Australian Senate Hears Open Source Is Too Expensive

schliz writes "The Australian Government Information Management Office says that a platform change to open source could cost more than it saves. It was pushed to investigate open source software to reduce its AUD$500m budget at a Senate meeting yesterday. From the article: 'Agencies are obliged to consider value for money on each occasion they apply a software,' spokesperson Graham Fry said. 'If the cost of assessing it [open source] was greater than the cost of the software, you would have to think twice.'"

Comments

Do this guys know the definition of user lock-in?

[ls671]

> Australian Senate Hears Open Source Is Too Expensive

Well, dear senators, this is a normal consequence of vendor lock-in:

"In economics, vendor lock-in, also known as proprietary lock-in, or customer lock-in, makes a customer dependent on a vendor for products and services, unable to use another vendor without SUBSTANTIAL switching COSTS. Lock-in costs which create barriers to market entry may result in antitrust action against a monopoly."

http://en.wikipedia.org/wiki/Vendor_lock-in

So, of course, there will be a substantial cost for switching ;-))

In the end, it all depends on how long you wish to stay locked-in. You have to consider the matter in the long term to see the advantages, and long-term thinking is seldom seen in modern politics ;-))

How about the other side.

[deniable]

Yes, there are costs to adopting open-source, that's the basic message when you use a bureaucrat to English translator.

How about these from TFA:

A 2007 AGIMO survey revealed that 68 percent of government agencies were either piloting or using open source software.

Centrelink, the Australian Bureau of Statistics and National Archives of Australia were known to use open source products;

Looks like it's getting a fair hearing.

Re:Do this guys know the definition of user lock-i

[Anonymous Coward]

Switching cost includes more the just the cost of the software its self. Just because you're using open source does not mean you don't face a certain degree of lock-in.

Re:Duh...

[shanmuha]

Do they really think it's better to pay for an egg every day than for a chicken today and then nothing for the foreseeable future?

....then pay a little for the chicken feed and pay a little for cleaning up the cage and pay a little vaccinating the chicken etc.. Clearly, you have never bought a live chicken for the eggs :)

Re:Do this guys know the definition of user lock-i

[Stuarticus]

Sorry, it's too expensive to even assess if there's any money to be saved by switching. Next item on the agenda, can we get some sort of magic machine that makes sure no-one is watching anything dirty in their computer?

It could be legitimate

[tjstork]

On the developer front:

If you have a lot of database stuff, Visual Studio can be much cheaper to develop for, so long as you ignore Microsoft's Architectural Group. For me, moving to Linux isn't just about saving money, really, its to break free from the corporate brain cramp that is Microsoft Architectural guidelines. Visual Studio and C# are great tools, but, if you have to use evaporate 2x as productive multiplier to do 10x as much stupid stuff, there's hardly a savings.

On the office front:

OpenOffice's spreadsheet is not even close to Office 2007 Excel. We developers can say Open Office spreadsheet is good enough, but telling that to someone who lives and breaths Excel is only for laughs.

Re:It could be legitimate

[nospam007]

"OpenOffice's spreadsheet is not even close to Office 2007 Excel. "

95% of the users I know use it to make phone lists and such with no calculation at all, because they never saw that Word and its companions can do tables too.

Re:It could be legitimate

[Sumadartson]

If you're brain dead, you can't use the advanced stuff in excel. If you're using excel for the advanced stuff in a critical application, you must be brain dead. Why, for the love of god, is the advanced stuff in excel ever used? I still don't get it.

Honestly, the amount of business critical applications buried in excel macro's is shocking. And, as we all know, the person who wrote the macro never leaves his/her job. This is especially dangerous for government who, for particular branches, have to be able to transparently show how they came to certain decisions. Any responsible official will stay away from excel for all but the most menial of tasks.

Re:It could be legitimate

[micheas]

On the developer front:

If you have a lot of database stuff, Visual Studio can be much cheaper to develop for, so long as you ignore Microsoft's Architectural Group. For me, moving to Linux isn't just about saving money, really, its to break free from the corporate brain cramp that is Microsoft Architectural guidelines. Visual Studio and C# are great tools, but, if you have to use evaporate 2x as productive multiplier to do 10x as much stupid stuff, there's hardly a savings.

