< link rel="DCTERMS.isreplacedby" href="http://sirhumphreys.com" > Sir Humphrey's: Rhymes with Microsoft Shill


SITE MOVED:Sir Humphrey's has moved

Please join us at our new site: www.sirhumphreys.com.

The RSS feed for sirhumphreys.com is now here.

Friday, October 28, 2005

Rhymes with Microsoft Shill

I'm a bit late jumping on this bandwagon, but I'm going to do it all the same. Just catching up on others blogs recently and courtesy of frog, came across Francis Till giving from the NBR having a go at the Greens support of Open Source Software (OSS). This is one of the few areas where I believe the Greens are right to support the promotion and adoption of OSS where practicable. I see that even Russell has commented on this issue.

For those of you that have read my blog before joining Sir Humphreys, you'll know that I have a particular interest in the adoption of open source software within Government. I believe that there are significant economic benefits for our Government to adopt open source software as each year many tens, or even hundreds of millions of dollars are siphoned out of the country to support the Microsoft Tax.

It is interesting, and has been commented on by other bloggers, that around 70% of the publically accessible web servers on the Internet run the open source Apache web server software, and this number is growing. Compare this to Microsoft's Internet Information Server (proprietary) which holds a lowly 20% of the publicly accessible web servers.

Fewer statistics are available on the Samba - an extremely popular open source file and print sharing solution.

Dominic on Till's own blog rightly points out that tillnet itself appears to be powered by the Linux operating system , the Apache webserver, most likely the MySQL database, and is running WordPress for the blogging software.

Anyway, onto the fisking.
Even in servers, its strongest point of contention, Linux holds only a very minor share of the market.
More, much of the non-server Linux uptake in the enterprise market during the last several years has been experimental in nature, with little sign of escalation -- often for good reasons that include both security and cost of ownership.

And if you care to read the IDC report linked to, it has some interesting figures about Linux...
Linux servers posted their 12th consecutive quarter of double-digit growth, with year-over-year revenue growth of 45.1% and unit shipments up 32.1%. Customers continue to expand the role of Linux servers into an ever increasing array of workloads in both the commercial and technical segments of the market.

That to me does not sound like "growing -- but only in a limited way and in back-end operations". Rather it sounds like a platform that is running from strength to strength. Back in January even, IDC analyst Al Gillen said that Linux was going mainstream.
Linux server revenue exceeded $1.4 billion in quarterly factory revenue in 2Q05 as Linux server revenues showed 45.1% growth, the fastest rate of growth since 2Q04. Linux servers represented 11.5% of overall quarterly server revenue – reaching an all-time high – as Linux servers continue to expand their presence in data centers around the world for an increasing variety of workloads.

What is interesting about this figure, is that whilst in terms of revenue Linux is generating approximately 25% the revenue of Microsoft platforms, this only covers the sale of new hardware which typically provides a measurement bias in Microsoft favour. It doesn't include the many thousands or millions of computers out there that have been retasked as Linux servers because of the free availability of server software. It is this 'recycling' of hardware where perhaps Linux is strongest, and also extremely difficult to measure.

Just a brief note here - I'm mainly referring to Linux as that seemed to be the target of Francis' attacks. In reality a number of these arguments apply to other server operating systems including the large BSD family.
... and only 1.7 per cent of consumers were opting for Linux on notebooks.

Many open source operating systems struggle with handling power management on laptops - even Rodney Hide has blogged on this very problem. On a personal note, this is why I run a Mac. It has unix underneath, a fantastic user interface, and everything just works - unlike Windows and Linux laptops I've used (and continue to use and support).
But for enterprises, interoperability with off-campus software is key...

If that is the case, then why do you support closed, proprietary solutions instead of the recently released OASIS OpenDocument standards.
Microsoft Office 12 -- the coming version -- will use an "open" XML code system, catchingly called the Microsoft Office Open XML Format, as a key component of its code engine.

