Talk:Network Computer

Rewrite
This article needed extensive rewriting, as at least 50% of it was literally /wrong/. An NC is not a dumb terminal, it was a computer in its own right, which happeded to use the network for storage, hence 'network computer'.

I've also added historical information and details of notable implementations, as well as links to some relevant external sites.

Phil webster 11:04, 23 November 2005 (UTC)


 * Did you fix some of it at least? 220.233.48.200 14:57, 11 December 2005 (UTC)


 * This and the Web Desktop article both need to be clarified and split into more concrete articles. The Contemporary analogy section is particularly bad, as network computers have nothing whatsoever to do with webtops or web desktops. A better analogy would be between network computers and PXE network boot, because the best selling network computers were just diskless, network-booting Unix computers. Thecorbaman (talk) 02:51, 22 July 2014 (UTC)

The table
The table looks soooooo ugly and isn't using wikimarkup for it, maybe someone which knows wikimarkup well can port it over?

220.233.48.200 14:57, 11 December 2005 (UTC)

Reasons for the demise
I helped develop Oracle's second-generation NC software stack in the mid-1990s, helped write the reference profile, and collected the T-shirts as well as a garage full of cast-off NCs. I have some insight into Oracle's intent.

The article's representation of Ellison's vision for the NC is somewhat simplistic. We knew the NC would not supplant the PC in all its use cases. We didn't intend that. We aimed it initially at cases in which a significant number of full-fledged Windows PCs at some site were employed in single applications. We conceived of airport kiosks, hotel rooms, hospital rooms, factory floors, and some typical business workstations (e.g., reception desks). The statement that the NC concept is "appropriate … in certain areas" is a validation of our original aim, not a recognition of some failure of it. Ellison deserves credit for realizing at the time the natural limits of the new computing paradigm he was proposing.

We hoped that NCs would eventually find their way into homes, but that was not part of the original marketing or technical strategy. We were well aware that dialup connections were far too slow for what we had in mind. File storage access was over NFS, for example — fairly unusable over a dialup connection. Our intended market presumed reliable broadband connectivity, so claiming that the NC failed to hit a market at which it specifically didn't aim is somewhat misleading at least in Oracle's case.

Oracle did not consider replacing dumb terminals any kind of violation of the "pure" NC philosophy. We expended great technical effort to ensure that terminal emulation was 100% compatible with hardware examples. Replacing aging and broken dumb terminals with cheaper and fixable alternatives (i.e., NCs running software terminal emulators) was expected to be a major source of revenue.

The notion that an NC would perhaps run only one application program in its entire lifetime was quite welcome in Oracle's NC philosophy. Of course we wanted the NC to be seen as a "real" computer, but the selling point was precisely that you could relegate the NC to a single task without a twinge of conscience, since you were presumably already relegating a fully-loaded Windows PC to that same task.

The market reasons for the NC's demise are relatively simple. Oracle intended the NC to reduce the total cost of ownership for computers that were primarily used in very simple ways. Technically we planned to accomplish this by using inexpensive hardware, open-source and open-standard software, and by managing the software centrally at the server. But in the late 1990s the average price of a Windows PC plummeted to the point where the NC could not compete with it on a price-performance basis. And Microsoft began to address the problems of centralized control.

Those who proposed the NC argue that the PC industry lowered its prices out of fear of NC competition, but I can't find any PC manufacturer who will admit to that. But the NC proponents say the NC concept fulfilled its goal (lower TCO) without needing to be realized.

The technical reasons vary from company to company. The Java Station, for example, was simply too sluggish to be useful as a computer. It relied on first-generation Java virtual machines, and so in that case it can be argued that it was ahead of its time. Java is now routinely embedded in handheld appliances such as cellular phones. Oracle's technical problems amounted to biting off more than it could chew.

