?) will carry significant weight. Whether any support advantages (i.e. time taken to maintain) justify the increased cost of these OS's will be for users and IT departments to decide.

Tony Hopkinson
Tony Hopkinson

and pass exams and everything before you understand that saving money = spending less. We are just too thick to get this business stuff....

Deadly Ernest
Deadly Ernest

account for all the systems that refuse to send that data or send fake data. It's very hard for anyone to have a Windows system refuse to send the data when queried, and even harder to have a Windows system fake it. Yet many Linux system and many Unix systems refused to send such data by default, and most of the rest are easy to set up to refuse, while a lot are also easy to set up to fake the data. One thing I learned many years ago was to set up Fire Fox to say it was MSIE on a Windows system. This was due to so many systems being 'optimized' (real meaning being screwed up to only work with) MSIE. Thus I'd have my Linux FF system tell the asking system that I was running MSIE on Win XP Pro and it never knew any better. What was interesting is a site I could see perfectly like this was often unreadable garbage if I didn't set the system to lie. BTW Some of the sources for the stats on what OS is in use are based only on sales figures from the major vendors. While some others are based only on querying systems accessing a long list of US based websites visited almost exclusively by business users, which makes them unrepresentative of anything but the USA business world as they don't properly represent the USA home users well.