Few topics in the IT industry are more contentious than the prospect of putting Linux on the corporate desktop. Opinions range from the religious view at one end, promoting a fundamentalist belief in open source as the saviour of mankind, to the reaction of corporate conservatives at the other, dismissing Linux as irrelevant to …
A pretty damning conclusion
> The view was that the relatively light and predictable requirements of [professional and transaction] users such as these ...
So according to this article, desktop Linux is only suitable for intelligent / experienced people (would that be because it's too hard for the rest) and those who don't use it much anyway.That seems to be to be pretty much who uses desktop Linux today, anyway - so there are no surprises there. In fact if that really is the case, you've got to wonder what the future for desktop Linux is (although we've already beaten this topic to death).
However, earlier in the article, it talks about internet delivery of applications and services. In which case all you need is a browser. There is already an initiative to burn a Linux kernel into a PC's BIOS, so the box effectively boots Linux off it's own hardware - rather than needing a disk with an O/S installed. Maybe this is the future: to turn Linux into a firmware product and PCs into embedded devices with no local storage (or just an SSD with a database on it, to preserve state information across reboots).
Will this be the death of operating systems?
The old desktop Linux chestnut
There are good reasons not to move to Linux on the desktop, particularly lack of certain applications. Photoshop/Illustrator/Quark/other high end graphics? Nope. (Don't say Gimp, just don't say it). For the Mac retards, there are plenty of people who use PS on Windows. That doesn't mean that PS couldn't be built for Linux, but that it is simply not.
But there's another thing which might be skewing the desktop, and that is the number of Linux distros out there. Which one could or should be chosen? One thing you can say about XP is its longevity, 8 years old now. With service packs, of course, but basically the same desktop for the last x years. "Linux" gets a new KDE or Gnome or XFCE or whatever when a new distro is released. Sometimes change is not a good thing. Sometimes people want familiarity. I got a copy of Deneba Canvas 7 from a cover CD a couple of years ago. It was released in 2000. It runs fine on my Windows XP box. And anyone anywhere can do the same. That's a good thing.
Unless Linux can run Windows programs users might as well use dumb terminals.
Office full of Macs
I visit my client's office once a month, which they share with a design company. I'm not sure why but there's something quite eerie about walking into a big office full of Macs.
As for me, I'm a true penguin head. ;)
Windows Programs on Linux
A real lot of windows programs run directly on Linux, using wine, some distributions will even open a .exe file in the wine environment for you.
The Only real problem is .NET programs, which are NOT supported under linux. Ironically .NET was touted as being cross platform, etc etc etc by Microsoft.
many games will install and run happily under wine.
I've even run quite a few applications that were bugged on windows, and these either ran more stable under wine, or were easier to kill, without having to reboot.
So it is a fallacy that windows programs cannot run on linux, a lot of them can.
It is just that very few people tries.
You can even run some programs that no longer will run on windows, because those programs are too old.
The Wine environment can be tuned per application to supply windows 95, 2000, XP etc support.
the majority of end users
aren't that important in a large corporate environment, they don't particularly use the operating systems anyway. Most have locked down access and desktop icons for everything they use, controlled centrally, so the users actual experience is pretty independent of the underlying OS. A few power users might be an issue, but i don't think the differences would even be notable by the majority.
What would be the main problem, are the vast number of customised applications out there, browser apps that would need to be re-worked for something other than IE6 (not a good example as this should be done anyway), custom built c/c++applications, both dos and windows apps, hardware drivers, networking components (our applications use some custom stuff, it was pretty fancy in the early 90's anyway), not to mention the current adoption of .net technologies for creating anything and everything quickly.
The sheer amount of stuff that would need to be re-written just to end up with the exact same functionality makes migrating a bit of a non-starter for most large companies. Not to mention the amount of time it would take, guarantees you would need the two desktops to co-exist for several months at least.
We did this with the migration to xp, a department at a time for several months, turn up, switch off nt pc, replace with Xp machine and give the user new login details for the different domain, It took years of planning and preperation and 99% of the apps ported directly with no changes.
Where's the apps eh?
It's the apps, schtoopid. Or the lack thereof. Where are the equivalents of (say) the Sage and Dynamics and Pegasus and whatever on Linux? Nowhere. This is what is stopping a serious corporate desktop presence.
narrowing niche ?
Interestingly Windows is being squeezed out at the high end, including supercomputing, servers, and use by developers, network professionals and computing scientists who need more control over and understanding of their systems. It's also being squeezed out at the entry level, including in netbooks, embedded systems and individuals who just want their computer to work without excessive cost or wanting to know how it works. Who can afford to throw computers out every 4 years and sacrifice much of the performance to an anti-virus program which they don't need on Linux due to the more reliable and trustworthy software integration, packaging and validation that comes from open source ? Then there are those who have power requirements more suited to the Atom chip than a quad core CPU consuming over 100 watts, which is likely to be a better fit to most people's electricity bills, particularly in developing countries where most computing users will be located within a few years.
