In other words, the chances of desktop Linux appearing in the Australian public sector in numbers is close to zero.

Con Zymaris, the chief executive of the Melbourne-based Cybersource, a company founded in 1991 to provide open source solutions, was speaking to iTWire soon after the deployment of Windows 7 netbooks by the Department of Education and Training in New South Wales.

Zymaris, who these days is mostly involved with training people in the use of OpenOffice,org, has a long history of involvement in FOSS, as a developer and advocate.

He spoke at length about FOSS in the country, the general state of the IT industry and the way forward.

iTWire: I was surprised to note that not a single company had bid for the Department of Education and Training netbook supply in NSW. I see a lot of companies represented on the Open Source Industry Association mailing list and I'm surprised that none put in even a token bid.

Con Zymaris: There are not enough companies in Australia with the muscle to put something in and have any real chance of success. The problem with something like this is the way bidding has gone in this country in the past, say, 15 or 20 years, where successive governments, both Labor and Liberal, have pushed towards larger and larger outsourcing components. It used to be that an agency would put out a tender for a small project - 50,000, or 100,000 or 200,000 - which was feasible for many small Australian companies to bid for.

But these big projects are beyond the reach of most Australian IT companies, certainly they are beyond the reach of any of the Australian open source companies. And it's coupled with the fact that it costs a lot of time and energy and money to put together a bid for a large project. It has to be tempered by the fact that you have to figure out what your chances of success are.

CONTINUED

---

iTWire: How would you define large?

CZ: If we are talking about the netbook deployments by governments we are talking about tens of thousands of computers and the projects would be worth hundreds of millions of dollars. I would define large as anything more than a few million dollars; even more than half a million is tricky for open source companies to bid for unless it's in a realm of the industry which is likely to have success with an open source solution when bidding - contact management systems, possibly some enterprise resource planning systems, some document management systems. Areas where government has shown a track record of acquiring open source technology to make it likely that open source companies would think, 'I'll have a shot at winning this business, therefore I'll spend the $30,000 to put in the bid'."

But for something like the laptops and netbooks for education, most open source companies, the local ones, would say, 'it's too big for me, I can't bid'. And the major players who have an opportunity - and there aren't many of them, there's probably realistically only one - if they don't bid, nobody else would.

iTWire: Which is this one?

CZ: The only player in the country who could bid for the netbooks contract would be Novell. And if they're not bidding, nobody else will. IBM moved out of the desktop market, the client market, some years back when they sold their desktop business to Lenovo, Red Hat does not focus on the desktop - certainly not in this country, it does in the Asian subcontinent and in parts of Europe - and Sun is not interested, they are not doing anything at all that's different because they are now in the pipeline towards being consumed by Oracle. That doesn't leave anybody else. I think Novell has spent some effort and money in trying to move into the Linux desktop business within government in Australia and it's only had modest success at best. From Novell's perspective, to put a lot of energy and money into bidding for something if they are unlikely to win doesn't make sense.

iTWire: In the current situation, Novell would not bid against Microsoft.

CZ: Not necessarily. There's no reason why they wouldn't.

iTWire: They're in bed with Microsoft, why would they bid against them?

CZ: It doesn't currently stop them from selling client and server solutions that compete with Microsoft's core product. They're not necessarily in bed with Microsoft but they have got a different strategic position. They saw that was an advantage play from their perspective, whether it is or not may bear fruit in the short term and may or may not in the long term, i don't know. But that wouldn't stop them from bidding. I don't think they would pull the reins on the bidding horse because of the strategic deal they have with Microsoft.

iTWire: Novell has hitched its horse to the interoperability factor. A lot of the interoperability possibilities come from Microsoft. Why would they want to rub Microsoft the wrong way?

CZ: That might be a factor. There may be strategic reasons. My guess is they wouldn't even need that as a reason not to bid. I don't believe they've had substantial success in the Australian public sector to justify enough of a confidence level to say that they'd put in the resources to bid.

CONTINUED

---

iTWire: Are you saying that they have had some success?

CZ: They've had some success, particularly in NSW, Department of Commerce and some other agencies have deployed Linux desktops. We're looking in the realm of the hundreds. It's a big jump to go from the hundreds to the tens of thousands, possibly the hundreds of thousands that we're looking at for the netbooks. But it boils down fundamentally to the fact that we do not have the model of technology adoption in Australia that is required to deploy a different technology platform.

