Slashdot Mirror
Birmingham Drops Open Source Initiative
eldavojohn writes "Birmingham, England put a stop to a half million pound project to put Linux and open source applications on library access PCs across the city. From the article, 'The council planned to roll out Linux software and applications on 1,500 desktops in libraries across the city, but in the end went no further than a 200-desktop project. Several industry watchers have voiced their concerns about the project, particularly around the number of PCs rolled out. Birmingham's expenditure averaged over 2,500 pounds per PC.' Why did they stop after 200 PCs? Because they claimed with Windows, the project would have been 100,000 pounds cheaper. One may wonder if they paid for initial training of their workforce making the first 200 more expensive than the rest but the article does not say whether or not this occurred."
This article brought to you by..
"Microsoft: Where do you want to go today?"
Tequila: It's not just for breakfast anymore!
A quick read thru the article reveals not a problem with Linux, but with the idiots trying to manage the deployment without knowing what they were doing.
I feel sorry for Birmingham. Not so much for having to use Windows, but for having to live with an IT staff like that one.
Learning HOW to think is more important than learning WHAT to think.
I used a couple of the Linux machines in their main library, and they were rubbish compared to the Windows ones. I think whoever set it up hadn't bothered using the machines themselves! They even had US keyboard layout set, did they just plough through the setup wizards clicking Yes to everything??
The actual article is titled: "Criticism mounts over Birmingham's Linux project"
This is a followup on the project being discarded, mainly focusing on critical comments of how the project was managed.
Notable quote: 'Mark Taylor, whose Open Source Consortium also exited the project in the early stages, said: "I have no idea how anyone could spend half a million pounds on 200 desktops, running free software".'
Anything is possible given time and money.
...when Microsoft is paying you to use Windows.
...the Birmingham city council is using gas lighting because of the cost of teaching their employees how to flip a switch to turn on electrical lights.
£333/desktop if they had rolled out the full number of desktops.
Its not surprising that they spent a lot of money to achieve seemingly nothing - Birmingham City Council BOASTS all over the place that they are "the biggest employer in the West Midlands". Probably cos it takes 10 muppets to do the same job that 1 competent employee should be expected to do.
SURELY NOT!!!!!
It depends, really. It depends if you have to retrain staff to use the new systems, if you need to hire extra support personnel, if you need to buy hardware that works with Linux... all sorts of things could affect the overall cost, not just the license cost.
By summer it was all gone...now shesmovedon. --
It's possible to save even more money, Birmingham. Here's how: First do what ever it was that apparently would make the project cheaper with winblows, next replace windows with linux. It's free, so ALL the money spent on windows would be gotten back.
Let's reiterate: 1. Save £100,000 2. Save £++ by using free software 3. ??? 4......you know the joke by now.
In other news, Birmingham, Alabama is doing the exact opposite. Open source has fluorished here, as low funds make one go to low-cost alternatives.
"No freeman shall ever be debarred the use of arms." -- Thomas Jefferson
The Munich migration is more expensive than the windows upgrade would have been. They just made a deliberate decision not to use Microsoft.
Do not try to read the dupe, thats impossible. Instead, only try to realize the truth
What truth?
There is no dupe
To me it seems that the biggest cost with Windows is not the upfront cost per seat, it's probably the cost of maintanance and data lost due to viruses and spyware, but hey, what do I know....
"It is our choices, Harry, that show what we truly are, far more than our abilities." -- Prof. Dumbledore
The RIAA has to win every every court case because, by the legal principal of non-mutual estoppal, if they lose once they cannot use the same legal arguments in any future case they might wish to bring (i.e. if P2P music sharing occurs through an IP address you pay for, you're automatically responsible, guilty, and owe them lots of money regardless of what you actually did, or didn't, do).
Microsoft has to win every desktop every time because, if a large-scale commercial Linux deployment succeeds as a viable alternative to Windows, it will be considered seriously as a candidate in every future large-scale deployment of PC's. Microsoft will have to fight for every future desktop contract, instead of being the de facto only option for 99% of them.
And both groups are willing to do whatever it takes to win at all costs!
"It's the height of ridiculousness to say for those 9 lines you get hundreds of millions."
Linux is harder to set up, thus it will be more expensive for governments to switch over to linux because it takes more time to set it up. Whereas with Windows, after being in the market for so long and having a lot of people at least be exposed to it once or twice in their lifetime, will not require as much time compared to linux. Therefore it is cheaper to use Windows both in the long term and short term.
Once linux has the same support, features, ease-of-use as Windows has then it has a chance of succeeding and taking a good chunk of market share. But as linux continues to be protrayed as the geekdom of software, it will not be a cheaper solution to Windows.
Previewing comments are for sissies!
If they were genuinely slow-witted enough to make such a calculation, how do you figure their chances of maintaining a large Unix install base. And if your figure is significantly greater than zero, what does that say about the intelligence required to be a Unix admin?
What I'm listening to now on Pandora...
Ignorance is a very big problem with Linux, which results in unsuccessful deployment and failure. It's not an easy job for making people to switch from one environment to another. Building similar GUI is not a resolution. Most computer users do not know what they use. And people should increase their knowledge if they want to use computer. Using windows is like learning driving on one car. Though you need to learn standards, you need to know what horn is and used for, instead of learning it like 'you push that place and it emit sound'.
We have to put Linux awareness on computer education. Else people would behave Linux as Windows and once they fail they would blame that on Linux. Administrators of Windows think that they know everything due to their computer training and once they encounter something different, they think its broken 'even though they did everything right'. Users thing applications 'do not work', or 'does not do something' because they can't see their familiar GUI in front of them. They don't even check other places, or don't even know where to look at it.
Technical personnel can't report bug reports, can't realize what causes the problem. They mostly get used to 'reinstall' or 'restart' to fix stuff never in need to knowing cause of previous problems.
And even worse, since they don't know deep working of some basic stuff, they design current systems platform specific. They don't use standards but rather using platform specific tools or ways to handle things due to their 'buggy training'. And when they need to change platforms they have to reinvent lots of other fixes they had before.