So the question isn't so much as to whether Linux is "ready for the desktop". It's been ready for most desktop purposes for years. The question is how quickly the middle segment of the desktop market currently dominated by rich westerners lacking in computing education will be squeezed out by the growing segment of the market more concerned with computing capability/budget ratios and open source code level as opposed to closed binary compatibility.
Problem is people don't know how to implement it correctly - the general consensus at least around here is:
- x terminal / thin client (ltsp or similar)
- 'chubby' clients that can run local apps seamlessly with the remote desktop - using ltsp as an example you can have firefox / flash run as a local application and video playback / audio works fine.
- dual monitors - with the money you save give users dual monitors with the thin clients - it really is a cheap motivational and productivity gain - users feel they are getting something instead of losing windows.
- HP and Termtek are examples of thin clients that will support the above.
Openoffice is good enough for most tasks, crossover office works ok if you really need it. Getting users to use a wiki instead of docs is even better. Most core business apps are web based or moving in that direction anyway.
I work for a european company which is starting to make a staged migration away from MS products. I guess eventually it will totally embrace Open Source, but not today.
The first thing was to replace MS Office with OpenOffice. This is now underway and seems to have been handled sensitively and efficiently. Very few people who have had the 4 hour introduction to OpenOffice have viewed it negatively. As they will not have the option to say "No!", they have just got on with it.
It's all about the approach taken by HR and MIS organisations. If they get it right, it's actually quite easy.
Mad UniVersal Warfare for Maggots ..... Simple Sane Salvation for Smarter Souls
"Participants in the study identified other groups,such as Windows power users, highly mobile professionals and creative workers, as being much more questionable in terms of targeting with Linux. This is because the number, type and mix of applications upon which such groups are dependent often translates to either high migration costs or an unacceptable degree of compromise in terms of end user capability or experience." .... Dale Vile, Freeform Dynamics
Given the presumably more intelligent nature of " Windows power users, highly mobile professionals and creative workers" , I would have though they were the Perfect Target Market for Linux.
And an even More Perfect Target Market would be Virtualisation and Cloud Manufacturers, which don't Plan to Play their Games for Free on or Deny Proprietary System for their Modus Operandi and Vivendi has Set Up ITs Bases in Rich Internet Applications, which are in Reality, Get Rich XXXXStreamly Quickly Virtually.
And that Creates a Problem for Earthbound Administrations, which would seek Proxy Pseudo Power to Tax and Control a CyberSpace Environment/Virtual Operating System System, with any Penny Pinching Parasitic Practices ....... for the New Revised System is IntelAIgently Designed to either Devour or Fully Expose Corrupting Systems which Feed off Virgin Endeavours and Perverse Practices.
There are Interesting ZerodDays ahead, El Reg, with Colossal Changes afoot........ and which is a much more constructive paradigm for journalism to sharing before the fact, than wasting so much time and effort in reporting on events, after the fact, for who cares what has been whenever the only thing that one can change is what is to be.
Spin a Yarn, Start a Tale, Lay a Course and Follow ITs Trail....... And IT is not Difficult at All to Do for All Those who would know what they are Doing, although for All Others would it be Simply QuITe CompleXXXX by CyberIntelAIgent Design ...... in Order to Create and Maintain a Crushing and Overwhelming Natural Advantage over Stupid Unwary Ignorance and Blind Studied Arrogance........ and therefore, a Most Definite Improvement Heralding Real Tangible Progress from Virtual Intangible Fields.
And a ProgramMING which has been Offered to Team UK, but who would appear to Playing for Foreign Opposition rather than PathFinding AI Leading Position ......... and that is a Treachery and Calumny well suited to House of Fools/Nest of ASPs and would thus then extraordinarily render them as a Future Irrelevance to be Outed and Bedevilled with Evidence and Questions.
Battle for the desktop
1. Linux doesn't need to replace Windows on the desktop, just provide an alternative. Every time MS make a decision based on a Linux alternative. e.g. extending XP support for netbooks, or leading the Vista marketing campaign on improved security. Linux has done it's job.
2. Cost of migration. I know a professional typist who uses MS Word all day but doesn't want to change to 2007 because it is different. If sticking with MS means there is still a cost of migration why not try something different.
Advantages from an IT point of view
In my old position we had about 400 users. Most of which simply did not care about the operating system. As long as they could get their email, log onto the corporate server and retrieve files and browse the web they were happy. I think this probably applies to most businesses. As long as linux can provide those things then it is a viable option.
The advantage on our side of the fence is the ability to lock down the OS. It is a relatively simple thing to do within linux which would, IMHO, reduce the nu/mber of callouts and downtime. If you simply restrict users capabilities with the machine so much that they really can't do anything to the OS themselves, throw in good a AV and firewall then you ave a pretty robust machine.
The question is, are the IT staff willing to make the change? In my experience the answer is no. Within most IT teams I would take a bet that at least 50% of the staff don't work with linux and wouldn't know how to troubleshoot problems when they arise. Once that hurdle is overcome convincing the end users to switch will be easy as you, if you wanted to, could make the new system look near identical to the old. It'd be an opportunity to train staff too. Just in the basics but stop them from being scared of the computers they work with.