On the office front:

OpenOffice's spreadsheet is not even close to Office 2007 Excel. We developers can say Open Office spreadsheet is good enough, but telling that to someone who lives and breaths Excel is only for laughs.

However people that use their spreadsheets for statistics will tell you that using Excel for you calculation is about as productive as using substituting rand() for your equations.

Here is one of several papers about the fact that Microsoft has no interest in fixing the broken nature of excel for statistical work.

As usually, price is the only criterion.

[SharpFang]

As usually, price is the only criterion. And I remember a letter of prime minister of Peru to Microsoft. He explained clearly and plainly that the TCO was moot. It doesn't matter if the analysis is good or bad. It matters that proprietary software is not suitable for government.

Government must not allow for vendor lock-in. It must not create a situation where their data is hostage to a private company.

Government must be transparent in all its processes. Their software included, being open for public scrutiny.

Government must use secure software. No black-box encryption can be considered secure.

Government's duty is to be as accessible to wide public as possible. That means, amongst all, open API for their services, and software available to all citizens no matter what their material status. No paywall of any kind to let only the rich have their way.

OSS is not a choice of "cheaper". It's the choice of "doing things the right way".

Re:Do this guys know the definition of user lock-i

[Sarten-X]

There is definitely a "certain degree" of lock-in, but it's like being trapped in a prison with a key-making machine and full details on every lock in the place. Sure, it'll take a bit of time and effort, but you can get out pretty simply.

Re:Do this guys know the definition of user lock-i

[rtb61]

The bureaucrats last far longer than that and ultimately they are often the ones that make decisions by undermining decisions more often based upon power plays and ego, rather than upon sound economic decisions. In this case one person was making statements full of if, could, necessarily, assumption, all to cover the fact that they had not bothered to conduct any research. The reason for the lack of research, that research could cost more than $500 million dollars a year, one could only guess that Graham Fry was intending to contract out the research into using open source software to a closed source proprietary software company.

Obviously Fry has no concept of foreign debt, no understanding of maintaining control over software upgrade cycles, no idea about monitoring historical trends and how many times they have bought the same software, no concept at all of life cycle costing, believes the lie that closed source proprietary software is free of maintenance costs and, fails to understand how governments choices in this sector impact upon private industry choices and further impact foreign by a nominal factor of 10 (500 million becomes 5 billion). A true asshat that does not belong in a role that legacy, longevity and, political astuteness has provided him, rather than expertise, national economic awareness or even basic common sence. Sounds like the Green Party in Australia is far more technologically aware than the rest (they also oppose censorship).

It seems that global trend of the right shifting to the loony bin and the left shifting to the right of centre leaving the humanity and environment (over greed and power) based parties, in this case the Greens, to take up the centre left position, holds true. With FOSS the bulk of the money in software can always be spent locally and that's down to state and city level, not just country.

There's also functionality to consider

[Sycraft-fu]

Different solutions, open source or not, aren't always as functional in the ways you need as what you have now, or what you are considering buying. Now, it's easy to say "Well it's open source! Just hire some programmers to write the functionality you need." However that is a problem for three reasons:

1) That costs money. All of a sudden the "$0 per copy" thing isn't true anymore. You have to factor in the cost of the development team. That is not cheap, at least if you want it done well. Good programmers don't work for minimum wage. So that cost must be factored in.

2) You have to support it. If you are doing major development to something you need, you'll then have to support that development for yourself. This means ongoing support personnel costs. While you might not need to keep the whole dev team on, you'll still need some of them because they are going to have to maintain the software. Again, most costs to factor in.

3) It won't be ready right now. If there's an off the shelf solution that meets you needs now, you have to weight that against the development time for what you'd need to add. It isn't as easy to put a dollar figure on, but it factors in. Saying "Oh just wait 18 months," isn't so easy to do.

One area I've personally seen this as a real problem is video editing software. The OSS solutions are pretty abysmal next to things like Sony Vegas Pro or Apple Final Cut Pro. Now those aren't cheap, but in most cases I bet they are way cheaper than trying to fix up an OSS solution. I mean say you've got a shop with 20 editors that all need their own copy of Vegas. That'll run you $12,000 for the licenses. You decide that the included 40 network rendering licenses are enough for the farm for the workload. You also decide that you want to purchase their yearly-ish upgrades, so about $5,000 in maintenance per year. This assumes no discounts.

