I realise I'm missing the thin client point but this seems to be very expensive. Could buy a conventional cheap unit, with Wi-Fi and get some change

This new thin client is so thin (cheap) that you can buy a standard desktop system with several times the grunt for less money. Such progress...

If 'thin' systems cost more like a one-laptop-per-child system (you dont even need to include the lcd, harddisk, and battery which are the expensive bits) then they would be a much more interesting toy.

At the bottom end of the market, the expensive part of a computer *is* the screen. In the absence of new display technology there's just no point in building a computer for less than 100 quid because you'll have to spend at least that much again before you can watch it boot.

I'm a big fan of thin/remote computing. At least until the line between the client and the server breaks. Wyse - a former king of the "dumb terminal" world - is trying to catch up to (or keep a hold on) their old market share, I guess.

Why don't we see, instead, more of a focus on "OS on a removable drive"? Commoditize the hardware to have no fixed storage beyond the boot loader, and just plug in your data - which includes your OS of choice - to the resulting "terminal". Your OS - on a USB or Firewire drive - might be a thin client that calls out over The Network. Or it might be a full-fledged OS, able to function with or without a connection to the rest of the world. It could be on a "real" HDD or on a memory-based thumb drive.

This would be a perfect solution for libraries, kiosks, and other public places. The only hard part is solving the security problem: we don't want malicious virii or trojans left from (or by) the previous user twiddling our precious bits.

Maybe I've been reading too much Vernor Vinge lately.

What is the POINT of a thin client computer packing a CPU clocked at over a GIG?! Do you seriously need that much grunt just to run a window-server?

Thin Client is just a gimmicky and insubstantial as any "Get You Thin" diet, you'll be hard to find any chipset (Cirrus's ARM based PDA/STB parts comes to mind) that doesn't have the capacity for far more than rendering windows. Why should we waste that?!

As for a thin client notebook, dream on. Why the hell would you want a portable computer that only works while it's connected to the internet, and crucially while your SERVER'S uplink is working. Sure the corporate website might stay up 24-7 but does your site's uplink stay online and below capacity all day?

It's bad enough that file shares and exchenge drop every now and again, but if they took everyone's desktop session as well?! Not just BAD, attrocious!

Dumb Terminal by name... numb idea by nature.

Companies like IBM and Oracle and Google dream of owning giant server farms where everyone rents space and computing. The end user is stuck with a non-programmable thin client. The good news is your work is always available from anywhere. The bad news needs a list:

1. Network down, you're hosed.

2. Sharing files and CPU time with everyone else sucks. If you're old enough to remember time sharing, you know what I'm talking about.

3. Openly programable platforms like Apple and Microsoft disappear, and so does our software industry.

4. Way too much power concentrated in the hands of whoever owns the servers.

Firstly, these dedicated thin machines have advantages over a stripped down conventional machine; passive cooling and lower power consumption might be important to some people and are hard to get elsewhere, so it might actually be worth a little more to someone who had the money to spend.

Second, the laptop actually does work disconnected; read the article a little more closely! Granted there's less point with a portable and I'm not at all tempted by it (XP embedded for offline use is barely more use than nothing at all) but it's not like it's dead without a connection to the server.

Great idea. Suppose you have a big room that you want to use for computer teaching, or seminars, or with a clear floor for dance lessons. With a non-laptop form-factor its a real pain to rearrange the furniture.

Keep a bunch of thin-client (and ideally wireless) laptops charged up in a cupboard and you can just deal them out when you need them. You also get the thin-client advantage of only having to maintain four rack-mounted servers for forty workstations.

The user who thought it was a dumb idea up there wasn't thinking outside the beige box.

"In fact, the company envisions a future where thin clients get thinner and thinner to a point where they are free and disposable. Madness, you say? Well, they certainly seem to be headed that way."

If you consider $480 to be heading towards a future of free and disposable, you must be thinking of "future" in evolutionary terms (millions of years) as opposed to reality (five, ten, or twenty years). Thin clients have ALWAYS been expensive. Why would a company want to pay $480 for something that does less than a $300 low-end system which, with a little work, could be configured the same way?

Thin client pros: Small, quiet, secure (since there are no hard drives).

Thin client cons: Expensive, speed is dependent on server speed and current server load, must be connected to server at all times.

Simply put, Wyse (and other thin client makers) are trying to revive their company to its former glory days by overcharging for what many people consider to be useless equipment. If that's what you want, and you're willing to pay it, great. But I have a feeling they could offer these at less than half the price and still make a ton of money.