How does this make sense?
Why would organizations add additional complexity by identifying pockets of potential Linux desktop usage, when they struggle to maintain their existing infrastructure today? This approach has worked quite well in the server space. But unlike the numbers of servers deployed, the sheer magnitude of desktops/notebooks deployed would never lead a sane person to thinking they should try to make a fraction of those users different.
Driving down costs by using free software is an idealistic point of view and one that I admire and adhere to wherever and whenever possible. But the real facts are that labor costs generally far exceed any expenses related to the purchase and maintenance of the software involved. IT's the people, stupid.
The apps are irrellivant sort of.
as most users don't use lots of apps. They use a word processor, email client, calendar/scheduler and a browser and some a spreadsheet. Many don't even need full apps, just need to read documents/sheets.
Compatibility is an issue though. Despite advances OpenOffice docs are not 100% compatible with MS Office and there isn't another app that really does all that Outlook does in a single familiar linked manner.
One point being made though is that the desktop is becoming more irrellivant as apps move to browser tools and then there are always tools like Citrix that can provide desktops within desktops.
You havin' a laugh?
"consider the value of creating a mixed estate of Windows, Linux and possibly even Mac"
you havin' a laugh ... or do you just like spending all month delivering patches and updates?
Most apps don't run on linux (without some sort of emulation / wine affiar) ... so for the sake of £90 .. I'll spend the cash rather than have all the wasted man hours with hundreds of patches, user training sessions, new standard builds, admin security courses, test environments (as the mac is so good at updates not breaking things) etc. etc. ... ... ...
What about the GIMP? I'm just kidding, I have to say I agree that Photoshop is one of those apps that makes it hard to migrate away from Mac / Windows. But I don't have Photoshop, and neither do any of the other several hundred desktop PCs where I work.
I'd say for most Windows shops, Office is more likely to be the sticking point. Migration to Linux is easier done piecemeal; migrating to OOo first (quite painful in itself if the company has a lot of templates and macros to port over) is probably a good starting point which saves a bit of cash on licensing and paves the way for a future OS migration. Switching away from Outlook / Exchange is another mildly painful but very worthwhile switch.
Companies looking at FOSS seem to often find one or two systems that won't easily port over and give up on the whole thing. I'd say the only sane thing to do is port what you can, when you can; there's plenty of time but it's never too early to take the first few steps. Start swapping in the occasional portable application in place of an unportable one now, and you can incrementally increase the company's freedom to choose further down the line without worrying about switching a single system's OS just yet. Who knows, maybe in ten years' time your company will still have two or three Windows machines, just for those legacy systems that never get ported; you can still save tens or hundreds of thousands on the other few hundred or few thousand systems.
The article mentions the benefits of running a multiple OS environment on the desktop - apart from, of course, the massive increase in costs associated with building, supporting and maintaining 3 standard operating environments.
Most enterprises I've worked in buy their machines from one manufacturer to a standardised spec in order to minimise the work in having a SOE. All machines that start life in a standard way makes desktop support easier.
I can pretty much guarantee that the desktop support people will not be experts in 3 OSes.
Also, I can't think of many corporates that would run 3rd party apps on WINE. Geez, some software isn't even supported on a VM let alone this. Bloomberg, Reuters 3000 etc are windows apps and WINE isn't an option. Other industries than finance will have their own must-have apps that won't move. Excel to Open Orifice? Please. The users that can't switch are the ones that finance the damn thing and need all the VBA shit that goes with it.
@Sooty - hallelujah someone that understands that just because it can be theoretically done doesn't mean corporates will.
Hell, even the Vista fuck-up hasn't moved them onto something else - just made them consider things. Doesn't bode too well does it? Why would you dick about re-writing all your software for a year or 2 just to get the same functionality on a different OS? Meanwhile your competitors are increasing the functionality of their bespoke apps.
There's good reason Linux made it into the enterprise on servers - most apps are loosely connected to the client and users aren't allowed onto them. Desktops have users pissing about with them and the standard operating environment rules.
You want desktops? Hunt down Government - local, state, federal etc. They have an obligation to be open and standardised as they exist off of tax dollars (although corporates have joined that party of late).
@Michael Fremlins and other comments
How many people in a corporate environment actually use Photoshop, Illustrator or Quark. Also, the fact that you mention Canvas 7 as an app you use shows that you are not a normal corporate user.
Most people in places I have worked use the following apps.
Word processor, Spreadsheet, Email, Calendar, Web, Internet/Intranet access and Some way to access file on the network
All of this can be done by any desktop OS today. It's just the inertia of Windows and Office and the fact that people don't like change.
If given the choice of keeping things as they are or changing people will keep things as they are. Case in point is the comment from @Hugh_Pym about the typist.
Changing will get resistance.
