mrflibble.net

Very, very excited about this new project.  Vindaloo will soon replace mrflibble.net as my preferred online location for photos, music demos, flibbtorials and flibbtube videos.

Check out www.vindaloomedia.com - pre registrations are now open.

And yes, i'm fully aware of the irony in light of my 'Corruption of Content' blog.  Just, just be quiet.

You're probably wondering why a guy who works in the industry and has his own website whinges about the internet so much.  My first question to you then would be, what are you doing here?  Only people i know personally know about this site, so they already know the answer to that.

Nevertheless, for the disappointed google searcher who actually wanted to find information about the Red Dwarf character Mr Flibble and has soldiered on through the site, here it is:

It's not that i don't like the internet.  I do, i use it every day and think the idea and the technology are fantastic.  What i hate, and hate is not nearly strong enough a word, is the internet hype.  It's not nearly as good as the average person thinks, or more to the point, as good as the media tries to make the average person think.  And it doesn't just come from the ignorant mass media.  So much hype actually comes from the industry itself, and the reality is the system can't hold a candle to it's hype.

I guess i've somewhat bought into that hype, and am daily disappointed by how limited the internet actually is.

Here's some examples.  "Snail mail" as it is so degradingly called, grew, yes that's right, GREW in 05/06 by 0.5% or 28 million letters.  So how many letters do you need per year in order for 0.5% to be a growth of 28 million?  5.6 billion.  Five point six BILLION, and this is Auspost we're talking about, not U.S. figures.  What?  You mean dispite the wonders of email, regular mail is still delivering 5.6 billion letters per year?  Yes.  Why?  Because no one in their right mind would actually rely on email to send important information.  Legal documents, bills, even advertising material.  If you want your message to actually get to the recipient, the best way is to pop (not pop3) it in a mail box.  And email is the best, most relied on part of the internet.

Retail sales.  Another area where the internet should be booming.  An area i think is fantastic and i love using.  I hate shopping, having to drive, find a park, walk around busy shopping centres.  Much easier to go straight to an online shop, click, click, click, enter credit card details, confirm.  Yet the Q1 retail sales figures for the U.S. showed that only 2.3% came from online sales.  Yes it was nearly $20b worth, so that does sound impressive on it's own, but not when you consider that the total was nearly 1 trillion dollars - and especially not when you consider how much google is worth, since google's only real value (and by value i mean ability to generate real dollars) is in google ads and leading people to make online sales through organic searches.  2% is pathetic.  20% would be worth some hype.  80% would be worth the hype the internet gets, and google's worth.

And here's the kicker.  Only 16.7% of the worlds population actually use the internet.  North America is nearly 70%, accounting for 21% of the total useage, while Oceania is 54%, but only makes up 1.7% of the total useage.  Basically the figures say this:  Most people don't use the internet.  Now of course these figures are always growing.  But as they stand, it's not a pretty picture.

What you don't hear about, and where i think the internet's true value lies, is internal business systems, intranets.  The ability for a group of people to have centralised information that is relevant to them, that they can access from anywhere in the world.  Internet useage statistics are irrelevant, because all you need to care about is if everyone in your group has access.  That's powerful, that's useful, and that's worth creating hype about.

Oh, and gaming.  Online gaming is awesome.  Worth any hype it gets, worth every dollar it generates.

So all in all i'm just agitated by misdirected internet hype, and every time i hear about google, youtube, myspace, second life or web smegging 2.0, i get annoyed, annoyed enough to write a blog that i doubt anyone will ever read.

Hehe, smegging.  That'll make things even worse for the google searcher trying to find the genuine Mr Flibble.  So will that.

Yes i know, my titles are very extreme, and the articles themselves are far more tame and reasonable.  I guess this title should be, "Google searches are shallow".

It's just that so many times i've searched for some information and haven't been able to find it.  Most of the time i do find what i'm looking for, but there has been many many times where i haven't.  And tonight i realised why that is.  Most of my searches are shallow.

A shallow search is a single dimensional, single query.  "Jessica Simpson".  If you search for that, you'll find a lot of information, images, videos, audio, pretty much all the information you could need (need? perhaps want) on Jess.

But that information is useless and not going to benefit me in any way.  Of course the search itself is useless, so the internet/search engine can't be blamed for that.

So lets try a search that is going to benefit me.  I have a Gibson Epiphone Les Paul and i need to buy new strings for it.  I want to know which brand of strings are going to give me the best sound.  HA!  Good luck searching for that.

Sure, you'll find plenty of results for the guitar, maybe some results for the strings, but a match of the two, including the best sound produced?  In fact, i wouldn't even know how to phrase the search.  I'd get millions of results, could click on all the first 10 hits, not find what i'm looking for and then do the unthinkable, the UNTHINKABLE... pick up the phone and call a music store.

I'd just like at this point to stop and clarify that i don't search for Jessica Simpson.  I don't know what made me use that as an example, right after stating that most of my searches are shallow.  I probably picked her because she's one of the top google search phrases.  Most of my searches are for PHP or javascript functions.  They too are shallow searches.  Ok, back to our main program...