Summing all that up, they stay in the middle of vendor lock-in. if we can't educate people well on computers, and they think they are educated enough, they would not blame their knowledge but the products.
Dude, you are underestimating the competence and worth ethic of Doozers when you trash Muppets like that. They working harder than illegals and their work tastes like candy.
Tequila: It's not just for breakfast anymore!
I can write malware for Linux. I threw out a script that hijacks sudo to get root access using a couple bashisms. Drudging up passwords is nothing.
Support my political activism on Patreon.
At current exchange rates, that is $4,736.75 per desktop
Not really. Anyone who knows *nix can adapt to Linux in a couple of days. And there are lots and lots of people who know *nix out there.
True, they might be more expensive than someone with an MCSE. But the MCSE you'd hire/contract for a migration of this size would be more expensive than the MCSE you'd hire to maintain a site that has already migrated.
Migration specialists cost the same whether they're Microsoft, Linux, Sun or whatever.
Again, not really. The problem is when people do not look at it as a real migration. If you've ever done an Oracle/Sun migration, you'd know the costs involved and the amount of planning. And those are the kind of experts you'd be calling in for a project such as this.
The strange part is how they could spend so much money, so quickly, on so few PC's.
Realistically, they should not have spent 1/20th of that before finding that Microsoft would cut their sales price to come under the Linux figures.
And most of that money would have been spent on identifying all the apps used and which could be ported and for how much.
Linux desktops are cheaper to run than Windows. Particularly if you're using them in a diskless environment.
The HUGE costs are porting the apps or migrating the data to Linux-based apps. This is because most vendors have spent time locking your data up in their proprietary formats in order to make it as expensive as possible for you to dump them.
Which is why migrations such as this are STUPID to rush into.
It makes far more sense to plan them over 5 years. That way, the cost of migrating/porting those apps can be compared to the cost of upgrading them (or migrating anyway when the ISV goes out of business) and the real savings can be seen.
And you can realize the easy savings sooner to off-set the more expensive projects later.
I rolled out linux on my computer illiterate mum's computer, from another country, on the phone, by telling my windows-only using little brother how to download and install Ubuntu (no dual boot, of course).
Cost of the operation => 0 (I don't pay for the phone)
Happy linux users => +1
So cut the crap already great britain !
One may wonder if they paid for initial training of their workforce making the first 200 more expensive than the rest but the article does not say whether or not this occurred.
It's like considering switching all traffic so that vehicles drive on the *other* side of the road. Even if it made more sense, it would be expensive as hell to do. And it's silly to take into account the learning curve of all those who had to initially learn to drive on the current side of the road. What matters is solely the cost of any changes going forward given that they have a staff already trained and familiar with Windows. If it's cheaper in the long run to stay Microsoft and everybody's already reasonably happy with it then, technical and ideological reasons aside, why switch?
Want to improve your Karma? Instead of "Post Anonymously", try the "Post Humously" option.
And looking at the general direction the comments on this story are going I'd say we have a winner. Another great day for Slashdot ad impressions and another "look at what teh evil Micro$haft did" data point to use in the next flamewar.
Web2.0: I love when people Flickr my cuil and digg my boingboing until my google is reddit and I start to yahoo
Especially if you roll out such a small number. You have to keep all of the overhead and support costs of your existing Windows infrastructure and add a new layer of support costs. You won't start to see savings until you reduce your support costs for Windows, this requires a bolder move on the number of seats converted to Linux.
cheers, ben
Never miss a good chance to shut up -- Will Rogers
Because your data is usually locked into the original ISV's proprietary format ... and that ISV knows approximately the cost of a migration ...
So that ISV will price their "upgrade" at a low enough point that the pain and cost of the migration is difficult to justify to upper management.
Remember, software isn't like a car. Once the time/money has been invested in writing the software, distribution is practically free. You make more money the more times you can sell the same code. Even if you have to "discount" it for some of your clients.
The real issue is on "TCO" and "ROI".
How long will it take you, running the new software, to make back the cost of the migration and start really saving money?
It's in the ISV's best interest to lock up your data so that the migration cost is so high that even with huge savings on TCO and massive ROI, it will still take years and years to make back the migration costs. That way, few businesses/governments will be willing to justify a migration that might fail.
I see two simple options. First:
1. Build each computer with 1GB of RAM and no hard drive. A fast CPU is not needed, but the ability to net-boot is required.
2. Set up a Knoppix image on a net-boot server, which the workstations can net-boot from. (The Knoppix image might need to be customized for this purpose, but even if no modified Knoppix image already exists with this feature, it shouldn't be overwhelmingly hard to make one.)
Thus, everything runs off a read-only NFS filesystem, and is impossible to vandalize a workstation on a software level (a reboot undoes any vandalism). Furthermore, Knoppix has proved itself to be very good at autoconfiguring itself on a wide range of hardware. And I don't know ANYONE who couldn't figure out how to use Knoppix if they tried.
Another option:
1. Build each computer with 512MB+ of RAM and at least a 5GB hard drive. The ability to net-boot is not needed.
2. Install Knoppix on one computer's hard drive, and copy the disk image to all other computers. Many tools exist for this, of which Norton Ghost is merely the best-known (open-source alternatives do exist).
Thus, you get the same advantages of the first solution, but with a local hard disk. The first solution would offer easier clean-up on workstations, while the second would result in higher performance. Of course, you could get even higher performance by configuring a net-booting Knoppix to load to RAM, but you'd need more RAM (I'd guess at least 2GB) on each workstation.
Why is it so hard for them? Did they get brainwashed by Microsoft's P.R. trolls, or am I way smarter than than Birmingham's IT staff?
also happens to be the 6th most dangerous city in the US. Open source adoption leads to crime? Hmmmmm?
