zlog 2.

online web periodical.

path: home / archives / 2004 / 02

Business as Usual
> written 29/02/04. 1 comment.

Some of you may have noticed that I am in the process of writing my own weblog content management system. I'm slowly, but surely, adding in various levels of functionality into my weblog via its service-orientated infrastructure. Now that I'm nearing the end of providing the bare requirements (posts, feeds, comments, templates...), I'm thinking hard about where I'd like to go with it next. Two areas I'm definitely going to be looking at are the design, and the methods of providing the said essentials. For example, generally I'm not the greatest fan of comments. Mostly my posts aren't really up for discussion, hence comments mostly are redundant. Don't get me wrong, I love to hear from you but I believe that with comment spam on the rise and random Google visitors stumbling across zlog the quality of comments is bound to suffer. Again, somethings just make better emails.

Another area I'm not so happy with is the method of posting to zlog. Currently I'm writing posts up in TextEdit, copying and pasting to phpMyAdmin and inserting the other fields by hand. It lacks finesse but it works. I'm about three-quarters of the way through implementing the MetaWeblog API service for zlog so the above process should be altered to posting via NetNewsWire. Still it isn't perfect. I want something like Radio Userland's auto-synchronised localhost posting system. You post locally, and via XML-RPC the post is "upstreamed" to the remote Userland servers. As this is a transparent process, it never gets in your way; and that's the beauty of it. This method also has the advantage of having a localised clone of the site which is good in a number of respects. For example when I'm working off-line -- something I do regularly -- I have a full copy of the site I can work on and when I reconnect it automatically synchronises (having a laptop without wi-fi is dumb).

Currently I'm posting to del.icio.us around three times a day. Mostly my links are fresh (previously un-posted to del.icio.us) so I'd like to think I'm adding something to the community over there. Now, I'd like to incorporate these links into zlog. Kottke's "a post is a post" method is out of the question as I believe it just fragments the site. On the other hand, in the current design there is simply no room for a sidebar to display the links. I'm thinking of doing a Monday/Friday dump of the links into a single post. Obviously this process will be automated, probably via the del.icio.us API and cron.

Business as usual then.

Simple Semantics
> written 22/02/04. no comments.

I've been thinking a lot about the the Semantic Web ( "Semweb" ) recently. You know, as you do.

As the Semweb will be both useful and plausible in the near future I'd recommend reading up on it; that is, if riding on top of the wave of the technological advances is your thing. And for some people it's not.

Specifications mean nothing to end users -- all they want is raw results, usually in the form of easier and/or quicker task management. They'll want to know how they, personally, can benefit from this new Semweb thing. It's fine all those "web developer" types telling us that it's useful but what we really want to know is what new features does it add? Does it make the internet an easier tool to use? And so on. Of course the arrival of the Semweb doesn't just revolve around the web developers integrating Semweb-rich data into web pages. Applications which use this new enhanced data must be developed. One of the most obvious applications for a Semweb revamp would be Google. As (none other than) Paul Ford has written before;

"At its [the Semantic Web's] heart, it's just a way to describe things in a way that a computer can 'understand.' Of course, what's going on is not understanding, but logic...", Ford goes on to explain, "Using a markup language called RDF [...], you could put logical statements like these on the Internet, 'spiders' could collect them, and the statements could be searched, analyzed, and processed. What makes this different than regular search is that the statements can be combined. So if I find a statement on Jim's web site that says 'Jim is a friend of Paul' and someone does a search for Paul's friends, even if Paul's web site doesn't have a mention of Jim on it, we know Jim considers himself a friend of Paul."

This is what Google, Amazon and every-other commercial online venture has been waiting for. A chance to understand it's customers in greater detail, to be able to infer information about it's potential customers; this would in turn, be more profitable for them. Imagine a world were a smart "more like this page" link reigned supreme over the current, pseudo-dumb "similar pages" link, returned under each result from a Google query. The data for the "more like this page" feature would have been inferred from other documents on the same type/topic, of the same length and/or by the same author. As this would be much more likely to produce a more relevant "similar page" then just a simple text comparison. You thought the web was good? It's going to get better...

zlog Mailing List
> written 21/02/04. 5 comments.

Ever wanted to discuss important tech/web issues with the other zlogians?

