RagePank SEO

Stopping Manual Spam on your Blog

Thu, 25 Jun 2009 00:00:00 +1200

I woke up this morning to a number of new comments on my blog. All of them utter crap from offshore SEO link builders.

I know they were posted by offshore link builders, because I happen to track referral URLs on all comments. These particular referral URLs all had search queries like web design "post comment" which is a pretty standard starting point for link builders looking for blogs to comment on.

Great post!

This got me thinking. When was the last time one of these link builders actually posted a useful comment? Their comments make it straight past the CAPTCHA because it's a human entering them, and because they don't usually contain spammy phrases such as 'WOW gold' or 'viagra' it's hard to algorithmically filter them too.

Does anyone else have problems with these 'low quality but not completely spam' posts? The ones where you have to stop and think for a bit before deleting?

Would it be so bad if I simply said no to all comments from visitors that arrived at the site from a search query such as web design "post comment"? Or any search query containing "add link", "submit url", "post comment" etc.

So, I thought I would put a bit of code together to put this theory into action.

Announcing...

So here is ManualSpamBlocker. It's a Wordpress plugin that closes comments for all posts if the visitor arrived via a dodgy looking search query. The link builder can still enjoy your quality blog posts, but they just won't be able to comment on them because they will think comments are closed.

Simple huh?

This is the first Wordpress plugin that I have released, so I would really welcome any feedback you might have on it. It's still to be considered Alpha quality code, so don't be rushing to install it on your high-traffic blog just yet. Once I'm happy with how it's looking, I'll submit it to the Wordpress plugin database and see what happens. Happy to give out SVN access to anyone who thinks they are able to contribute.

Source: Stopping Manual Spam on your Blog

POST is the new nofollow

Fri, 19 Jun 2009 00:00:00 +1200

Google has recently posted some trivia on the use of nofollowed links. I had to check the clock to make sure this wasn't one of their April fool's stunts. It's not, which is a bit scary.

Matt Cutts sums it up right here in an interview with Danny Sullivan.

Danny: So there’s 10 links on home page, 5 pages Nofollowed, where is that link juice going?
Matt: You can think of it as evaporating.

Wow, ok. Oh, by the way, Google has been crawling Javascript links too.

Danny: You guys have click-change, that now you are actually reading javascript, but you said "if you have paid links, you can use javascript and your good", now you're not good?

Matt: As Googlebot got smarter we started changing our advice on this. What we haven't mentioned is that elsewhere, even on the onclick, you can put a rel=nofollow on a link within javascript, you can do that if you want to be completely safe, I expect to see those stay safe.

Optimising your links

Right. So for those who have been busy sculpting their pagerank via the nofollow attribute (which is what Google designed it for) it now appears that we have been shooting ourselves in the foot. All our nofollowed links are actually working against us by reducing our PR, and our javascript links are being clicked on.

So - how do you link to something that is a bit average quality or you don't really vouch for it? Nofollow has now lost it's shine, and the javascript option is gone too.

Options

Which leaves precious few options. The obvious one is no links. I speculate there will be a plethora of Wordpress plugins released soon that give no link at all. In lieu of a link, you will need to copy-paste a plain-text URL into a browser window if you wish to visit the site. Not especially great for usability, but now one of the few ways to preserve your PR so you might understand webmasters going down this path.

The other option that springs to mind is a flash based link. This could be complex because Google's crawling of Flash sites is constantly improving, so you might need to embed your links into an external XML file or obfuscate them somehow. This approach would be an accessibility disaster, but I can definitely see webmasters going down this path too.

POST

The third option, and perhaps the best option I can think of is HTTP POST. Google doesn't follow through POST forms, because POSTing data can potentially change the contents of a website. POST forms often contain things like delete buttons, and you don't want spiders running around clicking those all over the place.

Google confirms this on their blog, illustrated by the following comments.

Gael Fraiteur said...

For good internet citizens (and according to the REST bible), the rule is that a robot should not invoke POST methods, because they may have side effects. GET methods are supposed to be programmed so that they have no side effect.

I hope Google respects these rules too.

Susan Moskwa said...
@Gael: As mentioned in the post, we'll only be retrieving GET forms, not POST.