I expect from a support and management standpoint there are a lot of Windows Admins out there, but far fewer Linux admins, not to mention desktop support staff. People go to your comfort zone. Back to inertia. Windows admins and desktop support teams will recommend sticking with Windows because that’s where they are most comfortable.
The shift in what we use is happening slowly. Firefox is growing, OpenOffice is starting to get its fans, Googledocs and Gmail are popular. The need for it to be a Microsoft centric environment is decreasing and so the potential to use something different is becoming more appropriate.
Finally, I heard a great comment.
How many of us are using the same operating system and applications as we did when we joined the great unwashed workplace.
I used Windows 3.1 with Office 3.0 Looks a bit different to Windows XP with Office 2003. From an admin standpoint it is completely different as well.
So, in my opinion people can change and get use to new apps and ways of working, the problem is the move from Windows to Linux on the desktop is sold as a big thing when for most people I think it really isn’t.
I bought a car recently from one of those massive car supermarkets in North West London. I noticed that the sales person was running a SuSe desktop, and when I commented on this, he had no idea what I was talking about.
He just sold cars, and used the web-based apps he was provided with.
The finance guy was using XP. Again, he was unaware that they had a mixed environment withint the same office.
The talk of PhotoShop, which is ALWAYS bought up by the wintards, is just so utterly irrelevant to the average desktop user (average? heck, I've never seen PS used in an office in 30years of computing).
Can I just say that 15 years ago, when I first started using alternatives to Windows, most people said "Linux is only for geeks, noone else will use it!"
Now, we have moved to "Linux is for geeks and people who dont use their computer much".
Sounds like an improvement to me...
Whats happening is that slowly, Linux is gaining more users. As it gains more users, more apps are developed for it, user interfaces are improved, people become more familiar with it, and it gains more users.
Take netbooks & nettops. People are taking them up and finding Linux is fine, not the scary beast they heard of. This means they try Linux on their desktop, probably using a Live CD. They like the fact that it runs quicker than Windows, and try out the alternative apps. They like it, so install it.
Now we have a user who would most likely be quite happy with Linux on their corporate box.
I think, as was mentioned above, the real sticking point will be the IT Pros, many of whom consider themselves experts because they took a 1-week course. They have only ever dealt with Windows, and will have to learn from scratch to move the company over to Linux. So they will dig their heals in, and spread scare stories to the management about how hard Linux is to use.
As for applications, well, it is going to be a slow process IMHO, but as more ppl move to Linux at home (and the number of ppl asking me to put Linux on their machine instead of Windows is growing), the more apps will be written for Linux. This will be prompted by many, many phone calls to their help desks saying "I bought your software and it wont work on my PC", to the point where they see that it is worthwhile to port it to Linux.
All this will likely take another 10 years or so, but I seriously beleive that MS's days as dominant power in the desktop world are numbered...
I completely agree about Office and Exchange. I know that some people have spent many painful days or weeks writing macros and other yarbles for their Excel spreadsheets, and moving away from that would be very painful.
Exchange is a little bit different. At my old company we had Solaris machines as the primary MXs, forwarding on to Exchange machines for the mail boxes for most people. Some preferred to have IMAP on UNIX instead. One of the things that the Exchange users liked to use was the shared calendar. But as all the staff sat in one big open plan office it was a bit unnecessary, and I suspect for many small companies it is equally unnecessary.
Most of our platform ran on UNIX, and the people who ran the platform ran UNIX. Most of the "office staff" ran Windows. It would be almost impossible to pull them away from that. Web developers - use PS, Dreamweaver, Illustrator, whatever else. Finance - use Sage. Management - er. Trying to make the techies use Windows would have made us unhappy. It's exactly the same for trying to make Windows users use Linux.
It is a bit frustrating, but I have no axe to grind.
Photoshop is a flagship for an application that will never come to Linux--Adobe has made that abundantly clear. Another potential sticking point is Microsoft Office. Now, simple documents, spreadsheets, etc. can be ported over to OpenOffice with little trouble (I've done it for many documents in a relatively short period of time). The trouble is when you get to complex files filled with VBA code or custom screen templates. This is simple fact, but MSO and OOo use two entirely different programming systems (the former is based on Visual Basic, the latter on ECMAScript IIRC), so porting the scripts is no easy task.
Another potential problem for migration is critical custom applications that have been in existence for a long time, cost a lot to acquire, and more than likely come from a company no longer in business. They're too expensive to replace, too critical to live without, and too touchy to move. So many companies have no choice but to "live with it".
I don't understand why people get so worked up about this...
Ah, this topic is like a dog with an old bone. He loves chewing it over and will never let it go.
I don't think anyone is trying to say that migration to Linux is a magic silver bullet that will solve all the worlds ills.
As said in previous comments, for those who solely use a machine for web, email, word processing, chat etc some of the newer distros of Linux are a very viable option. From my experience this makes up a large portion of people who work daily with a computer, whether that's in IT or not. The actual proportion of people who work in an industry where they specifically require a package that only runs on windows has got to be relatively small by comparison? And yes, without question, for these people Linux is definitely not the answer. Not yet at any rate.
