Cost? Don't make me laugh
The cost of all those committees and quangos alone will dwarf license costs, let alone actually writing the plans, and that's before we even get onto implementing the plans.
Iceland’s government is accelerating its move to open source at the expense of proprietary software from the likes of Microsoft and Oracle. The government has launched a one-year migration project moving public institutions to open-source software and is working on a call for a tender to buy services based on free and open- …
Actually, this is probably the best time to make those changes, because of the significant changes being introduced in Windows 8. Likewise if you havent switched to the latest version of office yet, Swtching to Openoffice will probably involve less Retraining.
Besides, i've worked at a number of large companies over the years and have yet to see any evidence of company sponsored "Retraining" due to changes in OS and Office.
"just extend service contracts" - That's the reason to switch.
Hello Mr goverment, microsoft here. we've just decided you are upgrading all your machines to Windows8 and Office 2012-turbo-ninja-cloud edition.
You don't want to?
Well in that case the service contract for next year is $1000/seat, then $10,000/seat, then .... stop me when you've had enough
We switched a huge project from VMS to Linux in the mid-90s when Dec informed us that the we were switching to Ultrix or the licence on our Vaxen was going up to around 10x the purchase price.
We figured they would later do exactly the same to us again so we bit the bullet , bought a bunch of cheap NT alphas, installed RedHat and rewrote everything. Anyone remember what happened to DEC?
There will indeed be significant costs relating with retraining, assessment, writing plans, support, implementation of change etc. But these are mostly one offs. As they are saying in the article - once the foundations are laid, and if they stick with it - significant savings will be there in time - once the initial cost is amortized and they have to pay no more regular license fees.
Interesting move from Iceland. I guess, like with any other IT project - it will all be down to how well and smooth it is implemented.
So no iPads etc. for the Icelandic government eh?
Wonder how well that will go down. I expect the government might just get a few calls from MS, Apple and friends.
Seriously, its about time that cash strapped organisations started to examine the feasibility of using Gnu/Linux and its associated ecosystem to provide IT services.
Open source is a laudable aim but I believe it is true open standards that are the better one first up. The ability to be compatible with others and switch out your provider upon them turning out to be useless tw@ts is far more important than someone having a huge pile of code they don't understand. Take the UK Government as a prime example.
Open source stuff is only free if your time is free, especially where nix stuff is concerned.
I have nothing against open source software, however a good percentage of the time it's not great for anything other than hobbiests or home users.
Like I say before the down votes roll in - I have nothing against it, but it has it's place and that's not always in business.
"I have nothing against it" & "not great for anything other than hobbiests or home users"
I think you might have something against it if you ignore the vast numbers of commercial Linux webservers and the large number of scientist and technologists who use Linux etc.
I think it's seen as quite good for supercomputers as well
« it's not great for anything other than hobbiests or home users. »
Wow AC, it's always impressive to see someone as clueless as you. Actually, the usages you mention is precisely where Open Source is the least used because most home users won't bother changing the pre-installed operating system, contrary to professionals who happen to know what a reliable system is.
Bravo, you just made my day.
"There are still quite a lot of Windows only tools in the office arena."
And you know the names of some that are relevant to Iceland? Perhaps you could tell us so that we could either tell you how to get around the problem or indeed let us learn of a real reason the stick with Windows.
"I have nothing against open source software, however a good percentage of the time it's not great for anything other than hobbiests or home users".
Yea it sure looks clunky .. :)
UbuntuTV
http://www.youtube.com/watch?v=7jYj1gio7qE
Ubuntu 12.04 LTS - Unity 5.0/HUD http://www.youtube.com/watch?v=q49P6czyPs0
Ubuntu 12.04 http://www.youtube.com/watch?v=hzfvUKgjJtg
While I agree that open source is only free if time is free, the story is about open source being _cheaper_, not being free. I can easily see how it would be cheaper.
Firstly, there's an open market for supporting software like Linux because it doesn't come from a single source. You can shop around to find the best support deal for your organisation.
Secondly, most system support in most organisations is provided internally by a department hired and trained for whatever software stack happens to be in use. It's relatively rare that you kick a problem back to the supplier and very unlikely if they're going to restrict what they use to the big name projects that have seen wide deployment, like the Apaches, OpenOffices, etc of the world. So in-houe support costs probably remain the same.
