New CMS

Details

Created on Friday, 27 January 2012 16:05

I have recently moved almost all of the content from my old static portfolio and from my Wordpress blog to one content management system. This mainly took the form of copying and pasting, and as one would expect, was tedious at times. Except for a few casualties of opportunistic quality control, all of my old material is now in one place. It can be managed centrally, and at no cost, thanks to Joomla.

While Dreamwaver allowed someone like me with a passing knowledge of HTML and CSS to build a site from scratch with complete freedom as regards layout and structure, the propitiatory nature of the software and reliance on FTP to work between a remote server and my machine were less than fantastic. The psudo-dynamic site I built with the aid of Dreamweaver's templating may have looked better than my present one, but it was unsustainable and ever more time consuming as content multiplied.

I have found the learning curve to be remarkably gentle so far and will begin further customizing this site as my familiarity increases. I will also be sure to take advantage of extensions that are likely to emerge as more time passes since the release of Joomla 2.5.

'The PC I Shall Build' or 'Subjunctive Technological Frugality'

Details

Created on Thursday, 02 June 2011 00:00

At present and on aggregate, I have a negative sum of money to my name; quite a large one if you count my outstanding tuition and accommodation fees. I have, in the past, possessed positive sums of money and hope to repeat this feat; my life would be wretched indeed if this was never again the case. If I continue to live with my parents while doing the inevitable minimum wage menial job, I might be able to avoid the archetypal young artist’s hand-to-mouth existence and accrue some capital.

As I use my PC to access information and entertainment, to communicate with others and to work — I think it’s fair to say a good PC is a high priority for me. I’ve had more or less the same machine for five years now, and it’s served me pretty well, bar a few minor hardware failures and a pretty major one when I was first putting it together. I’ve decided to put myself in the mindset of being ‘in the market’ and have been playing catch-up with technological progress. Here’s how I see the market for desktop PC components:

It seems that Intel rule the roost for high performance CPUs, their quad core i7s capable of outperforming some of AMD’s hex core Phenoms. Clearly it’s not merely about the number of cores, but the way they’re designed. As I don’t see myself having that high a budget, I’d almost be tempted to invest in one of AMD’s cheaper hex cores as the price is very good for high-mid performance. However, when Intel’s Ivy bridge CPUs start hitting shelves next year, they could leave AMD lagging further behind. I am quite set on a motherboard with a 1155 CPU socket (possibly a 2011socket , but they’re likely to be out of my price range) as this could bridge the gap between the Sandy Bridge(32nm manufacture process) and Ivy Bridge(22nm process) CPUs. In order to cool my CPU, I’ll most likely ditch the stock cooler and invest in a reasonably priced air cooler like this Freezer 13 Pro.

In terms of motherboards, I’ve been rather parochial in sticking with ASUS, as I’ve used their hardware before and their good reputation precedes them. It’s likely that by the time I’m buying, motherboards with an intel Z77 chipset or later will be available, so there’s little point choosing a specific motherboard now. If I were to buy now, I’d be torn between the P8P67 DELUXE and P8Z68. The first has stood the test of time since its release and the second may have compatibility for Intel processors using both Sandy Bridge and Ivy Bridge microarchetecture. As it is, I’ll have to wait and see.

For graphics, Nvidia’s GTX 4/500 series seem to have a slight lead on AMD’s Radeon HD 6000 series. As with ASUS, I’ve used Nvidia’s products before and trust them. The GTX 480 seems like a good bet for powerful card for someone on a medium budget, but by the time I’m purchasing, the 580 might be superseded by another card and therefore, affordable. It is also possible that AMD or another manufacture develop something that snatches the crown from Nvidia in the next year or so.

When it comes to RAM, I’m not particularly concerned who the manufacturer is, as long as they’re approved by my motherboard maker. When I was looking at motherboards with a 1366 socket, I was pleased to discover that many had 6 RAM slots, making it possible to have 24GB of ram using 6 relatively inexpensive 4GB cards. All those cards would have set me back £234.74 in total. The 1155 socket motherboards I’ve seen only have 4 RAM slots and no apparent support for 8GB cards, meaning that I’d be limited to 16GB. In all honesty, this should be more than sufficient for many years to come unless one were building a small server. That much RAM could allow me to do quite awesome things with high definition 3D sculpting apps like Zbrush and memory intensive video editing programs like Adobe Premiere. With 16GB, I could probably work with both both at once!

Hard drives… I’ll probably go for something from the Seagate Barracuda range — high capacity and moderately fast. Solid state drives, in my opinion, have some way to go before they’re a cost effective storage medium. If I do find a SSD for £100-150 with around 100-200GB capacity, I might use it for my OS and program files, storing most of my data on an external hard drive.