It is not open. Microsoft has been patenting the XML relating to its Office XML around the world. Here in New Zealand, the NZ Open Source Society has appealed the patent - where the appeal is currently at, I can't find out right now. But back to my point.

Any standard where a company owns of the intellectual property relating to the standard is not open - certainly not to the extent that governments, including own own, would like. The State Services Commission has warned New Zealand Government agencies not to utilise Digital Rights Management in Microsoft Office for precisely one reason - the only software they will ever be able to access the documents under will be Microsofts. And if they hold the patents on the technology used, then they will obtain lockin. Groklaw has a good article that provides more detail on this issue.
But all that aside, one of the most critical issues with Open Office (apart from interoperability) is how much grunt it takes to run it -- and grunt is not a feature on most corporate computers, where RAM tends to be doled out by the teaspoon.

Riiiight. Funny then how the Windows XP Professional systems requirements are identical to OpenOffice.org 2.0, no?
On the security front alone, study after study has shown that open source code is more vulnerable to exploits than proprietary code

You've linked to plenty of other reports Francis, how about linking to this one too? Or is the only copy hosted on microsoft.com?
The idea that government "has a democratic duty to provide information to the public that is in an accessible and open format" is a fine declaration of principle, but meaningless in practice.

Right, so you would also agree then that if a move was made to file all our tax returns online, and that these could only be completed using Microsoft Internet Explorer running on Microsoft Windows that would be OK? The New Zealand Government would just have granted an exclusive monopoly on all the computer systems used in New Zealand by people that needs to file taxes.
The upshot is that documents built on one system are unlikely to render properly, if at all, on the other. In fact, the new, OpenDocument-based OpenOffice 2 is not even backwards compatible with its own precursor.

And Microsoft Office is the same - for example, ever tried opening a Visio 2003 diagram in Visio 2002 or older? It won't work, unless you explicitly save the diagram as a Visio 2002 or older document. I've had to do this with a number of our clients projects. This is a common problem with software where the document format is tied to the office application platform, and it will remain a problem with Microsoft software. One of the benefits of the OpenDocument formats is that they they won't necessarily be tied to an upgrade of any particular suite of office applications. Any application that is compliant with that version of the specification will be able to read and save the documents. Naturally, this scares Microsoft as they will no longer be able to force upgrades on users as they create new documents using features which are specific to only the latest version of their software - features which can't be used in older versions, and hence 'encourage' other users to migrate to the newest release.
... and if users based decisions on whether to use OpenOffice or Office on which was likely to be supported in ten or twenty years, Office would have to be the winner on the day.

Once again Francis demonstrates his naivety around the issue. He is too busy focusing on the software application and not the mechanism used to store the document. The decision to make here is which format to use, not which office application suite. The great thing about truely open document standards is that a wide range of software can access and manipulate the same document format and this is why Microsoft and Microsoft supporters are running scared. Especially with regards to the recent announcement that the State of Massachusetts is eschewing proprietary office formats in favour of more open formats. Microsoft has been hopping mad about this decision, but really it is easy for them to work with it - all they have to do is support reading/writing of OpenDocument formats.

Put another way, if the New Zealand Government opted to use proprietary office documents and in 10 or 20 years found they were locked into expensive solutions that resulted in the syphoning of hundreds of millions of New Zealand dollars being syphoned offshore - that would make past projects like INCIS pale into insignificance.
As it stands now, the government is a rabbit warren of department level bespoke IT solutions with isolated IT units sworn to their own unique visions.

Yep, it is amazing how many projects coming up on the Government Electronic Tenders Service have Microsoft platform requirements. I don't hold this against them, because as I understand it, the Government IT environment is a very complex and complicated one in which to operate.

I am not suggesting that changes are made right away or even at all.

