Here it comes
2016 is going to be the year of Linux on the Desktop
VMware's released Horizon 6 for Linux, thereby making it possible to deliver virtual Linux desktops. And there the story ends because, as we all know, there's so much Linux on the desktop that it occasionally manages to get a whole 1.91 per cent of the market and desktop virtualisation is generally figured to be about five to …
"After all these years you still need Microsoft Windows to manage VMware."
Really, so why do I do *all* of my management from a Gentoo based laptop or desktop. I'll grant you that Update Manager has to run on a Windows server but that is it for management.
Somehow, I don't feel my job is threatened by you.
I'll also grant you a few other shortcomings in vSphere but doubt you'd even understand what I'm on about.
No, I used the web UI to manage vCenter. I use the VCSA. The VCSA is a vCenter appliance that comes loaded on Linux. I've been using VCSAs since VMware 5.1. They work fine. I don't know why you need Windows at all to run vCenter.
And yes, the VSCAs work just fine for very large deployments.
"where you had to install Windows and dot Net then connect to a service who would install the hyper-visor for you ?"
Don't know what you are on about but I generally PXE boot "naked" hardware and install ESXi from an ISO.
How do you install your hypervisor of choice?
Three shills in a row at this time of night? That's pretty impressive. It's a shame none of them managed a coherent comment.
Clearly they are all experts in the field: we generally don't see a hyphen in hypervisor, nor spell "shut" as "shout". As for what the first AC is on about is anyone's guess - it may have got confused and attempted to go off piste with a predictably awful result.
"1. Didn't it occur to you that your dusk may be someone else's dawn?"
Fair enough. It occurred to me just after I had hit "Submit" and I noticed TP wielding his mighty +1 gob of Pottyness and smacking ACs left, right and centre with his paddle. "Oh" (I thought) "he's several timezones to the left of me."
About those seats, and the people sitting in them. Which of the many variants of desktop Linux had you in mind? I'm interested from the point of view of staff (re)training, having had some experience of that and still occasionally having problems sleeping at night.
Or where these people already using a Linux desktop and the development described in the original article allows you to virtualise those seats?
I sort of expected it would be MATE. Judicious customisation of the panels (basically removing one) can result in something that looks very Windows ish thus reducing training cost/pushback on change. The Alt-F style keyboard shortcuts for LO and all work just the same as in Windows as well.
Slackware choice interesting: you have a plan for pushing out updates to the slackbuilds I imagine. Running RStudio from a slackbuild had me recompiling WebKit almost twice a month. Good luck and write it up when it happens.
Probably not for this application but have a look at Deepin Linux. Very win7 ish custom DE on top of Ubuntu LTS. Has codecs and flash.
This post has been deleted by its author
This post has been deleted by its author
About time we got this out there -
Yes X11 is a network protocol. Have you ever *looked* at the shit signalling and traffic volumes involved? - Perhaps ever considered the bandwidth your core network might need to support in order to have more than 25 clients on the network?
I've managed to muddle along with vmware using viperl and firefox on linux for quite some time now.
once the VCSA is up who cares where the windows went? ( apologies WvB )
"""Dare we mention just how the average user will react to Ubuntu, Red Hat Enterprise Linux or CentOS, the supported Linuxes under Horizon 6?"""
The average user's reaction will be directly proportional to how well can they accomplish the task at hand with the tools provided.
And when the task is producing a Word document or an Excel spreadsheet or a PowerPoint presentation you're back to Windows.
And no, other office suites don't compare and yes, the requirement is for Word, Excel or PowerPoint, and yes the training budget to switch and the following loss of productivity would be more than the Windows/Office license costs so the business will not fork over the money so just forget any impulse to try and change the desktop productivity suite. Quickest way to be shown the door known to IT.
Linux on the enterprise desktop is just a non-starter for the vast majority of use cases and will continue to be so for a long time to come, unless it is forced down the throat of users by politicians (see Munich).
The technical superiority of any specific platform has very little to do with adoption and always has, however much techies may hate that fact.
"And when the task is producing a Word document or an Excel spreadsheet or a PowerPoint presentation you're back to Windows."
No, you're not.
"And no, other office suites don't compare and yes, the requirement is for Word, Excel or PowerPoint"
Even if that were true - and you'd have to be a blinkered, half-witted, lobotomised idiot to honestly believe it - you can use the web versions of Office 365 to get your brand tribalism on. They work just fine under Linux.