To those who question the economics of supporting multiple OSs, there has to be a point at which the cost of support in different platforms is less than the cost of licenses for a given size of organisation. But this again is about targeting the audience and deploying where appropriate.
A fact of life is that some people are always going to be resistant to change, whether it's to a newer version of the same thing, or a different product with similar functionality.
Yes I work for a large IT company as a developer. And yes I use a Linux desktop. But it's a company developed client that they are happy for people to use instead of the Windows version. For me it's appropriate. When I bought a new computer for my Mum (no mum jokes please) did I get her a Linux machine? I thought about it, but in the end decided that the best option for her at the time was Windows.
It all about using right tool for the job.
A "normal" user
Whether Linux is usable for a "normal" user depends largely on what you describe as a normal user.
Does a normal user use Quark, Photoshop, Sage, Premiere and a bespoke accounting programme written in 1703 on MS Parchment 1.2? No, in fact, hell no.
A normal user uses a desktop, browser, word processor and a spreadsheet. My wife (Holy ****, a *nix geek who's capable of getting laid... whatever next ;)) is a normal user. I overwrote her annoying XP load with Ubuntu and after restraining her initial wave of anger, she quickly (in approx 10 mins) learnt that a Gnome desktop works *exactly* like Windows for the stuff she wanted to do. Connect to WIFI? Check. Open a browser? Check. Look at Videos on youtube? Check. Update spreadsheets with home finance information? Check.
Normal users don't recompile Kernels. They don't install drivers, on any platform. They call us, Reg readers. And if you don't get called it's because you're rubbish ;)
Personally, I don't care what OS you use, but it does make me laugh when people spend £100 or so on an OS as bad as Vista. There are bad Linuxes too, but at least you don't have to pay for them.
Most users, unlike geeks, don't like *any* change. They'll complain about Vista after using XP*, or MS Office 2007 after using 2003. This is due to the fact that:
a) many non-geeks don't enjoy learning new things, esp. to do with bloody computers. It's hard work, it takes time, etc.
b) A lot of people learn to use a computer by rote, as apposed to learning the concepts, thus even relatively minor UI changes will upset them, and affect productivity.
Things that can help are:
a) Coincide the change to linux with new hardware. People expect a new computer to look different to the old one, and a faster machine will take the edge off having to relearn a few things.
b) Running Firefox and OpenOffice on windows before changing the OS.
c) New employees get Linux only from the beginning, where possible.
A lot of users only require web browser, email, basic word processing and access to the file server, which Linux provides adaquately. Also given the trend to SaaS and web applications,
the underlying OS becomes less important.
The hardest part of Linux (as with any OS) is the install and initial config, which in corperate environments isn't any issue.
Turning your PC into a thin client
Linux seems to be sufficient for people who just want to use email and run a browser. Way to turn an expensive PC into a thin client. A programmer I know recently started working for a company that uses Linux. He had two reactions being there a few months. First that using Linux reminded him a lot of Windows 3.1. He was shocked by what a crumby OS it is, given the relentless hype of its fanboys. And second he was shocked by how expensive "free" software was for the company, who must employ a small army of programmers to constantly support and fix their software (not so much Linux, but the crapware that orbits around it, like Ruby on Rails).
@all the "apps unavailable" talk
So... what will happen when Windows 7 comes along?
You know, all these legacy apps won't work anymore, except in VM, or XP-mode or whatever in a premium version of Windows. Or so I've heard around here. Is it true? How well will it work? What will be the cost of that again? Just keep XP. What, you aren't allowed?
Re: @all the "apps unavailable" talk
... and I guarantee you that if push comes to shove, they'll take the bits that work and recode the rest for the next Windows platform, not the latest linux distro.
Paris, because she thinks that way, too.
I see lots of people talking about the mythical 'normal' or 'basic' user, I've been working in IT since about 1994 and I have still to come across someone who just uses basic office apps and email. Typically there would be a minimum of a groupware application in the companies that I have worked for, I am also struggling to think of a single department where there wasn't a key bespoke application or ten.
Also, I am only just posting from home again because my Linux laptop (acer aspire one) lost all abillity to controll its wireless last night. I'd just got to the point where I though it was just crap about wireless randomly failing, I assumed that sort of thing had been fixed. Now having spent all of last night trying to work out the problem, many reboots, hacking around in the command line, and general swearing and consulting of the internet, I'm not so sure. I got home tonight and it just works again. This sort of thing is not ready for business just yet. Geeks, yes, business, no.
Definitely not ready for power users, especially developers!
Wine is OK'ish, but it's not windows, so applications like the awesome Directory Opus don't run properly. IMHO, Directory Opus is the best directory management tool, on any OS; I have used it since it was on the Commodore Amiga, and would be lost without it, so one more reason I don't rate Linux for desktop use.