In fact, I am fully supportive of every organisation choosing how they wish to manage their internal IT infrastructure. The key, however, is in defining standards for communicating between different ecosystems in the IT world. I will defend to the death every organisations right to use and store documents in their internal IT system in whatever format they choose. However, the very reason we have the Internet, World Wide Web, blogging and many other fantastic technologies today is because they interoperate. There are many open standards at work just to deliver this page to your screen!

As long as each organisation can transfer their documents to an open format when it leaves the organisation - this is really all that matters.

I'll give you a brief insight into my business. Most of our document work is done in Microsoft Office on a mixture on PCs and Macs. Our server systems are Samba running on Linux and Mac OS X for filesharing, and we make use of the LAMP architecture (Linux, Apache, MySQL and PHP) for running intranet applications. We internally work on Office, but most of our documents are delivered in pdf formats as they can be more widely distributed and accessed. Hopefully in the next year or so, I will also be distributing documents in the OpenDocument format to increase awareness within government of this format.

As you can see, we use quite a mix of hardware and software and we've got a system that works well for us. The general trend in our business is actually away from both Windows and Linux towards Macs as the machines of choice for security, reliability and productivity. The only downside currently is that I can't run OpenOffice.org v2 concurrently with Microsoft Office on the Mac.

Open source software in general (and I'm talking about more than just Linux here) is becoming well entrenched in the IT world and its here to stay. In fact the very nature of open source software only begets more open source applications.

I'll close with one important point to make - unlike computers, using open source software is not not a binary decision. You can use both at the same time. OpenOffice.org actually can compliment Microsoft Office quite nicely - creating pdfs from Microsoft Office documents, fixing corrupted Word documents files that even Microsoft Word can't read. My own laptop is stuffed with a wide range of both commercial (yes, it's legit before you ask) and open source software.

It's all about having choice and a wide range of options in this exciting market - and this is what open source software provides. Microsoft and their supporters solely want one system to rule them all.

Posted by Bernard Woolley | 10/28/2005 10:57:00 PM


Blogger Antarctic Lemur said...

I have many thoughts on this topic.

First, Till's article has been incorrectly portrayed by some (e.g. the Biggus Twittus himself) as being some sort of pro-Microsoft anti-open source rant. I did not come away with that impression at all, though I do think the article was business-focussed rather than education or public sector focussed. Obviously the different areas have different priorities.

2ndly, Open Office is a pile of stinking crap if you actually run it side-by-side with Microsoft Office, Corel's older apps, etc. And of course there is no native OSX version of it yet. There was in fact a recent assessment of the system resources required by Open Office compared to Microsoft Office. It was not pretty (for OO).

3rdly, the fact Till's non-commercial private blog runs on WordPress(PHP) on Apache/Linux if anything strengthens his stance. In general, technophobes do not blog. If they do its with the close support of another technically minded person. Blogging software is not yet user friendly - though the next generation might finally gain the coveted WYSIWYG interface. Even Apple's implementation included with Tiger Server is only a very limited package.

4thly, Microsoft's dominance is assured until someone figures out how to rip off Microsoft Excel properly (macros, bugs and all) and parse older Office documents into an open format. There will be no dramatic move away from Office file formats if older documents must still be viewed within Office itself.

It's time for me to head off but I'll have more comments tomorrow.

10/28/2005 11:44:00 PM  
Blogger Bernard Woolley said...

Also, some more quick comments before bed :)

Re: 1. Yes, although the majority of the links posted were pro-microsoft. I didn't read them all, but keep finding myself at Microsofts website all the same. There was a distinct lack of focus or reference to positives coming out of open source - of which there are many. His article did need more balance rather than just bashing on open source in general.

Re: 2. From what I've seen of v2, it is much improved over v1 which is as you suggest not that flash. Agree on OSX too, I'd like to be running it as well. On the point of system resources, it is hard to make a true side-by-side comparision (which Till did link to) as it doesn't appear to factor in the API's and libraries built into Microsoft Windows that Microsoft Office takes advantage of to run. Looking at the memory of a Microsoft Word process for example won't tell you the full story because many of the API calls run under Windows libraries instead. So, not exactly a fair comparision from my quick glance, this could also lead to unfair speed benefits over other office software - all part and parcel of leveraging the Microsoft monopoly. Oh, even Windows Office 2003 looks horrible compared to Mac Office 2004.

