eWeek’s Debra Donston has penned down five ways the end user desktop in enterprises will look different in five years. While some of her ideas and predictions make a low of sense, there are a few things which are slightly debatable. Mostly, the reliance on virtualisation and web applications.In the article, Donston places a lot of emphasis on virtualisation, basically stating that within five years, virtualisation will play an ever important role. While most certainly its role will increase, I’m not sure as to what extent. Seperating the client software environment from the client hardware environment, as eWeek calls it, is not exactly a new idea; the thin client and the networked computer have been around for a long time, but never lived up to their promises. While performance of virtualisation products has indeed improved greatly over the past few years, I’m simply not certain if enough has changed between the thin client of yesteryear and the virtualisation products of today.
The second important development she mentions is the importance of web applications. While web applications may have their merits, there will always be the inevitable performance penalty – something you really don’t want when you are running Photoshop. In addition, the intertubes can get clogged, and break – how would you like your company to come to a grinding halt whenever the internet has performance issues?
Lastly, there’s the issue of application availability. When a traditional application vendor goes down, that’s bad – but you still have the company’s software on your computers, running safe and sound, giving you the time to find a replacement. If a web application company goes down – your applications are gone. Again, I’m not so sure if companies will like that idea.
Anyway, feel free to make up your own mind, and apart form the above, she makes some good points.
Fair enough, but a lot of the stress in the article on virtualization and mobility depends on something not yet proven – the cloud computing model. Corporations and governments will be busy sticking their oars in to the cloud and generally mucking it around for profit and control, so there is no guarantee the cloud model will be a great success, or not at least within five years which is only a short time away.
Generalizing always boils up the emotions.
Debra makes a lot of points, we will see in the corporate desktop in the future as these are all trends we already encounter.
Yes, there are applications, that absolutely make no sense as a browser-based application. Regarding efficiency, stability, user experience. Photoshop is a very good example with thousands more.
That’s why these five trends are not mutually exclusive.
What the article lacks, on the other hand, is surprises.
And I bet, the next big thing on the corporate desktop is not among those five trends.
I hope that summary is correct, since I gave up reading before hitting the second page. (Two full page advertisments and I only read a couple of paragraphs. WTF!)
I don’t think that many people are going to be running Photoshop as a web application anytime soon, but I would be surprised if a majority of corporate offices even have Photoshop, nevermind regularly use it. That stuff would either be in a different part of the business, or contracted out. But the questions you have to ask yourself are: what are the benefits of web applications and which types of applications is it viable to use?
I would imagine that most databases would be viable to transition to a web application model. Web front ends for database queries have been the norm for well over a decade. It is finally getting to the point where you can build a decent interface to manage databases too. The chief benefit is that the software is centrally managed to an upgrade or plain old patches are easier to roll out.
What about word processing. A few years ago, we were dealing with crude text entry fields or heavy Java based applications. We no longer live in that world and certain types of word processing can be managed right from your web browser. You achieve the same roll-out benefits as above, but you also get something more important: more flexibility in license management. Not only do you have central license management, but you can reduce the number of licenses that you need to purchase. A lot of businesses need a word processor at every desktop, but they don’t need to run these word processors concurrently. If you only need 50 concurrent licenses for 250 users, you could be saving a lot of cash.
There are two fallacies people make when discussing web applications. One is assuming that it will run on the publishers servers. While this is probably going to be offered as an option to smaller businesses, chances are larger businesses are going to have the option to run their own servers. A lot of them would need that option because of privacy laws, or simply to deal with laws that vary between jurisdictions. Besides, this sort of dependency already exists between businesses. Companies like IBM make a lot of money off of it.
The second fallacy is assuming that everything is going to become a web application overnight. That’s idiotic. It’s going to take time, just like it took time to get where we are today. You probably don’t think of OS News as a web application, but it is. In the olden days, you would have needed special software to read usenet (or fidonet) news. Email is another popular web application. If you told me that most people are going to end up using webmail even 5 or 6 years ago, I would have laughed my head off. The prime examples were slow and limited. Give it another 5 years, and noone is going to think of webmail as a web app because people will think that email was always that way. News databases and encyclopedias and dictionaries are another example. It used to be that you had to buy that stuff as applications that shipped on CD-ROM. Now we just go online. Ditto for web publishing (what do you think CMSes are for). Eventually almost everything will be a web application. It will be cheaper and more convenient for everyone involved.
I almost don’t read articles that have titles ala “X ways to…”, “X reasons why…”, “X top…”, “X best…”, “X most…” etc. any more. These are rarely truly informative and valuable..
