What I've discovered, though, is that my desktop PC, for standard development tasks, is astoundingly faster than my work laptop for just about everything.

Personally, though I really do want a new PowerMac G5 I can’t see myself investing much in desk-anchored computing anymore. Not since I got my first laptop, and later my first wireless ethernet card. What I can see myself doing, though, is maybe investing a little bit into a new PC for games, and maybe for a box with lots of storage and CPU power to stick in a closet somewhere and use via network.

<p>Sure, a dirt cheap box tied to the spot via a dozen cables should be able to smoke my lightweight personal computing device&#8230;  but what if I use that stationary box <strong>from remote</strong> with that lightweight device?  It&#8217;s client/server all over again, but this time I own both the server and the client.</p>

<p>See&#8230; that&#8217;s where I <strong>really</strong> think it&#8217;s at. :)</p>


Archived Comments

  • That's why I have my laptop, my desktop, and a Linux box behind the desk that I use as my server.
  • Whoa, now you're getting all ahead of me! :) Actually, that's what I pretty much have, too. And not to say that using the desktop from the laptop is yet as nice as sitting at the machine itself. It'd be nice if you could run Eclipse between the desktop and the laptop as a client/server app to get all the portability while offloading the memory hungry and CPU intense things automatically to your nice and swift game machine.
  • I think the main reason laptops are so slow is that they have very slow hard drives. With many Windows apps having obscenely large working sets, just fetching all that from the hard drive takes quite a while. That's why I have a thin and light laptop (Toshiba Portege 2000) for office productivity apps and email, but the heavy pulling I will do on a desktop system. I am saving up for that G5 :-)
  • .. and with Half-Life 2 coming out on Sept. 30th, you're going to want a new gaming rig.
  • Laptop speeds should be improving rather nicely now that Hitachi (ex-IBM division) is releasing 7200rpm hdd's for laptops (albeit 10k IDE drives are now on the market for desktops too); making that slow component at last up to par with the average desktop. Maxing out laptop RAM, helps a bit too. With mobile Athlon64's on the horizon, affordable portable 64bit computing isn't far away either. Apple will probably have a G5 (ppc970) iteration next year along similar lines. Meanwhile tadpole finally has a sparc(64)book that starts at $3k which is really something given their history. Too bad not many people care about Sun hardware with AMD and IBM really pushing things in performance and price fields of late. Laptops seriously need something to replace cardbus, as gigabit ethernet et al need much more bandwidth. That is why Powerbooks w/ gigE actually have the gigE controller placed in a rather nonstandard implementation manner. Newcard is probably still two years away though. Then again, the average desktop still only has standard 32bit 33Mhz PCI slots - so peripherals that demand extra speed haven't quite hit the mainstream yet I guess. PCI-X and PCI Express aren't really in demand outside of server applications these days - maybe they'll continue to be of the same class of performance that ECC & SCSI have become - necessary for servers, but not really important in the commodity PC world. Looking at the google approach, it may just be that the cheaper but OK approach is still a win over the more expensive, but better approach. Just like wired vs wireless; desktops will always have some advantages over laptops - particularly with regards to speed & expansion. But the situation continues to improve for laptops too, usually - a generation's pace behind. I don't think we're that far away from the point where people start to care more about the convenience of something rather than the reliability. Then again, that's kind of how the mass of consumers has always been, and so we get VCR's and TV's that are cheaper to replace than to repair. On the one hand I think that's great - because technology shouldn't be so expensive. On the other hand I feel pretty adamantly that it's a myopic view of technology that leads to excess and waste. It's great that people view computers increasingly as just another tool; but just like the world of tools - the cheap ones will rust and break quickly; adding to landfills - and a good one can last forever. Maybe only the craftsmen care - but if they are passionate about their profession, then I think it is something that they should care about.
  • Sorry, I'm a bit hardware obessive and thought of a couple of other things that are of interest in the laptop & desktop musings. One thing I've noticed increasingly in the workplace is the popularity of the PDA, I actually find it quite intriguing to see a computer (be it a laptop or desktop) set up, and right next to it - as though a pygmy imitation is some PDA in a cradle. All the synchronization software leaves me thoroughly dissatisfied, and tends to cause many headaches (whether due to speed over a serial UART, or synchronization being all stupid and giving preference to the wrong thing at the wrong time - average users change their mind of whether the desktop or PDA is up to date, and they only configure such preferences once). Anyway - enough of that - what seems really appealing, and not that far away is the notion of a PDAcore, that you plug into a desktop/laptop form factor. Things like oqo.com and http://www.antelopetech.com/ are friggin brilliant in this regard - too bad that since the year+ that they've been announced, I still see none for sale. That and the fact that while they're a very beefy PDA, they're an extremely poor desktop/laptop. Ideally the docking host system would let you address faster hdd's, more RAM, possibly even a faster CPU - while putting all of the core's hardware to sleep, only relying on the hdd as a base. Anyway, my iteration is a pipe dream, but I think even their direction is intriguing. Makes the client/server thing a little bit less necessary, and the PDA + Desktop thing downright foolish. I feel like a lot of these technologies are getting closer to where they need to be, but are still a long ways off. Another example I've seen recently is with cell phone & PDA combos. Quite intriguing, but they all suck. Some PDA's seem to be good PDA's and OK cell phones - but can't send email. Some can send email, but are crappy PDA's and mediocre cell phones. I could rant on and one about failed attempts at convergence (another good one to look at right now is PVR's, DVD burners and digital [satellite or cable] tuners - something that should all be rolled into one package, but even the attempts that go for two of the three tend to fail at one or both). Maybe it's the growing pains of nascent technology. Maybe it's held back by corporate interests (e.g. the PVR+burner). I tend to get the feeling that more often it's technology trying to solve problems that people have, but turning in the assignment 75% finished.