NAB walks path to Windows 7

NAB walks path to Windows 7

Summary: National Australia Bank (NAB) is testing Windows 7 to assess its viability for an upgrade from Windows XP, which it finished migrating to last year.


National Australia Bank (NAB) is testing Windows 7 to assess its viability for an upgrade from Windows XP, which it finished migrating to last year.

A spokesperson for the bank today confirmed to that Windows 7 tests with Microsoft were proceeding under the software giant's Application Compatibility Factory. NAB currently supports around 700 applications.

The bank is a key partner to Microsoft, according to the spokesperson, who said NAB was the "only organisation in the Asia-Pacific region" which was a member of the so-called global Microsoft Deployment Council.

"This provides NAB with the forum to provide input on product strategy and development and deployment plans," the spokesperson said. "NAB has a strong strategic relationship with Microsoft, so we can remain current and leverage Microsoft innovations to best support the bank."

NAB was one of the last Australian banks to move to Windows XP. Last year it completed the migration of 28,000 desktops from the unsupported Windows NT operating system. NAB's desktops were until recently managed by Telstra; however, it ditched the telco last year for IBM, under a deal that is set to expire in 2010.

Any future upgrade of NAB's desktop operating platform is likely to be a low priority compared to broader changes occurring within the bank. NAB last month announced it would replace incumbent CIO Michelle Tredenick, who was heading up its $1 billion core banking system overhaul and overseeing its IT offshoring efforts, which were thwarted by the financial scandal its major IT outsourcing partner Satyam recently underwent. Incoming CIO Adam Bennet is expected to start this month.

The spokesperson said NAB "welcomed" the announcement that Satyam's fellow Indian outsourcer Tech Mahindra will take a majority stake in Satyam, but added that NAB had no plans to hand over any further processes to the company.

NAB's Windows 7 tests put it ahead of the Commonwealth Bank of Australia in terms of testing Microsoft's new system. CBA recently revealed it had not tested the new operating system in its production environment because Windows 7 was still in beta.

Welfare agency Centrelink is the only other large Australian organisation known to have tested the new operating system, with the agency giving an enthusiastic response to tests so far.

Other federal agencies such as the Australian Taxation Office and the Department of Immigration and Citizenship are yet to test Windows 7, while the Department of Defence and the Australian Customs and Border Protection Service — the only major Australian organisation to adopt Vista — both claim to have no plans to migrate to Windows 7.

Topics: CXO, Banking, Government AU, Microsoft, Operating Systems, Windows

Liam Tung

About Liam Tung

Liam Tung is an Australian business technology journalist living a few too many Swedish miles north of Stockholm for his liking. He gained a bachelors degree in economics and arts (cultural studies) at Sydney's Macquarie University, but hacked (without Norse or malicious code for that matter) his way into a career as an enterprise tech, security and telecommunications journalist with ZDNet Australia. These days Liam is a full time freelance technology journalist who writes for several publications.

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
  • This doesn't make sense.

    The bit about the Commonwealth Bank "CBA recently revealed it had not tested the new operating system in its production environment because Windows 7 was still in beta." makes perfect sense.

    What NAB is doing doesn't. Why would you test an unreleased beta operating system in your production environment? What's the rush to move to it? Surely they've got better things to do.
  • Windows 7 Beta NAB Test

    Gavin -

    I understand your point. Though I would see some benefit in a limited production test (eg. half a dozen non-business crtical machines) to get an early indication of likely issues and to be able to raise them with Microsoft. I suspect there is also some pressure on NAB to participate in this given their close relationship with Microsoft, which undoubtedly offers some pricing concessions for their beta participation
  • It does make sense.

    The article pointed out that the tests "with Microsoft were proceeding under the software giant's Application Compatibility Factory". The testing isn't happening in a production environment at NAB, rather in Microsoft's lab's and is to determine application compatibility.
  • A better idea

    Why don't Comm Bank and NAB et al migrate to Linux, like the German government and the French Police Force, and save themselves a packet?
  • Actually,,,, a worse idea

    Linux is not always the answer, unless of course you are a linux fanboy, and NAB are not. Total cost of ownership is significantly lower in Windows environments when it comes to desktops.

    Linux support technicians are harder to come by and hence more expensive, supported software is thinner on the ground (and likely to be totally incompatible with anything they are currently running, so there are HUGE migration costs), and retraining staff to use an unfamiliar environment is VERY expensive.

    A corporate linux environment only makes sense on the server side, not the desktop side - they would be saving NEGATIVE millions of dollars by running linux.
  • Where is the evidence?

    "they would be saving NEGATIVE millions of dollars by running linux."

