Mar. 17th, 2009
04:56 pm - Job finding out delay
I recently went for a Job interview in the oxfordshire area, they said they'd let me know early this week.
Just found out, I won't find out until next week. They thought I interviewed well, but are interviewing other people.
Mar. 5th, 2009
I don't like facebook or livejournal, in terms of the centralised technology they use. Keeping in touch with friends is good though, which I don't do often enough.
They do two useful things, store and allow us to share information and give us good tools to manipulate that information for presentation.
But we can share information with out a centralised server, P2P showed us this. We should also be able to build tools to manipulate the information.
So what I think is the way forward TM, is a program a bit like a ghetto version of dropbox (a program that allows you to automatically share files in a folder) that uses P2P for the file sharing. Combine this with a sandboxed web server/scripting language that can only access the files that have been shared.
An example, Alice would have a browser window that presented a application that was like facebook, but instead of sending messages to a centralised server saved messages/images to the shared folder. These would automatically start to be downloaded by Bobs P2P client, and the next time Bob started his friendship app he would see the new messages/image.
We would also want a way of saying we trust a server to post on our behalf, so that we could post/read from places we are not allowed to install programs on.
I'm glossing over a number of hurdles, the ability to add friends that are allowed to download your stuff in an easy and secure fashion is nontrivial
Feb. 10th, 2009
10:37 pm - Economy Stuff
I don't normally post stuff like this, but these two articles actually tell us things that I'm not seeing on the main stream media about what happened.
Dec. 30th, 2008
11:54 pm - Php and Ycombinators
Dec. 6th, 2008
Mar. 28th, 2008
01:34 pm - Science: The Discovery
I've been working at a crap job to get some money (I may even tell people what it is one day), so I have been limited in my online time. So apologies for not reading peoples LJs, communicating etc.
I have however got into the habit of reading a certain rationalist blog, part infuriating, part sublime ( I will not link to it as I post under my real name there). Recently one of the people that posts has been bemoaning the lack of interest in humans in the real world/science.
Now my own take on this is that science especially the learning of it is not a social activity. You don't really get a few mates round and discuss newtons laws. Where as my time playing computer games has shown that geeks are quite happy to spend some time chatting about the rules/stats and other minutia inherent in some game or other.
So is there anyway to make science a social game, so that it can be a shared interest, like WoW or sports?
So a game is needed. I'm imagining something like Conspiracy X. Where you are a group of scientist tasked with solving problems.
Perhaps sent to an alternate world/dimension so that the rules are similar but the constants are not, so that they would need to be rediscovered. An alternate world would probably be better as creating a consistant universe where they change things like the planck constant might be tricky.
To GM it you would need a variety of simulators (chemistry, physics) so that the scientists could run detailed experiments to try and uncover and fix the crisis or dilemma of the day, a long with trying to secure funding/materials (perhaps in more exciting means than writing to funding bodies*). These could vary from who-dunnits to
saving the world from meteorites or cracking enemy codes.
* I'm now imagining the pirate expansion, where you plunder the Eleven Seas to fund your tokamak experiments.
Yohoho and a bottle of liquid hydrogen.
Jan. 14th, 2008
I'll stop my one track blog for a while and give you some other interesting stuff I have found
This is an interesting video from a biologist to a group of hackers talking about how hacking biology will become cheaper and easier in the future.
The danger Biohacking is mucking around with dangerous viruses, or making bacteria that eats concrete quickly or other crazy arse things that can muck up the real world TM. With cheap DNA synthesis it should be fairly easy to create your nasty virus/bacteria of choice and let it loose. What was interesting and a bit naive on his part is asking whether there could be a culture of not working with dangerous crap in the biohacker community that emerges. For one, most hackers aren't freaking insane so they wouldn't play around with anything remotely dangerous to their own health/health of their loved ones. Secondly hackers aren't very good at policing themselves anyway, strange that for a bunch of introverted geeks. So no singular biohacking community will emerge, and probably no policing.
I reckon most biohackers will restrict themselves to mucking about with highly artificial stuff that can't exist outside lab conditions.
The danger will be fanatics, as ever, and they are not likely to be stopped by any biohacker code of conducts.
