Larry Ellison has once again demonstrated how he always gets what he wants in the end, by floating the possibility that Oracle could get into the stripped down client PC business. OK, it looks like he's latching onto the latest greatest thing - telling the JavaOne conference that now Oracle is swallowing Sun, he could well …
Funny how things turn out. Sun's dreams have been quietly booted aside by Citrix and possibly even the up-and-coming improved Windows Terminal Services.
If Sun want to make a Java-based OS-come-Citrix-like client, fine. But I don't see any compelling reason behind it, unlike say, Android combined with other Google goodies such as Gears, running Chrome which would give handy access to Citrix and other web-based client services.
JavaStations and NCs
The JavaStations were an interesting idea, but the Java-heavy software environment (HotJava for browsing while everyone else was getting better browsers) and the pricey hardware made them uncompetitive (and the hyped "Java chips" were mostly rebranded SPARCs that didn't make Java run magically fast).
The Oracle-related NCs were interesting, too, if you consider their heritage: the first generation apparently came from Acorn and were less ambitious variants of stuff the Olivetti (and Oracle) Research people had put together for video-on-demand back when Oracle wanted to own that market. After Oracle gave Acorn the brush-off, the second generation NCs apparently ran NetBSD, but by then Larry had really lost interest.
I for one wouldn't want to wait an hour for my Java based net book to boot up.
"If Sun want to make a Java-based OS-come-Citrix-like client, fine. But I don't see any compelling reason behind it, unlike say, Android combined with other Google goodies such as Gears, running Chrome which would give handy access to Citrix and other web-based client services."
Personally, I feel an OS with lots of support for Java and a Java-like Gears would be really useful. You need some kind of synchronization to support offline and online use (e.g. Gears). But otherwise, I would much rather write java applications than ajax ones. So you can have your android, and I'll have my ElliPod. (Ellison + iPod, get it?)
Takes me back
I used one of the very first JavaStations. The company I worked for bought one to see if it was worth investigating partnering to make some hardware addons for it. We went and had lunch while it booted up, and when we got back we looked at it and decided it was shite and doomed to fail. And there it sat, gathering dust until several months later when the power supply burned out and set the alarms off.
Network client computing is a fabulous idea, but the whole point of it is that it needs to be a lightweight client. The only thing lightweight about Java is the useability. What's wrong with X?
What is the point of having no hard drive?
I could understand it if hard disks where expensive but they are not.
A hard disk is one of the least expensive bits you could remove. Cost of internet will have to be reduced and quality increased before this is useful and people running BT have made it very clear that isn't going to happen any time soon so it's a no go for the UK.
And when all is said and done, MS do make good (enough) products and past experiences makes me sceptical of a fully java OS.
Sun Netbook ?
Only if they come in Sun Mauve with a Sun-style keyboard and run OpenSolaris! The temptation to NOT buy one then would be hard ... :D
Spot on. The most expensive capital cost of a cheap PC is the screen or the battery. The most expensive running cost is the connectivity. If Larry can significantly reduce any of those three, the world will beat a path to his door and they won't (particularly) be asking about thin clients.
Set top boxes
When the UK cable industry went digital the software it was on the Liberate DTV platform. Liberate was originally called Network Computer Inc and was the product of Larry's foray into the thin-client business. So far from falling completely flat the technology did make a lasting impact in the world of TV.
JVM version dependency hell
More Java crapware (see Oracle Forms / Java Applets) that hardly ever works because the client doesn't happen to have a compatible version of the JVM installed, exhibits dire performance, a lousy UI and crashes frequently.
Not just the hard drive
The typical justification for thin clients has been that it would be so much better if everything was done by the server guys in the computer centre (i.e. let's get back to centralised data processing, like in the 1970s). The main snag with this is that the necessary servers, storage, licenses and the folk who support them tend to cost rather more than the desktop systems with software and their corresponding support people. I've seen several arguments about total cost of ownership but the thin client proposals (together with the server side) always came out to something way more than the entire IT budget at which point management put a stop to it.
Re: Takes me back
"What's wrong with X?"
Hoooo boy, there's a loaded question. It may be more productive to ask what *isn't* wrong with X ...
I used X Terminals back in the day when I was at University and they were cool and all, but it was only when I started working as a developer that I realised what an absolute abortion X11 is.
Larry Ellison looks like Julius Caeser.
I knew he reminded me of somebody.
Thin Clients, the road to serfdom
Handing control of data and software to big server-based companies like Oracle, IBM, or Google is a bad idea. My concern is not just data privacy, it is the possible lost of an openly programmable hardware platform. How can we be sure that the independent software industry and the open source movement can continue with total freedom on an oracle-controlled platform?
I don't think so
We've been down this "network computer" road before. Doesn't matter if it's a "Netbook" with a browser based OS or any other sort of "cloud computing" nonsense. Cloud computer or "Network Computing" is entirely dependent on expensive back end Server hardware and software. In this economy, who's going to buy those big boxes? This is why Sun was bought by Oracle. They aren't selling the big iron any more because no one wants or needs it. That isn't going to suddenly change because Larry bought the company and wants a netbook run by Java. No one really cares about Java any more. Linux is having a tough enough time gaining any traction.
Previous failure was predictable
The earlier javastation concept was fatally flawed for one simple reason; compatibility. Businesses use computers to perform processes, and those processes typically involve Word and Excel documents. The Java thin client just couldn't handle that. So businesses didn't want.