"Linux on the enterprise desktop is just a non-starter for the vast majority of use cases and will continue to be so for a long time to come, unless it is forced down the throat of users by politicians (see Munich)."
Wrong. It's all about the applications. Nothing more. And yes, more and more people are willing to work with "equivalent' applications. Smartphones have actually made people more willing to try new things on computers.
"The technical superiority of any specific platform has very little to do with adoption and always has, however much techies may hate that fact."
You're right. Cost is the driving factor. Your personal blinkered predjudices have fuck all to do with anything.
"So how that going running all those Excel BI addins for your company's business applications going under Linux?"
Perfectly well, thanks. Probably because I'm not a complete idiot and I didn't allow my company to be locked in with such obvious idiocy, so I don't have any Excel add-ons, and have completely and utterly forbidden office add-ons with teh exception of ubitmenu.
Mind you, I don't use Oracle for my database either. Also because I'm not an idiot. It's my company. Why the hell would I voluntarily hand control of it over to someone else?
>> Your personal blinkered predjudices have fuck all to do with anything.
And neither do yours.
Despite a decade or more of claiming clear technical and moral superiority (and as many "year of Linux"s), Linux hasnt made any significant inroads to the desktop. Calling everyone else stupid for not agreeing with you and living in your own little dellusional bubble is fine if that works for you.
Its not just about the apps Trevor - there is no compelling reason for the vast majority of users or businesses to move to Linux. Its change for the sake of change (or ideology in your case). Its that simple. Delivering it as a VM doesnt substantially change that equation, and wont bring Linux crashing through the lofty heights of the 2% ceiling it currently aspires to (across all distros).
"Its not just about the apps Trevor - there is no compelling reason for the vast majority of users or businesses to move to Linux. Its change for the sake of change (or ideology in your case). Its that simple. Delivering it as a VM doesnt substantially change that equation, and wont bring Linux crashing through the lofty heights of the 2% ceiling it currently aspires to (across all distros)."
It absolutely is all about the apps. The apps are what matters. They are why we have OSes in the first place. It's that simple.
Change to Linux just for the sake of change would absolutely be stupid. Change to Linux however, brings massive benefits. Not the least of which being cost. Especially now that we can properly do Linux in a Desktop as a Service model.
Microsoft's licensing is comically atrocious. Especially as regards Desktop as a Service. But Linux - aha - no such problems here. I can pay for a handful of licenses for my development instances, then use the same scripts and configs to cascade to free version of the same distro and I'm good. Especially in a "golden master" scenario.
What this means is that I can build really low overhead instances to deliver individual apps, and do so cheaply. And it doesn't matter what OS is on the client. And yes, this will see an increase in use of Linux by end users...even if they don't know that they are using Linux to have their app delivered.
Oddly enough, it probably will mean OSX gains traction on the actual endpoint, not Linux.
It also means that lock-in isn't a concern. Microsoft's lock-in is fierce, and they have no problem whatsoever with squeezing those vice grips they have around our collective testicles. But lo! Now there is a real alternative. Proper remote desktop support was all that was really missing to get Linux off the ground in many cases.
Competent Linux administrators aren't nearly so difficult to find as they were 15 years ago. And with Horizon's delivery mechanism it's really hard for end users to bork anything.
Also, for the record, you're not an idiot for "not agreeing with me". You're an idiot because you can't get your head out of your own brand tribalist ass long enough to assess things objectively.
It has nothing to do with loyalty, or corporate brand, or stupid little IT religious bullshit. It has everything to do with "good enough", cost control and the business risk of suppliers that simply can't be trusted.
There are absolutely going to be lots of use cases where Windows will be required and nothing else will do. There will also be a whole bunch of cases were inertia and brand tribalism prevent people from even considering anything different or now.
But by the same token the tools available for use on Linux absolutely are "good enough" for an ever increasing number of individuals and companies. More to the point, perhaps, projects like enabling Horizon to work with Linux make consuming Linux based apps and/or entire desktops as risk free as possible. It allows for smoother transitions and it can dramatically assist with cost control, even if that cost control is simply "another stick with which to beat Microsoft at the negotiating table".
More options is a good thing.
You don't seem to every be able to get that. More fool you.
>> Change to Linux however, brings massive benefits. Not the least of which being cost.