Re: 3. Only the web interfaces aren't WYSIWIG - aren't there some nice applications available that post via XML-RPC?

Re: 4. It depends really on the level of use. If the macros are being used entirely within an organisation - I don't have a problem with what office suite is used. I would have a seriously problem if such a macro-encrusted spreadsheet was required to be used, especially as part of complying with government legislation. It also begs the question - is a document with embedded code the best way of managing the information? Would it be better to manage this information using a web interface (as a quick example).

In my business, we have all sorts of hassles managing some business information in spreadsheets because the multiple version we end up passing around. We can better consolidate and manage this information using a web application rather than using an office document with inbuild business logic. But thats just us.

10/29/2005 12:35:00 AM  
Blogger Chefen said...

I like to have everything available. MS stuff is fine and easy for simple stuff like normal wordprocessing and things like that. It is diabolical for doing any technical typesetting, but then LaTex is a pain for ordinary docs. Real number crunching isn't a good idea on MS Windows either. Programming, again depends on the app. So, right tool for the job as always. A dual boot machine is easy, a couple of machines even easier. In a business environment the quality of the IT staff is more important most often, no matter how swish the software. But what struck me about the Green position was this statement by Turei:

‘Robust’ competition does not inevitably lead to the best outcome in an area such as this because multiple suppliers and standards often lead to wasteful duplication.

Which is plain dumb no matter how you cut it. The problem with the Greens is they come close to being sensible about something for once and then go haring off into idiot-anti-competition-land just because they can't bring themselves to admit competition in software development does give better results.

10/29/2005 01:05:00 AM  
Blogger francis said...

The story really is not a bash of OSS at all. OSS is fine, in its place -- and there are lots of those places. OOo is certainly not a worthy standard bearer for the movement, though, and I think most fair minded persons would see it as too limited for general use in a complex business environment. It was also not shilling for Microsoft, but written in response to an series of statements that specifically reject Microsoft because it is such a bastion of capitalism -- and so it had a Microsoft focus. It could have had an Oracle or IBM focus, but those weren't the companies the Greens named. Finally, I do not think piecemeal, bespoke systems, siloed off at the department level, are the way to go in any business, including government. The Fonterra-EDS model is an ideal solution template for government -- and if Linux servers wound up behind the web serving, application serving and cluster computing in that model, I wouldn't have any objection, per se. I do object, per se, to the idea that a thing is made better for business use than a proprietary solution just because it's open source. I'm working on a follow-up piece to the original commentary (it wasn't a news story) and I do appreciate all the lucid reaction that has been thrown my way since the piece first went up.

10/29/2005 02:07:00 AM  
Blogger Antarctic Lemur said...

wooaaaaaaaa there Francis.

EDS are a bunch of inefficient fools [is that a legal statement to make?] who profit from the foolish outsourcing of IT functions by corporates like Fonterra and Telecom.

It might be a neat easy idea on paper to let some other organisation run your IT functions. But in reality the contracting company does not experience any direct and immediate financial pain from lowering the quality and range of service provided, and it inevitably tries to "standardize" all x,xxx employees to some easily supportable (from EDS's point of view) collection of software.

I know for a fact EDS is hated by Telecom employees. Everything requires a form, and when the jobs are finally done it's often in a spastic and inefficient manner. For example if your laptop stops functioning correctly EDS will essentially pull the damn thing apart and replace each component. Why? Does that sound like efficient IT practices? I don't think so. Every year at the annual Telecom staff conference people ask when EDS will be dumped. I believe a couple of years ago someone managed to ask Theresa herself this question in front of a packed auditorium (she was looking in to it).