In this case the brilliance of the internet is reduced to an online phone directory, so that i can find a music store to call and they can tell me what i need to know.  The information super highway become the information super directory.  There's so many areas where it doesn't cut the mustard, i've mentioned music, but there's legal, financial, automotive, health... actually, about the only areas it does cover quite well are the internet, software and hardware.  Perhaps i should have called this article, "The Internet is still just for Geeks".

I just did a google search for 'mr flibble' and sat there pressing next on the page numbers til i found mrflibble.net, 3rd from the bottom on page 22.

I'm taking over the world people!

A new internet.

Bring it on!

So this is how i dealt with the problem of weeds.

It's quite simple, just take the top 100 matches if there are less than 1000 matches, take the top 10% of matches if there are between 1000 and 10,000 matches, and take the top 1000 matches if there are more than 10,000 matches.

The system just ranks the matches from most matches to least matches, then cuts of everyone below 100 - 1000 depending on the total number of matches.

I think this will prove to be quite effective.  There will always be enough data to make sure the average is a true representation of the average rating among people who think like you, and the accuracy will get better and better as the number of total matches increases.  So the more users and the more ratings, the better for all.

About 3 years ago a friend of mine came up with an idea for a movie review site where the system would learn your tastes based on your reviews and then be able to predict whether you would like a movie or not.

Back then i had the programming skills of a dead camel, but last night i built it - in 3 queries.

It works like this.  If i review the Matrix and give it 8 / 10, Joe Bloggs gives it 7 / 10 and John Smith gives it 2 / 10, then Joe is considered to 'think like me' and john is considered to 'not think like me'.  It uses a tolerance of 1, so anyone who reviews with a score of 7, 8 or 9 are considered to think like me.

Then when it comes to a new movie, say Transformers, I go to the site and see what people who think like me thought.  In other words, what rating did Joe give it.  If Joe gave it 2 / 10, then there's a good chance that i won't like it either, since we think alike.

However, Joe is considered to think like me because of 1 match.  When i review another movie, say, Titanic, i give it 1 / 10 (only because 0 is not an option).  Joe gives it 10 / 10, he saw it 7 times at the cinema, and owns the extended addition (kill me) DVD.  So what the system does in this case is say, well, Joe thinks like me, and gave Transformers 8 / 10, so i'll likely enjoy it, however, the accuracy of that prediction is only 50% because Joe is only 50% like me.

So it does the above for all people like me and comes out with an average rating.  People who think like me on average think 7 / 10.  Obviously the more people use the system, the better it gets at predicting what you'll like, because it can weed out the anomalies, like people who are into horror movies and romantic comedies, or people who like hollywood action and french arthouse.

But, the weeding bit is the bit i haven't done yet, and am not sure how to go about.  At the moment it will match everyone who thinks like me.  This could mean that John Smith is matched to me because 1 out of 100 movies, he thought the same as me.  This will mean that the prediction becomes useless.  People who think like you thought 8 / 10 (1% accuracy).

Over time, with enough data, it will come out with around 50% accuracy which is not good enough.  I've thought of 2 ways to fix it.  The first is take the top, say, 100 matches.  This just means ranking people who think like me based on the number of times we agreed, then rank them from most agreements to least agreements, and cut off everyone past 100.  The problem with this approach is your not working with as much data as you could.

The second option is to choose an accuracy value for each match.  So, if Joe doesn't agree with me more than 75% of the time, then he doesn't get counted as someone who agrees with me.  The problem with this approach is choosing the right cut off value.  50%, 75%, 90%?  Who knows what would be the best.  Obviously the higher you go, the higher the accuracy of the prediction, but the less of a pool for the average.  For example, 2 people might think like me 90% of the time, one gives it 1 / 10, the other gives it 10 / 10.  That leaves me with an average of 5 / 10 with an accuracy greater than 90%.  It doesn't help me, because i still don't really know if i'll enjoy the movie or not.

So, this is where you come in.  Firstly, i've implemented the rating system on this blog, so you can rate the posts and start to see how well it predicts for you.  It's a really bad test case (movies, music, books etc would be much better) but it will give me more data to play with.  Secondly, you can post comments about what direction it should go from here to get the predictions more accurate.  Maybe there's an option 3 i haven't thought of.

Because the rating system needs to know who you are, you need to login, but unfortunately the signup process is busted at the moment.  I can create a login for you, so send me an email and i'll set you up with an account and you can start reviewing.  I've also got to come up with a domain, something like www.yourmoviereview.com (or .net since .com is gone).

People who think like you think... Mr Flibble is AWESOME!

While arguing with my work colleagues about the problem of rubbish content on in the internet, the solution presented itself.

You don't need a regulatory body censoring the content, all you need is to:

1)  Remove anonymity
2)  Remove free publishing

See, the reason people get away with publishing utter rubbish is because:

1)  We don't know who they are and
2)  It doesn't cost them anything

It's very simple.  Make every piece of content traceable to the author.  Address, email, phone.  Then, charge them for publishing content.  If youtube charged even $1 per video, you'd cut out 99% of the rubbish right there, the lack of anonymity would take care of the other 1%.  Does a corporate website care if the content can be traced to the author?  No, they want you to know who they are.  Do they care if it costs them $1?  No, they've already paid thousands to get their website up and running.