Like I said, claiming "massive benefits" and other hand-waving is only convincing to the small minority that are already convinced. For everyone else, the "massive benefits" are either not massive enough to bother, or the switching cost is too high. Either way they arent going to bother - physical / virtual or any other way you want to deliver it. No-one likes paying a fee for something, but the license fee is a pretty small component of the overall ownership/opportunity cost. No business is going to think you are a hero by saving them a pound for every 10 you make them spend on your alternative.
"Like I said, claiming "massive benefits" and other hand-waving is only convincing to the small minority that are already convinced."
Except that's not true. The majority of endpoints int he world are, in fact, running a Linux kernel in the form of Android. The majority of embedded systems are a Unix derivative. VXworks, Linux, WindRiver, or so forth. It is only the traditional desktop market that has been slow to change, but the change is occurring, no matter how adamantly you stamp your feet.
More and more businesses are convinced that the benefits of alternative desktop OSes are many. What's really interesting though is that I don't need Linux, or OSX, or anything else to be "the majority" operating system in a given niche in order to see value from it. I'm capable of doing objective and independent assessment of customer needs on a case by case basis and picking the best tool for the job.
Unlike you, I don't have an obsession with systems administration via what's most fashionable.
"No-one likes paying a fee for something, but the license fee is a pretty small component of the overall ownership/opportunity cost."
Few people mind paying a fee. The problem is that we're not talking about one-off fees anymore. Microsoft wants into you for subscriptions. Now that's fine in the USA where the economy is more or less stable and it makes perfect sense to bet your business on the fact that you'll always be able to pay your subscription fees to, well...fucking everything.
But the rest of the world isn't the USA. A significant chunk of the world is "boom and bust" economies with very few large enterprises. In these economies, dominated by SMBs, ownership of assets matters. Including digital ones. Front loading costs during a boom ensures you can survive the bust. It's a lesson hard learned, but one that billions of individuals in over a billion of the world's SMBs understand quite well. Microsoft is actively hostile to this model.
What's worse: you don't save money with Microsoft's "subscribe to everything, forever" model. Especially if you exist in an economic climate where there willinevitably be points where you need to sweat assets to survive.
What you don't ever seem able to grasp is that embracing Linux, OSX, BSD, Unix et. al is about more than trying to dodge some small fee. It's about having the flexibility to grow your business on your own terms...and to make it through rough patches without firing people. (Or by firing fewer people.)
You also make rediuclous false assertions that somehow switching to Linux - in whole or in part - will cost you more than simply submitting to Microsoft's demands. This isn't true. It hasn't been true for some time.
Larger organizations with more legacy cruft - Excel macros and plug-ins and so forth - may well have a high hurdle to jump. And I can only imagine that the process of moving a Fortune 500 or a government to Linux would be painful and expensive.
But IT isn't homogenous. Just because it's going to be an expensive, painful process in one area doesn't mean it will be in another. And, niche by niche, SMB by SMB, Linux absolutely will make inroads.
Horizon - and many of the open source alternatives that I know are emerging over the next 18 months - making Linux DaaS viable over WAN is a much bigger step towards this than you want to admit. It means that Linux desktops can be delivered by industry-specific MSPs, CSPs and VARs to their SMB customers. It offers a whole new model for application delivery and even the ability to provision entire desktops at prices that are far - far - more affordable than a Microsoft-based solution.
So we're not going to see a massive turnover tomorrow. But piece by piece we will see uptake. And the best part is that uptake will probably be innovative. Free of licensing restrictions, we may well see new models and new approaches emerge that simply aren't realistically feasible under a Microsoft regime, and they may exist alongside traditional Windows desktops.
And hell, why not? Windows on the physical endpoint because that's what available at PC-world, but push out the individual applications via browser SaaS, Linux DaaS or RDS as required. Reduce the RDS as much as humanly possible in order to cut costs and simplify delivery and eventually you're free of Microsoft.
Sure, maybe you have Windows endpoints, but if your application delivery is all over the wire they don't have to be Windows. They can be whatever the end user wants...they don't actually have any apps on the damned thing anyways.
Get run over by malware? Restore the endpoint back to it's clean slate and run your 5 days and 500 reboots worth of updates and you're good.