I have recently realised why I am fascinated by this and adaptive computing. I think they are both fire types of technology, immensely powerful with the chance to change society hugely, but with the power to burn and hurt as well. I'm half wondering why people aren't legislating against this. The future will be a wild ride, and I would rather be on the crest of the wave than in the tube, if you'll pardon my stretched metaphor.
Also my vote for coolest looking virus. Bacteriophage
Jan. 1st, 2008
12:16 am - I am an Anti-PC crusader
I have realised, in one of those sparks of inspiration we all sometimes get, that a portion of my life can be packaged in a humorous and somewhat marketable format. So here is my propagandised screed. Read with tongue somewhat in cheek, with perhaps a bit of it sticking out. Also some points are brushed over to create a more forceful message. But the underlying message is something I believe in.
PCs have transformed our lives and will continue to do so. Yet the way we interact with them hasn't changed much since the 80's when they first started to pervade the home and office. They still have the keyboard, monitor, mouse and they are controlled by arcane programs, that the average user does not have the time or inclination to get to understand. And with the way things are at the moment this lack of understanding means they cannot control these most flexible of computing tools. And so we get malware, botnets and spam.
The keyboard, the mouse and the monitor all serve to tie you in one place, in one position. They are admittedly the best solution we have at the moment, but they are also the reason why RSI exists so prevalently among computer workers today.
Computers are tools and tools are are meant to be used, not to stress, injure and constrain the user unnecessarily. So there needs to be a concerted effort to get rid of PCs and replace them with something better. Something that busy people can control how they act without having to understand their internals. Something that does not rely on a people sitting down, glued in one position, repetitively tapping away. Computers should be intuitive, controlled only by the users desires and virtually invisible.
The intial phases of changing the interface will be easy compared to changing the architecture. But to an extent advancing the interface relies on the architecture, it should be personal to the users desires.
This cannot not be done alone, we need to sift the hillside of potential computational devices to find the pebble that will start the landslide that sweeps away the ossified and constraining mode of computing we have at the moment.
Unfortunately we have to use PCs to defeat them, but we must not get lured into the easy trap of complacency that leads only to continuing frustration and pain
Remember love computing, hate the PC.
Dec. 27th, 2007
Christmas has been quiet but busy, due to my Father's continuing illness and having my Grandma down. My brother and girlfriend will be coming over for a bit before new year, which will be good.
Other things I have been interested in recently have been reprap (including being able to get a set of components for £300 in the near future from bitsfrombytes ) and potentially getting a new mobile phone, the Samsung e250. My basic requirement is phone + bluetooth and ability to work as a modem + java. A camera is an optional luxury. Anyone know of good phones that fit that kind of description apart from the e250?
Should really get on with applying for PhD places and random non-odious jobs.
01:05 am - Logo reborn faster better stronger
Who remembers Logo? Yep that thing that was used to draw shapes and teach basic programming back in secondary school (at least for me).
Well it turns out that it has been turned into a nice easy to use system for experimenting with multi agent systems. Now you are no longer restricted to 1 turtle you have as many turtles as you want, and they can be made to do many interesting things. The name of this super advanced space age logo, is netlogo. You can even get it in the third dimension!
I'm currently prototyping my system, working on the utility sharing and message passing systems. A little bit about it, I want to the system to have completely distributed memory, processing and bandwidth, so that means I'm working on a grid system where a patch (this is a term from netlogo for a bit of spatially distibuted ground) is a little processor and memory store and does a bit of processing then passes on a message to the patches around it to continue the processing.
The many problems with it are slowness and timing issues. Although the timing issues will be there to plague me in whatever system I do, netlogo has its own brand of timing issues.
I am dreading having to create a decent programming language (or possible multiple) to deal with making it self-programmable. Most programming languages try and abstract the machine away, yet as self-programming depends upon getting the right instruction to the right place in the distributed memory, you have to know where each bit of code is stored and set up lines of communication to get there. I am using simple Virtual Circuit communication so that programs can be moved around and duplicated without conflict, whilst keeping each node simple in terms of routing. It will be a good test bed for random and pseudo random initialisations until that point though.
Navigate: (Previous 10 Entries)