I also found it annoying, that Linux has so many ways to manage devices, that it can be a pain to get settings tweaked, get them to stick, and get the right settings, this is especially critical for input and display devices. I find it quite annoying that distributions like Ubuntu refuse to provide the same flexibility of auto-device configuration as demonstrated by the 'impure' Debian-based distribution KNOPPIX. The killer for me is the apparent requirement to reinstall the OS for a new major version of a distribution (e.g. UBUNTU, this is stupid. With Window, I just upgrade the current OS in-place, with windows updates, a windows service pack, or a new Windows version.
"Take netbooks & nettops. People are taking them up and finding Linux is fine, not the scary beast they heard of. This means they try Linux on their desktop, probably using a Live CD. They like the fact that it runs quicker than Windows, and try out the alternative apps. They like it, so install it."
You've made two big leaps of logic there.
1) People actually do find Linux fine. In my experience they don't- I know at least two people that bought original eeePCs and then took them back and traded them for Windows ones. They didn't like Linux.
2) This means users will want to try it on their home PCs. Netbooks are used for browsing and little else, home PCs can be used for all sorts of things. The average user never even reformats and reinstalls Windows in their PC's lifetime, it's quite a leap to suggest that they'll try out a Linux live CD.
Good article desktop Linux
RE: Definitely not ready for power users, especially developers!
"The killer for me is the apparent requirement to reinstall the OS for a new major version of a distribution (e.g. UBUNTU, this is stupid. With Window, I just upgrade the current OS in-place, with windows updates, a windows service pack, or a new Windows version."
Yesterday, I fired up a machine with Xubuntu (8.10) on it that I hadn't used in awhile. I ran the Update Manager, which said "A new version (9.04) is available" and provided an Upgrade button. I pressed that button, it downloaded some stuff, rebooted, and was in 9.04.
Therefore, I really, really have got to wonder WTF you are talking about.
DOS apps are not an issue. Indeed, one could argue that DOS apps run better under Linux/FreeDOS than under Win7 command shell. The only caveat being that if the DOS app is tied to hardware (a scenario that is more common than you'd think) then you are screwed, but then that would also apply to Windows anyway so . . . .
It's over, not for you, not now, but it's over. Linux and FOSS won. They won for the same reason academia was able to effectively outpace private industry in the race for "computing" the human genome, because people will always ultimately opt for the most open, flexible means of communication (given the chance). Because faster nervous systems eat slower, nervous systems, and, far more importantly, because the desktop is about gossip, facebook and youtube, and Linux and FOSS ultimately offer the most flexible, extensible means of gossip. Your $XXXX app you use to convince your clients you're worth a few hundred bucks an hour is a skewered to the thin end of the tail and "GNU" Linux offers the mean and the mode all they need. Windows is dead, like a charging elephant shot through the brain it'll just take a while to hit the ground dead. In 5 yrs you'll be exercising 20/20 hind vision and saying you saw it all coming.
PH? You'll figure it out.
Re: Good article desktop Linux
Crap article actually. Linux is no more a "square wheel" than Windows is the "perfect wheel" but nice try though.
To use Linux
some brain power is required.
As apparently about 14% of users in Britain, can't even turn their computer on, the chance of getting them to use Linux is very remote.
The icon is guaranteed to turn me on .
Users shouldn't control the desktop
The last time I worried about the desktop was determining if the users really needed access to the command line in VMS. I searched all of the logon files and found that the most common app was the happy birthday Fritz. The users didn't even know enough to change the date and name in the file to their own. I switched my wife to a linux desktop back in 1999, my daughter has an OLPC. When I need a windows application I run it in vmware or connect to another machine via vnc, or citrix. One of the biggest trouble I've is Seen is users writing applications in Access or Excel then expecting IT to take over support when they leave. 0
@Definitely not ready for power users, especially developers!
You can upgrade Ubuntu in place. Trust me, I've done it twice so far with no ill results.
"Photoshop is a flagship for an application that will never come to Linux--Adobe has made that abundantly clear"
Maybe but Photoshop runs very well on OSX. In-fact most serious CS3/CS4 users I know use it on OSX. Again this is a niche product - most corporate desktops users don't need it (or the Gimp for that matter). Adobe have apparently talked about Creative Suite as SaaS - can't see it happening but Picasa runs on Linux / Mac / Windows and version 3 is getting up to Photoshop Elements level.
"Another potential sticking point is Microsoft Office. Now, simple documents, spreadsheets, etc. can be ported over to OpenOffice with little trouble (I've done it for many documents in a relatively short period of time). The trouble is when you get to complex files filled with VBA code or custom screen templates"
I think I last saw a 'complex office document filled with VBA code' in the 90's. They are such a security risk (Excel 2007 switches off macro code be default, Outlook occasionally kills them) that they generally don't work well when passing documents around. The best version of Office is probably the Mac version - which has no VBA support so even transferring such files between MS office versions is problematic. I know the latest versions of OpenOffice has increased support for macro's but seems pointless; the world has moved on to sharing content via the web rather than complex documents with active code. Remember this sort of thing is going to hit Microsoft Office Live users as well.