OK, so how about we add a POST form with a single submit button. The form POSTs to redirect.php?r=www.domain.com which then does a 301 redirect to the actual link.

This isn't a link, so it doesn't need to be nofollowed.
It's POST, so Google will respectfully ignore it.
The user will end up where they want to be.
You can use CSS to disguise the submit button and make it look more link-like, perhaps even using javascript to edit the statusbar rollover information. Unfortunately, I don't think you can get the context menu for 'open in new tab' to show up.

I'm giving serious thought to making this change to the blogs I administer.

Nice one Google

Google has their head in the clouds if they think webmasters are going to ignore this change and simply go about their business of 'creating great content' and not worrying the fact that half their PR is heading down a black hole. I think the way websites link together is going to change, and expect usability/accessibility of websites to go backwards as a result.

Source: POST is the new nofollow

robots.txt vs sitemap.xml - Who will Win?

Thu, 18 Jun 2009 00:00:00 +1200

On-page SEO is largely about getting search engines to visit your site so that they can index it nicely. But it's also about stopping them from indexing things you don't want indexed.

So today when I came across some indexed content that shouldn't have been indexed, I was a bit surprised / shocked. You see, the content had been specifically blocked from spiders using 2 different methods - robots.txt and a noindex/nofollow meta tag.

However, upon further inspection, for one reason or another this content had lost it's noindex/nofollow meta tag and had somehow found it's way into the XML sitemap. Doh!

Game on

So the situation here is that robots.txt is saying "Block" and sitemap.xml is saying "allow". I verified that the content in question was in fact blocked using the tool on Google Webmaster Tools.

Sitemap.xml is the winner - the content gets indexed regardless.

Makes sense, I guess

I had thought that robots.txt was acting as a failsafe to make doubly-sure that the content doesn't get indexed, but this isn't really the purpose of robots.txt. Robots.txt is more designed to instruct crawlers that are blindly trolling through your content. It will stop a crawler from visiting a page if it happens to find a link to that page on it's travels.
However, when you write a sitemap.xml file, you are specifically inviting the bots to visit and index the page. It makes sense that they ignore the instructions given in robots.txt because they are serving a higher authority.

So, this does mean that extra care needs to be taken with the creation of your sitemap.xml file. If search engines are treating this as gospel, then make sure it's correct.

Source: robots.txt vs sitemap.xml - Who will Win?

Are you using Google screenshots illegally?

Mon, 13 Apr 2009 00:00:00 +1200

SEOs and web design agencies often use a screenshot of Google results to show off their wonderful rankings. I hadn't really considered the copyright implications of doing this, until I stumbled across the official page on how to use Google's brand.

Because most of us are too busy enjoying Gmail, Chrome and other cool freebies, it's easy to forget that Google are actually a multinational behemoth and have an army of legal-weasels dedicated to copyright infringement, EULAs, and all that yucky stuff.

I'm hardly surprised that such a page exists - of course it does - but the following piece I found interesting.

This use must be unaltered: You can't superimpose graphics, photos, or ad copy on the screenshot or change the look of the screen-captured image in any other way.

And this also...

Please note that using a screenshot of a Google search results page in connection with advertising your products or services (for instance, showing a top ranking on Google) is not considered instructive or illustrative, and therefore is not permitted.

Pretty clear. SEO and web design companies who use screenshots as 'advertising' is a clear-cut violation of these terms.

I know I have done this in the past, and I'd wager that there are some pretty big SEO companies out there using Google screenshots as advertising, or altered screenshots to illustrate a point clearer.

Source: Are you using Google screenshots illegally?

OSS Bar Camp, Dublin, 2009 - Review

Sat, 28 Mar 2009 00:00:00 +1300

I have just spent the day at the thoroughly enjoyable OSS Bar Camp in Dublin. It was a much needed dose of geeky goodness that I haven't done for a while, so well worth the effort.

I haven't been to a bar camp before, so wasn't totally sure what to expect, but the presenters were entertaining and I came out this evening knowing a few things that I didn't know this morning.