It doesn't work to just do the same-old same-old. Substantial difference in procurement has to happen for any real deployment of Linux desktops to occur because it's a major shift in strategic positioning for any one of these government agencies. It's not a case of 'it'd be nice to deploy a wiki, we don't have a wiki, look there's an open source wiki, that looks reasonable, let's deploy it'. That's the way open source is being deployed in the public sector here, that type of path, but not 'we want to change desktops to something that's different from what we currently do'.

iTWire: Are you talking about the numbers only or the complexity of the project?

CZ: The biggest thing from the public sector point of view is risk. Companies can take it up but the agencies wouldn't be interested in having them take it up. The block is the agencies themselves. They are just not interested. The government agencies are not interested in deploying desktop Linux, client-side Linux. Where they have deployed Linux is in areas where it is seen to be completely commensurately risk-free from the industry perspective.

iTWire: Are you then saying that a Microsoft solution is risk-free?

CZ: It's risk-free in the sense that if it fouls up, that's what the industry selected, so the industry is at fault rather than the agency.

iTWire: So it's a question of passing the buck?

CZ: Passing the buck as with all things. It's more so with agencies than anywhere else that nobody ever got fired for buying,  insert name of player here.

iTWire: This may have been a plausible argument five years ago. But given the progress that free and open source software has made, and given that government should look to implement the cheapest and best solution, don't you think FOSS should be getting more chances?

CZ: I would agree, but obviously they don't see it in the same way. From their perspective if they either haven't done the serious risk analysis or the risk analysis has been done and it shows for them that they would get more howls of protest if they went with a Linux solution than if they stayed with the status quo, then that's what they would do. They are just staying with the status quo. From their perspective, that's whatever Microsoft pushes out, whatever it's called, it is the Microsoft desktop o-s in its various guises. 

The only way that they could seriously move away from that is to make a decision at some point that they will move away and implement a strategic multi-year framework and all the steps that are necessary to achieve that. That's what has had to happen in other countries. Tn other countries, they can't just say 'next week we're shifting to Linux, let's sign the deals with the appropriate vendor and off we go'. It doesn't work that way. But the agencies, in whatever way, shape or form, make that decision, whether it's internal people or external consultants who are brought in to do the analysis that shows them they can benefit from a number of different financial, strategic and possibly even security perspectives, at the end of the day it's the agency that makes the decision.

At no point do they go to market like our agencies do and say, anyone who wants to bid for a desktop do so and then we make a selection. It doesn't work that way. If you look at the path towards Linux desktops and open source client-side computing in all these other government agencies, it's all based on this flight path of the agency that makes the decision 'we will do this, what do we need to do to achieve it?' 

CONTINUED

---

iTWire: So there has to be a fundamental decision to change over first?

CZ: Yes. And it's not happening here. It hasn't happened here in any substantial way, in any government agency that I know with the exception of fairly small, but fairly innovative entities like the National Archives and segments within the military research establishment. DSTO, for example, has been a longtime user of major Linux client-side resources as well as supercomputing resources and whatever supercomputing resources that the government agencies fund which is once again considered the mainstream.

iTWire: Don't these government agencies learn from each other?

CZ: None of them are really doing it. The agencies that are doing it aren't really on the radar in any way, shape or form as far as the other agencies are concerned. When you look at the public sector, the last time I did a headcount NSW had 30 public sector agencies, Victoria has about 17, the Federal Government has something close to 200 - we're not counting most of the other states, Queensland has a big public sector, probably as big as Victoria - you tally them and you're looking at several hundred. Only the really large players are on anybody's radar -  defence, education, health and so forth. And none of these major players are shifting away from the status quo.

iTWire: You don't see any chances of this happening in the future either?

CZ: Tomorrow, if large tracts of English-speaking public sector agencies around the world started shifting to desktop Linux, then there would be pressure on the local agencies to do something similar in the near future. I don't see that happening in the US or the UK...

iTWire: Why does Australia always look to the US and UK and not Europe?

CZ: It's what we're focused on - whether it's good or bad.

iTWire: In Europe as you know, there are a large number of deployments of Linux which have been successful...