:)
No, not trolling... low funds lead to crime and open source... just kinda funny
Well, I think I'd like to extend what you said. I think that people should start making their own microprocessors. Otherwise, how do they know how they work? We should mandate microprocessor design in public schools. Otherwise, how can you debug your own kernel panics? After that, we need to make sure that people can make their own hard drives. After all, if you don't know where the 0's and 1's go, how can you fix the problems?
I've already done the same for my car. I won't drive one, until I know how it works. Right now, I'm busy growing rubber trees so that I can make my own tires so I can change one myself. I'm pretty excited. Only another 5 years to go, and I'll have enough rubber to make a tire! After that, I have to learn how to mine iron to make steel for the steel belting in the tires. But hey, I'm not ignorant! I figure in another 200-300 years, I should have the know-how needed to drive my car.
Does anybody know how to make a tire stem and valve? I can't put air in my tires until I know how these little bastards work.
- I doubt that you have proof of that.
- I also doubt that the long-term costs of Windows would be cheaper than Linux.
Any time you make an initial jump on a platform, it costs more. But it is the long haul where you make your costs/money.I prefer the "u" in honour as it seems to be missing these days.
"That's ridiculous," said Eddie Bleasdale, the owner of open-source consultancy NetProject and an early participant in the project. "It's an unbelievable cock-up... They decided to do it all themselves, without expertise in the area," he added, saying that a lack of skills in open source and secure desktops would undoubtedly have raised costs.
Id also wonder about which distrobution they were planing on deploying. Was it RHEL or something as equally expensive?
Not all conservatives are stupid,
but it is true that most stupid people are conservative.
- Hume
Unit cost for a Linux desktop = £2,5000
.. :)
Unit cost for a Windows desktop = £2,433.00
Where did the money go
davecb5620@gmail.com
Of course this is only news because its a Linux IT project failing. There are so many over-budget, behind-schedule public-sector IT projects involving non-Linux systems that they dont make the headlines any more.
Oh, except the new UK Health Service IT system which has just gone waaaay over budget....
http://www.realmeme.com/roller/page/realmeme?entry =linux_meme
The buzz factor of Linux has been falling for a couple of years.
The odds are good that this will reflect back into the actual adoption rate.
What learning curve. This is for library access so I assume they mean web browsing and word processing. What training do you need to use Firefox as against Iexpolorer. I do know of at least one library that has gone the Open Office route on Windows with no complaints. To say Windows is cheaper than Open Source is to use different mathematical functions than the rest of us.
Re:Initial training?
davecb5620@gmail.com
I know I have trouble finding a PC for less than $5,000.
Not necessarily. I'd attribute it to incompetence first.
Reading between the lines, it sounds like the team tasked with this were 100% Windows folks with no Linux experience. Ask such a team to deploy Linux to several thousand PCs, and I'm not surprised it all fell over horribly.
It doesn't help that the problems you encounter when dealing with 100 or 1000 PCs bear little or no resemblance to the problems you encounter when dealing with 1 or 2.
It is truely amasing how technically incompetant some people are. That they can walk and talk is sometimes a surprise. I sometimes think the human race has already split into two streams. If so the problem is we have the worng people running the show and there is little accountability.
If they spent $2500 pounds on average on 200 machines then imagine if only 10% of that money had been made available to technically competant people! The rest of the money could have been donated to the welfare budget.
Q: How can windows be cheaper than a free OS?
A: *nix requires real hardware, boatloads of cheap pseudo hardware that offloads all the work to poorly written windows drivers has flooded the market over the years.
Check out which side of the road Sweden drives on. Perhaps the average Swede is a little smarter than the average Brit?
To cope with the new system. Duh.
By summer it was all gone...now shesmovedon. --
Most companies wont go to fedora, opensuse, ubuntu or your other free downloadable distros they will go to redhat or another vendor and buy the software... Reason being no real support comes with fedora and the other free distros.. Yes the community support is very good but when it comes to businesses and enterprises as wonderful as it being free having 0 support is very risky if a problem should arise its nice to know you can get support... I dont know how the licencing works for redhat but a workstation with std support runs 299 if its another OS provider that adds specific stuff that cost can be higher...which is the same as XP Pro or Even the Vista business is virtually the same price... Now add in your DB will likely have to be rewritten for Linux since its a library im sure it has databases for books and other resources one would assume being found in a library... Even more cost there... Im alittle suprised that it cost so much for only 200pc's but not suprised that end cost of rolling out 1500 would be more expensive than upgrading windows... Im All for free software and linux and don't want to sound like a troll, but when it comes to businesses, governments, and other enterprises I personally don't think that they shouldnt risk their data to a free distros...
The problem of course is that around here it's commonly understood that because I installed Fedora on the two Celeron boxen in my room and didn't spend a dime, that deploying 25,000 desktops across an enterprise should be no more complicated or expensive.
This is a strawman argument. No one else said it did not cost money to roll out 25,000 desktops in an enterprise. The discussion is should it cost as much as they claim to roll out 1500 desktops as workstations in public libraries. The consulting firm that they parted ways with called their costs "ridiculous" and they have a lot better idea of what the project entailed than anyone here.
And looking at the general direction the comments on this story are going I'd say we have a winner. Another great day for Slashdot ad impressions and another "look at what teh evil Micro$haft did" data point to use in the next flamewar.
Who's talking about Microsoft? We're talking about the incompetent shmoes in charge of this project who decided to stop working with two different Linux deployment consulting firms and "do it themselves" with current staff who had no experience and questionable purchases.
Let's not forget that most governments have unionized employees, which (if true) is material to any massive IT redeployment. In true Slashdot fashion, the following post is pure conjecture and generalization. But I think it's plausible.
Ideally you would want to hire expert sysadmins on contract to conduct a pilot project such as this one. However, there is likely to be language in the union contract forbidding a contract employee from taking a job that might be done by a unionized employee. Unless a sufficiently far-sighted employer included specific language covering a Linux deployment, the deployment would necessarily default to the in-house IT people.
And you had better believe that the union folks would be vocal about it. Especially if they -- as Windows experts -- could be replaced by Linux sysadmins in a wholesale system turnover. In fact say they believed that Linux might require fewer sysadmins, thus threatening their jobs. Maybe they wanted it to fail for that reason? Again, pure speculation, but plausible given my previous interactions with unions.