That all being considered, you hopefully end up spending slightly less on support in total and nothing whatsoever on licences.
On the assumption that someone reasonably intelligent has set up the computers and locked them down in the same way that most corporate machines are locked down (so, e.g. for a desk staffer it'd boot up to a GUI desktop with a browser, a Word-like word processor, etc, and all customisation and package management would be disabled) I also don't imagine you're looking at any real extra training costs and in any case training costs are a one off. You'd budget for maybe a month of slowed productivity as a switchover cost, which probably would pay for itself within a year.
@Ivan 4 - I'd be really interested to know how that went for the Gendarmerie, I consider myself pretty much agnostic when it comes to Lin/Win arguments, each is better for some things, each is poorer at other things. I have had a good old hunt round with Google and can't find any articles about how it went, it's all just that they were going to do it. This is odd, because when Munich changed over there was lots of fanfare and then with the exception of a few fanboy web sites (on both the MS and Linux sides, so discount both for fairness) the only thing that I can find out is that they haven't finished yet and the project was initially conceived in 2003.
Re the Gendarmerie,
The Gendarmerie seem to be very happy with Linux, They can't play games but as the Big Boss said they are not there to play games.
Munich decision to change to Linux invoked a panic in Microsoft with visits from top execs, and bribery to prevent this mutiny.
Nice to know it is still proceeding though.
Hmm, that somehow contradicts with the many hours I see my Windows loving co-workers dealing with some bizarre Windows only problem. Or some Windows admin, who couldn't even find the file permissions dialogue on Windows XP. Or the fact that said dialogue doesn't exist on _some_ versions of Windows.
@Eulampios - Supported software is expensive, be it FOSS or COTS, if you need support you have to pay and most companies need support. You can pay for that by purchasing some off the shelf software or you can subscribe on a per install basis for FOSS, it sill costs.
(Although I do agree that the "hobby" comment is transparent fanboy bollocks.)
It doesn't contradict with what I meant. Although, say a Lenovo lappy with XP has an hdd owee. It refuses to boot. no backups could be retrieved from the machine (with lenovo 7 installation/rescue cds!!) ubuntu could not fix the ntfs, but could successfully save the data onto an external media.
In this case Windie tools were overly complicated hence expensive and ... did I mention ? -- useless. Booting a live Linux distro is so elementary and cheap. So you can take it from there to extrapolate.
Who's bollocks did you mean?
Probably not Apple or Oracle, but MS requires mindshare which insists that no option is perceived as credible.
With nothing to lose, cue freebie lunches, windows and office (360?) licenses.
Also, Iceland is so small, the complexity and hw requirements are probably not too great.
A lot of commercial software doesn't support Icelandic characters or dictionaries, so open source is one alternative for a country where words like 'haestaréttardómari' should come with a health warning.
There's also a bit of an urban myth about Icelandic Windows. The traditional word for a window is 'skjár'. Icelandic windows traditionally weren't made from expensive imported glass, but were usually stretched sheep's stomachs or placentas. Those of us who had a Vista machine, can fully understand the experience. Usually though, most people say 'Windows' and everyone else just nods.
Dictionaries, I'll grant you, but Icelandic is fully coded by ISO 8859-1 or Windows-1252 (The Icelandic-unique characters Þ,þ,Ð and ð are at 0xD0, 0xDE, 0xF0 and 0xFE respectively in 8859-1 and thus Unicode too). Basically, if you can't write Icelandic in your character set, you can't write other "minority" languages like Spanish, French or German either.
The only living Western European language that cannot be coded in 8859-1 is Welsh. The reason for this omission that I heard will really, really annoy Welsh speakers, so I won't repeat it, but you can probably guess what happened...
What characters are used in Welsh that aren't in the Roman alphabet? I just Googled "Welsh Alphabet" and didn't find such, just different pronunciations for things like w and dd. (Also scripts in which d g t and some other letters are written rather differently, but isn't that a font not an encoding issue? )
Welsh (can) use a caret (circumflex) to denote a 'long' vowel (e.g. Siân) - there are other, less common, diacritics as well. This is no problem for a,e,i,o,u - but Welsh also has w and y as vowels and there are no appropriate accents for these characters. See Welsh orthography.
