Brief outline of project
=========================

  - User enters keywords used to get list of URLs
      generated by Google

  - URLs with direct link to document (i.e. URL.../page.html)
      are extracted from search results (maximum of 10)

  - URLs are 'visited' and code is extracted from each page

  - Chunks of text (i.e. content within <p>, <td>, <span> etc.),
      and image SRCs are extracted and stored into respective
      arrays. This will be done for each of the URLs extracted from
      the Google search.

  - Text and image arrays are used to create an SVG document that
      is essentially a collage of content from each of the 'visited'
      pages.

The idea is to create a piece of art that is generated entirely from sampled bits and pieces of the web. The final product generated would be an abstract representation of a collection of data put into a completely different context. To make things more interesting, I think it would be neat to play with the keywords entered by the user. For example, mixing up the order of the keywords would prioritize earlier words, or adding boolean keywords that would randomly make some words required whereas others would be excluded from the search. I think there are many ideas I'm trying to illuminate in this project. Among them are:

1) The web is a chaotic entity. There is essentially no order to the web outside of self-contained sites.

2) Search engines are entirely based on algorithms. There is no guarantee that the keywords searched for will return relevant content.

3) A reminder that the nature of the web as a whole is a completely dissociated body of information.

4) Art based on samples of samples of samples... The final rendered document will be based entirely on content that is loosely based on keywords, which will not necessarily have any relevance when put together. The creation of something new based on existing content. Call me lazy.

Anyways, just a quick brainstorm on some ideas that have been floating around in my head. I think I will use this as a basis for my project. If anyone has any thoughts or comments I would be glad to hear them out.

One of the nicest things about object-oriented programming is it's ability to abstract, or encapsulate complex structures. If done properly, this method of programming can be rather powerful as it allows for a flexible interface that is both easily used and understood. It often makes reading and writing code a more pleasant activity as in abstracting code it makes for a more semantic-oriented structure, where objects are in themselves complex structures but whose complexity remains hidden elsewhere. This is the beauty of OOP, but it is also a great challenge to do properly. How much control over an object is necessary? How many roles should a single class encapsulate?

The Internet is a great medium for experimenting with new languages, forms of presentation, and sharing content. One of the best things of the Internet is its capacity to share information globally and instantly. Since its inception, there have been many interesting ideas around how information should be organized and presented. But I fear we are facing some sort of language epidemic where there is too much choice between languages and many times these languages heavily overlap. For example, PHP and Ruby can easily do most of the same things when used for Internet-based applications. It seems to me that the choice really comes down to personal preference. But how long will these languages survive? Sure, some have been around for so long that they're practised because people have so much experience with them, but these tend to become overly cumbersome over time. Perl is one such language. It is interesting to contrast this with the evolution of HTML, which started as a bloated language but is now being streamlined and highly standardized... Different levels of languages... Branches and leaves...

+++ Some thoughts on open source +++

1) Open source does not necessarily mean free of cost. The core philosophy of open source lies in the idea that the source is freely distributable and must remain so if it is to remain as open source. Code written by other bodies may be modified and used in any fashion as long as the new code allows the same to be done to itself. This freedom allows people to create useful software and give or sell as they see fit. Monetary value is not the issue; freedom of use and distribution is.

2) What makes open source projects more effective than closed source ones is that it allows a dynamic pool of developers and users pore over code and features. This is advatageous because: a) debugging is a faster and more user-oriented process, b) programming insight is easily shared and implemented, resulting in more robust and usable code, and c) features are more user oriented and cater primarily to usability rather than functionality, which allows for greater flexibility in the final application.

3) Although open source projects can be made for sale, there is often little commercial interest. The community of open source developers seems to be one in which quality is more important than economic value.