    Really? Facts please.
  • Actually a worse idea

    Sure, in the short term and if you don't count productivity losses largely specific to Window's environments which continue to be incurred from Windows generation to generation, platform to platform. Changing to Linux is a hard pill to swallow, operationally and (short-term) financially. The real question is how the ongoing indirect and distributed (i.e. user productivity) costs are quantified - or not.
  • Bwahaha

    And all linux users are long haired, smell ridden, acne invested social failures, right?

    You're way out of touch, mate. Linux is not just popular, it's growing in popularity. The latest Ubuntu desktop is on a par with/better than the Windows 7 RC 7100 release for end user usability. The next release of Gnome promises to eclipse any current user experience on the market, including Windows 7.

    Database support? Postgresql.

    Web Serving? Apache

    Web Applications? Glassfish, Tomcat, Websphere, etc etc. All enterprise quality.

    Firewalling? Built right into the kernel as a module, along with a superior network stack.

    Enterprise directory services? LDAP, or if you want to use exchange, Evolution or Crossover.

    Kerberos mutual authentication? Right there buddy.

    Advanced scriptable system maintenance and scheduling? Above and beyond.

    Secure environments? Runs off a live CD.

    Enterprise quality spam filtering? Yep. Virus scanning? yep, not that it's necessary.

    File systems with journaling, encryption, and built in ACL support? yep.

    Windows application compatibilty? Yep, Xen or KVM parallised if you want to virtualise it, or Wine which will run pretty much most stuff the banks would be using without any issues.

    TCO? WAY less than windows environments. Most of the cost in deploying Linux are one off costs of training of staff, ironicaly, about the same price as deploying windows environments (those licenses aren't free you know, and support contracts? Whooeee MS charges by the pound of flesh!). Linux ongoing costs are marginal, not so for windows.

    Yeah, I think that's enough for now.
  • Fan boy

    Nearly everything you just listed is server platform technology and not relevant.

    The original poster is right, Linux is staggeringly expensive in manpower and resources to implement and manage. I know, I just watched it happen at our org in the server space and that involved relatively good IT staff.

    Centralised group policy style management of the OS is a must (RHEL Satellite). A consistent, controlled SOE that can be managed by low skilled techs reduces costs like you would not believe. When HR says they want their innovative "motivational screensaver/wallpaper" pushed out to 3000 distributed desktops you need a solution that gets it out there by the end of the month. No you can not "Just write some scripts"!

    If the C class guys dont like it because his favorite app to integrate his nokia isn't there or his kids cant run WoW you have no chance. If their secretaries dont like the formatting options and font support in Open Office you have no chance. If the sales guy cant get his windows video to run in his powerpoint presentation you're toast.

    HR and the CFO better see the money. Retraining 3000 people to do basic activities they can already do because the last employer showed them is going to cost a LOT. Its going to put a lot of strain on the corporate helpdesk and those IT staff are going to have to be retrained. Whats more you are going to have to train EVERY NEW EMPLOYEE you hire because no one else is teaching them. If you have even moderate churn your costs go on forever.

    IE is a fact of life for most corporates. The Intranet they built on .net/activex or Domino that cost a million dollars developing must run.

    The collections of bizarre legacy apps the company has must run without migration costs. Having just dug through config files full of references to C:\ or entries in the registry I've seen this first hand. Now imagine doing that for hundreds of apps. No thanks. The corp dev team will kill. If they (and they will) have an investment in the existing apps/language you have no chance. Migrate developing platforms? Retrain the dozens if not hundreds of developers? I dont think so.

    Finally Linux is not free. No enterprise is going to take a free unsupported application. They must have risk management and a legal scapegoat to escalate to. Which leaves you with RHEL5 and they bleed you dry with licensing, scoping, migrations and 3rd level support.

    Linux desktop might be feasible for small business but even then your stretching it. I cant imagine the pain of working with it in a place that has 28,000 desktops.

    Stick to using it for Apache.
  • Interesting Points

    I wonder how the ongoing costs of deploying M$ is going to cost. NAB & the other banks have all spent huge volumes of money over the years developing their own software & each one almost replicates the others. They used to do it on Unix & most of the processing software would still be available if necessary. Developing a central enterprise control application wouldn't be cheap but it'd be a damn site cheaper than M$'s ongoing costs. It would also be a sales outlet for them if they wanted. I'm pretty sure there are versions around anyway but as you say you have to know how to use them & that takes training. The difference here is that you can make a script & upload it to every one of 23000 computers if need be & with a little less running around from site to site as is needed by the M$ version.

    The training required is minimal for the teller operators. Most of this is done by server anyway & works on the screen. IE is not a must as you say. That statement is full of it. Firefox is more secure for starters.

