Article and followup

[Jason Greenwood]

Hi all,

I had an interesting discussion with several people yesterday regarding this article:
http://computerworld.co.nz/webhome.nsf/NL/D15D1179C0CAFE78CC256CE10008DC54

It quoted David Lane and so I emailed him to get his thoughts while simultaneously submitting the story to LT. I also rang the company in the article to do some Linux advocacy. I also forwarded David's rebuttal to the company as well just so they had both sides of the story. As it happens, I also forwarded the rebuttal to Brian Profitt (Managing Editor of LT) and he is going to post it too, once he gets David's permission (which was forwarded to him this AM). Thought it might be of interest to you all.

Here is David Lanes rebuttal:

The Datasouth URL is at http://www.datasouth.co.nz/case_studies/pacific_wide.htm

My critique is below.

Cheers,

Dave

Hi Andrea - just looked at their site.  The case study related to Pacific Wide makes a lot of vague statements. which range from being unsupported assertions to outright falsehoods.  For example, Linux can (and does) do just about all of the things they state it doesn't do, including the ability to provide "remote network access."

I'd say that they might have run into a poor supplier of Linux services, but even then, I'm not convinced that the case study wasn't simply politically motivated - i.e. the former "Linux champion" at Pacific Wide, or whoever got Linux in there to begin with, might have moved on, and the next person wanted to "make their mark" by undoing what had been there before regardless of whether or not a change was warranted.

It's also unclear how recent a version of Linux was being used.  If it was, say, 5 years old, then yes, it wouldn't have all of the capabilities required, but then again, neither would any Microsoft product of that vintage.  If it was a relatively recent Linux system, then this quote by Anthony Washington is uninformed and/or patently false:
“The Linux system also didn’t provide us with remote network access or Web site hosting, which became increasingly important as we expanded..."

I have built a business around Web Site Hosting on Linux - if anything, that task is the reason companies switch to Linux!  Linux is the most widely used web serving platform on the Internet (see www.netcraft.com).    60% of websites are powered by the open source Apache web server, and the majority of those installations are on Linux.

As for remote network access, it's quite trivial to set up - I have set it up for clients personally, and use remote access (on various levels - from command line to full desktop access) daily for all manner of system administration, document management, and software development tasks.  The remote access tools available for Linux equal or surpass those for Microsoft in nearly every context.  They are also available for free or a much lower cost than Terminal Server - e.g. commercial products like Netraverse's Win4Lin Terminal Server allow a Linux server to provide Windows sessions to remote users just like Terminal Server, at a much lower cost (see www.netraverse.com/products/)

I also find it hard to believe that their Linux system was an "unstable and unreliable platform."   You might ask them
a) when was their original Linux platform installed?
b) what version was installed?
c) what services did it provide?
d) what was the hardware specification of that platform?
e) did they ever upgrade the Linux version after it initially installed?
f) does the new Microsoft system run on the same exact hardware as the Linux system?
g) who was their Linux supplier?

If they are running the MS system on new hardware
h) did they get any proposals for functionally equivalent systems built on a current version of Linux?
i) do they know what sort of hardware requirements a Linux system providing equivalent functionality would have?

Also, what does "strategic advice" or insight on "technology mapped to their business" mean?  I suspect  that they are utterly subjective. They simply give Datasouth an opportunity to say things like "we think you should go with Microsoft because that's the only vendor we deal with.  We've bet our company on Microsoft retaining its monopoly hold on NZ businesses, and as such we believe it is a platform that will be around for a long time, and is therefore worth of your investment."  A Linux vendor such as Egressive might say something like this:  "We believe that you'll get good value from a Linux-based server solution because it allows you to leverage your existing investment in Microsoft desktop systems and software, while providing you with huge increases in capability, lower server hardware requirements, and a much better security record.  What's more, you no longer have to worry about server software licenses, you can use as many desktop clients as you like with no additional cost, and because Linux is an open platform, you're not locked into any one vendor or software package. If you don't think we're providing you with a good value for your IT dollar, you can go to any one of the hundred or so other Linux vendors active in NZ, including many small and medium sized local vendors or larger ones, for example, Gen-i, Computer Concepts, HP, and IBM."

The statement by Aarron Spinley that "Microsoft is the industry standard" is also very much an assertion without proof.  For better or worse, Microsoft is the defacto *desktop* standard, yes.  But it is by no means the standard server platform.  Issues of unreliability (especially related to webserving), virus vulnerability and security holes (see this recommendation from the Gartner Group: http://www3.gartner.com/DisplayDocument?doc_cd=101034), high maintenance requirements, and significant expense (especially in the face of MS's new licensing regime) make it very unattractive compared to Linux.   Many businesses and organisations have shifted to Linux as a server platform, and many others still use Unix and Novell-based networks.

A number of Microsoft partners, including Datasouth, install Microsoft servers because they don't have Linux expertise (and Microsoft licenses and the intensive support requirements of their products afford them the highest margins).  That lack of expertise, however, is their limitation, not a problem with Linux.

The most telling comment is this on from Anthony Washington:
 “Having a standardised Microsoft platform has lowered our Total Cost of Ownership (TCO) and helped us become smarter about our licensing and asset management. Furthermore, because we will now upgrade our IT systems every three years, we expect a quicker Return on Investment.”

I can't imagine how he can make a statement like that.  How can he compare the cost of an equivalent Linux system in terms of TCO if he hasn't got one?   What aspects did he take into account to determine TCO - it's a lot like a "how long is a piece of string" type problem.  I have clients who would say that their Linux systems cost them less than half, in total, than what they were paying for the Microsoft-based systems that our Linux systems replaced...  and enjoy much higher flexibility and reliability, and reduced administration costs, in the bargain.   Oh, and one of their favourite things it the fact that they can upgrade... when they need to, and not when Microsoft starts shifting the playing field by
a) making it impossible to purchase new versions of the software they're still using quite happily (i.e. Windows 98 and Office 97 in some cases), thereby forcing a mix of new and old software into the company which inevitably causes compatibility issues between new Microsoft software and their own older stuff...  Strong arm tactics? You bet.
b) forcing clients to register for new ways of extracting revenues like "Software Assurance"... which provides those clients with... nothing.

As for becoming smarter about licensing and asset management, that's a red herring - how hard is it to manage something with a license that doesn't place any limitations on you at all?!  Unless it's used with proprietary software, having a Linux server does away with any server and network software per-client license management requirements full stop.  Do businesses like Pacific Wide ever consider how much easier (and therefore less costly) their lives would be if they didn't have to manage licenses?   I suspect they don't, nor, I suspect, does their TCO statement...  I wonder what else they forgot to take into account?

There you have it, Andrea - there're probably a dozen other holes in the article, but I think I've covered the high points, and I haven't got time to say more.  You must really enjoy the fact that I rise so quickly to these challenges you send me :o)  (I thought my Effusion Group colleagues might also enjoy reading the "case study"... )

Regards,

Dave

======================

Regards,

Jason Greenwood