+++ Comments on the benefits of projects using an open source model +++

Development of open source projects seems to have parallels with concepts such as emergence and chaos. In a closed source operation projects are worked on by a selected group of developers who define rigid guidelines as to how the project should progress. The idea is that such a project would create a robust and usable product - but this is not always the case it seems. Because the minds working out the various problems of a closed source project remain static (i.e. the same people work on different problems day in and out) the final product is often one that fits the developers' ideal application and not that of the end user. This may not be true for applications that have passed the test of time, like e-mail clients, for example (though still, most are still quite horrible!), but when it comes to new applications or the need for innovation on existing types of applications, it becomes increasingly important for the users and not just the developers to have a say into which features the application should integrate.

Because development in open source projects does not follow the rigid model used by closed source projects it would be fair to think that their lack of organization would eventually lead to messy and highly buggy software. The truth is quite the opposite--this is where one might consider pluggin ideas of chaos and emerging properties... The constant development, feedback, patch processes in an open source project allows for bugs to be eliminated as the application is in development. Like swarms, users of the early software are quick to identify the cracks and because they have interest in using the sofware, are often quick to bring the fault to the developers' attention. This is in contrast to a closed source project that relies on a set of 'professionals' to oversee quality assurance, who want to impose on the user (indirectly or not) how the software is used. So where closed source follows a linear process of development, open source seems to follow a more chaotic process but in the end is a better application because it is shaped by the ideas and feedback provided by those who use the software.

I was astonished to see that MIT is offering free courseware from thier web site, especially considering how much we undergrads have had to pay for them in the past. I wonder if the quality of the content is equally good, and what about sourcing and copyright infringement? Does this mean that all content needs to be original if it authors do not give the rights to make thier content available free of charge? Could this be the start of open source education or online library?

This is my first of hopefully many entries regarding my upcoming multimedia thesis project. I hope to have comments posted to my notes in hope that they will help me bring a stronger focus on the subjects I wish to examine. I will try to update these notes on a regular basis and I would greatly appreciate any commentators to do the same. This is of course experimental and may totally flop, however I am very interested in seeing how this will turn out. The format will be fairly informal, notes being organized on an entry/comment basis. I will try to keep my comments readable but I can guarantee a number of spontaneous notes that will be purely stream of conscious style. That being said, I will make an effort to revisit my notes in order to clarify the idea(s) I was trying to express. Comments are open to everyone so don't feel that you need permission to post, just try to keep any contribution on topic. This is a growing journal and will continue to expand over the next year until the project is complete. Below are some topics that I am considering to include in my thesis.

Topics I am interested include:

• open source
• complexity
• emergence
• chaos
• the Internet

Books I am currently reading for informational, inspirational and sourcing purposes include:

• Just for Fun, by Linus Torvalds and David Diamond
• The Cathedral & the Bazaar, by Eric S. Raymond
• The Hacker Ethic, by Pekka Himanen
• this list will inevitably expand...

What's with apps like e-mule sucking out all my bandwidth on this ADSL connection? Back when I had cable it was easy to have e-mule (or any other P2P soft for that matter) open and still have decent transfer speeds in apps demanding bandwidth. This is a 300kbps connection, but rarely does it ever feel like it. Maybe it's the USB modem/router they gave me but I really can't say for sure. Would a non-USB router/modem help this situation? Or is it simply because the connection is ADSL and not cable? What's with this?

quick and short while i can remember the conversation i had earlier with a friend...

time does not connote value, it is a myth that we often associate with the preciousness of an object, be it an artifact, text or idea. something's value is not determined by its ability to survive in cultural circles nor does anything's 'value' mean anything in particular. this is not to say that everything is 'equal' but simply that we should not equate popular appreciation and values to our own perspectives on something. there is something to be said about classical items but there is so much that has been overrated and underrated as well.

how can i argue that a weblog that follows the mundane steps of someone's day is worthless? i find that although weblogs do offer this opportunity they are in this form no different than a bad sitcom. why would i be interested in such things? blogs should be made up of things that can be discussed or at least read with some amount of thought going on. i couldn't care less if you had an egg and orange juice to eat this morning, why do people write such things? keep it private.