Yep, the circumflexed letters ŵ,Ŵ,ŷ and Ŷ were relegated to ISO 8859-14 ("Celtic") along with other "obscure" letters like the dotted consonants that were used to write Irish before the orthographic reform of that language way back in 1958 (modern Irish is encodable in 8859-1, of course)
The lack of these codes in 8859-1 (and thus Windows CP1252) is the reason you sometimes don't see them in online text, but they're part of the language. Similarly, Germans used to write "AE", "OE" and "UE" on some computer systems, but that doesn't mean that Ä, Ö and Ü were not part of that language's orthography.
All these other coding systems are being relegated to the dustheap of history, hopefully sooner than later, by the increasing popularity of Unicode and the coding system UTF-8. Unicode defins a certin "code point", or number, for each and every character of each and every language used by humankind, plus ones for many symbols, too. I'm not sure if they've gotten around to making codepoints for the most obscure languages, but you can be sure all characters of all the living European and Asian languages are included.
UTF-8 codes Unicode characters. Each Unicode character (or "code point") is converted to one or more bytes, the number of bytes depending on how obscure the character set that the character belongs to is. An ASCII text file is already the correct UTF-8 encoding of the text.
More here: http://www.joelonsoftware.com/articles/Unicode.html
While it is true that going open source won't save the entire cost of the licenses. It is also true that the transfer costs reduce over time and use+input improves the breed. Licensing cost just keep going up.
Also some companies like change for change's sake to keep the product looking 'fresh'. New versions proprietary products are not free from 're learning' costs and can come bundled with 'features' that the buyer doesn't want. Would anyone here bet against idea that if MS Office were open source there would be a Non-Ribbon fork of the latest release by now?
This post has been deleted by its author
Ah but to put the counter-point:Open Source stuff tends to fragment. How many versions of Linux are there now? How many UIs? Has the KDE v. Gnome battle ended yet?
There's more than one version of Windows but aside from the changes wrought by UAC most developers don't have to concern themselves too much with that. Even the UAC changes aren't seriously breaking since the OS does it's best to accommodate you. The result is an ugly kludge under the bonnet but it works. Code for the Windows version on your dev box and chances are good to excellent that it'll work on any other Windows box in the world. Same with the UI. You code for 'Windows' and that's that. There might be a wide choice of frameworks (which is good) but you don't have to worry that customers are going to start bleating that your application doesn't run on their particular Windows UI.
It's a complex equation all round but all I'd advise anyone thinking about this is that they think about it hard and impartially.
"You code for 'Windows' and that's that."
I'm not so sure that that is really a problem.
Apart from the obvious security issues, most office type workers would be just as happy and productive with WfW 3.11 and MSWorks, in terms of MS use.
If a Govt chooses to go the open source Linux route, then there's almost certainly no need to fragmentise their own internal market with multiple desktops or underlying OS versions or to keep updating it every year.
So what if there are different interfaces available? There will be a standar configuraton for all PCs. You can run KDE programs on Gnome UI very easily despite what you would like to imply, most package managers will pull dependencies automatically, never mind that actual it professionals should have no problem with that. But yeah nice beating that strawman.
> How many versions of Linux are there now?
Linux is the kernel. There is one current version.
What you may be referring to is that there are many distros, and many windows managers. This is not an issue because software developed for one framework (Qt, GTK, etc) will run on any of the window managers on any of the distros.
The trouble with forking is that you get two versions the system ceases to be one as a whole. Some people will chose to follow one fork, some the other. The practical upshot of this is that you can end up choosing the wrong fork and ending up in a dead end, possibly even with abandonware, which you have to extricate yourself from. Commercial software doesn't suffer from this as, while sometimes unpopular decisions are made, it carries on in the same direction and your company doesn't end up risking making the wrong choice.
(This is not to value FOSS or COTS more than the other, just a comment on forking.)
> Commercial software doesn't suffer from this as, while sometimes unpopular decisions are made, it carries on in the same direction
Taking some examples: Windows Mobile 6.5 was totally dead ended, there was no path to WP7.
Visual Basic 6 was deadended. Stuff had to be rewritten to move to VB7.