Luis de Bethencourt - Ubuntu

Luis is a core developer for Ubuntu, based in Dublin. I use Linux for my servers but haven't really used it for a desktop OS, though I did have a play with an Ubuntu boot CD and was suitably impressed. Luis stressed that Ubuntu is not so much a technology as it is a community, and that it's the quality of the people that make Ubuntu special. His keynote speech was a nice reminder of why open source is cool.

Jaime Hemmett - Git

I have had the pleasure of meeting Jaime at the 2 PHP meetups I have attended since being here, and since she is Australian, we have something in common (that is, everyone in Ireland thinks I'm Australian). Git, a version control system written by Linus Torvalds is something I had heard of but hadn't really had the time to investigate further. The room was basically filled with SVN users (myself included) and Jaime's objective was to get people to give Git a try, nothing more. SVN has been a godsend for the work I do, but it does tend to be a bit slow and unreliable. As I write this, I haven't been able to do a proper checkout of Jojo CMS because the JS Min SVN repository (which we link to) seems to be down. I'm not totally sure that Git is a better option for my purposes, but I'll be giving it a try.

David Coallier - Get Ready for web 3.0

Shit, I thought web 2.0 was all about buzzwords for the sake of buzzwords and AJAX for the sake of AJAX. I didn't know there was a web 3.0 until this morning, but I guess it makes sense.
Adccording to David, web 3.0 is all about standardising web APIs. David has contributed to the PHP core and is a developer for Echo Libre who where sponsoring the event (thanks for a great day team).

The whole discussion of APIs is really interesting. I use APIs on a daily basis, and understand the frustrations of having to learn something new every time I want to connect to a different web service.

Web 3.0 is apparantly all about standardising how we connect to web APIs, but I'm still having trouble visualising how this might work. The issue in my mind is this - how do you standardise data formats when the data is completely different across each industry? I have spent the last week or so connecting to different Bible APIs in order to display the bible on a website, and it's not trivial. Aside from any technical considerations, the solution we eventually chose has a number of legal restrictions on it, so I'm not sure how this might be standardised.

Stuart Langridge - Advanced Javascript

This talk was pure comedy. We looked at some new whizz-bang javascript features, things that make our lives easier and make our javascript faster. We were told that this was available NOW. Some of the developers in the audience were concerned about browser compatibility, but this isn't something that bothered Stuart - most of the features he showed us worked fine in Safari, Firefox, Opera, and all good modern browsers.

This is all well-and-good, but there is that small percentage of the market (80%) using Internet Explorer which doesn't support any of the new features he showed us. Hmmm. However, the good news is that a lot of the functionality can be replicated using a javascript library like jQuery.

The take-home information from this presentation: Keep using jQuery, it's bad-ass.

Paul Biggar - Optimizing PHP with phc

PHC, or phpcompiler is a compiler for PHP (yep). Paul is a developer for this project, and has a passion for compilers. He discussed some of the challenges in optimising PHP code, and left the audience feeling that PHP was actually a bit of a lost cause. "real" programmers often give PHP grief for being an untyped language, and say that it encourages sloppy coding, which I do agree with. But it's also simple enough to get something running quickly, and powerful enough to do whatever you need it to do, so PHP does deserve a bit of credit.

I enjoyed this talk, which was coming from a unique angle.

Alan Burke - Drupal

Drupal is similar in soooo many ways to Jojo CMS. Obviously Drupal has a much much larger community / budget, but the concepts behind the products are similar.
Personally, I have struggled with Drupal being a bit limited out of the box - if you want Drupal to be 'just how you like it', you need to install the right modules or do lots of customisation, which can be a bit daunting. However, there is no question that it's a powerful platform, and I guess it would have been nice to see some example Drupal sites in detail. There is a 2 day Drupal geek-fest in Galway next weekend, so I'll be attending this, and are looking forward to it.

The pub

And then we all headed off to the pub for a thoroughly enjoyable discussion of all things geek. I stayed a bit longer than I intended, and ended up missing my bus and having to faff around in Dublin for an hour before heading home. Wouldn't be the first time I have gotten lost in Dublin.

I'd like to extend a big thanks to Laura Czajkowski and Jaime Hemmett for organising the event, which was by all accounts a success. Open source is alive and well in Ireland.