    The cost of security within an M$ shop is incredibly high whereas the same is not true in a Linux shop. There is no comparison here other than the dollar cost. The absence of real issues here for Linux is immense in the security game & most of the issues that Windows brings will never be contained because of networking not being the basis of the computer. The UNIX-likeness for Linux means that many of the vulnerabilities are inherently not there because these have been dealt with over the past 30 to 50 years by the UNIX folk & Linux folk since then.

    If you want the support from a major corp then that is par for the course that you will pay for it. I still believe that you will find out that using Linux improves upon the expense of using Windows to a large degree. It will take a bank to make Linux perhaps or maybe a Govt in Oz.
  • Expense, I'll show you expensive!

    Microsoft's math libraries for have significant errors, particularly in the tails of the statistical distributions. This is one of the reasons financial institutions using Microsoft products underestimated risks in their quant (derivatives trading desk) operations. Now that was expensive!

    Key banking operations should never have an operating system and software monoculture. It is truly a false economy.
    Having some different systems to run backup calculations would have revealed the discrepances which then could have be tracked down. Instead they may broke the bank and they don't even know (or admit) it.
  • What a load...

    "A consistent, controlled SOE".. Consistent? In what respect? Its ability to reduce users to tears? Controlled? Yeah, but by whom...

    Retraining staff to move from MS office apps to e.g. Open Office? Missing formats and fonts? What arrogance. Anyone who can live with MS Word et al can be fully productive with OO. IE a corporate fact of life? Ever wandered down an aisle and seen how many of these implied 'need to be retrained users' run other browsers? Retrain hundreds of developers? What do you think they are? Stupid?

    Linux for small business users? Alas, here we see the true arrogance of the self-proclaimed ivory tower IT guru. Get a life, look around and divest yourself of the fear of becoming an obsolete MS advocate.
  • Interesting comment

    Did you have a price comparison for:

    A) The Linux solution + support you mention in your post

    B) A Windows solution which Implements the same functionality?
  • Rexie Babeee Strikes again

    Interesting assertions Rex. I'm assuming this is based on your massive experience? I keep finding lots of Linux people writing this crap are flat out supporting some small business. Ever worked in an IT dept supporting more than, say, 100 users? Or are you a hobbyist?
  • Long Bow

    Now that's one long bow to draw. Methinks if this was the huge issue you say it is there would be lawsuits aplenty. Muppet...
  • ignorance and accuracy are mutually exclusive

    You have one, and not the other.

    Linux had group policy management and in-built access control long before enterprise even thought of bringing it to windows. Hardened linux/SEClinux goes not just one step futher, but many steps.

    Linux supports hosted desktops natively, through the XOrg graphical server, and can handle connections by named pipe, tcp, ssh'ed tunnels and a number of other technologies. In this day and age, there is no reason nor excuse for providing local desktop services to the majority of staff.

    Training people to use Linux is NOT the cost it used to be - in fact, reviews of Ubuntu 9.0.4 put it at a grade of usability above that of the incoming Windows 7. Most people who were raised on XP or 98 still have to be trained to use Vista, by the way, and that device that won't work under linux? Sorry, no Vista drive for it either, better hope like hell you're not using Vista 64. Linux device support is superior to every other operating system on the market, unless you're buying right on the bleeding edge, in which case as we are discussing enterprise "You're a damned fool!".

    Open Office presents an interface that requires no retraining. It is an accurate match to the comfortable Word97, PP97 etc interfaces that everyone is used to, and that are almost universally prefered over the confusing Ribbon design of the later office products. A lot of companies had to soak the costs of retraining their staff to use the new office products after the interface changed so radically.

    IE can be run through WINE if you really need it, but again, in this day and age, there is no excuse for running an well acknowledged buggy and inferior browser. Domino? Domino on Linux has been available for over 7 years. Apache can run almost every .NET web application via mod_mono. If you're coding active-X then that is your fault. No one does ActiveX any more. Spend the money and upgrade your tech, rewrite what you have to - it is cheaper in the long run than trying to maintain a flawed system in an increasingly security hostile environment. But hey, if you want to deploy an ActiveX enabled web application, that supports neither accessibility, nor Mac, Linux, or Mobile users, in the modern digital world, well... hey, it's your corporate suicide, not mine.

    Where I work, we built our networks on Netware, and have just migrated to a new infrastructure - managed exchange and document hosting, support for secure VPNs from home and on the move/on site, virtualised desktops and application serving. Not only can it be done, but it should be done if you value your company's ability to compete in the modern digital economy. Dinosaurs will be left in the dust as their insecure, out dated, out moded, and disfunctional software and hardware infrastructure becomes a legacy consideration to the rest.

    The end point of all this is just this: "You don't have a clue." - tell you what, get some real time experience working with the latest linux products and distributions, and linux based solutions on the market right now, then come back and discuss.