This is not to say that unions are useless or evil. Or even that any of this happened or was a factor in Birmingham. But unions do form part of the institutional culture, and if not taken into account, they can cause projects like this one to fail.
Toronto-area transit rider? Rate your ride.
Unless they decided to use Redhat or some other pay linux distro which requires support agreements. Redhat is more expensive than windows in some cases if you consider how cheap it is to buy a PC with Windows on it. Others have already pointed out that other operating systems require certain hardware to work properly due to lack of driver support. For instance, they could not buy an intel 965 chipset system unless it was all sata and the distro of choice had the absolute latest kernel which is unlikely. When dealing with one PC, its easy to rebuild the kernel but deploying a large number of desktops that may have slightly different hardware, its not as easy. One could argue for netbooting or something but in reality they probably want it to work like windows with a local OS install. Its also entirely possible they did not come up with a good system for imaging. Not everyone is a linux expert and if they used Microsoft friendly employees who aren't familiar with linux, it could be a very difficult, slow and expensive process to switch.
Remember in GNU land, its free as in freedom not in price. Redhat can charge whatever they want for their distro and support agreements. Even if they used a free (price) distro like fedora or gentoo it may still have been more costly with personnel or they simply didn't know how to automate certain tasks like patching that Microsoft tries to do. Most free systems are not polished enough to make mass rollouts easy and knowing how to do it isn't always clear for Microsoft users.
MidnightBSD: The BSD for Everyone
"I have no idea how anyone could spend half a million pounds on 200 desktops, running free software".
Its Birmingham fucking council, the council tax (a tax on your house that does not vary with income, only with size) goes up every single year, yet they still cannot even pick up rubbish bags without making a mess of the whole area.
I love my city but everyone knows that Birmingham council are a bunch of absolute losers, so this does not comes as any big surpise.
My little Linux and tech blog
The article itself does not say nothing. It is quite possible that such migration does not make any sense. If they have loads of Windows-only applications they would have to rewrite it all to web frontends or something. Also tech support and administration can be quite costly if they have no unix background.
But on the other hand the outcome of this case does not provide any specific information where they failed? Was it lack of apps? What distro they used? What strategy? What were main problems? Etc.? Etc.?
For this money to be not completely thrown into trash they should publish the outcome and what they have experienced. For others to learn but also for general public to justify themselves. Right now it is like big question mark on what exactly they did. And this leaves their incompetency as an option too.
What on earth did they try to pull off? So with Windows installed it was 67 pounds cheaper to purchase? I dunno but this sure smells baloney!
this guy actually has a good point. anyone who has futzed about trying to get poorly supported wireless cards or a soft-modem to work under Linux can attest to this.
upon the advice of my lawyer, i have no sig at this time
Also have a read of this, in which they reveal that Birmingham was using suse 9.2 *without* a yearly support contract. Also, they weren't upgrading the actual hardware as far as I can tell, otherwise an XP upgrade doesn't make sense. And what's "an equivalent XP upgrade" anyway ? Did they take into account all the extra software needed to run XP at an equivalent level - anti-virus, MS Office, etc. That's without taking into account that they will have to "upgrade" windows again at some point, and thereby spend *another* £429,960 ! (Except it will be more than that because the hardware won't be up to running Vista).
http://www.ubuntu.com/support/paid Databases formats can easily be interchanged. They don't have to be rewritten. And the training cost is one time only. In fact, if they have a gradual change and not a sudden one as they are trying, they would not feel the pinch.
"..."
I'm not making an "argument" here...
You made an unsupported assertion about what you claim people think. That is an argument.
After 30 minutes my predictions are correct.
Really? Not according to the posts I read in this discussion.
For both my points, take the time to go through the comments posted so far.
I did thanks, I just don't see how they support your assertions.
Now did you have a specific point about what I posted or are you just looking for a scrape?
Here's my point, you're making baseless assertions about "what people think" and are ponderously close to being a troll. You throw around random large numbers and apparently did not bother to read the article being discussed. Just because you say, 'Slashdotters all think this" does not make it so and does not mean there is anything valuable at all in what you've posted. How about instead of generalized jabs at your opinion of the consensus here you try addressing specific posts from someone or *gasp* the article itself.
The real question to me is why their Linux workstations were costing that much. 2500 pounds sterling is about $5000 American- I can get pre-loaded, somewhat functional, Linux workstations at Fry's for $199. I could throw in an additional $200 worth of work and $50 worth of memory to make them as functional as any Windows workstation for $450- under 1/10th the amount they were spending.
SJW: a person who perceives an injustice, and while correcting it, commits a greater injustice.
It depends, really. It depends if you have to retrain staff to use the new systems, if you need to hire extra support personnel, if you need to buy hardware that works with Linux... all sorts of things could affect the overall cost, not just the license cost.
Everything you say is true, but then to have a fair comparison you would need to include the cost of training support staff on Vista and including the cost of upgrading the computers to being Vista, unless the libraries were going to maintain multiple versions of Window's operating systems (which would increase costs even further).
With the limited information in the article, it is difficult to determine how the cost could be so much higher for linux. One would assume that the hardware costs are comparable. There is the cost of Windows, office, etc. versus linux, openoffice, etc., which most likely would have been less costly going the linux route. So the only other area where cost seem to be an issue is on support and training.
To arrive at the numbers they are quoting, one would have to assume that they are spending next to nothing on support and training for Windows and a lot for linux. However, in a true cost accounting model, that cannot be true. The ongoing support and training costs should be similar, so we are really talking about the support and training to impliment the project.
Now, it does stand to reason that installing new technology is going to cost more in terms of support and training than the current technology you are using. However, in their circumstances, that is very short sighted given the fact that Vista is being released within a month and during the middle of their deployment cycle. They could mitigate the training and support for Vista by foregoing it and paying for it out of their regular operating budget, but that is just shifting the funding source. If they were going to do that, they could of done the same for linux.
