The PC Is Not Dead

Belle writes "Bill Gates has an op-ed in this morning's BW Online, in which he responds to the magazine's question Is the PC dead? with a resounding "No!" and argues that the most revolutionary years for personal computing are yet to come." From the article: "The result is that the personal computer has become far more than a cog in the machine of corporate computing -- it's an essential tool for every individual in the organization. Take the personal out of computing, and most companies would grind to a halt."

Comments

Not dead but very sick...

[ites]

... the PC as an island of personal data is facing real threats:

- invasion from parasitical software
- competition from smaller devices
- competition from web-based services
- ever cheaper hardware

Of course I'm typing this from a PC and I can't imagine any other way of working, but still... in 10 years' time:

- would I have to move physically to a box somewhere in order to read slashdot?
- would I have my data sitting on a single hard disk somewhere under a desk?
- would I be surfing on the public Internet using the same infrastructure as I use to (e.g.) access my bank accounts or write contract proposals?

The PC as "personal computer" is running out of reasons for being... ... the future belongs to secure virtual infrastructure, secure distributed data, and redundant portable devices.

The PC will eventually be relegated to a keyboard, mouse, and screen.

The PC isn't dead, but PC innovation is

[pocari]

Typewriters were around for a long time virtually unchanged. There is no doubt that the Intel/Microsoft platform has become the Wang word processor of the 21st century, essential to every office.

The circumstances that led to the PC revolution are long since past. When the anti-trust case against Microsoft was settled four years ago with no consequences, investors and entrepreneurs were told that there is no reason to bother to do anything Microsoft might have an interest in, because Microsoft would be free to use the Windows monopoly to crush them.

During the dot-com boom, almost all software talent went to Internet development, sucking the oxygen out of innovation meant for the PC. Bringing things on-line is important and valuable, but the 10,000th brochure website, or even the second on-line bookstore, is not innovation.

The dot-com crash in Silicon Valley has meant the loss of 400,000 jobs there and 400,000 people moving out of the valley. It's debatable how much of this is due to outsourcing, but for every job lost to some other location, that's one fewer young engineer cooking up ideas in a garage. India and China have gained, but the software industry has lost something by the scattering of young talent; the disappearance of tech veterans has long-term consequences, too.

There are still business opportunities in cleaning up security messes and customization of enterprise software products, and there always will be, but none of this really counts as innovation.

When I moved to Silicon Valley in 1995, it wasn't obvious that Microsoft was going to dominate the way it does today, or that the Internet would suck the oxygen out of other kinds of software projects for a while. The smart money and adventurous people have moved on to other things. Forever.

Re:Maybe next year, eh?

[BoomerSooner]

Not to mention administration. The biggest time-waster at my company is fixing users computers (hell sometimes mine included). Updating, upgrading, trying to hunt down and unreg all the gator entries, ...

Administration costs are insane for large corporations. Thin clients make that task a little more manageable. Only problem is when the main servers go down you're killing not just one user but a whole organization.

And there is much of my quarrel with BillG

[mwood]

Certainly some computing should be personal. But some is not and should not be. I have to work ten times as hard on Windows PeeCees as I do on other computers to get them to do impersonal things, like send me a summary of their own activity for the last week without my having to push a button.

Some very useful computation is not personal, interactive, exploratory, or "an experience". And Microsoft traditionally just didn't "get" this. Like the old robots in Asimov's "Runaround", supposedly automatic processes just won't go without a human in the saddle giving orders. They are getting better at this, but still have far to go in order to catch up with the 1960s, let alone the 21st century.

I often laugh bitterly when I hear about the "increased productivity" attributed to gadgets that make me do everything manually rather than just doing the work and sending me a note on how it went.

If you want my recommendation for your software product, ask yourself, "would there be any point in having this run automatically when nobody is around?" And if the answer is "yes", *make it easy to do so*.

Re:Maybe next year, eh?

[eno2001]

Well... I moved my family to a thin client system based on RedHat 9 a few years back. So far it's worked out great. There is very little functionality that most users need that require a fullblown PC sitting in front of them. The current list of apps we use in thin client model are:

VNC +GDM - Remote Desktop Functionality
GNOME - Desktop Environment
Firefox - Web
Thunderbird - Mail
Sunbird - Calendaring
OpenOffice.org - Office Apps
GIMP - Image editing
Xine - Media player
XMMS - MP3/OGG player
WINE - For those "must have" Windows apps/games
GAIM - IM
DOSBox - For old DOS games
OpenVPN - To remotely access our VNC desktops

Printing is handled by the centrally attached Epson Photo printer and the "thin clients" are laptops with wireless NICs, custom scripts and VNC clients.

It works very well for our needs. I would say that the only needs not met by this set up are things like scanning photos (since the server is headless in the basement, putting a scanner down there would be inconvenient) and 3D games that need fast screen performance. This would be better if I moved to 802.11G probably. (hehehe.. I've played Quake 3 using VNC over an SSH tunnel viw a DSL line. Too slow to be playable, but it works) My point with all of this? It's possible to do this sort of thing. The fact that a non-geek like me can set it up indicates that it can certainly be done by experienced developers. It's just that no one has tried hard enough or had a decent plan to do it. Realistically, if the bandwidth was available on a wireless device and it was no more than a display, kb, mouse and audio terminal for a really powerful backend box, this WOULD take off for the home user. Why should our desktops be married to one location? That's just stupid. Your desktop should be accesible everywhere with all functionality available. The only thing that needs to catch up is bandwidth.