"What is the POINT of a thin client computer packing a CPU clocked at over a GIG?! Do you seriously need that much grunt just to run a window-server?"

Spot on, scruffy! I have a system at home running a 500MHz CPU, a 5 gig Compact Flash hard drive (the Seagate itty-bitty-actual-hard-drive-in-a-CF-format type), one gig of RAM, and Ubuntu Linux. The only speed issues I ever have are caused by slow Web sites.

These units have a very long life span and very low cost of ownership, that's why they are considered inexpensive.

It costs much more to keep a typical desktop up and running than just the purchase price. Most thin clients have at least a 7-10 year life span, you would replace a typical corporate PC 2-3 times in the same time span, and still have to deal with the support costs generated by having windows installed locally on the device.

The idea of the thin client laptop is particularly interesting given the concerns of data security in today's climate.

The best of both worlds would be a desktop appliance that "Cached" the system locally to utilize the processing power of today's chipsets and the centralized support systems of the thin client systems. I think we will see Google offer such a system in the next 5 years, then will will really see a catfight between giants. . .

I'm a big fan of Thin Client. There are a lot of advantages over normal computing - remember that no call center weasels are going to be nicking these, or able to stick USB keys into them and nick your data (ok, they can, but you have to enable USB) You can run network hungry/talkative apps inter continentally, which is quite important if your datacentre is in the UK and your staff are distributed round the world. Everything works a lot faster as well, because the data is next to the servers are next to the network and the storage, no running gigabit wans everywhere.

That said, they are shite at multimedia, but how many of your staff can justify multimedia capable desktops and can those few people get a normal PC? Probably.

If Apple had used the same pricing structure....

8g model for 399 for the 1st month then raise the price by 100.

I think they would have sold twice as many units.

Because it saves a fortune, by getting rid of the desktop-dependent IT department folks, and the Microsoft Select agreements, and consequently maybe even most of the IT outsourcing bill ...

Yes thin isn't the right answer for everyone, especially not for l33t lamerz er sorry gamerz.

Yes thin is likely the right answer for 80+% of corporate desktops.

Timesharing went out of fashion because distributed computing hardware started to get cheaper than centralised computing - processing, memory, storage, bandwidth. Now, that trend has continued so far that the situation is sort of reversed - hardware (and bandwidth) is dirt cheap, but people's time (and some software) is expensive. Solution: get rid of the desktop management nightmare, which automagically also cuts down on security risks and other trivia like that.

If you are an IT bod and your skills are focused on Windows desktop support, better call the Jobcentre and see what's in demand, before your employers start looking at the cost implications of a Vista rollout. Or read up on thin client and be the one that tells your management how they can save big money on IT, not just by avoiding the Vista rollout, but by taking it further, without the risks of offshoring or outsourcing. You wouldn't want the big-iron server vendors and resellers to get to the management first would you (and they *will* get there, eventually), because some of them will want your jobs as well as the server sales.

I'm not in a position where I have use for thin clients, and for damn sure not Citrix.. but these do peak my interest as just quiet low-power computers. Not the desktop so much, my current small form factors use much more power but are already quiet... and I'm not running dozens, hundreds, or thousands so I'm not worried about power use so much.

The notebook though.. It sounds like a crap thin client to me unless it was strictly used within a single building (perhaps carrying it around within a hospital?), but it'd be fabulous to have a notebook that draws 13W of power! That could bring back the days of 5-10 hour+ notebook battery life, which would be fantastic in my book. Just as Morely Dotes, I run Linux on my systems (gentoo on some, and ubuntu on any fresh installs) and, as he says, a 500mhz system is plenty; a 1.2ghz would be just great.

If thin clients will be disposable, then I'll save the trouble of carrying one with me, and just look in the bin when I want a laptop to check my email.

What's the point of making them that thin? I'd wager that within 2 years flash 'hard drives' will be competitive enough that you'll be able to buy a 500GB laptop drive for about the price of a current HD--which will make battery life similar to a stripped down thin client 'notebook.' With properly implemented encryption, data/device theft isn't that worrisome anyway. And as others have mentioned, the pricing of these is currently atrocious, when you can buy a cheap desktop or laptop with more power and complete autonomy. Probably DVDs will be a thing of the past soon enough as well, with all movies offered on some sort of flash RAM or ROM devices.

We have been using Citrix for over 10 years. We have expunged fat clients from our system and deliver all applications via WAN and Internet to 14 offices throughout New Zealand and one office in Australia. We also deliver applications via Citrix/Web Interface to 3rd party client i.e as an ASP.