Ok, now i just need to start my campaign to youtube, flickr, blogger... and get them to charge their users and publish personal contact details.  Should be implemented by next week.

A simple question, do we really need pictures of teenagers puking after drinking too much, or videos of dogs humping their masters legs on the internet?  Obviously some people think we do, and therefore the question becomes, do we really want these kinds of people being able to decide what goes on the internet?

I think the current state of the internet could be described as a successful experiment.  We can sit back and say, “Ok, so it works, now let’s start again and do it properly”.  What I would love to see, and I doubt I ever will, is a new internet, one where content is controlled.  See, in the real world, if you have written something, you take it to a publisher to get it published.  If you’ve filmed something, you take it to a TV network to get it aired.  Yes, I know, that’s the ‘beauty’ of the internet, it allows anyone to be heard, not just those with talent, or money.  True, but in reality, ‘being heard’ means uploading videos of your dog humping your leg.

What’s the solution.  I think, another internet.  A different protocol, a different port.  Same idea, but done… much, much better.  Google have started the process of separation between complete rubbish and authoritative content.  My wife sells education material and google have published a paper encyclopaedia for children.  In it are details of a special google search page you can go to, and they tell you the key words to use for some information you may be looking at.  The results that come back have been verified as authoritative and safe for children.  See, even google realise that a child searching for “fun toys” is not necessarily going to return them what they are looking for.

Safe search, net nanny, content filtering.  All these things are bandaid solutions to a serious problem.  A problem that is caused by allowing anybody, whether billion dollar corporation or teen with webcam, to publish content on the internet without some regulatory body controlling it.

And it’s not just keeping it safe for kids.  I’ve found that often the answer to my google search is contained in the google ads, the organic search results are full of rubbish.  I don’t care about somebody’s opinion of what I’m searching for, I want the authoritative information.  I want information from people who know what they’re talking about, not some 15 year old who thinks it’s “totally dope”.

But who will control it?  That I don’t have an answer for.  There is already some level of control with domain names.  Joe Bloggs can’t register a .gov or a .edu.  In fact, he even has to have a legitimate company to register a .com.au.  I’m not sure how it could be done, but I think it has to be.  Imagine if in 50 years time the only form of publishing is the internet.  TV, books, movies, magazines, newspapers – all online and only online.

Sure, keep the current internet, let that be the place for ‘personal expression’, and create a new internet, one where you know that the content has been stamped with a seal of approval, where the content is authoritative.  Do we really need blogs about the corruption of content clogging the internet?  It think not.

I would just like to start out by saying that the idea of ‘Web 2.0’ is one of the stupidest things I’ve heard of in a long time.  Changes to the web are, and have always been, incremental.  The web is not some piece of software that gets released every few years, so the idea of assigning a version number makes absolutely no sense.

Besides which, the name itself is completely post modern in nature.  Web 2.0 means whatever you want it to mean.  I’ve read that Web 2.0 is AJAX, XML, User Driven Content or New Media and I’ve also read that Web 2.0 is not AJAX, XML, User driven content or New Media.  So what is it?  It’s nothing.  It’s the web.  AJAX, XML, User Driven Content and what makes up New Media have been around for a long time, years before the term ‘Web 2.0’ was ever thought up.

But who cares?  So what if someone wants to whack a label on the internet.  It’s just one of those buzz words for marketing people to get excited about.  And I say, let them.  What this blog (part of the Web 2.0 phenomenon!) is about is the delusion of ‘freedom’ within user driven content.  I’m so tired of reading articles of how users are changing the shape of the internet, it’s all about the users, user content, user opinions, user relationships, user, users, users.  Unfortunately this is a delusion.  The web is not about users, it’s not about content, it’s about what it’s been about since the first banner ad was ever created, making $$$.  And big dollars at that.  The biggest sources of user driven content are owned by multi billion dollar companies.  In fact, read the terms of agreement properly and you’ll find that the content is owned by multi billion dollar companies, your content.  Sure, you might write it or upload it, but you don’t own it, or if you currently still do, I doubt it will be long before you don’t.  Most people don’t realise that if they’re using some kind of web based email client, they don’t even own the content of their own emails.

What I guess I’m getting at is that there’s a fairy tale of a wonderful land called the internet, where people are free to express themselves and find out about others.  The reality is, we are pawns in a multi billion dollar game.  We, the users, do the work of populating these websites with content.  We are not paid for our efforts.  We hand over the ownership of our thoughts, photos, videos – our very personal lives and in return, we get to feel all warm inside because we’ve helped make obscenely wealthy companies even wealthier.  Sorry, that doesn’t do it for me.  flibb-tunes, flibbtube, flibbkr and flibblog is the place for my content, where i still own and control it.

Wow, my first socialist rant.  Boo the commy!