Kiss your Fat Client Goodbye?

Story: A Future of Instant-On Cloud ComputingTotal Replies: 13
Author Content

Jun 15, 2008
9:51 PM EDT
I've said it before and will say it again: before everyone gets all excited about this new Web 2.0 technology, think about the implications it will have for Linux users.

If Web 2.0 gains momentum, and every application transitions to the web or has a web analog, the need for Fat Clients dwindles considerably. If you lose your Fat Client then you lose all or most of the things a Linux user holds dear in their day-to-day computing.

You'd lose the ability to chose your Linux distribution. There would be no need for one as everything is handled over the web and is brought to your thin client appliance/net device.

You would no longer be able to learn Perl or programming or system administration. Or if you could why would you need to? Everything is handled by an outside supplier via the cloud.

I know others in this forum believe that those days are well into the future but I believe that maybe it'll happen sooner than we'd like to think. And if the desktop computer as we know it (you know, the one with CD and DVD trays sporting its own OS) is soon to be replaced by the appliance and UMPC and 'net tops,' it kind of takes all of the fun out of Linux.

I know I am determined to enjoy things as they are until any transition takes place, but it isn't a pretty prospect when you know that your current daily routine's days are numbered.

What say ye all?

Jun 16, 2008
1:59 AM EDT
I say that, for every push behind "cloud computing" and "thin clients," there is an equal or greater push for "smart clients." The trust required is simply too great. Microsoft and Google (the two biggest contenders) have demonstrated all too well that their intentions are at cross-purposes with those of their users.

Jun 16, 2008
2:12 AM EDT
What say ye all?

"The Cloud" doesn't exist yet. Today it is more like the real weather than anything else. We don't have a seamless network, with global coverage and economically feasible flat fee rates. We have monolithic Telco's extracting their pound of flesh and their services combine to a scattered, non-global and unreliable network.

MID's are nice devices if you can get online somewhere, but if the connection to "The Cloud" is severed, they are totally reliant on off line apps and data. I don't think that the current market is flexible and agile enough to make "The Cloud" happen anytime soon. "The Cloud" needs ultra low pricing, seamless and vast coverage. Switching between providers on the move needs to be possible, transparent and billing needs to be non-prohibitive.

Mobile connections are premium services in Europe. Don't know the situation in oversea regions, but being online everywhere in The Netherlands is expensive. Slow Internet via KPN (Royal Dutch Telco) costs 35.00 EURO a month ($ 54.34) and coverage is around 90%. Furthermore you can't share that connection with other users OR devices, you can't be online 24/7 and you can't download more than ten times the amount of the "average user", whatever that means. They don't tell you what an average user downloads per month. Don't think about crossing the border, because then KPN charges extra.

I don't want to pay $ 652.08 a year just to be fleetingly online in The Netherlands and still have to have a "mid-range" cable connection too (at $ 558 a year). As long as there aren't any economically viable connections at reasonable prices, meaning 24/7, 99% coverage, usable at home and on the go, no download limit and no surcharges for roaming in the European space, we won't see much "Cloud Computing" in The Netherlands.

Fat clients won't be going anywhere anytime soon, simply because they are independent from whatever online services. If you have your applications and your data locally on the machine, a network outage doesn't leave you dead in the water. Also, having sensitive information off line is a good idea. You can't know for sure that your online provider is always ethical and won't abuse their position (or be forced to), exposing your data to third parties. What is accepted today, could be your downfall tomorrow.

Jun 16, 2008
10:21 AM EDT
While I think that cloud computing's presence will only grow in the next few years, I agree that we just don't have the kind of connectivity needed to make it ubiquitous. For the U.S., anyway, a major initiative would need to be undertaken to make extremely high-speed networking (and NOT what most of us have, which is slow DSL) available to all at a reasonable price. And I don't see that coming together in the next five or even 10 years.

For the next 5 years or more, expect things like the Amazon cloud to be a part of the desktop experience, but not a replacement for it. For me, having Google's offline Gears component is what makes the online Google Docs finally work for me. They had to add the ability to use it while NOT being connected in order to "complete" the service.

But we will be doing more and more work over the network. Much of it will be in traditional Web browsers, but I have a good feeling that the opportunity exists for all sorts of applications to interact with the cloud, and for the cloud itself to be a resource for the desktop and not replace it entirely for some time to come, if ever.