Well, your wish is my command; presenting the zlog mailing list. Subscribe by emailing 'list-subscribe hat zlog dot co dot uk'.

Minimum requirements: a working email address and a brain.

Distributed Thinking
> written 19/02/04. 4 comments.

People often muse about the possibilities of distributed computing, a phrase that was coined long before the web had even started to take off.

The idea behind distributed computing is this, you have a problem; one that requires raw computing power (like finding prime numbers or simulating protein folding, for example). As building a supercomputer tends to cost a fair bit, people began to seek other methods of acquiring the vast amounts of computing power needed. The internet provided the mechanism. By its very nature the internet connected computers together, simultaneously. All that was needed was a way to make the computers work separately, together, on a single problem.

People viewed distributed computing over the internet as such a good idea because average computer users weren't even using the raw computing power that was available to them. Browsing the internet, checking email etc only used a small percentage of their raw power. For example, many users have screen savers that activate when the computer is deemed idle. Wouldn't it make much more sense to use the computer's processing power instead of displaying a pretty picture or a Matrix scroll?

All the above technology is old but it provides a context. Distributed thinking, while it has the same elemental methodologies, differs by the fact that it utilises raw human brain power being effectively used instead of idling away. The only real life examples of this I've seen recently were in the form of Wikis but with the availability of programs like SubEthaEdit, iStorm and others I don't know why we don't hear about distributed projects. Take this example -- writing a book. It takes time, it needs copy-writing and it requires checking (spelling, grammar, clarity/sense etc.). If several people were to collaborate (in real time just for kicks) wouldn't that solve some of the time aspects of a project. One person could write freely while another checks spelling and/or grammar in his/her wake.

I'm not sure really why it hasn't taken off. One can only assume that it just doesn't have a high enough profile and that people just aren't aware of the possibilities.

UK Apple Store
> written 17/02/04. 1 comment.

Apple Store Ginza

Three days ago, Apple announced they were planing a new European flagship store located in our very own London, England. The planned twenty-thousand square foot store is "expected to sell Apple's entire product range". This means no more anxious waits for ordered products from the online store.

If anyone has had doubts about Apple's financial situation, then this could dispel them, as the rent of the building alone is expected to cost one-point-five million pounds sterling each year.

It's been a long time coming.

The Periodical Returns
> written 09/02/04. 7 comments.

I'm back from my holidays (editor's note: his holidays. I've been working on this...). Did you miss me?

I've been written up from scratch (save a file or two) and am now one-hundred and ten percent better. Guaranteed or your money back. I'm more efficient, more extensible, much faster and I'd like to think I've lost a pound or two in the process (editor's note: I have if he hasn't...).

So what can you expect from the all new zlog? Quite a lot actually. It really does have "loads of great new features". Like RSS feeds for every page, a new design (well, --ish) and much more streamlined and functional code -- we got rid of the nasty Perl! Plus there are a few upcoming features that are a little hush-hush at the moment, but trust me, I think you'll like them.

Most of the other improvements are benefits for myself. The whole site is now governed by a single template; not a template for each page-type, Movable Type style. This will allow me to deploy site wide changes much more easily and I can only assume will entice me to do so more often. Stupid things like changing the order of navigation items or the wording of the copyright can now be done in five seconds instead of the usual five minute slog as I pace through Movable Type template after Movable Type template. Not that I'm bitter or anything.

I'm a little bare at the moment (and sans-syndicate-able) but as the weeks progress things should start slotting in to place. Literally.

Special thanks to Jesper, Daniel and Jez.

Social Boom, Software Bust
> written 01/02/04. no comments.

I'm suffering from a case of web-hysteria. Everywhere I turn I'm being informed that social-software is where it's at, but as far as I can see social-software is just another 'dot com boom'. What does it really provide us? At most it's an amusing toy, a gimmick -- and who will devote great amounts of time (and energy) to playing with a mere toy? Children. There is a fundamental flaw in social-software and that is the fact that there is always someone who will wreak havoc. Always someone who will bend the rules and always someone who will ruin it for everyone else. Imagine for a minute that you had to pay to access a social-software website. Wouldn't that alienate the ones who would chose to abuse the system? I can't see people paying to ruin something.