IE6 is still being used because MS specific code was used in the intranet, and created by FrontPage, and it won't work in later IE versions.
When I look at my Apache logs I see Windows machines advertising that they have several versions of .NET installed. They do this because they are all incompatible with each other and need different runtimes.
> The examples you cite are all of a product ceasing, not forking.
They were example of NOT "it carries on in the same direction" (to quote the original). The replacements were complete disjoints.
In fact if they had forked then the old VB6 code, or IE6 stuff, or WM6.x would _not_ have to be thrown away and/or rewritten.
@Richard, I agree, but: If a package is ceased, you know that you have a certain amount of support left - it's usually in support for several years after the "end date", therefore planning smooth transition is more of an option, with no following the wrong path for any time. However forking means that you make a choice of which fork to follow, which could mean several years of development before something is dumped, this means that you get much further into system development and therefore have more to undo before re-doing it. Also, when a package is ceased - there is usually a good upgrade path and tools made available to assist the upgrade.
@AC "The examples you cite are all of a product ceasing, not forking."
And you were talking about the faults of picking the "wrong" fork. To quote you:
"you can end up choosing the wrong fork and ending up in a dead end, possibly even with abandonware, which you have to extricate yourself from"
This is not different that choosing a commercial product from a supplier that eventually goes to the wall. And now that you have been caught with your pants down you are trying to change the argument.
You are wrong - man up and admit it.
The argument doesn't wash. If you pick the wrong fork, you simply switch, generally with very little pain or change. OTOH commercial software often just dies on the vine and you get stuck with migrating to something totally different. Then again as a person who get's paid more to implement the more painful change. . .
@AC Your "fork" argument is such utter FUD it is incredible.
Let's say you pick the "wrong" fork, so what? You have the code base, you have the file specs, you have another (probably compatible) code branch. At worst you and your fellow community members pay a few devs to transfer your files over. Heck, if there's enough of you it might be worth your while to keep the "dead" fork going, at least until such times as there is a better transition point.
What happens when your "COTS" software provider dies? If you did not have an escrow agreement, you are screwed. If that escrow agreement did not include the file specs, you are screwed. If those files specs are not complete, you are screwed. If you don't have a list of other customers who you can band together with to build migration tools, you are screwed. And by "screwed" I mean "looking at a very large bill, for a one time job that may not work properly".
So if I had to pick a hole to find myself in, I'd pick the fork hole as I would be in it with a bunch of other people and the tools would have been lying around to get ourselves out.
You, AC, are nothing but an anti-F/OSS shill; what's that pay by the hour?
A company who aren't software developers are no more helped by access to the source code than any other - Indeed at my last company we knew of a problem with RHEL, we knew how to code around it, but it required a re-compile of the Kernel - RHEL told us that if we recompiled the Kernel we would receive no support.
As it happens I am certainly not anti-FOSS, I use FOSS every day at home and work - at work I have a desktop running Windows and one running Linux, at home I have various Linux, Solaris and Windows servers and desktops/laptops. There are relative advantages and disadvantages of both, it's the black and white, polar attitudes that "if you criticise FOSS, you're a COTS shill" or vice versa that I really object to. They both have their place and both are not perfect, however both are held up by some people as perfect, which they clearly are not.
@AC "They both have their place and both are not perfect, however both are held up by some people as perfect, which they clearly are not."
This from the person who claimed there was no risk with making the wrong choice of commercial software! So not only are you deluded, but you are hypocritical too.
" Commercial software doesn't suffer from [fork risk] as, while sometimes unpopular decisions are made, it carries on in the same direction and your company doesn't end up risking making the wrong choice."
Of course there are risks; companies can drop support forcing a migration or go bust. No real difference from a fork drying up, except that with F/OSS you already have the code, specs etc and at least have a small chance of saving things and a much greater chance of an easier transition.
Give it up already, or would you like a bigger spade to go with that hole?
"You do realise that it's not necessarily the same AC posting don't you?"
Quite possible, which is why AC comments have to treated with some distrust.
Even if it was the same AC, they clearly don't believe in what they are saying to put their El Reg username against it, something that a shill is likely to do.