If I was a Telecom shareholder or Fonterra sharefarmer (correct term?) I'd be furious. Telecom can afford it of course, because of the massive subsidies from its fixed-line monopoly it uses to support some business units and inefficient business practices in general. But I'm very surprised Fonterra went down this path.

10/29/2005 09:17:00 AM  
Blogger francis said...

Sorry, Al, I wasn't clear. I'm not recommending EDS as an outsource vendor for government IT. I'm saying a unified architecture is better than a shantytown.

10/29/2005 10:31:00 AM  
Blogger francis said...

That said, your take on EDS is interesting to me from a story POV. I've been looking at EDS from a corporate heads POV and haven't gotten to a user POV yet -- maybe you can set me up backchannel with some lines of inquiry. Sounds like a good story.

10/29/2005 10:35:00 AM  
Blogger Colonel Tux said...

imho, I would suggest Linux and open source has penetrated many businesses and state organisations under the radar for the mere fact that you don't require a purchase order to deploy it - and therefore it doesn't show up in the cfo's accounts.

The other point is that the internet exists today only because of open source - ie. apache, mozilla, bsd internet stack in windows, sendmail etc... Once again no flashy sales people in suits pointing that out (I guess because there is no commision)

10/29/2005 10:58:00 AM  
Blogger JamesP said...

Just a couple of points.

1) OO is a dog for anything other than reasonably small documents. In my experience application load time, document opening time, and graphing time are all noticably worse than Office.
2) Having said that OO has improved and will continue to improve but in the near future it won't be an Office beater.
3) It looks like MS will include an Open Document export filter in Office 12 to go with OD import and PDF export. Whether it gets the old embrace and extend remains to be seen.

10/29/2005 11:41:00 AM  
Blogger Antarctic Lemur said...

check out this microsoft site:

Pretty cool use of AJAX. Works fine in Firefox.

10/29/2005 12:13:00 PM  
Blogger Bernard Woolley said...

A few more points...

1. Microsoft/Telecom/EDS can be seen as the unholy trinity of IT in New Zealand. As a shareholder at some point in the past of all these companies, I made a fair effort following their actions, and recall them forming an alliance in the late 90's. Not sure if this still holds today. But these three companies have strong relationships with each other. I'd love to see big business examples that are not linked to these companies.

2. 95% of NZ business is still in the SME sector. They don't necessarily do that much VBA coding, customisation and the suchlike. They stick predominantly to simple office documents, email, intranet applications etc. To most IT users in New Zealand, it is irrelevant what big IT users are doing because they are an entirely different scale to the remainder of organisations.

3. Word is absolute bollucks for managing large documents. I'm actively looking at moving towards a solution whereby document content is managed in a wiki, then exported via XML and produced using DocBook, LaTex or similar.

4. Word also introduces people to extremely bad formatting habits, and users end up spending more time trying to make a document look pretty, or even just get it to work how they want, rather than developing the content. I've certainly seen this in our business, and I expect it is fairly commonplace.

5. OOo currently has the most potential for home and small business users. Those that are writing simple letters, and managing simple spreadsheets. I do have it deployed on most of our laptops at work (currently going to roll out OOo v2) but most users still use MO by default

6. Francis - I agree that 'piecemeal, bespoke systems' as a means of reinforcing silos are the way to go. However, it really only needs balance. As I mentioned in the main article, I don't care how data is stored internally, what is important that it is communicated between organisations using truely open standards which are not encumbered with intellectual property rights. Spatial data is a good example - it doesn't matter whether you store the data using ArcSDE or MapServer with PostGIS or MySQL with spatial extensions. What is important is that these servers communicate and deliver the data in open formats, as provided by the Open Geospatial Consortium. I believe the same should be true for office documents - internally, use MS, Open or whatever office document to manage your data. What is important is that OpenDocument is used for the transfer of this information between different organisations. It should be noted that VBA/macros etc are generally not required in this context - they are mostly for internal usage.