I almost don’t read articles that have titles ala “X ways to…”, “X reasons why…”, “X top…”, “X best…”, “X most…” etc. any more. These are rarely truly informative and valuable..
Same here, I hate them as well.
Edited 2008-06-01 10:42 UTC
Really, TFA is just pointing out what’s already in the process of happening. It’s just that few IT managers see the whole picture yet.
Companies like Oracle are going after visualization as a means to streamline support and improve quality. You can download VMware images ready to go for certain products right now. That means you worry about getting VMware on your hardware and they only have to have drivers to support the VMware model..reducing the number of potential failures.
The big trend TFA missed is security. Security means keeping bad people out, information in, and having backups and recovery options. Visualization helps those goals because you have one “box” that does one task and it’s easily replaceable… matching hardware and drivers don’t matter when you have hardware failure and you need to run now. Separation of systems also means less cross-product bugs. Your Oracle install doesn’t have to have any other unnecessary components, but the hardware can still have a browser or whatever else needed for usefulness.
Mobile users are becoming the norm. It’s not so much “users” as “bosses” that want to be out golfing at 5pm and not just sitting around waiting on reports. There’s also the fact that they can be notified by the plant equipment itself when trouble happens so they can be available to make decisions… very important.
To make that possible data will be sliced in to neat mobile-size pages. As that is built over time, the need for heavy desktop versions to get exactly the same data goes away. System monitors are now web page reports. Reports are now web pages, error messages are sent to mobiles… etc. Applications run next to the hardware/processes they are important to and feed the database.
The future is giving users access to their data and letting them share their “owned” data on pages they make. Sort of how online we have blogs, company managers will have their “process” pages and allow others to share them.
In the end, data becomes most important. We’ve severed ties to desktop apps by making it available to web and mobile users. Not all at once but in stages. We’ve severed ties to heavy computing hardware and software by using replaceable virtual machines and SQL links to web servers. The new backup tape becomes a “company in a box” with the right technologies in places the data can be reconfigured however the business owners need it to be.
Much how flexible windows hardware replaced over priced, proprietary mini-computers, OSS-based business processes will replace the complex licensing schemes as the established companies are still trying to tell you how to USE the software and the ways are about to drastically change again. Sure, not all apps will be web based, Some are too complex, but MOST will. Companies like Google sell “search-in-a-box” as well as some of their other products so a business can rent a system from Google and put it in their server room.. and still access it from the web… expect that to continue, selling software how the CUSTOMER wants. I don’t like the meme of “thin clients” because that’s not really going to happen in a centralized manner like they promised. The trend I see is small “disposable” machines. Powerful enough to run disconnected, but data synced well enough that techs replace one with software trouble to the user and re-flash them.. and most/all of the users data will already be captured to the network. An eeePC with a 19″ monitor and standard keyboard would fill 80% of office users needs, and be mobile for work on the go. Companies are starting to understand that they buy software to fit THEIR business process, they don’t change business process to fit the software vendor. Computing will be fun when that starts to catch on.
”
Security means keeping bad people out, information in, and having backups and recovery options.
”
Not a really important point but your level of security (it is a verb rather than a nown) includes keeping bad people out, keeping information in (or with authorized users). It also includes keeping good people from causing damages by accident (or being able too cause them purposfully of course) plus all aspects related too disaster recovery. Protecting data from bad guys is the most public and exciting concern of info security though so it get’s the glory.
The problem is that the definition of “bad person” is defined in all the wrong ways. All “security” I’ve seen so far just brings fear, not peace. When thoughtcrime becomes reality, it’ll be introduced all in the name of “security”, just like all the CCTV cams in the UK that haven’t actually done anything beneficial but instil useless fear in everybody.
I try to distinquish between security practices and street theater productions.
Airport “security” = street theater, it’s not slowing down professionals with bad intentions
There is definately a lot of street theater going on and much of it justifies polititions side projects. We’ve recently had six lovely video cameras installed in the nearest subway station; three pointing each way. It’s a perfect face recognition trap for scanning the crowds and each of the three cameras watches the other set for damages.
Awfull lot of “feel secure” these days but I’m not see much of that “security” it’s supposed to provide. If you see something, say something (hopefully, someone will listen).
I think she was focusing on the business desktop. She certainly mentions it repeatedly, so I’m not sure you’re thinking in the same context.
The vast majority of business desktops (eg tech/support rep computers) are probably already running web apps almost completely. Obviously, only a subset of personal desktop tasks make sense in a browser or as ‘cloud computing’ (why do I hate that term?) apps.
Also, I don’t think this is merely a technicality: You’re already using virtualization on your desktop. If OSAlert runs JavaScript, anyone browsing the site is using layered security and providing a ‘virtual’ environment for an application to execute in.