Ok, you think you can develop OSS to be the same level of quality for that price? Not likely, you can't even hire a programmer for that, never mind that it'd probably take more than one as well as other people (like designers to make it nice and usable). Never mind that your work either has to wait until its done or you need to buy something now. Makes much more sense to just buy the commercial solution.

So while OSS can be a cheaper solution, and can be a better solution, there is no guarantee it is. All the costs have to be evaluated and that includes things like "Does it do everything we need?" and "Is it easy for non-technical users to make use of?"

Re:There's also functionality to consider

[TheRaven64]

It's worth noting that the economies are slightly different if you are a government. By nature, a government is a monopoly and benefits directly (in terms of tax revenue) from increases in the local economy. If software doesn't do what the government needs, they hire local programmers. This means that the money is staying in the local economy, rather than going abroad, and so they get more tax money: They'll get some percentage back immediately in income tax, while they'll get nothing back from foreign purchases. Then they'll get more back from sales tax, and so on, as the local programmers buy things. If they then release their changes, then that means that the software is now better and will benefit companies. Some of them will then be able to use it unmodified, and spend money on other things, rather than send it to a foreign corporation.

Overall, spending $1m on Microsoft software might, for a government, be a worse decision than spending $2m on hippyware.

Re:Hidden costs of open source

[Alex Belits]

So for a long term saving, it's often cheaper to stay with what you've got (or for a new installation, choose the same as everyone else) and pay a lot of licensefees, than to change to something that's cheaper in licensing and have a shitload of other costs.

In the long term it's NEVER cheaper to follow a vendor's lock-in.

That said, I LOVE linux, open source and free software. But for commercial use, it just isn't always optimal.

Oh, the hallmark of Microsoft astroturfers.

Re:Do this guys know the definition of user lock-i

[mrjb]

Switching cost includes more the just the cost of the software itself.

Yes. in this case it also includes the cost of training people that have never worked with anything but Windows. That is, of course, if you assume you *have* to retrain your existing admins, rather than firing two of them and replacing them with a single Unix admin. In the end, it all depends on how you make the calculation. Sure, a switch *could* cost more, but it *could* also cost less depending on the scenario you choose to follow.

Re:Hidden costs of open source

[SharpFang]

But for commercial use, it just isn't always optimal.

We're not talking about COMMERCIAL.
We're talking about GOVERNMENTAL.

In this case cost is a far secondary issue.

1) while License is usually cheaper than full IP rights for one item, when it comes to deployment of thousands it's often cheaper to purchase IP rights and be free to deploy as much as you wish (one per every citizen of the country...?) Also, starting your own support dept. in this case may be desirable, especially if the problem is in the software and the vendor is not willing to fix it.

2) Cheaper. Safer? More available? Without creating dangerous lock-in? Without danger of losing backwards compatibility?

3) Yes. It doesn't have to be gratis. It must be open.

4) The life cycle of a well established product ends when the vendor says so, and that's the final end. The life cycle of an open-source product ends when you're not willing to support (pay for) its development. Nobody can force you to upgrade if the current version is better than the new one.

5) A lot of software could be written for the cost of licenses of purchasing software that is already written. It's taxpayer's money better spent if the taxpayer gets a piece of software they can use in return, than if a foreign firm gets to sell some licenses.

Re:Hidden costs of open source

[wrook]

I'm curious. Who do you get for support for Microsoft products. Does Microsoft offer support? And by support I mean, if there is a bug in Word corrupting your mission critical documents, will they promise to fix it? And will they give you a projected time for completion on the work. And will they give you periodic updates? And will they send you a patched version as soon as it is fixed? How much does that kind of support cost? Are you sure it's really cheaper than an open source project?

And what happens when Microsoft "End-Of-Life"s a product? Can you get support from a third party? Can you develop internal resources to provide support and add small features? Or do you have to simply buy whatever Microsoft replacing it with, regardless of whether or not it fits your needs?