Source: OSS Bar Camp, Dublin, 2009 - Review

SEO Counter - jQuery character counter

Fri, 20 Mar 2009 00:00:00 +1300

For the last several years, I have used a little javascript character counter to keep my page titles and meta descriptions to a certain length. When titles are less than 70 characters and meta descriptions are less than 155 characters, Google won't truncate them in search results, and this looks nice.

Today, I decided to bundle up a version of this script so others can use it in their CMS or website. This variation of the script is based on jQuery, and is pretty straightforward to install.

Why?

If your page title is longer than 70 characters, Google will shorten it when it is displayed in search results. This means visitors don't get to see your full message. I like my titles to include the keywords I'm targeting, but also have human appeal so they are clickable. If they get shortened by Google, they have slightly less human appeal in my opinion, and you have less control over how your listing is presented.

Above: Titles longer than 70 characters and meta descriptions longer than 155 characters will usually get shortened.

SEO Counter

SEO counter is a simple script that adds a counter to your text input / textarea fields, so you know how many characters you have left to work with. For titles, the counter starts at 70 characters and counts down to zero. This makes it easy to know when your title is likely to be truncated in the search results.

Example

You can see the script in action at the SEO Counter example page.

Download

Download seocounter.zip

Installation (quick version)

Installation is fairly straightforward.

Include jQuery in the <head> of your document.
Include seocounter.js in the <head> of your document, after jQuery.
Add class="seocounter_title" to any title text inputs on your page.
Add class="seocounter_meta" to any meta description textareas on your page.
Make sure your form fields have a name attribute set (this also needs to be unique).
Call SEO Counter using the following code in your <head>...

<script type="text/javascript">
$(document).ready(function(){seocounter();});
</script>

Installation (detailed version)

Include jQuery

Add the following code between the <head> and </head> tags of your page...

You can also use Google's copy of jQuery...

<script type="text/javascript" src="http://ajax.googleapis.com/ajax/libs/jquery/1.3.2/jquery.min.js"></script>

Note that if you are already running jQuery, you won't need to include it twice.

Include seocounter.js

Add the following code between the <head> and </head> tags of your page. This must be added after jQuery.

Add classes to your input boxes

You won't want the counter on all of your input fields, so you need to specify which fields where you want the counter. This is done by adding a new class to the input or textarea.

If you want a text input to behave like a title, add a class of "seocounter_title".

Make sure the field also has the name attribute specified, this is required by SEO Counter.

If you want a textarea input to behave like a meta description, add a class of "seocounter_meta".

<textarea name="meta_description" class="seocounter_meta" rows="4" cols="50"></textarea>

You can add counters that countdown from any number as well. Add a class of "seocounter_XX" where "XX" is the number of characters to countdown from.

<input type="text" name="foo" class="seocounter_50" />

<textarea name="bar" class="seocounter_200" rows="4" cols="50"></textarea>

Call SEO Counter

Add the following code between the <head> and </head> tags of your page. This must be added after jQuery, and after seocounter.js

<script type="text/javascript">

$(document).ready(function(){seocounter();});

</script>

This will cause the SEO Counter script to appear after the document has loaded. If you happen to add extra input elements to the page dynamically, you can call this function again to add the counters.

Title suffixes

I pretty much always use a suffix on my sites. For example, on RagePank.com every page title ends with " | RagePank SEO" which I like from a branding perspective. However, this means that this is 15 extra characters that needs to be included in the equation. In effect, my titles can't be longer than 55 characters (70 - 15 = 55), so the counter needs to account for this.

To workaround this, when you call SEO Counter, include the title suffix as an argument to the function. eg.

<script type="text/javascript">

$(document).ready(function(){seocounter(' | RagePank SEO');});

</script>

Easy.

License

You are free to use and modify this script free of charge, subject to the terms of the Lesser GPL license. If you like this script, please consider linking to http://www.ragepank.com

Testing

I have not done extensive testing on this script, however it is quite simple and I wouldn't expect to run into many problems. Feel free to get in touch if you find any bugs, or have any comments.

