When Is A Good Time To Upgrade?
Andru Edwards writes "In an article which looks at the techie's mindset as it pertains to upgrading, Hector Martinez takes a deeper look at what makes us want to buy the latest gadgets. What are your options, and when should you actually just keep what you already have?"
To fight the urge to constantly be upgrading the last 6 years my rule has been I can upgrade motherboard and CPU every other year and on the off year I can get a new video card. Also, never spend more than $150 on a video card, or $200 on a motherboard and processor. Of course some upgrade require new RAM, so it isn't a perfect rule.
Other than that I get a new hard drive about every 3 years and I've had to get a new power supply in there somewhere. This has worked well for me.
Its much more impressive to me to see someone downgrade, albeit with new non-bloated software, than it is to see the 'latest and greatest' ricerbox being sliding off someones credit card ..
I dislike badly written software, (for example most MMORPG's have very poor 3D engines compared to most single player games - as an avid gamer this is a bug bear of mine) and I think it's been clear over the last 20 years that there are diminish returns in each upgrade cycle (Mac OS 6 through 9 were all faster to use for regular desktop bound tasks than Mac OS X for example, and it's not at all clear that the additional features are worth the disproportionate drop in speed - similar things can be said of the Windows platfom in comparing Windows XP to Windows 3.1).
I think it's fair to say that both Mac OS 7 and Windows 95 were faster for navigating directories, opening word processor documents and checking email on average hardware for the time than either Mac OS X or Windows XP are on significantly newer hardware. Linux on my 1.5 Ghz PowerBook is *staggeringly* fast compared to Mac OS X for example (so much so, that I'm very tempted to keep buying the hardware, because I like the feature set, but switch back to Debian). The very apparent lack of optimisation on Mac OS X is just staggering.
However...
I can honstly say that I find it much more impressive for me to see Doom 3, PlanetSide or Unreal 2 running on high end new hardware and it is to see that after months of hard work they have managed to port a rather mundane title to a 10 or 20 year old system. The price of not having bloated software is sometimes just that - a mesurable monetary cost which someone has to be willing to bear. Time, money and the very real resultant possiblity that if it can't be made quickly it can't be made at all (which in the case of some software, could be be bad for overall productivity).
I do think OS vendors have a lot to answer for - they are responsible for massive amounts of bloat (particularly Microsoft and Apple), but other than the period release of a new OS like XP or OS X, it's games more than anything that drive the cycle, and most developers are keen to do as much as they can to keep the performance as good as possible within reasonable limits.
I hear David Braben is working on a sequal to Elite, I'm sure it will be very tighly coded. I think perhaps it will be a long time coming though - if, dare I say it ever.
I do not think coding in assembly is very efficient, as it takes significantly longer to write complex software, which is why we have C (and other higher level languages). We should be picking the right battles I think. Perhaps by attacking poorly implimented and very inefficent high level languages and ensuring we have good compiler technology (and run time engines) to appropriatly optimise software to make use of the hardware it's running on.
Hardly any programs (certainly hardly any of the programs I use) take advantage of features like Altivec (on PPC G4's) or Hyperthreading (on Intel P4's) - having them do so by being appropriately written and having compilers that can do a good job of optimisation would be a big step in the right direction. It was the efforts of Motorola (donating code to GNU/FSF) and the work of companies like Red Hat and independant individuals that Altivec optimisation was added gcc for example - yet Apple rely on it, even ship it on CD/DVD with every Macintosh, as such many would think perhaps they should have been leading contributions to such a project. Which brings me to perhaps one of the best targets for bloatware critisim - vendors of commercial operating systems (i.e. Sun, Apple, Microsoft) - some of there software is frankly appaulingly slow given the hardware it runs on, and what they do in many ways sets the tone for the rest of the industry.
In Minnesota, garbage collectors won't take computer monitors in the trash. They won't take TV's either, and generally don't want things like microwaves and other computer hardware either. I don't know if the laws are different up here, or perhaps due to Arizona being a drier state they don't worry about the chemicals leeching out.
By the way, it costs about $15 up here to get a computer monitor recycled.
"I'd love to buy a computer that was manufactured in a not-environmentally-hostile manner. However, with our current technology, that is not possible."
I think things are getting a bit better. I work for a company that produces hardware. We've just done a complete redesign across the board to get rid of the parts that contain lead. We had to do this because of a law that will soon be going into effect in Europe but it will help everywhere.