"Another potential problem for migration is critical custom applications that have been in existence for a long time, cost a lot to acquire, and more than likely come from a company no longer in business. They're too expensive to replace, too critical to live without, and too touchy to move"
For an IT manager, these sorts of applications cause grief from a security point of view and would be a problem even if upgrading to Win7 (I have also seen older VB/.NET problems killed by service packs). They need to be isolated and run on a VM, (and earmarked to be replaced with better flexible and portable solutions in the future to avoid the same issue). MS are supplying a VM with Win7 (at least the expensive versions) exactly for this reason; which makes you wonder why you could just run the VM over VMWare ESX or on a Linux box?
Basically use the right tool for the job is the answer. In our (small) office,we have a mix of OSX, Vista (x32 + x64), XP and Linux. Opensource vs Windows is becoming less important all the time. I have Office 2007 including Outlook, but mail is google apps not exchange. Our Windows 2008 server (required for a specialist app) runs on an VM.. which is running on a Debian server. Some people use OfficeOffice and Thunderbird, others MS office. The decision are made on cost benefits not on fear of anything that is not MS or rabid open-source advocates.
I suspect at least some comments in threads like this are brought & paid for astroturfing or fan-boys but reality (at least for us) is that we are living in a heterogeneous world.
.NET and Ubuntu upgrades
Good article! I enjoyed reading it...
"Ironically .NET was touted as being cross platform, etc etc etc by Microsoft."
Not ironic at all. This was I think intentional -- they wanted to provide LIP SERVICE to portability while ensuring everything but Windows is a second-class citizen. I think Microsoft was (is?) quite terrified of a popular truly cross-platform environment, after all if significant software ran under an environment like this, people could feel free to shift away from windows and know their apps will work. So, 1) They really aimed .NET to harm Java, since it was well down this path,and did take the wind out of it's sails a bit. 1.5) With Silverlight, same for flash 2) Make sure .NET and Silverlight works best on Windows.. I've run a few things under Mono, but Silverlight for Mac & Moonlight are junk, they're just enough so they can claim cross-platform support while really being a wedge trying to push people onto Windows "Oh hey this Silverlight-based video site works on Windows!!"
"The killer for me is the apparent requirement to reinstall the OS for a new major version of a distribution (e.g. UBUNTU, this is stupid. With Window, I just upgrade the current OS in-place, with windows updates, a windows service pack, or a new Windows version."
Don't know if you're serious or not, but Ubuntu specifically has facilities for going from one version to the next in-place.. 1) Open Update Manager. 2) If you are running an LTS (Long Term Support) version of Ubuntu like 8.04, it'll only offer to upgrade to another LTS release by default (which won't be out until April 2010). So go in preferences and set it to show all Ubuntu releases if you want 9.04. . 3) Text shows up near the top of the update manager similar to "Upgrade -- click to upgrade to Ubuntu 9.04". Click that, click "OK " to tell it you're sure, it downloads the updates and installs them. If you have to interrupt the downloads they do continue where they let off. 4) Reboot. Done! .........OR...... plug in an install CD, ubuntu seems to offer to upgrade to the version on the cd. I expect it's similar except it saves some downloading by getting most files off the CD.
Can't any of you lot read?
The sub-heading says: "Selective targeting is key"
Linux (BSD, UN*X in general, also see OSX) will work quite nicely for probably 95% of all office workers if, and I stress the *IF*, the folks in charge of the roll-out know what they are doing, and understand the exact needs of that particular (group of) office(s).
THAT said, the other ~5% of users need a specific software product that only runs on a specific OS. For that small percentage, let 'em use the OS required.
If your IT staff can't handle more than one OS, I submit that you need new IT staff.
For people complaining about end users having issues with software installation and/or driver/hardware configuration ... THAT IS THE JOB OF *IT*, you idiots! It is OUR job to ensure that the hardware & software is compatible, and does the job for the end-user!
To date, with over 100 companies and schools, large (75,000+ seats) & small (a couple seats), moved to mostly Linux with a few peripheral Macs & Windows machines, as well as larger hardware, I have yet to have a single company or school ask me to switch them back.
It can be done in the real world, and it is being done in the real world. Successfully. And the numbers are on the increase, although perhaps not exponential (yet?). Get used to the idea, because it isn't going to slow down any time soon.
Did you just make that up? Really did you? Does Steve Ballmer pay you to blow his trumpet?
@jake & Alastair
"If your IT staff can't handle more than one OS, I submit that you need new IT staff."
BRAVO! I wholeheartedly agree!
"You've made two big leaps of logic there.
1) People actually do find Linux fine. In my experience they don't- I know at least two people that bought original eeePCs and then took them back and traded them for Windows ones. They didn't like Linux.