And when you say that finding people able to do internal support (I assume first level support, since you can't really do anything else with proprietary software) is easier and cheaper with more popular software, isn't this simply a training issue? Do you really have such a high turnover rate in your company that most of them were trained in using software at their previous job? Or are most of them trained at your company, meaning that it doesn't matter if it's the most popular software or not -- It just matters that you can find initial training at a reasonable cost?

Certainly I think it's a good idea to get support for software you buy. However, I have never worked at a proprietary company that offered anything resembling what I think of as support. "Support" in the industry means get the off the phone as quickly as possible because every minute on the phone eats your entire profit. Sure we did special one-off deals for customers who bought 10,000 copies of our software, but we gave them a bloody hard time of it. If they didn't threaten to not upgrade to the next version, they wouldn't get anything at all. We might fix their bug in the next service pack, or maybe not, at the whim of the program manager.

Real support, meaning having someone who is contractually obliged to help you when your software doesn't work for you only seems to be available for custom built software. And if you aren't getting source with your custom built software, you're getting ripped off.

Or at least that's been my experience. It would be interesting to see how your experience differs.

Re:Duh...

[sznupi]

Users of open source software don't exist in isolation; the economy of scale is huge in their case, too.

Re:Do this guys know the definition of user lock-i

[TheLink]

Yeah, you also have vendor lock-in with reiserfs.

Exactly right

[HangingChad]

In the end, it all depends on how you make the calculation. Sure, a switch *could* cost more, but it *could* also cost less depending on the scenario you choose to follow.

Having actually replaced proprietary systems with open source alternatives, I can tell you none of the expense talking points that usually get thrown around by people invested in Microsoft products have ever materialized. There are always minor disruptions, but no worse than moving to the next version of a proprietary product. The license savings have been huge, but it's more than that. You don't realize how often proprietary companies come back and back for another drop of blood until they're gone. It's like Little Shop of IT Horrors. The up front license costs are only one layer of cost savings.

This may not be a great example, but the last company I worked at saved big when we replaced Exchange with Gmail, which I don't consider an open source product. Not only did we scrap Exchange and the associated server OS licenses, we let the Exchange admin go and replaced them with a lower cost developer. That saved a ton of money and we were able to channel that savings into increased productivity. Double bonus. Gmail is simple enough the help desk could manage the administration.

Really, it's all in how you implement the changes. The barrier for most companies is that their IT decisions are being made by people invested in proprietary technology. They'll never get out from under it.

Re:Exactly right

[HangingChad]

How'd that Gmail outage work for you?

You must be an Exchange admin. There have been three disruptions in two years, only one had any impact on business operations and that was only for a couple hours. Not enough to justify the cost difference.

How's that new cubicle working out for you?

Re:Exactly right

[Vu1turEMaN]

Similar scenario.

I'm interning at a 25 person non-profit. They were putting thousands into Exchange. I did 4 things:

1. Switched them over to Google Apps for free. Saved them loads of money, and they all love having the ease of access. To the exchange admin below, suck it. Seriously, that one outage was nowhere near as bad as the spam problems and other hassles an offsite exchange server created.

2. Got the people who "just couldn't" use gmail's web interface copies of Outlook 2007 through techsoup. Which, after 3 months of switching, was only the secretary and the president.

3. Switched our 5-computer lab for visitors and program members over to linuxmint. It needs no configuration, let alone administration, and its better than the prior windows 2000 by far.

4. Set up Hamachi for remote file access, because nobody used the VPN anyways (cause "my home computer is so slow and full of WeatherBug!").

5. Set up an open source phone server. It was a PITA, but it was WAY better than renting terrible equipment from the phone company.

Re:Hidden costs of open source

[TheRaven64]

Project foo stops being actively maintained. You hire 5 developers, 2 testers to work on it. You pay them ~50-60k/yr. (5 devs, 2 testers - This is for a smallish project.)

If you're the only one covering development costs, this implies that you are the only one using the software. How long do you think a commercial software company would keep supporting a program with one user, and how much do you think they'd be charging you? If there are other users, then you can share the maintenance costs with them. Typically, you'd get a company like Sun or IBM to adopt it and sell support to the current users.