All the programs relevant for office-workers such as Firefox, OpenOffice,Gimp and package tools have been developed nicely during the last couple of years. Thousands of developers continue to improve them. For example OO can now nicely open and convert MS Office documents.
I call your post FUD because there are certainly as many commercial packages which are end-of-lifed as there are OS projects which are no longer actively developed. The difference is that you can get the source for any OS project and use the program indefinitely on an unbounded number of machines. Commercialware vendors will simply dictate that this software must not be installed on new machines any more. MS simply forced tens of thousands of happy companies to switch to Win7, despite the fact that they were completely happy with XP.
It's no great surprise that the various tech companies mentioned can successfully transition to Linux desktops.
Does anyone have real world examples of corporations or government organisations in other sectors (i.e. where MS Office was the main windows usage) that have successfully transitioned and made a financial saving?
I believe Omlet <www.omlet.co.uk>, proprietors of posh chicken coups are an Ubuntu house. Also seen other small enterprises running Gnome-y looking window managers for the office staff on my travels and would imagine cost is the most likely driving factor, coupled with some tech savvy individual in management perhaps.
I hope they plan this out and do the transition right. If it works, it will help foster more expertise and interest in *nix rollouts.
That in turn might make MS get off their a$$ and tighten up the code (and the costs) for Windows.
That makes it better for all platforms.
Fair competition is good.
I actually think MS did a reasonable job with 7, and I think that is because their pocketbook got hurt with Vista and the subsequent race to alternatives. People realizing that they don't need a pc for watching youtube videos and facebook hurt them too.
Hopefully Iceland calls them. Who knows, maybe some techie in Munich will get to move to Iceland to fulfill their lifelong dream.
Heck, Iceland might have a suitable homegrown *nix administrator/user base to hire from with no need for imports.
I'D like to move there, at least for one offroad racing season.
In my experience, Windows Server works reliably and without "special" maintenance needs since at least Windows Server 2003. It also appears that Linux desktops built on top of any of the more popular distros are also fine for office use since about or around the same timeframe (since 2005 or so, at least). But old prejudices die slowly.
When it comes to iceland, just the reduction in trade deficit probably pays for the transition costs. And there should be much less bureaucratic overhead than there would be in a bigger country - how many layers of government do you need with a population of 300k, after all?
Can we have an update about what's happening in Munich? I seem to recall that Munich planned to move 14,000 seats to Linux? How did that work out? Did they save money? Were there other benefits?
Here at PreppyCentral we've been a Red Hat/Fedora shop since 1999 with no problems at all. And we've cut our licence costs to close to zero.
About Extremadura. I don't know if they have reached that number quite yet but typically that region is one of the poorest in Spain so the incentive is the same, saving money, in the first place.
But I doubt it could ever lead to poorer results than using Windows.
Linux would fit many of the Eastern Europe countries (and any country) very well too but a fairly common opinion is that the governments in the east have been successfully bribed.
And, as was already pointed out, "open standards" is one of the greatest benefits of using open source.
http://linux.slashdot.org/story/12/01/24/0416236/spanish-extremadura-moving-40000-desktops-to-linux
A friend of mine has been using one called Zarafa with ClearOS for small shops. Seems to work quite solid - and the web interface is quite nice. Not sure about licensing - the usual game in this space was charging for MAPI (Outlook) connectors, but hadn't heard about it for push mail to phones.
I find it amazing how many organisation claim that the retraining costs would be to high to convert to linux/open source, yet when they upgrade their Windows system they don't bother doing any retraining anyhow.
Also the number of people who insist they need retraining with software updates, yet can easily manage to use new tech like IPads (this was also the case when Psion, Blackberry came out) without any form of training.
Its just how you approach the transition that matters. I would personally like to see that all data is held in an open standard, then we can switch to the system that works best.
Otherwise their plans would have been opposed in the EU courts by MS as being "anti-competitive" or similar, just like they managed to get the requirement for open standards crushed in the UK and and keep them off the statutes within the EU.
MS as an entity utterly opposes F/OSS and does everything within its power to undermine the efforts of the community to help itself.
making a statement like "the goal is not to move everybody over in the space of just 12 months but rather to lay a “solid foundation” that institutions can use to plan a move."
Normally project leaders say things that make my arse feel like it has been violated with a baseball bat...