CZ: There have been but we don't pay attention to Europe. They don't speak English therefore they are not on our radar. The further someone is from English, the further away they are culturally, the less we pay attention to them. Ninety percent of our IT industry focus is the US, which is silly, because in many ways the US has stopped being the trendsetter for anything but dot-com related and Web 2.0-related technologies. For many, many years now, whether it's in telecommunications or infrastructure, a lot of the technology has shifted to eastern Asia, southern Asia and Europe but yet we are still US-focused. We pay as much attention to other countries as we do to their celebrities - we pay utmost attention to American celebrities. It's a similar mindset. It's not good but that's the reality.

iTWire: Isn't Australia supposed to be some kind of leader when it comes to open source?

CZ: In terms of development and smarts at that level we punch above our weight. We're not anywhere near the top of the stack. In some ways we tend to kid ourselves that we are right at the top, but if you look at the outputs of the Scandinavian countries and Northern Europe, the Dutch and the Germans in particular, you see that they do a brilliant job, they are right at the top. If you analysed you would find that Australia's position in the IT industry as a whole, globally, based on the proprietary side of our software and products side of out industry is woeful. On the open source side, we are doing very, very well.

If our local IT industry as a whole had the same output if you like, as our open source sector, then we'd be doing a fantastic job, But we lack a number of things that are fundamental to the proprietary side of the industry, the marketing and the management skills to make global companies, we completely lack that, with a few rare exceptions. But many of those requirements are not needed in the open source space. You don't really need the same kind of marketing, you don't really need the same kind of capital injection to be able to reach global markets. which is why when we do open source stuff, technologies like LAMS and Moodle and so on, we do it well.

CONTINUED

---

iTWire: Apple, I suppose, can't get into this desktop thing because of the cost?

CZ: If the market wasn't for netbooks, if it was for laptops, Apple would be competitive. When you factor in the lifecycle cost of an Apple product as compared to a Microsoft product, Apple could be competitive. But Apple still has the same issues that Linux has - it may be more seamless, easier to use, but it's still not Microsoft. Decision-makers only feel comfortable with Microsoft, because that's what their colleagues are doing. If they do anything different, they stick their necks out. If they do what their colleagues are doing, even if their colleagues are wrong and they are doing what they're doing because it's what they've done in the past, then they are safe.

We are talking about something at a macro level which can only be analysed at a micro-level. I have the skill to look at an organisation and say 'this is what you need to do for the next 18 months to clear the road for moving away from, say, Microsoft Office, or some other proprietary software to a Linux desktop, an open source offering.' In most cases it is fundamentally feasible for most Australian industry and public sector organisations today to move some amount of their infrastructure to Linux desktops and certainly to Linux servers. They are not doing that for any number of different reasons - mostly vision and political will or the lack thereof. If you talk to most IT people nowadays, the question they ask is 'how can we reduce costs but not by doing something fundamentally different?'

iTWire: Who will bell the cat?

CZ: In this country if the public sector does not lead, then it will be just a case of osmosis, small victories here, marginal victories there and so on.

iTWire: Wouldn't it be private businesses which are more focused on cost reduction?

CZ: Sure, but then businesses are also pretty risk-averse. If a large enough organisation has enough layers of coverage and there is no shying away - the cover your arse syndrome - something may happen. You need someone with some degree of intestinal fortitude to move away from the mainstream. It is happening in smaller businesses - a large part of what I do is OpenOffice.org training for for 5, 10, 15, 20-person businesses, every now and again for a 200-person business and every now and again for a multi-hundred person government agency. But the bigger cases do not happen that often. If you sat down and did the rational analysis, you would ask why this is not happening more often. Lots of organisations do not need a full-on office suite. But they find it politically easier to fork out the next hundred thousand or so for the next set of Microsoft Office licences - even when the next version of Office is going to cost them a non-trivial fracture in terms of staff usage patterns, retraining issues and so on. 

iTWire: You paint a fairly bleak picture.   