So, the cost issue, seems to be a red-herring. While I have no doubt it is a real number, it is hard to determine what the cost actually represents and how it relates to the overall project. It's also unlikely that it is a software issue or that would have been used as the reason -- "we have stopped the project because the rewriting of all of our applications turned out to be more monumental and costly than we were led to believe." But that wasn't the reason given. Nor was the cost of support staff or any other measurable reason.
So, by throwing out a number and saying it is too costly, they have 1) silenced critics to the decision and 2) most likely have made a political based decision.
Tah, dah:
# ngrep -q "user"
# ngrep -q "pass"
Excuse me, but please get off my Pennisetum Clandestinum, eh!
Any Government project expands till it exceeds the budget. Therefore the deployment cost of a Linux project simply depends on the size of the available budget. If they budgeted for a total expenditure of $100, then it would have come in at less than $500 for all 1500 machines.
Excuse me, but please get off my Pennisetum Clandestinum, eh!
Mass rollouts? AFAIK all distro's have dd and netcat. It is just a simple matter of incompetence, lack of experience, lack of knowledge, whatever you want to call it, it is not Linux that is at fault.
The trouble is that first of all, you need to know about dd and netcat and then you need to know that you could use Google or type 'man dd'. It takes a certain amount of UNIX familiarity, which a Windoze d00d simply doesn't have.
Excuse me, but please get off my Pennisetum Clandestinum, eh!
Okay I followed each and every one of those links. Most did nothing at all to back up your assertion. Of the few that did, one was modded as "funny" and one was modded to 0. I hope you had fun making all those links, but they don't support your opinions as far as I can tell.
Now, let me save you some keyboard lubricant. Go back to my original post, and then read it very carefully. Understand what I said in that post.
Yeah, maybe you need to take some English classes again, because I read your post. You made numerous assertions about the opinions of people here, and no, I don't see how you can claim most of those links support it. Address them specifically if you have a real point.
Knowing the way these things tend to work in the UK, they'll have had to source them from a single approved supplier, who will have subcontracted it to somebody, who will have subcontracted it to somebody else. And all of them knew it was a public-sector job, so they would have more money than sense and hence were ripe to be ripped off.
The person signing the order's primary concern was probably not "is this value for money?", but rather "will I be able to deny all responsibility for this?"
What matters is solely the cost of any changes going forward given that they have a staff already trained and familiar with Windows. If it's cheaper in the long run to stay Microsoft and everybody's already reasonably happy with it then, technical and ideological reasons aside, why switch?
This would be a good argument if the Win2K/2003/XP line was going to be around for another ten years. It starts getting phased out next month.
Retraining time.
My God, it's Full of Source!
OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
Birmingham's expenditure averaged over 2,500 pounds per PC.
That is just total bs. 2,500 pounds for 200 pc's? Get real. With the help of one other skilled person and a couple weeks of planning I could have rolled 200 pc's in one night. What could be on library PC's that couldn't be replaced or framed to run as a network application? That's just absurd.
Here's a story from 2004 where a library rolled 200 workstations over to Linux and that included public kiosks!
I've seen development projects send millions down a hole when managed poorly. Now we see the same thing is possible when desktop migrations are run poorly. The operating system isn't the core issue here. Bah!
That's our life, the big wheel of shit. - The Fat Man, Blue Tango Salvage
Birmingham, Alabama or Birmingham, England what's the difference they can compare fruit and vegetables as equals, and use numbers equally well. I mean, y'all ain't got no cents, if'n ya don't know Equal is as Equal does.
Who the hell is Equal?
Unaccountable leaders are masters, and unrepresented people are slaves. How do US and EU fare?
While it is fun to lay the blame outside of Linux, the community should really be looking at the product provided and working out how to make it deployable for every one of the 6.2bn folks on the planet if it is going to get the pervasive desktop deployment that some seem to be looking for.
I've been using Linux as my primary (and only at home) OS since 1996, and I code on it for a living now.
I know this argument sounds reasonable, that "the community" should put in "more effort" to make Linux pervasive on the desktop, but it hasn't worked this way, and will not.
"The Community," in the guise of various volunteers and companies, (e.g. Ubuntu) have done a lot already, and this pervasive adoption hasn't happened, and it won't.
People will not just use Linux because they don't want to. They don't care. They are not interested. They like Windows because it comes on their computers by default, "everyone else uses it," they didn't see how much it cost, and it looks pretty, even though underneath it's pretty ropey.
"We" (whoever that is) should stop wasting our valuable time casting pearls before swine. OK, that's maybe a bit harsh, but the work has been done now (shiny user-friendly distros and Microsoft-compatible apps), it is up to them to take it if they want it.
What is far more important to me, and I suspect most of "us", is a healthy and diverse hardware and software ecosystem where everyone can play and compete, through open standards so that no one is left out if they don't want to be, and healthy progress can proceed.
"We" do not need Linux (as only one flabour of *nix) to be pervasive, to replace one monoculture with another. It would be better if everyone ran a better OS (i.e. not Windows) but that isn't going to happen.
"We" should be quietly confident and work to improve "our" software, and when any of the Heathens feel ready to convert, we should offer them our patient and friendly support.
If they don't want to convert, respect their decision, whether is is due to ignorance, laziness, fear, legitimate need or personal taste.
There ends my rant for today.
Stick Men
For the life of me I do not understand how each PC could have cost so much. I was the admin for a public library for 25 years. I have installed many hundreds of library public-use PCs. My most recent full-PC installs, on Windows XP, were running about $800 each for brand new fully capable machines complete with per-seat security software of various kinds (the public is REALLY hard on machines) such as Centurion Guard and Fortres. Just before I left I installed thin clients which were running about $400 per seat (including the servers).
I realize lots of folks here see this as a Linux vs Windows issue. It's really not. The OS in this equation just isn't that much. The issue is total cost of installed base: dollars (pounds) spent divided by number of machines. These were 2500 POUNDS! That's got to be something like $4700 per machine.