As for a chassis and PSU: I think I will to go for something fairly pedestrian, like this:

This is just my personal opinion, but some of the more expensive and flamboyant chassis resemble sports cars and only seem to serve as over-designed ego extensions for gamers. Below are a few images of not very discrete chassis for your amusement:

If you're skimming through this, I must point out that I have no intention of buying these.

CPU: Intel Core i7 2600k 3.4GHz – £236.37 (quad core, a good overclocker, Ranked 10 by Passmark, losing to 9 of Intel’s most prohibitively expensive CPUs)
CPU cooling: Arctic Cooling Freezer 13 Pro £29.43 (Can withstand up to 300W of heat, max TDP(Thermal design power) of the i7 2600 is only 95W. This means that I can pop my overclocking cherry without fear of overheating my CPU)
MB: ASUS P8P67 DELUXE – £171.94 (good manufacturer, great reviews. Sadly, I’d miss out on Ivy Bridge)
GPU: Nvidia GeForce GTX 460 – £148.99 (1GB GDDR5, ranked third by PassMark, losing by a narrow margin to the slightly more expensive GTX 570 and by a significant margin to the the overpriced 580)
RAM: 2x A-Data AX3U1600GC4G9-2G 8GB (2x 4GB Kit) – (16GB of RAM in total would mean I could use it for many years without worrying about falling back on virtual memory too much when using demanding applications)
Storage: Seagate Barracuda 7XT – £109.90(2 TB, 7200rpm, SATA 6; specs pretty much say it all)
Chassis and PSU: Coolermaster Sileo 500 with Coolermaster 500W PSU – £69.58 (Clean, minimalistic chassis design, hopefully cool and quiet. As for the 500W PSU, well I’ll have to do a bit of research on overclocking to find out if that’s sufficient for my rig)
OS: Microsoft Windows 7 Home Premium – 64-bit – £68.15 (This price seems suspiciously low for the full version, but this is only a theoretical exercise, so I won’t lose any sleep over it. By the way, Windows 7 licensing is notoriously complicated )

I think that’s it, the whole rig would only set me back £931.02, which isn’t all that expensive. It could be my designated workhorse for gaming, rendering, transcoding and so on while I could get a cheapish laptop for university — when I eventually go back.

Survival Machines

Details

Created on Sunday, 01 May 2011 00:00

In my last post I said that I was going to put most of my energy into learning to code for Unity3D’s API. What’s the point of using games development software if I can’t at least code to some extent? For just under a week, I’ve spent most of the day trying to structure my ideas into algorithms, and those algorithm into UnityScript (a Unity-specific variation on JavaScript).

Instead of learning by making some generic game, I decided to pursue my casual interest in evolution simulation.

The survival machines(grey, robotic-looking bipeds) blindly move around the environment in a manner defined by several variables local to them. They will turn if they touch an obstacle such as a wall or block. They will be rewarded some resources if they touch the golden resource pickups. The energy they expend as they move drains their resources and they will die if they run out.

What I have at present is a partially random environment and simple actors. If a machine’s variables are suited to the environment, it will accumulate more resources than its competitors. I have yet to code any interaction between them or a means by which they can replicate their variables(genes) to offspring.

I plan to introduce a system of interaction where if one survival machine touches another they will both be called upon to act. Both machines will have a variable to determine the likelihood of either fighting or fleeing. The outcome will have a direct effect on both survival machine’s resources. I will probably use as prisoner’s dilemma type system like this:

• Aggression against a passive other is the most rewarding(costing one unit to carry out an attack but stealing two from the other).
• Passivity in the face of a passive other is the second most rewarding, neither gains or loses from this.
• Aggression against an aggressive other cost would each actor 1 resources.
• Finally, passivity, when faced with an aggressive other would lead to the greatest loss of 2 units.

Ideally, passivity would be the best action, but a rational actor would realise that they stood to lose at most 1 unit and gain at most 1 units if they were aggressive. If they were passive however, they would stand to gain at most, nothing, and lose at most 2 units.

The main assumption of the prisoner’s dilemma is that given an isolated, simple scenario; aggression, trickery or lying are more rational alternatives to passivity, cooperation and honesty. It gets more complex when there are other factors involved, or it is played out over many iterations.

Passivity would be very advantageous trait in a population dominated be other passive survival machines, but this trait could easily be exploited by a minority of aggressive machines. If the there was an explosion of aggressive machines, then they would be at a disadvantage due to the fact that they wasted 1 unit of resources every time they met. I’m not sure which would turn out to be the most stable mix of passivity and aggression overall. It might vary from simulation to simulation.

I am aware that for example, real predatory animals can recognise prey, and vice versa, making the situation far too complex to be represented using a prisoner’s dilemma type model. I think I will leave those kinds of complexities until I’m more experienced, if I retain an interest in this sort of thing.