    We just moved from Delphi7/J2ee to bare metal virtualisation of C#/.NET appliances, from Netware4 and Lotus Notes to hosted exchange and document management, from individual desktop machines to thin client infrastructure served over gigabit networking, 3G modems, and individual laptops that are automatically upgraded every year as part of the supply contract.

    Has this had a cost? Yes, of course, but that cost is nothing compared to the ongoing maintenance costs of legacy networks, legacy software design and development infrastructure, legacy security controls, et al.

    Finally, you opened with "Nearly everything you listed in server platform technology and not relevant". That is the most ridiculous statement I have read in this discussion. Enterprise is almost exclusively concerned with server technologies - whether it's mobile services, exchange and email, web applications, client-server software archite
  • Linux Total Cost of Ownership gets LOWER in volume

    I've used every M$ OS since PCDOS 1.0 in 1982,.. but I agree that it was wrong to claim that Linux distributed to a large number of workstations in a large organisation would be more expensive than the M$ proprietary route. The larger the number of workstations involved, the MORE sense it makes to run a friendly GUI Linux like Ubuntu.

    With an M$ 'rebuild' you need to source original OS disks (where are they kept?) or dig around for just the right sub-species install disks with a Certificate of Authenticity number.. it takes ages to re-establish just the OS. Then you have to do the same with each proprietary application. And the whole process requires an IT tech to be present. In comparision, with Linux, you can have a $12 device (4GB USB stick) in each branch, and if the PC has major problems, you simply reboot the PC with the stick inserted and it will boot off the stick and then dupe install Ubuntu and all the apps, including open office to make the PC 'like new'... and if that doesn't work you know it was a hardware problem. With M$, you can't do thinks like that in an identical way, simply because you need to have each installation done 'de novo' to put in the serial numbers as part of the install process.

    And on user training, I've found that Open Office with the Australian Spelling Dictionary plug-in, and one-click PDF conversion is FAR better for users than saving things in DOCX format. As to OS training, Ubuntu is truer to the familiar Win98/XP interface than any later M$ OS.

    And teller PCs in a bank branch do almost ALL of their functionality as dumb terminals (or at most have 'smart terminal' status), with all the transactions actually stored elsewhere, and all on-screen (except some viewing templates) being retrieved from elsewhere... so they are in fact perfect candidates for a simple/robust Linux use....

    But kiddies, why worry about what the banks will do. Let's remember that they will never be hot-beds of innovation... (arguably for good reason). Remember the banks changed from EBCDIC character coding to ASCII some decades after most people had changed. Banks are slowly changing from point-to-point 'leased line' telecom links to IP-based networks about 15-20 years later than most. And for decades after others had shifted to 'servers', the big banks were running almost exclusively on mainframes, claiming servers could never play a role in the type of work banks had to do. Banks have in fact always been at the very tail end of IT innovation. Yes, they have IT budgets many times larger than most, but for them to change a major database or accounting system, they report development costs in the hundreds of millions (with a negligible hardware component), whereas other organisations with comparably complex environments might feel embarrassed to report a similar changeover costing a hundredth of that cost! So don't ever assume banks are about IT efficiency - yes they have economies of scale in sheer transaction volumes, but hopeless efficiency ratings on most other measures.

    So in short, it is purely academic to argue if a bank ought support Linux.... They will do so about a decade after it first seemed sensible to do so... but when do make the change, they will big-note how innovative they've been!

    Well NSW DET came to a far different conclusion Graeme! Time to get off that horse and into a car..

    Original source OS disks .. product keys.. bollocks .. you haven't kept up with much in the past few years have you? With half a brain & a day up your sleeve - all dealt with.
  • Huh?

    Why is not saving to docx better for users? I'm happily using the Office 07 pdf plug in and saving files to .docx ... and the problem is??
  • cost of features

    What most people ignore in TOC, is that it differs from it department to it department. If you have a brilliant windows admin he is going to do his best to get the lowest TOC for windows. Likewise for linux. Now, i am biased but i'm going to say this -> windows is built so that the admin follow the "microsoft" path (there are little options there is the microsoft way or the highway). On linux, you *can* choose what you want how to you do it. You can deliver customised solutions that are a perfect fit for the customer. I will bet you any user who has never used a computer would find linux (ubuntu) easier to use than windows 7. The mac is around / above ubuntu (gnome) 's level of ease of use.

    If you don't believe me try both out for your self. I tried windows 7 yesterday for a few hours. Its better than vista but not better than windows xp for a number of reasons. (in terms of the user interface).

    The total cost of ownership MUST include how nice a system is to use. I personally would never given anyone a windows other than xp. Vista + windows 7 do not really improve on the user interface from what xp can provide.

    Summary, user interface is important + toc depends on your admins