CZ: From an industry perspective, it's bleak because things aren't happening fast enough. They could be always happening faster. The flip side of the coin is that the open source industry has become big enough to support the existing players and to nurture and grow them. I guess as an industry we would like see more innovation, more change, a greater growth rate. For any one of a number of reasons we seem to be lagging in adoption. If you look at our adoption rate, you might say that we're half of what Europe is. That's the minus; the plus is that we're sustaining a viable industry in the open source space now. Imagine what will happen when our adoption rate doubles.

CONTINUED

---

iTWire: How much of this backwardness would you attribute to the efforts made by lobbying groups - open source lobbying groups, organisations, whatever - and do you think people have gone about it the wrong way?

CZ: One of the bigger issues that we had - and I can sympathise with government certainly - is that when they go to talk to the car industry, they talk to the car industry. There is a co-ordinated interface between the government and the car industry.

When they go talk to the IT industry, there's six different organisations, and 38 different pressure groups. They talk to the ACS, they talk to the AIIA, they talk to OSIA, they talk to the .NET developers in Queensland, and they talk to the local software consortia, and all these different entities often have fundamentally different requirements, different pressures. From the government's perspective, this is an issue.

I think and I guess what's at play here, what's at stake here, is a broader more fundamentally broken issue than just whether or not government adopts open source software. It is the Australian public sector outlook on industry development. We have across as far as I know, all the states - certainly the states that I've had dealings with - and at the federal government level, a fundamental barrier between industry development and procurement.

We seem to have a completely allergic stance, an allergic reaction to aligning government procurement with industry development. There are myriad stories of government departments not buying local product for fear that it won't be supported in the long term. Open source has answers - where do you go to when you have source code? You can go to anyone who can support PHP or C++ or what have you. All these things that are advantageous in deploying open source, the government has not as far as I can tell, pursued that.

There's a large combination of things. We don't want 'Choose Australian' for all these things - but the government is not above choosing Australian, you know, in some industries. Because if you go to departments, government agencies - state and federal - and you look at their car pool procurement policies, my understanding is that they almost always fundamentally buy local product. They would only ever buy Australian cars. But in software, it's almost always the opposite - they would almost never buy Australian product, Australian technology. Why is that? Can you imagine what would happen if we had a scenario where the public sector in this country, which is (spending) something like $30 billion a year, if some portion of it was earmarked for local IT products?

This was what was intended when the big outsourcing contracts were begun by the Federal Government in 1996-97, a certain portion was supposed to go to local companies. But nobody ever followed up on that. This meant that the local companies couldn't bid for the big projects anymore. They had to become underlings to all the big outsourcing companies, and that's just where the services were.

But very few offered development, and very few of the product purchasing deployments went to local companies either. You can have an industry where you can nourish local companies, if there's services companies. But generally you cannot have, and you cannot build enough of a layer of fat, in services companies, to allow them to grow big enough in this environment to be able to reach out globally. It won't happen. In a product company, you have a chance of that. Because if you hit the right kind of product, and you are successful at selling it, you can build up enough fat to take that product international. It's not at all easy to do that in technology, in the services sector.

That's why we have a fairly paltry services sector for technology as well, by comparison to many nearby countries - Singapore, certainly India, and I believe East Asian countries have a fairly big telecommunications services export industry as well, which we certainly don't have. It's a big complex area. It's not something you can simply point your finger at and say "presto x, y and z". But there certainly doesn't seem to be any onus, even with the Rudd government - they've been in power now for almost two years - I haven't seen any indication that there's any interest in pushing anything beyond the consumption of technology.

There's an old, possibly apocryphal, story about our satellite launch industry. We were, after the Americans and after the Russians, the third country in the world to launch a satellite. Most people don't know that. And the government, the Liberal government, said 'we're going to be digging stuff out of the ground and shearing sheep. We're not in the league of launching satellites, got nothing to do with us.' They completely missed the point about land management and remote sensing, and the possible export of such technologies to the region, you know, Southern Asia, and South-East Asia and New Zealand and the South Pacific. Not of interest.

They essentially scrapped and didn't really fund our space programme until the late '60s. Similar things happened with regard to our technology and computer industries in the '80s. Barry Jones has many stories about what happened when he was minister of technology. I don't think we've changed. I don't think we're fundamentally any different. That's part of the bleaker picture that you mentioned. But there's enough to sustain what's there. The hope always is that something will happen to accelerate the adoption of more local product, more local skill, and the adoption of open source product and skills as well.