7. A unified architecture also presents a greater security risk to your IT infrastructure. An organisation will me more resilient with a greater mix of operating systems running. If you have a homogenous system, then often if you can crack one you can crack them all!

10/29/2005 01:20:00 PM  
Blogger Antarctic Lemur said...

1 - I believe Telecom Xtra is trying to separate itself off from Microsoft. Unless they've changed their minds again.

2. I agree, but these are also the very guys who rely on some key Microsoft Excel spreadsheets for accounting functions, project management, simple databases, etc. Getting OO to imitate Excel is very important.

3. Very true. If you have found a solution for this then please email me as I need it for my company. We're also interested in customizable web-based pre-formatted reports in which users can check a couple of boxes to change the content of key paragraphs, add some of their own numbers and text, then convert the report into a PDF for printing and archiving.

5. I think OO still needs another major version change and a native OS X version (ie lots of TV and print media users) before it will gain wide acceptance. And any small company with an archive of Office documents will still need to ensure they can read and edit them if necessary.

10/29/2005 01:52:00 PM  
Blogger David said...

I'm in the ICT strategy group for an Aussie state government. We're all aware of the availability of open source software. All our technicians are aware of it. We don't have a preference for either open or proprietry products... provided they meet our requirements, interface with our other systems, and have a growth path, then we'll pick the best software for the job. So we really don't need a policy to promote open source, as the Greens would like. Or members of the public writing to recommend that we consider open source software, as if we'd never heard of it or were (for inexplicable reasons) opposed to it on principle.

And I'm confused as to why the ONLY IT-related issue that causes people to write to us (or to our Minister) is open source. No one writes to recommend that we consider the use of a meshed rather than a hub and spoke network. Or to enquire whether we consider the advantages of appliances over servers. Or to ask what our policy is on integrated identity and access management. It's always open source, as if there is a conspiracy among IT professionals to give money to evil Microsoft. And that requires intervention by the public and the Cabinet to foil our evil plans.

10/29/2005 03:31:00 PM  
Blogger Bernard Woolley said...

David, I think the main difference is that supporters of open source are driven by philosophy, where as most of the other issues you suggest, whilst valid, are more technical than philosophical and hence don't attract the same attention. It is probably this same philosophical underpinning that attracts the Greens to the open source cause, as some of their other philosophies.

10/29/2005 03:53:00 PM  
Blogger David said...

Bernard... I don't see what the philosophical issue is. It really doesn't make any difference to us whether the software we use comes with source or not. Realistically, there isn't much you can do with it anyway. We might be able to tweak small things, but large changes would leave us with large ongoing support issues. And our guys have too much real work to do without spending time getting to grips with complex products such as office or operating systems, at least to the extent that they'd be able to make coherent supportable changes without breaking existing function.

So we're back to using software, rather than changing it. Selecting according to best function, rather than on the basis of provision of source. And that we really don't need to be told to consider open source, because why wouldn't you... it makes no sense to rule out use of a software product just because the source is available.

And if it is a philosophical issue, then why isn't the philosophy applied to other areas? Why don't people lobby the government to prefer the purchase of motor vehicles that come with blueprints, and software listings of the onboard computers? And where the design of the vehicle was done for free by enthusiasts, with the blueprints placed in to the public realm? I think the "philosophy" is mostly an anti-Microsoft thing.

Oh, and I'd disagree with your earlier comment "A unified architecture also presents a greater security risk to your IT infrastructure". A unified architecture means that you build up enough expertise to manage and secure it properly. A non-unified architecture means that you spread your skills so widely that your staff are jacks of all trades, expert of none. And introduces a whole lot of integration problems with associated security implications. For instance we run nothing but Cisco network components... we have a group of Cisco experts who keep all the IOS up to date, install and configure IDS functions, ensure that everything is administered professionally using Tacacs and other management and authentication tools, and ensure that all our wireless gear is using WPA and other modern protocols. It's tight. As the security manager, I'm happy. But if we mixed network components from many different vendors then the guys would spend all their time fixing integration problems, and wouldn't build up the skills to ensure that all the bits were configured to maximise security.