I've been reading a lot on network theory lately and it has come to my attention that power laws play a deciding role in the distribution of information within the Internet. Research has shown that the few sites are are linked to by many, similar to hubs with manny links going in the site's direction. This has been a source of fear for some as it seems to suggest that the once sought-of democratic web is not what we thought it would be. Of course, this is natural network phenomena and as such I don't think that there is anything that we can do to eliminate this behaviour. Would it be possible to work within these constraints and somehow manage to go beyond the inevitable hyper nodes in a network? How would we mange this? If this is a naturally occuring phenomenon of netoworks how is it possible to create a network without this occuring? Do we even need to fear these hubs? It's not that our access outside of these hubs has been cut off but that the less links you have to the rest of the network the smaller the chance you have of being seen. Would it be viable to pursue a strategy in which you attracted attention to yourself via these supernodes? How would you avoid being lost amongst an array of other people also attracting attention? At first thought it seems that the best way is to create some sort of stratified system that sorts content but I find that this defeats certain aspects of the Internet. If not categorization then what? To create democracy it would be necessary that everyone has an equally strong voice but how is this possible when dealing with millions of voices? Perhaps a middleperson or some sort of aggregator is the key. Three layers: user, aggregator, content. Of course this would mean that content should be in a highly structured format and that in itself is quite imposing I find. It would require resources that not everyone could afford. This is certainly a large question to be answered. I just wanted to play with it a little...

Bought a nice LaCie external USB hd today, beautiful design by GmbH Porsche. 40Gb on a USB 1.0 though (I still need a USB 2.0 PCMCIA...) could make a backup take quite a while. Well it's not like I have anything else to do ;) Once the backup is done I will be making the move to Mandrake 9.2. I'm hoping that my laptop devices will all function properly as I have read that laptops and Linux are not always the best of friends... I'm going to search for a Linux installation review on a Compaq 2190 (my model) and if I can't find one I'll just write one myself :) More on this as I make progress...

19/02/04

I'm sad to say that my laptop didn't completely agree with Mandrake, at least not out of the box. I would spend more time with it but I grow weary of not having an Internet connection available at home (my laptop is the only computer here and I could not find drivers for my Cayman 3342 USB modem). So it's back to Windows, I'll just have to live with it. It's not that there's anything terribly bad with xp: it allows me to perform the basic things I want to do, and then some. But I must say that if I had the choice I would certainly go with Linux. Oh well, maybe next time I'll buy with Linux in mind.

I have started a blog at http://jing.blog-city.com/ I will be focusing on doing all of my writing at this new site. I plan on creating a new one, however, which will deal primarily with programming and various other computer-related experiences.

Linux

• What are the Real Differences Between Distributions?
• Filesystem Hierarchy Standard

Interesting reads

• The Microsoft Killers
  Open source software has come of age, and open source working methods are spreading beyond computers
• A Quantum Leap in Cryptography

• Visionaries are using photons to develop data-security systems that may prove the ultimate defense against eavesdropping hackers

• Microsoft, monopolies and migraines: the role of monoculture
• The New Face of the Silicon Age

attempted to install suse 8.2 last night, failed miserably. despite the reviews i had many problems with the installer. not because i didn't know what i was doing but the installer simply didn't work as it should have. initially i couldn't get my network card working but that problem was quickly resolved. (note to self: dlink dfe530tx works using via-rhine.o module). the next problem came when trying to download the suse tree via ftp. the first server i took down didn't seem to have all the neede d files. no big deal, just use another ftp (note to self: take down ftp ip and path to suse files _before_ commencing install!). once i found a working server the files began to download at a slow, but steady rate. after 3% was completed the installed halted and seemed to go nowhere. it claimed that some files were not available from the server. my guess is that it was just busy. after troubleshooting for a few hours just to get a busy ftp server i gave up, frustrated from the slow process. i am now in the process of installing gentoo 1.4rc2 and it has gone remakably smooth up till now. the installer comes with extensive and easy to understand instructions for each step in the installation process. they give some good insight and examples throughout - very impressive i must say. made it easy for an amateur linux user. the install even comes with a copy of lynx so you can browse the net while installing if you need to troubleshoot or are just bored. i am very impressed so far with the ease and clarity of the installation of gentoo. conclusion on this to come after the installation is complete.