2) This means users will want to try it on their home PCs. Netbooks are used for browsing and little else, home PCs can be used for all sorts of things. The average user never even reformats and reinstalls Windows in their PC's lifetime, it's quite a leap to suggest that they'll try out a Linux live CD"
WRT 1) I also know of people who did the same, but then they were idiots expecting it to be a fully functioning laptop. They were just as dissappointed with the XP versions. Those I know who bought netbooks for what netbooks are designed for (i.e. a bit of surfing, email, a few card games, IM...) were quite happy with Linux. Even those who bought the Maplin Minibook (Also CCL minibook, made by CnM) were happy for what they paid. The problem is the people who buy an eee and expect it to run the latest 3D games etc, and then whinge that they bought the wrong tool for the job.
As for 2, I know plenty of people who just use their expensive desktop PCs just for browsing & email. And the reason I walked this line of logic is that people have asked me to install linux for them, after trying a live cd that either a colleague gave them or they found given away for free. I discount the ones who actually downloaded and burned it themselves, as they actually know about computers to some degree. The low-end, non-computer-literate ppl liked the siomplicity, the speed, and the fact it was free (most of them were running pirate versions of XP before that).
What I wrote was actually based on my own experience, and that of people I know.
Colour me wintard.
> Does a normal user use Quark, Photoshop, Sage, Premiere and a bespoke accounting
> programme written in 1703 on MS Parchment 1.2? No, in fact, hell no.
I visit offices full of them on a continual basis. While not being a default Redmond-hater (therefore being a 'wintard' presumably) I couldn't give a hoop *what* OS they use. In fact, I'd rather they used Linux since it would give me something new and different to play with.
> Directory Opus is the best directory management tool, on any OS; I have used it since it was
> on the Commodore Amiga, and would be lost without it, so one more reason I don't rate Linux > for desktop use.
The lack of a Linux version of a fairly obscure directory manager is hardly justification for damning Linux as a desktop OS.
@Dr. Mouse again
"I also know of people who did the same, but then they were idiots expecting it to be a fully functioning laptop. They were just as dissappointed with the XP versions."
The people I know weren't. They could browse, send e-mails... but also load up iTunes, change the music on their iPods, run Skype without issues (when my friend got an eeePC the Linux version was piss-awful, might not be the case now), install MSN/Yahoo (not a crappy Linux client with no webcam support) basically do everything they want to do with their laptop. Their needs were not demanding, but Linux could not provide for them.
Plus, there's the issue of doing stuff differently. Yes, Ubuntu has great repositories, but what if you're used to double-clicking on an EXE? It's worth 20 quid to a lot of people just to not HAVE to learn new ways of doing things they already do.
We handled marketing strategy for SuSE before the Novell acquisition, and in part promoted their desktop Linux.
Waaaaay back then we were telling enterprises that rolling-out desktop Linux was a multi-phase process. Start with IT (sans Windows support). Once IT had formed Linux support internally, then roll out to transactional users. Only then could an enterprise consider non-power user info workers.
This dovetailed well into what our research into CIO/CTO Linux strategy was (see our white paper "What CxOs think about Linux".
"For the Mac retards, there are plenty of people who use PS on Windows" I doubt any "mac retards" have ever denied it. I wish people like you would just disappear, you nasty little shit. What is with this Photoshop/"high end graphics" obsession? Is it some badge of honour? Have you heard of Pixie or Aqsis? Blender? InkScape? All fine pieces of software if you understand the basic principles of computer graphics. Photoshop/Quark et c. just aren't the killer apps you make them out to be *BECAUSE* they are available on both platforms. How many of these people are using PhotoShop as a proverbial sledgehammer to the peanut that is the image editing that most of you hobbyists do. I have been using Photoshop *professionally* (on a "retarded mac) for 18 years. In those 18 years I have seen it become a bloated and expensive POS. The talent ISN'T IN THE FUCKING SOFTWARE. DO YOU UNDERSTAND THIS?! The GIMP is a perfectly capable image editor - always has been. So quit with "Don't say Gimp, just don't say it" lines and get a fucking clue you talentless twerp. For the record, Pixelmator on the Mac is my weapon of choice at the moment. It's like PhotoShop release 5.5 - IMHO the best release of PhotoShop.
The Linux kernel on the desktop has been ready for a couple of years - infact Ubuntu 9.04 is extremely easy to install AND extremely easy to use and configure. What is worrying to me is the amount of so called IT "professionals" (I'm sorry, if you only know 1 OS - you ain't no pro in my book - a blagger more like) that fear change. These so-called experts can't even cope with an interface change in the only OS that they *can* (barely) use - let alone their office suite! *How* can we expect them to use a different OS?! No, the only reason that Linux "isn't ready for the desktop" (tm) is because of the amount of pretenders that *really don't know what they are doing* working in IT today. How many of the 'features' that these idiots cite, do the average user use? I'm surprised AutoCAD hasn't been mentioned. There are people out ther that will prefer one OS to another FOR GENUINE REASONS - I accept that, and unless the reasoning is stupid ("I just don't like Jobs/Ballmer/Stallman/Torvalds et c.) I respect their opinions and move on.
Insert obligatory "I do run Windows when I need AutoCAD" here ;-)