Somebody screwed up.
How about a moderation of -1 pedantic.
The councils in the UK are notorious for employing the unemployable. Really. I am surprised it didn't cost twice as much. Let's look at where some of the money probably went:
1. 3 people employed instead of one, so the council can claim it is filling it's job share quota
2. Overpriced training for everyone - who cares if they didn't need the training, there is a training quota to meet
3. Never estimate how dumb some of these staff are - a task that should and would take an average IT person 1 week, most likely took 1 month
4. Meetings, political awareness, and general laziness generally win out
Overall, not a great picture. I experienced first hand the horrors of working for a council IT department, having come from the private sector. I was very glad to return to the private sector as fast as possible.
Remember - how many council IT staff does it take to correctly install Linux? No one knows, it hasn't been done yet!
I don't think you need to script it - by looking at how many comments you've posted so far to Slashdot I'd say you have more than enough time to rough it out and do it by hand.
Web2.0: I love when people Flickr my cuil and digg my boingboing until my google is reddit and I start to yahoo
in your statement:
"Linux is harder to set up, thus it will be more expensive for governments to switch over to linux because it takes more time to set it up. Whereas with Windows, after being in the market for so long and having a lot of people at least be exposed to it once or twice in their lifetime, will not require as much time compared to linux. Therefore it is cheaper to use Windows both in the long term and short term.
Once linux has the same support, features, ease-of-use as Windows has then it has a chance of succeeding and taking a good chunk of market share. But as linux continues to be protrayed as the geekdom of software, it will not be a cheaper solution to Windows."
1 - Linux is harder to set up. No. Since there isn't a registry, or machine locking, it is very easy to create a single bootable image and deploy. Indeed, only a bootloader need be present on the client machines.
2 - Windows has been in the market longer. No, Unix has. And Linux is a reimplementation of Unix.
3 - You conclusion based on these two facts. Neither of the facts holds up.
4 - Once linux has the same support, features, ease-of-use as Windows has then it has a chance of succeeding... "Linux" has arguably BETTER support, more features, and (because of lock-down), better "ease of use" (in a library environment). Linux is in quotes, because we must talk about a particular distribution -- Linux is just a kernel.
5 - linux continues to be protrayed as the geekdom of software. This may be true. But, you will find Linux in your router, in your TV, etc. more often than Windows. And, it has nothing to do with your conclusion (being a "cheaper" solution than Windows). Linux *is* cheaper because there don't have to be license fees (all other things being equal).
Personally, I don't care if Birmingham uses Microsoft Windows, Suse Linux, or IBM OS/2. But, of those, the ONLY choice that may be benefical to me is Suse. The reason? Birmingham may have been in the position to contribute back to the pool (with code, documentation, artwork, or something else). Since they have elected to REMOVE themselves from the pool, they cannot participate; will not receive the benefit, and (worse); I will not receive the reciprocal benefit.
This saddens me -- I would like to see the feedback of a few large public installations.
The one thing YOU should learn is: Linux does not compete with Windows. Asserting that is ludicrous. Suse or Redhat (etc) may compete with Microsoft, though. "Linux" and "OSS" welcomes participation -- without which it couldn't survive. OSS here is deeper than money; we will mourn the lost contribution of Birmingham; its too bad that they aren't going to make the next vital contributions in "library and government usability in OSS". But someone else will take up the gauntlet. That's the beauty of OSS.
Ratboy.
Just another "Cubible(sic) Joe" 2 17 3061
Also, when it comes to support Microsoft kicks ass. Really. If you have paid for support from MS, you really do get it. Seriously. Big time.
True -- but you have only a single source of support. If MS comes through, great; but if they don't, you're sunk.
Think about it: you can have the largest consulting firm -- Accenture, EDS, even IBM -- support your Windows-based development and rollout; but when a bug appears (and one always will), when a security problem is found (and one always will be), when push comes to shove, even the biggest consulting outfits must wait for MS to issue a fix. They can perhaps lean on MS a bit, or hunt for a workaround, but that only takes you so far. If you instead deploy on top of Linux and other Open Source software, your consultants (or in-house staff) can fix that inevitable bug themselves and then submit the patch upstream.
When your support provider has the source code, there's no one that they can blame: the buck stops with them. If they can't handle your needs, you can walk with the source and find someone who can. All support providers are competing on the basis of ability (not secrets): this in turn can lead to real competition, and, therefore, competitive support pricing.
Open Source reminds me of the Nike logo: Just Do It(tm).
...are the companies who come muscling in - "oh you're doing a Linux migration, you'll need our help, we charge £500 a day and will take several days to do anything at all".
It's largely not the staff "on the ground", they are enthusiastic. It's poor choice of distro (SuSE 9.3), requiring EXTREMELY expensive development to attempt to make it work as expected and general dithering.
But it is both a shame and a scandal.
We deployed 120 new desktop pc's which we built ourselves from parts purchased from a trade supplier. Spec was AMD 2800 Semprom, 512MB RAM, 40 GB IDE drive with no CD-ROM as we're trying to encourage the use of USB sticks. Each PC came in at £105 and the build took place in summer 2005.
We installed XP Pro on a volume licence (£35) and then duel booted with Ubuntu Breezy.
Total cost £16800 + the time to build. Without XP these would have been £12600.
Installation of XP consisted of install, update, install all applications and create disk image to be rolled out using Dolly. Install of Ubuntu consisted of popping the disk in, booting, clicking a couple of buttons, upgrading and imaging. The Ubuntu install took much less time as all the apps and drivers were installed at the same time. At the time of building a script was added to run a prompt for a machine name followed by winbinding to the domain.
The image is easy to roll out via our Gigabit LAN using Dolly. Network wide software installs can be done on Linux using a script that checks a directory on the server and after doing an md5 check uses apt to install whatever we want it to.
Given the ease of all this, the Birmingham thing just has to be down to incompetence. Excluding people who know what they're doing from helping is an arrogant act but ultimately one that probably caused the laughably huge bill.