Just because things have always been one way does not mean they will continue to be. I fully agree that the biggest corporations and governments will be slow to change, simply because they are traditionally conservative organisations that take forever to embrace new technologies.
But SMBs won't be so reluctant. The midmarket absolutely will jump on this. More to the point, I have already started to see a lot of excited chatter about it and meetings are scheduled to plan POCs.
You can rail against the dying of the light all you want, but Microsoft does not own the future. Gods Christ man, even Microsoft themselves know that. That's why they're cranking out their applications for non-Microsoft operating systems now!
We get it. You love Microsoft. You will always love Microsoft. We've always been at war with Eastasia and you absolutely will have them do it to Julia. Do you want a golf clap or something?
The rest of the world is moving on. We will make use of new technologies and IT will diversify. The hegemony is broken. Cope.
"More and more businesses are convinced that the benefits of alternative desktop OSes are many"
Why are pretty much zero of them migrating then? Microsoft still have a ~90% desktop market share.
This will only be of interest to organisations wanting to remove legacy Linux systems from the office and stick them in a data centre...
"Why are pretty much zero of them migrating then? Microsoft still have a ~90% desktop market share.
This will only be of interest to organisations wanting to remove legacy Linux systems from the office and stick them in a data centre..."
Because they're deploying their applications to non desktops. Browsers/aaS, smartphones and tablets are seeing exceptional growth, for example. There's more Android (which is Linux, BTW) out there than Windows. Apple's OSX market share isn't huge, but their iOS market share is impressive.
Also, there's no huge rush to move yet. Windows 7 was a great operating system. Windows 8 as ass, and Windows 10 is mostly ass, but we have 5 years before we all have to pick a path.
Android is moving onto the desktop. ChromeOS is more and more capable and seeing some pretty significant consumer uptake. OSX is growing significantly in the US and Linux is slowly, but surely, getting it's act together regarding desktops. (Now that X11 is finally on the way out.)
Over the next 5 years the options will change pretty dramatically. The availability of options through things like Horizon's support for Linux merely broadens the possible post Windows-7 migration paths.
Microsoft won't be maintaining it's 90% desktop market share past 2020. It won't be growing it's non-desktop market share by much either. Microsoft knows this. That's why they are developing applications for multiple platforms now.
The only person who can't accept the inevitability of change is you...and I'm not sure why you care so much that other people have access to - and choose to take advantage of - better options for endpoints. If you care anything for systems administration you should be happy that there is increased choice, not fighting it tooth and nail.
Which begs the question: who are you, and why are you rabidly opposed to anything except Microsoft's absolute dominance of all things? What's your stake in it? Most importantly, why can't you see the inevitable when it's heading for you like a freight train?
They used to say that about migrating from Windows Server to Linux and now Microsoft is quite rightly shitting its pants....
Linux owns the server world and in making inroads on the desktop (and if it is good enough for these guys....):-
Quite honestly, if you still think that Windows is a great desktop then bully for you but use it and the rest of us will enjoy something MUNCH better!
"Change to Linux however, brings massive benefits. Not the least of which being cost."
We already know from Munich that the cost of migration and building the desktop environment ("Limux in this case) vastly exceeds any short and medium term savings. And the end result sucked so much that they are looking to reverse course after over a decade of desperately trying to make it work...
"We already know from Munich that the cost of migration and building the desktop environment ("Limux in this case) vastly exceeds any short and medium term savings. And the end result sucked so much that they are looking to reverse course after over a decade of desperately trying to make it work..."
Except none of that is true. Munich has saved quite a lot of money by going to Linux. Moreover, the 10 year review of their infrastructure is part of a regularly scheduled and per-planed process that has nothing at all to do with any dissatisfaction that may or may not exist regarding their Linux rollout.
If you can't be honest and objective about publicly available and easily debunked facts, why should anyone believe anything you have to say about Linux or Microsoft?
I am intrigued by the idea of a virty Linux desktop, configured to look like Win 7, running web versions of Office - Outlook web access, etc.
For example, Outlook 2013 on Windows looks so much like Outlook 365 ("equally bad", perhaps) that the value in the native Windows desktop app is less and less reason to stick with the underlying OS.
Any case studies out there?
Have been going on for years.
IBM moved huge swathes of staff onto linux desktops over the last four years.
- Lotus notes is java based, it made no difference to that, in fact given better memory management in the OS, the app behaves better.
