Thursday, July 10, 2014

A new perspective on old Operating systems


I see a lot of uproar over Operating systems lately. 

Not the normal Windows versus Mac versus Linux fare.  No, the wailing starts around the time an older version is no longer supported meaning no further updates are produced for it and every request for help is met with the maddening phrase...

"Upgrade to the newer version."

Now truth be told, we rarely hear anyone complaining much about the abandonment of an old version of Mac OSX or a Linux distribution.  Devotees of those platforms are usually breathlessly awaiting the next release. 

It's usually the Windows world that complains the most and with good reason.  Their customer base is always wary of a new Microsoft release especially if it follows a successful predecessor.  That's because success is usually followed by failure.  A short review of recent history bears that out..

  • Windows 2003/Windows XP - Stable, easy to manage and user friendly at least after the first service packs.
  • Windows Vista - I don't care what Paul Thurott says, it was/is slow and almost completely unusable before service pack 2.  Microsoft's "my way or the highway" stance didn't help it gain many fans either.  It's the reason why Windows XP only recently entered obsolescence.
  • Windows 7 - Everything Vista should have been and for many the worthy successor to XP.
  • Windows 8 - A sign of things to come in interface design but a bit too much too soon.  Great for tablets but lousy on anything without a touch screen. 


So the trend is evident.  Expect the release of a successful Microsoft product to be immediately followed by a failure. 

Perhaps a change of perspective is needed here.

Operating systems are just "products" and products have a shelf life.  That means after a certain amount of time they go "bad."

So I'm proposing we think of operating systems in the same way we think about cars. 

When you buy a new car you can expect that for at least 5 years you're going to have full support from the factory that made it.  Any little thing that goes wrong will be quickly and cheerfully rectified and you can bet that all the latest features will be there for the taking.


Now moving past that 5 year mark we're getting into the realm of "mature" products.  You may still have a warranty but you're going to find that if the sun visor falls off it's going to be fixed but it's probably going to be on your dime.  If the engine blows up, however, it'll probably be covered.  Don't expect any feature updates from the factory though.  They've moved on to this year's model which more likely than not isn't much different than yours so there's no real compelling reason to "upgrade."


Once we get to 10 years you now own an old car.  You're probably not visiting the dealer to get it repaired anymore mostly because it's too expensive.  If you hold on to it you've either got a trusted mechanic close to home or you've suddenly taken an interest in automotives.


Over 10 years and the words "End of Life" start showing up.  At this point you're into legacy status and the factory could care less about your car.  If you really like the car you'll find all sorts of "aftermarket" parts and services to keep it running.  Chances are the factory and dealer won't be able to do much for you at this point. 

I think the model works pretty well for operating systems too.  At some point you have to accept that nothing is forever.  A 12 year old operating system is like a 12 year old car.  Obsolete, poorly supported by its maker and dependent on an "aftermarket" to support it.  There's nothing wrong with the aftermarket so long as they have good standing but it's a roll of the dice.  Also know that just as an old car may not have the safety features or fuel efficiency of a new model, an old OS may not be as secure or perform as well as a newer one.


So run Windows XP or MAC OSX Tiger if you want, just know that from now on you're going to be on your own.

Wednesday, July 2, 2014

Web hosting for cheap on a virtual machine


The thing you have to remember about working in IT is that no two projects are ever alike.  Even if you're being asked to do the same thing for 10 different people you're still going to be surprised.  Sometimes even on the same project.

So it was with my latest foray into virtualization on the cheap.  The client could barely afford to pay me let alone invest thousands in licensing fees.  So we had to get creative without sacrificing stability. 

That can be a tall order especially when everything you're using is Open Source. 

Now I have my issues with the way the Open Source community does things but a good product is a good product regardless of who made it.

Of course, "good" is a relative term. 

It's always a trade off.  A bit of pain to save a lot of money is fair but too much pain can cost more than if you'd just went with a commercial option.  And I do mean "commercial" because I still firmly believe that any product that relies on a fractured support community or high priced "experts" to make a product work is just this side of an amateur effort. 

Not that all open source products are that way, however.  

Some communities are better than others and if they put together a solid package with "readable" documentation then I'm all for it.  If we're just stroking somebody's ego so they can get a guest spot on Floss Weekly I'll take a pass every time.

I put CentOS, the open source version of Red Hat Enterprise Linux, and Z-panel, the open source clone of C-panel squarely in the "good" category.

Together they offered a cost effective and relatively stable platform for web hosting.  Add in a virtual platform for them to live on and you've got a web host that could fit on a keychain.  Not bad...

Instead of bore you with 4000 words of text describing my latest open source virtualization adventure I've created a video that takes you from creating the virtual machine to administering your new web host. 

As you're watching you may miss a few of the links in the video.  I've provided them below.