Now today we have netbooks... and what's happening? Linux is being dropped and replaced with Windows XP. Why? Compatibility! Home users rarely can afford a real machine and a network-access machine; they're buying netbooks as cheap laptops so they can edit their Word and Excel documents on the move!
Sure, there are some people for whom the original netbook concept fits nicely (me!) but we're in the minority.
So why does Larry think his new javabook will succeed now? OK, OpenOffice and Google Docs are now around, but they're still not the same as having a Windows machine in your bag.
So even without trying, Microsoft will kill off this new idea. Blame Microsoft. As always.
Re: hard drive
Hard disks may be cheap to buy, but for a company they're one of the most expensive components from a TCO perspective (hardware failures, backups, rebuilds, etc) . Remote desktops (eg Citrix) running Windows with thin-clients actually makes sense. Historically, though, the pricing has been all messed up. Today pricing is becoming a little more sane and a number of large companies are seriously looking into virtual desktop configurations.
You have GOT to make a icon image of Larry in the money pile a selection in your list.
Get rid of the heart, nobody that comments here uses it you know......
Anon 'cuz I work for him now........
DOH! I had too!
Two Out of Three Isn't Bad
Larry got it mostly right. Thin clients are 10% of installed base of PCs and they save users a bundle while giving improved performance on GNU/Linux or UNIX systems. The Java thing did not pan out but we have everything we need: speed, power, low costs.
More than 80% of users of PCs could use thin clients very well. They work well in place of the desktop so about half the market for PCs could go thin eventually. I would use thin clients anywhere folks did not need mobile computing or video/heavy graphics. Browsing/editing takes care of most tasks in business, government and education.
Ellison is on crack
He has done it again. Pulling out of his ass the latest revision to the network computer.
This is a solution no one wants, to a problem no one has. I look forward to watching this fail yet again.
Yeh, I like the article, I like people having a dig at Larry, the hairiest, richest man in the world at least.
But Larry's notion of the Network Computer was never about the hardware. It was about shifting more databases into massive data center's that ran the applilcations everyone had delivered to their device.
It was cloud, saas, whatever you want to call it, before everyone rebadged it.
Not all of the vendors or the market quite realised that unfortunately, which is why we still have the tiresome legacy of Java - which unquestionably benefitted from the hype of being the Internet's OS (or at least it's execution environment).
He called it right.
@AC 03Jun09 18:25
Finger on the pulse.
Larry has a more acute smell of the trail of $$ than Bill Gates...
NC? Javastation? I eval'd SunRay in '98... nice concept, but it needed arrays of Sun 10k tin for a small population of users. And of course needing Oracle db's for everything.
Others are bang on too:
@JohnG - TCO with centralised is higher than distributed.
@Frizzl - opex/network far too high (and is it possible? - Korea maybe, UK no).
@Ian Ferguson - why Oracle when you've got Citrix?
Maybe it's a bit of FUD to disguise what part of Sun he's going to can next.
Smiley because it is always fun to try and second-guess Errol Flynn type COOs.
It was called the NIC
They used CD-Roms and booted in Linux. They ALMOST worked OK, but when you pay silly sales droids (I know someone who worked there!) to have pipe dreams and not sell anything (a good market was Libraries) it is going to fail.
If you want to know about the hardware, go to http://www.nicrevival.com where they try to support the hardware, but the links are broken (*SIGH*).
Oh, I have one of the machines as well.
Re: Previous failure was predictable
"The earlier javastation concept was fatally flawed for one simple reason; compatibility. Businesses use computers to perform processes, and those processes typically involve Word and Excel documents. The Java thin client just couldn't handle that. So businesses didn't want."
I'll only agree with that partially. I saw JavaStations in the wild in businesses which didn't need Word and Excel to run their processes. Indeed, if your business has the capability to develop or acquire decent software for managing processes, Word and Excel are poor choices: they are the equivalent of paper pushing (in its most derogatory sense) for the digital age.
"Now today we have netbooks... and what's happening? Linux is being dropped and replaced with Windows XP. Why? Compatibility! Home users rarely can afford a real machine and a network-access machine; they're buying netbooks as cheap laptops so they can edit their Word and Excel documents on the move!"
I guess you didn't see the survey that said that a lot of netbooks were being bought as second machines, then. And I see yet more projection of the "business laptop" mindset ("that's what I use them for, so that must be what everyone else uses them for, too") when home machines are shipped with the generally incompatible Microsoft Works. At least OpenOffice reads and edits most of the Microsoft Office formats, if compatibility is your thing (and your business, even though we're talking about home users, is not still at the level of paper pushing), but I'm inclined to believe that the "Microsoft-empowered" home user is one of those myths that sits so conveniently that people rarely bother to check its veracity.
And such myths are money in the bank for Bill and Steve, although I'm apparently a "zealot" for saying so. Sheesh!
Java is always a problem
it is just not that great a concept or language to code in.
A lot of bad developers like it because it is sort of accessible but takes a long time to develop solution so they get paid over a longer period of time, but it is so dull a language.
The functional languages I think is where the smart money is, there development times are swift, the language is enjoyable, the results are stable and scale.
On the client side, then Python makes a bit of sense, or C with Lua is quite sexy.
Sun was killed by Linux and open source; first the UNIXs will fall.