I think that writing to the National Audit Office would be a good move by those Open Source Organisations involved as someone really needs to be held accountable for such a blatant waste of public money. Then again, maybe it was an overtime fiddle by those involved with or, more likely, another public body using Linux to beat Microsoft down on price.
Hmmmmmm..... Deep fried and look like Squirrel.
This is why. Sit there for a moment and consider just what exactly is "Total Cost of Ownership"? In a nutshell it is hardware+software+training+support. You will always have the same formula be it Windows or Linux. So let's break it down...
Windows:
Hardware: $$$$ (As needed to support new OS)
Software: $$$$ (Yearly license fees)
Training: $$$$ (As you add new people. If you are going to save it may be here since Microsoft's dominant position in the market make it feasible that employees already know the system)
Support: $$$$ (Yearly or on an "as needed" basis)
Linux:
Hardware: $$$$ (As needed to support OS but older hardware will work and in fact may work better than newer hardware)
Software: ---- (No cost at all if you stay solely FOSS)
Training: $$$$ (As you add new people. Here is where the cost may go up)
Support: $$$$ (Yearly for the first few years then "As needed" basis. Cost can be high here to start but should go down as you progress)
So you see, right from the start you have a savings in Linux. The only way to save with Microsoft is if they "cheat" and undercut on their license fees. The assumption (often wrong) that is made in the Windows TCO studies is that training costs will be low since "everyone is likely to know Windows" already. It would be interesting to follow a large organization as they deploy Vista. By that I mean follow the TCO from day one. That will give you a huge clue on the true TCO.
B.
This is a sig. This is only a sig. Had this been an actual sig you would have been informed where to tune for more sigs.
TFA I read was talking about desktops not servers, it sounded like they were having a labor action to me.
Apocalypse Cancelled, Sorry, No Ticket Refunds
England must be a strange place. Because I can't understand how a free operating system like Linux can cost 100,000 pounds more on 200 computers than a proprietary system that costs $100 per computer when bought retail.
... correct.
Is Microsoft UK actually paying people to install Windows on their PCs? Is there some magic turning point where $100 (@60 pounds) becomes cheaper than free? Do English librarians actually know the concept of money?
What is it? Training costs? It costs the same to train system administrators to run Linux as it does to run Windows. They are both systems that are equally complex.
Is someone being paid off? Surely not in England! If I bribe an English library system to adopt an expensive proprietary operating system over a free one and the bribee comes right out and says that the free system costs 100,000 pounds more than the proprietary system, do the people in England just shake their heads and agree that it must be so because English librarians are just, so,
I'm just so clueless here. Could someone please explain this?
Birmingham is North of the Watford Gap, so what else can you expect?
Therefore, these folks must be retarded
I'm just guessing you haven't dealt with councils / government departments in the UK then...
These are the same sort of people who told a company that welsh dragon sausages should be made of welsh dragons
Although there is merit in the thrust of your argument, PC hardware does cost more in the UK. You would be very hard pressed to buy a somewhat functional equivalent Linux workstation at the UK equivalent of Fry's for $199 (around £104). For example the cheapest machine from www.morgancomputers.co.uk is about $211 or around $400. Perhaps someone can find cheaper, and bulk buy would get you a discount but those are reasonable indicative prices.
News flash: I.T. projects sometimes run hugely over-budget and get cancelled. Linux is not magic pixie dust that solves this problem.
Oh wait, that's not news.
http://outcampaign.org/
Yeah mind you i have relatives who live down in London who get an equally poor service but look at my Brum council tax bill with sheer envy.
So the whole project costs $16 with Windows and $12 with Linux? Something tells me that these figures need firming up and may, in fact, just have been pulled out of thin air.
"The RIAA has to win every every court case because...."
The RIAA, like Microsoft will never actually go to court. As soon as it looks like they might lose a case they settle quietly, out of court, under NDA.
No sig today...
There must have been a covert ops double agent for Microsoft/SUN on a grassy knoll outside the Birmingham Library with a targetting device to being the sattelite rays down on the heads of those unsuspecting techies who were not wearing their foil caps.
The £100000 that could have been saved by moving to Windows would just about offset the pay of one council employee in an overpayments scandal that's been floating around Birmingham in the past few weeks. See http://icbirmingham.icnetwork.co.uk/birminghampost /news/tm_method=full%26objectid=18018775%26siteid= 50002-name_page.html for some of the details. With appalling mismanagement of council finances in general, it's no wonder they've done this too.
>Yeah mind you i have relatives who live down in London who get an equally poor service but look at my Brum council tax bill with sheer envy.
Yeah okay, good point. But did you get to look at their pay checks? I am sure they were earning several times more than the difference in council tax bills,
My little Linux and tech blog
Even so, you can even get decent laptops for £500 these days, so £2500 for a desktop to sit and use some library catalog software or browse the net is ludicrous. Does TFA explain why the cost was so high? Training? Heated leather seats, mousemats and plasma screens? I think I'd start crying if I read the article.
which is totally what she said
how so? i know of a few types of hardware that won't work in Linux (without a considerable amount of futzing), but i don't know of any that will only work for Linux, unless you're talking about something truly ancient that windows has dropped support for.
upon the advice of my lawyer, i have no sig at this time
It seems to have very little to do with the purchase cost of the PC's. An analogy that might help comparison is if you decided to run a fleet of vehicles and decided to do in house maintenence. You would have significant costs in hiring the staff to work on the vehicles, the garage space for that work to be done, the tools for the staff, and other sundry costs. Now it might make far more sense to outsource that work but the article doesn't throw much light on it. From the end of your comment you don't seem to have read the article. Why not make a habit of learning a little about whatever it is you are commenting on in public, that way you will appear to be more informed.
Well I already knew from other comments that the article was less than informative, but of course you're right. For now I'd better not go read it since I have work to do! And it really does seem too depressing to read. If they're doing everything in house then the maintenance really should not be that high (I'd like to see how much it is for the Windows alternative).
which is totally what she said
Software, as explained previously, is immensily cheaper if you use Linux.