It’s also clear, based on browser vulnerabilities (and the fact that people run other applications) that this is not enough. Intensive apps like Photoshop and games will probably never be virtualized, but IM clients, browsers/Office/PDF viewers would benefit. I’d also love to separate checking/credit card/savings accounts further from my regular browsing to lower the risk of them being compromised.
For me, it’s the obscurity of the term.
The Net has a place, I can “look” in a way and see that my computer is here and on the other side of this divider is The Intertubes; I’m good with that. My data is here but I can travel over that to get to my destination.
The Cloud takes my data out of my possession and places it out nebulously on that Intertubes thingy. Where is my data? Who knows, it’s in the cloud. Who do I know my data is secure; meh, cloud says so.
Even as a security type, I’m not comfortable with having my data out in The Cloud. I’m willing to listen if I’m missing something in this concept but under my current understanding; not comfortable at all.
In a business setting, all the data belongs to the business and a centralized software/data management system that provides both through any authenticating terminal makes perfect sense.
many people don’t want viruses trashing their computer but they have no problem giving out their entire identities to the whole world.
Ahhhhh, light, airy, bubbly and refreshing – nothing like a nice content-free ad delivery vehicle to cleanse your web-surfing pallet for the afternoon.
Following is my opinion, with or without reasons as I please
I do not like, do not trust and do not want “Rich Web Applications” and neither should anyone else. I want a physical installer medium, I want my own personal machine, and I want control over my software and hardware (which I paid money for, which in turn I worked hard for). “Not So Rich Web Applications” for trivial stuff, like Google Maps is nice and usefull. (You could do without, but it’s helpful)
Windows, Linux, MacOS? I don’t care. I am a tinkerer, I like fiddling with Linux and it will always be around, thanks to Open Source. However, I do believe diversity and competition brings innovation, so I would like to see more Linux or even more MacOS desktops. Maybe more Linux than Apple since Apple’s just as ruthless and corporate evil as Microsoft just has less marketshare.
Virtualisation is complete crap for desktops. There is not one decent solution. Most often performance is bad, making for sluggish and unresponsive experiences. Then there is hardware support; until there is full, complete, 100% equivalent 3D accelaration in virtual machines I use virtualized desktops for quick testing of linux distros and that’s it.
Mobile desktops on mobile devices? I’d rather stick to pencil, paper, ruler, etc. than use tiny keys to type on tiny displays like on cellphones. The iPhone? The iPhone is like a person offering you candy and then beating you up for taking it. Make it open, make it developer friendly! Right from the start! Without catches, fine-print, EULAS, NDAS and similar crap. Let’s see what Google will make of Android…
Laptops? Make ’em run at least 10 hours on battery with full system load. Don’t make them cheap by using cheap components, make them cheap by excluding windows, the usual crapware, and less bling-bling. I want black, letter sized bricks, good keyboard layouts and 4:3 non-glare displays!
Security? Oh, please… go lock yourself in a closet, I’ll throw away the key for you. Security and Web Apps is like lips on a chicken, NOT THERE! Certificates for applications that you can use to verify who/what to trust? Yeah, like with SSL and HTTPS? “Damn, popup. <click>” They can be faked and modified. And who certifies the certificate creators? How about you take those certificates and policies, build nice walls around yourself and see how far you can move your arms, while I race circles around you laughing.
Fortunately, freedom is I can jump off the wagon anytime I want to, and if this corporate control thing and the other stupid/lazy computer users thing get much more out of hand in the future… maybe I will.
1. It Will Be Virtualized
Huh? You mean I need to delete my VMs and wait 5 years?!
2. It Will Be Mobile
NO! I have to wait to get my notebook back too?
3. It Will Be Secured in Depth^aEUR”And By Users
Crap, must remove encryption, firewall, and TCP Wrappers, cron for reinstall in 5 years.
4. It Will Be Running Windows^aEUR”Maybe
Dang it! Foiled AGAIN! (goes and finds Windows restore CD)
5. It Won’t Be a Desktop at All
Darn it I already gave up my notebook, now I don’t even get a desktop?!
hah!
Edited 2008-05-31 00:35 UTC
5 years is not a terribly long time in the history of operating systems. 5 years ago we had Windows XP, Mac OS X, good Linux distributions, Windows Server 2003… Does that picture sound familiar?
10 years ago things were a lot different than now.
Thus we should try and imagine how things will be in 10 years time. Surely in terms of hardware we could have a tremendous computing power available to everybody.
The one OS feature I’d like to see is decent voice recognition. In 10 years time we should be able to get rid of mouse and keyboard if we wanted to.