10/30/2005 03:12:00 AM  
Blogger softwarefarmer said...

I could not agree more with the sentiments about EDS and I also have Telecom contacts at management level who do indeed ‘hate’ EDS. Same with Fonterra.

I also had to work beside EDS for three years in the US and we simply referred to the company as ‘nazis’ because of the way they chewed up their employees. Not a place I would ever recommend to work at and their NZ outfit is no different. One guy I know here found himself loaded up with ever increasing responsibilities as people around him left and were not replaced. He’d complained about this in the late 90’s but I only just found out that he left last year after being hospitalised with heart problems – at 38 years of age!

If the company using EDS needs to make frequent changes to their systems as part of their product development and marketing world then EDS is certainly a problem because of the standardisation they try to impose and the bureacracy that goes with it. As long as you have a status quo situation they’re okay but as soon as change comes along they’re a nightmare to deal with. I’ve first-hand experience of that with Fonterra over the last couple of years as a contract PM. The forms AL speaks of are becoming net-based but still – it’s a hell of a process to get through. I found the only real way was the usual path of working the bureacracy and the people who did the jobs on separate paths – meaning a lot of emails and phonecalls!

All that said, they made perfect sense for Fonterra at this stage of the company’s existence precisely because of their standardisation and processes. These were exactly what Fonterra so desperately needed. The IT flexibility they need can be supplied by the likes of SAP-type packages – they just don’t need (yet) the sort of software flexibility that a teleco or insurance company needs. As bad as EDS might be you had to see the Fonterra IT infrastructure environment that existed prior to their arrival (and I have knowledge of that too based on working for the NZDB briefly in the 80’s. After combining the old Dairy Board, Kiwi and NZ CoOp organisations there was a godawful mix of standards in hardware platforms, telecoms, databases, tools, packages – not to mention the ubiquitous bespoke systems developed in various dark corners over the years. And don’t get me started on the processes that surrounded all this!

As a result there were huge costs hidden from users – a factor that only reinforced a lot of these issues. Multiply all this across all the global sites and you can see a tough problem to solve. Even tougher to change when you try to use the existing culture. There’s something to be said for making a big hit! Most of the IT infrastructure people I knew argued strongly that outsourcing was not cost-effective (certainly EDS charges like a wounded bull for people whose experience often does not justify it) and they may turn out to be right in the long run.

But right now, as well as cleaning up the cost sinkholes, there are a lot of good strategic reasons why Fonterra would go with EDS. In a few years time, with most of these issues straightened out, it may make sense to change the model – maybe even back to an internal group. But until then, to fix problems that were not being fixed, it’s worthwhile – and I can say that as a Fonterra sharefarmer as well!!!

10/30/2005 08:53:00 AM  
Blogger Bernard Woolley said...

Hi David,

I agree about the philosophical issue - not a lot of people understand, and of lot of people in the industry don't really care about it as you suggest. Its the difference between those that want source code managed as a community, and those that purely want the best product - commercial or otherwise. Neither of these methods necessarily produce the best product for any given user though. It the people at the coalface like you being pragamatic, vs the open source being dogmatic and there is still a glaring divide between them. The open source crowd, especially, needs to recognise this if they want their products to gain wider usage.

And I think the reason that the philosophy hasn't been applied to other areas yet, is that many are still focusing on computers. That said, people are looking around now, and finding all sorts of other schemes - such as not being able to connect to your car these days without using expensive proprietary equipment and software, and I believe people are starting to kick up a stink about this as well. So I think is may be starting to spread slowly. Also, open source techniques are being applied to encyclopedia (en.wikipedia.org) and our very own TopSpin project to document our politicians (topspin.sirhumphreys.com). So I think the philosophy is certainly starting to expand, but it will gain so much more through pragmatism rather than dogma.

