I’ve had a busy couple of days with computers, what with 2 machines going done in the space of 24 hours to mysterious symptoms. That, coupled with an interesting interview on The Linux Link Tech Show stimulated a couple of thoughts in my mind.
10 years ago this was one of the buzzwords that got a lot of airtime, particular in companies, but also used in PC sales shops. The idea was that if you spent extra money now you would get a machine that would last for longer. I had a think about this over the last couple of days, and here is the conclusion I have come to:
Future Proofing is a Myth
- Regardless of how powerful your machine is today, it will be out of date in 12 months time
- Regardless of how powerful your machine is today, the next software release (from Redmond (when it eventually arrives)) will suck the life out of it
- No matter what you buy, the lifespan of the components is unlikely to increase much – the chances are that _something_ is going to fail
Now – maybe I am pushing a point here, but if you _really_ want to ‘futureproof’ your next purchase, make sure that it (and by extension you) can run Linux. Why? Because a) Linux is the future and b) Linux will always run better on old machines than Windows. I am still running the PC that I bought in 1999 (well – the new motherboard and CPU that was required when I fried the old one – PII 300) as a firewall, and it does the job admirably. I’m about to install a file server, which will be in another PII box. The cost? Â£0. Because apparently old boxes are worth nothing…..
We all know that software and hardware are different, don’t we. And yet so often I see people acting in a completely opposite way – software failures, or even just slow and unstable software, is regarded as a hardware problem; either it’s broken or it’s ‘old’ and needs replaced. How frequent are real hardware failures in computing? I reckon that it’s probably a lot less frequent that we would first imagine. Yes, hardware fails, and yes, when it happens it’s bad news. But we have become conditioned to attribute certain things to hardware, when, in fact, they are software.
This is just one example of the conditioning that we have received when we have entered into the “Wintel” cycle. Here are some more:
- Old = useless (as mentioned above). Even more ‘destructive’ is the belief that new software cannot run on old machines. While this may be the case with certain kinds of applications, particularly games, there are many cases where this is not really the case; when I put Linux on an “old” machine, it is not “old” Linux I am using, but rather the most up-to-date available.
- Rebooting is a necessity. Whenever a (Windows) user experiences an issue, they attempt a reboot to resolve the issue. It’s become ingrained in our thinking and culture. <insert ironic laughter from Mac and Linux users>
- Linux will never be better than Windows. Sometimes, when I talk to people, there is this obvious belief that Linux will never be better than Windows, because it is made by “hobbyists”. But there are 2 questions that I have about this now:
- Is Linux not already better than Windows (at least for some things)?
- Now that/when*Delete as applicable Linux is better than Windows, how will Windows ever catch up? It’s not like they aren’t pouring huge amounts of resource into it, but the pace of development is still so much quicker for Linux because of the sheer economies of scale when it comes to developers
- Expensive = Good, Most Expensive = Best. Not true – so not true.
Laptops and Desktops
People like laptops. People think that laptops are cool (which is true). But, and let’s be very clear about this, laptops are not, and never will be, desktops. While there are some very powerful laptops out there, the fact remains that there are constraints on laptops that will never go away – they need to be small and light, and they can’t get too hot. As a consequence, laptops will:
- Never be as fast as the best available desktop
- Never be as adaptable or upgradable as desktops
- Will always be more expensive than an equivalent desktop
To keep a balance, I must admit that laptops are a) very convenient, and b) look cooler, than desktops. And I still really want to get hold of an Oqo……
So what is my point here?
I guess my main point is that, while I like having a laptop for my work, I’m not sure I’m ever going to replace my desktop machine with one. And, more importantly, I think that it is important to recognise that if you are wanting to run intensive applications, you’re probably better off with a desktop, unless you really need the portability. Ideally, you should have both I think that we are going to begin to see a decline in laptops, as more people move to palmtops, especially now that palmtops are becoming more and more powerful, and able to offer all the kinds of applications that you would want to use on the move. Make sure you make the right choice when you decide. Which leads me nicely into:
mrBen’s guide to buying a computer
Here are some of my top tips to buying a computer:
- Know what you need, not necessarily what you want. Be realistic – it’s all very well paying an extra few hundred pounds for that fancy graphics card, but then never playing games.
- Don’t fool yourself into thinking that you will do new things with your new computer – buy a computer for what you do now. (If you are buying a computer for the first time, then look at the most common usages for computers – word processing, internet (email, web, chat) and games)
- Know what your negotiables are, and what are your must-haves. Again, be realistic.
- Set a top figure, and work backwards from there – as a rule of thumb I usually would set a top figure, and then knock a chunk of money off for warranty and memory, and then look for the range on offer for that price.
- Always buy more memory, you will never regret it, and the default amount is never enough. If buying a laptop, always buy extended warranty – it’ll probably set you back >Â£100, but it’ll be worth every penny.
- Unless you _really_ need Microsoft Office (and 95% of the world don’t), don’t buy a package with it, as you will pay. OpenOffice.org will more than likely fill your needs. Better to have to buy software later that you need, than buy software that you never use.
- Stick to the minimal hardware where possible – much of it is cheap kit, and, let’s face it, scanners are a lot like toastie-makers – endlessly fun for about 2 weeks……
- Avoid high-street retailers – costs will be high. But you are likely to get better warranty and more security from a brand name. Especially relevant for laptops.
Oh – and you may as well check whether or not it’s Linux compatible, a bit of futureproofing never hurt