That said, having the choice of running completely in the cloud is just that -- a choice. One I find quite attractive, by the way.

Jun 16, 2008
10:24 AM EDT

You raise a good point about service providers charging for usage.

One thing cloud computing must have is low cost, unlimited bandwidth connections. Here in the U.S. the big cable providers appear to be moving in the opposite direction imposing limits to the amount of bandwidth one uses per month. And at a time when Apple and others are setting up on-line services which require a lot of bandwidth for movies and music.

My hope is that any "cloud" computing model will live peacefully with our current fat client desktops so we could have the best of both worlds.

Jun 16, 2008
10:30 AM EDT

I agree with your sentiment. I am not against any convenience that the cloud computing model would offer us; I am against the notion of cloud computing as a replacement for our current desktops and laptops.

I like the idea of choosing whether to compute in the cloud or off-line. It would offer the best of both worlds without changing what we have come to enjoy about computing: the ability to administer our own system.

And, as I've pointed out before, those companies who sing the loudest about the benefits of cloud computing seem to be the same companies that are poised to make the most profit from it as service providers.

Jun 16, 2008
11:44 AM EDT
An rough analogy of "cloud" computing versus the fat client is that of VoIP vs. standard PSTN. From, proponents of VoIP claim the following benefits over PSTN
Quoting:* The ability to transmit more than one telephone call over the same broadband connection. This can make VoIP a simple way to add an extra telephone line to a home or office. * Conference calling, call forwarding, automatic redial, and caller ID; zero- or near-zero-cost features that traditional telecommunication companies (telcos) normally charge extra for. * Secure calls using standardized protocols (such as Secure Real-time Transport Protocol.) Most of the difficulties of creating a secure phone connection over traditional phone lines, like digitizing and digital transmission, are already in place with VoIP. It is only necessary to encrypt and authenticate the existing data stream. * Location independence. Only an Internet connection is needed to get a connection to a VoIP provider. For instance, call center agents using VoIP phones can work from anywhere with a sufficiently fast and stable Internet connection. * Integration with other services available over the Internet, including video conversation, message or data file exchange in parallel with the conversation, audio conferencing, managing address books, and passing information about whether others (e.g. friends or colleagues) are available to interested parties. * Advanced Telephony features such as call routing, screen pops, and IVR implementations are easier and cheaper to implement and integrate. The fact that the phone call is on the same data network as a users PC opens a new door to possibilities.
OTOH, there a few minor services-oriented drawbacks to VoIP as listed at
Quoting: * Some VoIP services don't work during power outages and the service provider may not offer backup power. * Not all VoIP services connect directly to emergency services through 9-1-1. For additional information, see * VoIP providers may or may not offer directory assistance/white page listings.

Furthermore, Skype, acquired by Amazon's competitor eBay less than 3yrs ago, has undergone mixed reviews of its own VoIP services (see

Both VoIP and PSTN are therefore weighed according to cost, provided sevices, and need for special devices. PSTN is certainly govermentally regulated, taxed and monitored. Both VoIP and PSTN have limited competition among carriers. In the U.S. at least, one is currently free to choose either voice technology or both of these together with cellular service.

In ways similar to VoIP over PSTN, proponents of "cloud computing" also claim many computing-power benefits over that of fat-clients. The work-in-progress lists many of these.

An enormous discrepancy that stands out between "cloud" computing and the full-capacity end-user client is the degree of competition. The two biggest current "cloud" megalith providers are Microsoft and Google, as previously described above. Build this on top of the aforementioned connection-providers' increased limits to use (e.g., cable providers) and that's arguably very little Web 2.0 competition. With less unregulated competition (or especially with less competition which ends up becoming heavily regulated and taxed), costs to consumers go in one direction only...... WAY UP. It's claimed by the network carriers as "passing the costs along to consumers". In contrast to this on the fat-client, each self-contained or networked device is VERY competitive. A wide choice of which OS to install, which apps, which non-locked configurations,... etcetera.

Further negatives of "the cloud" vs. the fat-client will no doubt arise as bona fide "cloud" computing becomes more widespread.