So a Windows only solution starts with a substantial disadvantage when considering the last two items: training and support.
I have lots of experience giving, receiving and organizing training, for run of the mill products like OSes and desktop applications the rates are pretty standard regardless of the product, given the applications mentioned in the article, I don't see anything that would suggest using Linux would increase training costs enormously if at all.
As for support you can get many companies to take in bulk whatever you have, regardless of OS, and if you do it internally then you will not raise the salary of your techs just because now they have to support Linux (they may need training, but that should be budgeted yearly any way, so frankly I don't see how support would be a factor in a differential in price).
So the only way I can think of that a MS solution would be cheaper is if licenses, training and support are all offered in one package and then the price cut below cost in order to keep a foot firmly in the door.
Most companies can't afford to do this, unless somebody is prepared to lose money (or some of part of what they are offering is way overvalued).
Draw your own conclussions, but the only way I see this working is if MS offers a great deal for licenses (which are terribly overvalued, piracy and confussing pricing schemes pretty much probe that) and they do the support and training for cheap. Or they could strong arm the company providing the services to do so, it is not beyond the realm of possibility that the persuade a company to lose money in a high profile company in order to keep a healthy business relationship in other projects mutually profitable.
IANAL but write like a drunk one.
If you are a corporate or institutional buyer this is a non issue, your provider will ensure all your hardware works or will give you the advice so you can go and buy hardware that works.
This nonsense about comparing the problems faced by a hobbyist with the ones faced by an institutional Linux adopter are frankly comparing apples and oranges. No, like comparing oranges and caviar. Or something like that equally nonsensical.
IANAL but write like a drunk one.
People will not just use Linux because they don't want to. They don't care. They are not interested. They like Windows because it comes on their computers by default, "everyone else uses it," they didn't see how much it cost, and it looks pretty, even though underneath it's pretty ropey.
[...]
"We" do not need Linux (as only one flabour of *nix) to be pervasive, to replace one monoculture with another.
I've noticed something about personal computing that is quite disturbing and you've put the finger right on it. A lot has been said about how the IT "monoculture" has made us vulnerable to security compromises, viruses and such because the vast majority of PCs share the same platform with the same weaknesses, but it goes much furhter than that. The software monoculture has slowed progress and innovation very substantially. While hardware computing capability marches forward on Moore's Law, progress on the software side has almost ground to a complete halt. Let's look at a few five-year periods of recent personal computing history shall we?
1975 to 1980: Went from Altair (came standard with a few bytes of RAM, LEDs and switches for input) to Atari (800 came with 48K, had a full keyboard, 128 colour graphic display and multi-channel sound)
1980 to 1985: Went from Vic (and its you like competitors from Atari, Apple, TI, etc) to Mac (16/32 bit computing with desktop model GUI on high-res bitmapped display that you could go and buy at a mall)
At this point Microsoft's dominance starts to develop and the monoculture takes shape:
1985 to 1990: Went from Mac to Windows 3.0 (black and white GUI to...ummm...well colour GUI that was a but awkward to use)
1990 to 1995: Win 3.x to Win95 and NT (colour GUI to a prettier colour GUI and uhhh well I guess TCP/IP stack built in...whatever)
1995 to 2000: Win 95/NT4 to Win 2000 (nothing really notable...stability and scalability improvements)
2000 to 2005: Win XP (fisher-price theme on same old desktop.
What is most disturbing is 2001 to present: NOTHING AT ALL...we've been hobbling along on just service packs and patches. And the "great unwashed" JUST DON'T CARE. The only computer users who have enjoyed any innovation or notable improvements are those who are passionate enough about computers to see them as more than a tool and opt to run Linux or MacOS. Furthermore, because that market is so small any innovation that DOES happen happens slowly dur to either lack of resources or fear of being TOO different from the dominant monoculture so as to repel potential new users.
"We" should be quietly confident and work to improve "our" software, and when any of the Heathens feel ready to convert, we should offer them our patient and friendly support.
While I think you're observations are astute, your solution is far from the way to go. We can't merely improve our manners and accessibility towards newbies and be "quietly confident" becasue we'd never get on anybody's radar. Free software advocates have had "quiet confidence" and continuous improvement for decades now and it never stopped the formation of the Microsoft monooculture. We need to boost our volume considerably in the advocacy of Free alternatives to Microsoft (and even Apple, who have survived and are even thriving a bit from being very visible). We do have a few "loud" advocates, however they are idealogues like ESR and RMS--extremely intelligent and capable people but with eccentric personalities who are fringe thinkers. ESR and RMS are poor at marketing and PR and compromise and negotiation which means that if a person is aware of them at all they could be perceived as crackpots.
We need advocates that can effectively talk to the "common man". We need to shelve talk about the merits of the GPL and freedom and such for a PRODUCT-based approach. The philosophy discussion makes more sense amongst the newly converted, not for the unconverted. There are the Linux television commercials? The only notabl
This is the most ludicrous posting yet. 1,500 desktops at roughly $4000 USD each?
Here's a realistic brakdown of what this would cost in the USA:
1) Hardware: $1000 (let's get something really nice) (Could be as low a $400)
2) Setup the template system. 1 month. Say, $15,000 in manpower
3) Dup disks. $50/PC is the going rate for assembly and SW load.
4) First boot each PC (clearly this could be coded but lets add in this anyways). Have the library staff follow the directions and enter a computer name. I'd venture to say most library systems in the USA already have people who have used Linux or Unix at least. Not like this is the first time they are using a computer or something run by a CPU! Say $50/PC.
5) Shipping to site. $50.
Grand Total $1160 per PC. And I'm sure they used crappy hardware so it would be more like $560/PC. If they are paying any more then fire the idiot in purchasing and the idiot managing the project. If they are paying less, then give the genius in purchasing free movie coupons and the genius manager a bonus.
That's just common sense.
TimJowers, http://www.serviza.com/ Open Source Linux Computers and Training Bundles
Expect Freedom.