I'm aware of a (middle sized) financial trading house that tossed MS desktops to the side of the road in the last 8 months, reduced their MS server population by 85%, and are now working off a combination of linux workstations and Mac laptops. (not sure I can name them, I was a subC) and a huge fleet of virtual linux and Solaris servers.
To my knowledge a middle sized police force in Europe are engaged in switching to Andriod/Linux endpoints and minimizing their MS server fleet, moving to Virtualized windows apps only where there is no alternative.
Guess what - other than PowerPoint, Project and Visio - LibreOffice/OpenOffice/(there's another fork of this somewhere) has covered the office suite territory. (Yes it has a presentation document engine, but I'll admit *that* has a learning curve that is a bit longer) - Project can be swapped out for one of three options - again with a learning curve, but they do work reasonably well.
Visio I *personally* have not found a *decent* replacement for - but that may again be a learning curve and *I* may be being stubborn.
My current group of co-workers don't typically have wacky crap added onto the Office Suite components they use, since our analysis tools are in the TeraData/Informatica/MicroStrategy territory, and with the addition of Hadoop and the herd of OSS tools that brings, I've been asked to provide linux desktop access for them. The Desktop team have been trying for years to get it adopted, and with the work that RedHat has done, the virtual desktop tools we have already, its starting to look like we may have a valid reason to cut almost $8MCA out of the IT budget, *and still* be able to present everything the end users need, whether it be MS, MAC or linux.
However, in the long run, there will be folks who are *petrified* of moving off of MS's platform, either because they've bought the FUD that "nothing works in linux" or simply because they've no faith in their ability to change their tools.
As much as Trevor is apparently foaming at the mouth (to some of the commentards) and as much as that AC is apparently ignoring the real world, I'm one of those rational types that goes for logic and reason. Trevor is quite right in many of the things he says, and I'm inclined overall to agree with him, and the AC *does* have a couple of points, however they may well be based on the FUD that is *still* being spewed about "linux sucks as a desktop".
Right tool for the job at hand is my take, and what we have here from vmware is quite simply something that may well be the right tool for *many* folks that want to reduce costs, and improve their working environments.
"Visio I *personally* have not found a *decent* replacement for - but that may again be a learning curve and *I* may be being stubborn."
No, there's no good substitute. There are a few SaaS apps that are browser delivered that are getting close, however, they are likely still a year or two away from "good enough".
"linux sucks as a desktop"
Quite right. But then, so does Windows 8, and Windows 10 isn't really all that great either. The real issue is "are the apps i need available?" Right now, the answer for most SMBs exploring Linux is - surprisingly - "yes". Enterprises will not get the same answer to that question.
Fast forward a few years. X11 will have been replaced by Weyland/Weston in production and available as a first class display system in enterprise Linux environments. This will have FreeRDP server baked right in to the display layer, offering full remote access as good as anything Microsoft delivers.
The Gnome team have - after years of acrimony - found their own asses, KDE shows faint hopes of maybe one day being able to find theirs, XFCE has seen a surge of development and both Mate and Cinnamon have exploded in uptake and development.
Red Hat is pouring muchos money into making sure that if your application runs on Windows it will run on Red Hat. You'll be able to make your Linux behave however you want. Application developers are increasingly embracing OSX and/or Android, both of which make the jump to Linux trivial.
Perhaps most critically, Microsoft doesn't really seem all that interested in making a desktop environments that's actually good anymore, they're only interested in making one that is passable enough that enterprises might consider migrating.
So just as Windows 7 is about to turn into a pumpkin Linux looks set to have it's shit together. Will Microsoft?
Today, Microsoft is the dominant player, but they have nowhere to go but down. Linux's adoption is so low any upward gain at all is a victory.
So, as you say, "right tool for the job at hand". And if you've two tools to hand, both capable, pick the one that costs less and you can trust more. It's up to you, the customer, to decide which that is for any given workload.
I have worked with Lotus Notes for a couple of years and liked it overall. It is very different from Outlook and Thunderbird but it has its strong points.
Being written in Java is more of an advantage than a negative. You run the exact same software on all different platforms, no annoying differences like Office on Windows and Mac. Moreover, JVMs are pretty fast nowadays and Lotus Notes is typically started once a day so Java's cold-start issue is not very visible. I use several mid/big Java apps daily (like FreeMind, yEd and they work OK.