Jun 16, 2008
12:33 PM EDT
Quoting:For me, having Google's offline Gears component is what makes the online Google Docs finally work for me. They had to add the ability to use it while NOT being connected in order to "complete" the service.

Problem is, Gears isn't the cloud model. With the cloud you need to be online because that's where your data is. Gears is just a fancy caching mechanism that allows you to use a web application without getting connection errors, but the documents reside on your computer (until you connect). With the cloud (the real cloud, not this badly patched ad-hoc network we have today) everything is online. Gears won't help you there. With the cloud, Gears is just stale content and nothing more.

Quoting:One thing cloud computing must have is low cost, unlimited bandwidth connections.

You're forgetting "high speed", "flat fee" and "wireless". For the Cloud to go anywhere, you'd need something like global unlimited wireless 4G coverage at 1 Gbit/s or more, for about $10-$20 a month flat-fee (not counting inflation). Take one of those 5 elements away and the Cloud won't be more than a puff of smoke.

* No flat-fee or no unlimited bandwidth means no always-on connection * If it costs $50 or more then a lot of people can't afford it * If it's not wireless then all your mobile devices can't access the cloud. And mobile is taking a far bigger flight than the Cloud is. * Anything slower that 1 Gbit/s won't suffice in an age of streaming HD media everywhere.

It will be at least 10 years before this is remotely available. Probably more with all the big telco's pushing back and trying to keep their walled gardens. Your fat client will be with you for a looong time :-)

Jun 16, 2008
1:24 PM EDT
For next-generation broadband to be available on the level that the cloud requires, governments are going to have to get very involved.

Jun 16, 2008
6:36 PM EDT
Quoting:You're forgetting "high speed", "flat fee" and "wireless".

Hi Sander.

When I mentioned unlimited bandwidth I assumed high speed connectivity as any other kind wouldn't make sense, really. And I agree that unless the cloud has all of these components the cloud will never get off the ground. It's beginning to appear that the ingredients necessary for the cloud to operate may be hard to come by as it would take buy-in from different tech sectors--some of which will not be willing to concede to high speed, unlimited bandwifth. (i.e. the cable companies).


Jun 16, 2008
6:46 PM EDT
I think we might need to look more at the "fog" model, bringing the cloud down to ground level. While a lot of businesses might not like having their data, etc, on Google's server, the idea of having everything for their company located on a central server in their office is very appealing, and makes mobile computing SOOOOO easy for their remote employees. If you have a fat ERP app that encompasses everything from docs to time entry to email to accounting to...whatever...then you have a very easy system to setup, backup and maintain, which lowers cost a lot. You can even give your employees cheap, thin client laptops/desktops that really only have a web browser so they can't load the darned things down with crap programs, viruses and spyware, and limit their web browsing to your network and only your network. It really gives businesses a whole new level of control over their employees, which is really a great thing.

And if someone has already said that, forgive me...but these posts lately are turning into freaking novels...who has time to read it all???? Seriously, condense the posts, ya'll.

Jun 16, 2008
9:15 PM EDT
Quoting:I think we might need to look more at the "fog" model, bringing the cloud down to ground level.

You mean, like standard thin client networking we already have? Like FreeNX, VNC or Citrix/PowerFuse? If you ean web based, then the big missing component is a Google Apps that you can install and use from a local server. I was looking for that a few days agoo, but none exist yet. Lots of upcoming projects but no running code.

For companies, a better method may be fat clients that have /home mounted from an NFS share. That's what we're looking at in my company for out Linux boxes.

Jun 17, 2008
12:47 AM EDT
Quoting:...but these posts lately are turning into freaking novels...

Cheer least they aren't Tolstoy's War and Peace. :-)

Jun 17, 2008
1:21 AM EDT
I mean web-based, Sander...and hopefully something FOSS will fill that void soon. I think it will, assuming that there is a growing need for it. However, though there might not be one application that does all of that, many (or most) companies would be satisfied with having several different apps that cover their needs. I think that the only thing missing is the web-based documents that Google has, as you have said. Everything else is covered in one form or another, and there are several FOSS web-based ERP programs out there.

Posting in this forum is limited to members of the group: [ForumMods, SITEADMINS, MEMBERS.]

Becoming a member of LXer is easy and free. Join Us!