Re: security - that will teach me to make throw away comments and not back them up :)

In large organisations, I entirely agree that where you have personnel with security responsibilities a homogenous environment is likely to be the easiest to manage. However there are a great many organisations out there that don't have staff/contractors with security resposibilities (95% of business in NZ is Small/Medium Enterprise). And for those that do contract IT expertise - their value is mostly likely going to be less, as they are less specialised than the specialists you have in-house.

Homogeneity of networks is one factor in security - and it your case it sounds like it isn't an issue. But for organisations that don't have the security resources at hand that you have, I still believe it is a significant problem, because you can potentially lose many machines at once - particularly if it is a Windows exploit that hasn't been patched yet, due to the widespread use Windows computers. Especially as Windows is often chosen for 'Ease of management'.

10/30/2005 09:05:00 AM  
Blogger Antarctic Lemur said...

Francis - if you want to find out more on EDS then you should try contacting some engineers at Telecom Advanced Services. They'll have plenty to say. Oh yes they will indeedy....

10/30/2005 12:07:00 PM  
Blogger francis said...

This has evolved into a MOST interesting conversation. Maybe after the beginning of the year I can round up some or all of you in a chat inverview format. Al: I will talk to some of those engineers, thanks for the pointer.

10/30/2005 02:15:00 PM  
Blogger Antarctic Lemur said...

whoops - that should be Telecom Advanced Solutions.

Isn't it odd Telecom is building up its IT divisions yet still outsources its own internal IT infrastructure. I'd be worried if I was the EDS NZ boss.

10/30/2005 02:23:00 PM  
Blogger ZenTiger said...

Sadly, I too weigh in with dismal news about EDS. I worked as a third party software provider for a mutual client.

EDS managed their network. Whilst they had good people, (and they did), they tended not to stay long or be limited in what they could do.

I found them difficult in that if they didn't supply the software, they effectively cut out their competition from doing so (in this case, my company). The customer at user level seemed disatisified, impotent and ignored. Managing the SOE and desktop might work for the basic apps, but under EDS it progressively led to shutting out specialist apps.

They didn't have a good rep with this particular customer, but every now and then you'd see a nice fluffy press release with manager of company X saying how happy they were with EDS.

This may not be purely EDS' fault - perhaps a symptom of getting too carried away with the amount of services you outsource.

10/30/2005 03:04:00 PM  
Blogger darren said...

Yes, there's some cracking stuff here.
You go first with your story Francis.
Now, for the December issue of MIS I am writing a feature on work-life balance issues and how IT executives relax.
My initial inquiries have not progressed too far as the IT executives I have contacted so far are too bloody busy to comment!
But if there are any IT executives (IT managers, IS managers, MIS managers, CIO's, etc) who would like to take part, please let me know.
Furthermore, in the New year, i will be looking at outsourcing, so some comments on that issue would be appreciated.
MIS has its own "rentaquotes" but it would be great to get comment from fresh individuals.

All the best


PS I look forward to seeing your piece in NBR first, Francis.

10/30/2005 03:25:00 PM  
Blogger David said...

Darren... If you don't mind talking to Aussies, we're an outsourced environment with about 15,000 core users, and probably as many students on our network. We've recently let a $Au150m contract for network services, and are currently out to tender for about the same amount of desktop and associated services. We've spent a lot of time looking at outsourcing models, working out what was wrong with our old contracts, and finding out what other organisations are doing. I think we'd be happy for a bit of publicity.

10/30/2005 11:17:00 PM  
Blogger francis said...

Thanks, Darren, butyou go ahead for MIS if you want. Inside NBR print, it's going to be PAINFULLY short -- not the kind of reflective feature MIS would be likely to pull together. And two stories will probably throw more light than one.

10/31/2005 12:26:00 AM  

Post a Comment

<< Home