We have had a couple of thin client notebooks (MaxBook 810s) for a couple of years and just recently purchased another 2 NeoWare m100 units.

I see the argument that you could take an entry level laptop and pare it back to resemble a thinnote. However, why waste the time? It took me about 10 minutes from delivery of the m100s to get a 3G card installed and configured so that the user can work as they would at their desktop. I had to install PowerPoint viewer as one of the main uses my users have for the thinnotes is to put seminar attendees to sleep with their Powerpoint presentations. We can either hook the presentation off the server onto a USB drive and run it through the thinnote or run the presentation in a Citrix session.

My time is too valuable to waste on configuring new PCs or notebooks. Once I have the application installed on the server I can deliver it over and over with very little effort. I never want to go back to the days of configuring user workstations and installing applications locally. Break/fix with thin clients is rare, but the neat thing is that it is just a case of swapping out a device. The user doesn't have locally saved files or configuration settings that they want to preserve off their dead PC.

Similarly, the security of these things is worth the price. Because users cannot download and install garbage the devices cannot threaten my network. Even with a pared back entry level notebook, I would always be nervous that someone could potentially infect the network with malware. If someone leaves a thinnote somewhere there is no data at risk (apart from their boring Powerpoint presentation on the USB drive which no one is going to get any information out of anyway - apart from the cure to insomnia).

If I need a bit of processing capability when disconnected from the network or out of 3G coverage I use PortableApps OpenOffice on a USB drive. It works brilliantly.

nice that ps2 has returned to the desktop models - for a year now we've been getting usb only wyse terminals bundled with ps2 mice & keyboards.

What's missing is some sort of routing algorithm, perhaps OLSR or B.A.T.M.A.N. so the system could do it's own routing.

Just like previous posters noticed, if the network dies you're in trouble. Therefore it should be important that the network has as much redundancy as possible. A conventional wireless or wired network is hard to get redundant. In such a network you could essentially connect one or two terminals in an office to the server via wire. If one should fail, there's still the other one. If both should fail, the network will reroute itself to go over the next office. It will probably be somewhat slower, but it'll still work.

Obviously you should keep your connections encrypted.

Why not Ardence, cheap PC, No hard drive and fast network. Much cheaper, quieter, easier to manage if you sort common hardware and the drivers. Low Power will not be as good as ChipPC or Wyse buy much better than a thick client.

I always liked the look of the thin clients at http://www.ndiyo.org/

And they're British, whoo hoo.

Unfortunately, they're not really available, though you can get a starter kit of 5 prototypes for £1499 +VAT.

So, it's still a bit pricey

Nice posts. After reading them all, I was thinking that one of the most valid points which relates to disposability is the swap-when-needed, along with the no-time-waster.

All the other views on possibile similar aproaches, with regular systems will most likely have some time consumption of some sort, along with the eventual need for maintenance, being extra time consumption or expense.

While there are isses related with the choice, regarding network dependencies for one, in order to keep it viable and working, extra efforts need to be carried out as a form of backup support.

While not being the one and only solution, it is one, along others, with it's on pros and cons. The cat fight will come in the form of "how to's & what's" to try and make it stand out in the current reality.

To note: I don't know about other existing companies but, Intel and AMD current desktop CPU technology is already capable and further focussing on virtualization.

When you're a corporate IT department. Fixing a terminal server is rather less work (and much less frustrating) than fixing a maze of 300 Windows PCs, all different. Software breaks several orders of magnitude more often than thin client hardware does, anyway.

For home users, less so. Still, I'd consider one if they were cheaper and not locked into an expensive proprietary back-end technology. (But I'm a geek, so my house has IT infrastructure.)

are missing the point of thin client. I've just installed 50 into a teaching area. Mostly they'll be used for office and presentation work not high-end computing. Space saving is enormous and there are far less maintenance and troubleshooting issues. My ISS department does all the management remotely.

I'm not much convinced by these Wyse jobs though. I looked at these among others and went for JackPCs.

http://www.jadeintegration.com/jackpc.php

These fit into the same size space as a 13 amp socket. In fact I've got 6 fitted into short vertical power poles on each of 8 island benches. As for power, they pull about 5 watts max which does kind of put the 13 watts for the Wyse numbers to shame, and they get their power over the ethernet.

Not saying these are perfect or even the best, but they seem to be doing everything I wanted and they certainly look cool. Most folk look and ask where the PCs are.

Steve

Nothing wrong with this except the price and possibly security.