Ernest Miller pursues research and writing on cyberlaw, intellectual property, and First Amendment issues. Mr. Miller attended the U.S. Naval Academy before attending Yale Law School, where he was president and co-founder of the Law and Technology Society, and founded the technology law and policy news site LawMeme. He is a fellow of the Information Society Project at Yale Law School.
Ernest Miller's blog postings can also be found @Listen to the weekly audio edition on IT Conversations:
The Importance Of ... Law and IT.
Feel free to contact me about articles, websites and etc. you think I may find of interest. I'm also available for consulting work and speaking engagements. Email: ernest.miller 8T gmail.com
Balkinization
As part of its series on Microsoft's recently unveiled next-gen OS, Longhorn, C|Net News analyzes the strategy behind the software (Plan A for Microsoft). The strategy, according to C|Net, is a return to a "fat client" architecture, where lots of processing takes place on the user's PC or portable device, as opposed to a remote server.
Frankly, I don't really get this thin/fat client debate. The debate most often seems to be not whether there should be a server/client architecture, but how robust the client should be for the consumer. Personally, I've never really understood the value proposition for most thin client architectures, given the relative inexpensive of silicon and magnetic storage. The question for me is why consumers aren't running both clients and servers.
Many consumer electronic devices are basically computers, but they don't need to be that smart, they can be specialized. What I need for my consumer electronic devices is not easy internet access, but home network access. If Microsoft is looking for new markets for software, why not develop and sell a server for consumers? What I would like is a machine that homeowners can easily stick in a closet, but will provide media and applications functionality throughout the home. Consumers don't really need a new version of Office, they need something that will let them more easily manage their increasingly gadget-filled home.
Oh yeah, and it would be nice to easily run a server from home that handles all my internet publishing needs, but that is more of a telecom issue.
> I've never really understood the value
> proposition for most thin client architectures,
> given the relative inexpensive of silicon and
> magnetic storage.
From the customer perspective, the principle value proposition of a thin client architecture is not reducing the cost of the client, but the ease of system administration. The more software, data, etc. that resides on the server, the easier the tasks of upgrading, patching, backing up, etc. Of course, this is of greater value in a corporate environment than it is in a small office or home environment. From the vendor's perspective, on the other hand, the value proposition is in the customer lock-in if both the client and server software are proprietary and do not interoperate with third party products.
> Many consumer electronic devices are basically
> computers, but they don't need to be that
> smart, they can be specialized.
If you mangle the conventional definition of a thin client a bit, you might already own some. An iPod, for example, could be considered a thin client, and the Mac to which it connects could be considered a server. In this case the value proposition is very definitely in minimizing the cost of the client. An iPod *could* contain an analog audio input, MP3 and AAC encoders, a modem with which to connect to Apple's music store, and software with which to maintain one's music library. Needless to say, its cost would rise considerably.
> What I would like is a machine that
> homeowners can easily stick in a closet, but will
> provide media and applications functionality
> throughout the home.
You could, of course, install Microsoft's server products to accomplish this -- if you're willing to put a significant dent in your wallet, that is. The margins on a low-cost consumer version of these products (compared to their enterprise counterparts) are probably too insignificant to generate a lot of enthusiasm at Microsoft. But if you are willing to bite the Linux/BSD bullet, this is eminently achievable, and at no cost to boot. (Well, no *monetary* cost, that is; there are still the temporal costs associated with the learning curve, system administration, etc.) With the exception of the media components (which are available, but of little interest to me personally), I have run a network like you describe in my home since 1994 on Linux-based systems.
> it would be nice to easily run a server from
> home that handles all my internet publishing
> needs, but that is more of a telecom issue.
There are several Linux/BSD-based virtual private network (VPN) implementations out there that would allow you to securely connect to your home server via a dial-up connection and have the remote system appear as a trusted host on the local network. Once again, there are commercial products that allow you to do this in a Windows environment, but they are oriented toward the commercial market and priced accordingly.
About Us |
Contact Us |
RSS |
Advertise |
© Copyright 2000-2006 Corante. All rights reserved.