Trying to displace Microsoft from its entrenched monopoly over the business desktop is arguably a losing proposition no matter how technically advanced your solution might be. Very few competing companies put enough resources into their NCs to match Microsoft's efforts at improving the desktop at the time. Jpwindley 05:59, 19 October 2006 (UTC)

Maybe you never intended it but the implication in many news articles I read at the time is that NCs would eventually overtake PCs. I could never see the remotest possibility of that, not because of Microsoft or cheap hard disks but because of human psychology. Having your own autonomous computer is always more desireable to most people (even a receptionist) just as driving a car rather than taking public transport is. 13:15, 2 December 2009 (UTC)


 * Unfortunately this can't be included as it's original research. A list of published press reports (preferably with online links) would be helpful for the future. Thanks. --Trevj (talk) 14:30, 4 May 2011 (UTC)


 * There's an interesting paper in Int. J. Technology Intelligence and Planning. --Trevj (talk) 12:17, 2 June 2011 (UTC)

Non-contributing text removed
I just deleted the text "In The last few years (2006-2007) there has been a "ComeBack" to NC's and Thin Clients for some reasons and certain Applications. First because most advanced computing has become "Web Based Computing" from Internet Cafe's Corporations Education and Mobile Devices that all need Fast Internet with minimum Client needs. Also Important Reasons are Security, Low Maintenance costs, and Server Operating System and Hardware independance. Since Microsoft stopped going against this trend in the 90's and have joined with WEB-TV and are Partners to Citrix and started "Terminal Services". Also the Large "Third World Market" want to extend the life of "Thinner" Clients and Old PC's or inexpensive Laptops and 3G Cellphones." because, frankly, it is awfully written. Motmot 18:29, 11 February 2007 (UTC)

Still supported?
Are Network Computers still supported? The Netra j 3.0 server for NCs still has docs on the Oracle site. --Trevj (talk) 14:43, 9 June 2011 (UTC)


 * Network Computer, at least in the context of this article, was an Oracle product (long before the Sun merger). The Netra servers were specifically for JavaStation thin clients from Sun (if I recall correctly), although it was compliant with the Network Computer standard.  I don't believe Oracle supports the Network Computer standard any longer, and I don't believe any clients or servers have been made for the standard in many years.  I suppose you could say the (unintended) successor, first at Sun and now at Oracle, is the Sun Ray line of thin clients.  jæs (talk)  17:41, 9 June 2011 (UTC)


 * Thanks. So does this mean that the Netra servers weren't used with NCOS? --Trevj (talk) 11:00, 10 June 2011 (UTC)


 * The Network Computer standard was so vaguely worded that it included virtually any computer, but the best selling were just diskless network-booting computers, similar to PXE today. Thecorbaman (talk) 02:53, 22 July 2014 (UTC)

Concept cloudcomputing/ Tagline
Actually the Network Computer concept is what we call Cloud computing now. I did an interview with Larry somewhere around 1995, where he presented the concept during the Macromedia convention in San Francisco. Anything, anytime, anywhere where the main ingredients of his Network Computer concept. When you look at Netflix now, that is how Oracle was providing tv on demand systems based on that concept back in 1995... The man is a visionair, not a failure. — Preceding unsigned comment added by 82.168.179.252 (talk) 08:07, 14 February 2014 (UTC)
 * Was the interview published? If so, where please? As for comparison with Netflix, including that would be original research unless reliable sources have made such a comparison. -- Trevj (talk · contribs) 09:38, 14 February 2014 (UTC)
 * I'm extremely dubious about any such claim. The Network Computers do not resemble "cloud computing" as exemplified by AWS or other major cloud computing platforms. Grid computing itself has been around since the 1960s.  Any such comparison is sloppy thinking at best. Thecorbaman (talk) 02:56, 22 July 2014 (UTC)

External links modified (February 2018)
Hello fellow Wikipedians,

I have just modified one external link on Network Computer. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20070404210105/http://www.byte.com/art/9704/sec6/art1.htm to http://www.byte.com/art/9704/sec6/art1.htm

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 11:45, 16 February 2018 (UTC)