Source: SEO Counter - jQuery character counter

How to stir up your insurance company

Fri, 09 Jan 2009 00:00:00 +1300

When we bought our first house, about 5 years ago, we didn't have much cash and went for a 95% loan. Because of this, the bank made us get life insurance - which we didn't really want, but didn't have much choice. Now that we have kids, it's a worthwhile thing to have, but at the time is was yet another new expense that we had to deal with.

I'm overweight, and the insurance company wanted to do a full medical checkup before they would insure me. A doctor came over to my house and interrogated me, held me by the balls while I coughed, then took 15-20 goes at trying to find my blood vessels to take a blood sample. Literally - he just kept jabbing the needle into my arm, muttering about how hard it was to do. I was ready to kill him, then someone suggested I go to the local nurse for a blood sample - she got it right first time, bless her.

After all that, turns out I have really high cholesterol plus a whole lot of other things wrong that I didn't understand; such as uric acid and liver function - and the insurance company turbo-charged my premiums as a result. I have been unhappily paying them each month since.

So, now it's 2009, and over the past 15 months I have been busting ass doing lots of exercise and eating properly, and I'm 20kg (44lb) smaller as a result. It's not miracle weight loss stuff, but I'm pretty pleased with the result and definitely looking and feeling better. A few weeks back, we were selling stuff on TradeMe and noticed a SureSlim voucher for sale - these normally go for $1000, but we picked it up for $100 which is a bit of a bargain. I didn't really need SureSlim as I'm doing ok by myself, but I figured a bit of professional nutritional help wouldn't be wasted.

Sureslim required another blood test before we could started.

Blood test results

So I get the results back, and it looks good. Cholesterol is waaaay down (still higher than normal, but a massive improvement) and all the other things they check for are within normal limits.

So the first thing I do is phone the insurance company to sort those premiums out. I email them the blood test results, and tell them what I did to reduce the cholesterol. And this is where it gets interesting.

You see, I BCCed a copy to a competing insurance company, to make sure I wasn't getting done. Turns out, I am. Current insurance company are charging $97 per month, and the quote from competing company is about $30. Hmmm.

I'm currently in negotiations with the existing company - but in order to get them to look at my premiums, I need to complete a mountain of paperwork. I said I would print, scan and email the form (I don't own a fax) but it's 25 frikin pages - my scanner doesn't have a sheet feeder.
In order to sign up for competing company, I also need to complete a mountain of paperwork, so this isn't something I can avoid.

Here's the bit that surprises me though - My current provider actually wants my business. Normally with these faceless behemoths you tell the call-centre person that you are looking at going with another provider, and they couldn't be less interested. However in this case, I have had phone calls + emails from 4 different people trying to keep me happy. I managed to get their attention.

It seems that my case is moving further and further up the food chain, but nobody has any authority to actually do anything without the mountain of paperwork being returned. But it's nice to know they are trying their best.

What I think

Here's my take on all this - it's not about price, it's about respect. They have been enjoying my triple-charged premium for years, and now it's my turn. I have worked really hard to lose the weight that I have, and if the insurance company wants me to jump through further hoops, fuck them.

Part of the weight-loss process is getting a new outlook on life - you feel better, you look better, and you are confident again. The healthy food fuels you with more energy and zest than ever before, and dealing with this shit is now easy. When it comes to negotiating life insurance, you are no longer on the back foot and can start demanding what you want.

I am in control - and when Sureslim do another blood test in 6 weeks time, I'm likely to have even more leverage to work with. I'm contemplating my next move - this is one monthly bill I'm really excited about lightening up.

Source: How to stir up your insurance company

Ireland beckons...

Fri, 02 Jan 2009 00:00:00 +1300

We have recently decided to move the family to Ireland for a while, which I'm hoping is going to be a fairly exciting change.

We have tickets, and it's all happening quite soon.

The basic idea behind the move is pretty simple:

The kids haven't started school yet.
We don't currently own a home (having sold earlier in 2008), and the NZ housing market doesn't seem to be improving anytime soon.
My wife isn't tied to a flash corporate job, she is currently doing the housewife thing.
I'm a freelance website developer and internet marketer - if there was ever a job that could be done remotely, this is it.

So given all these things, why not?

Ireland is an obvious choice of locations for us - my wife Kirstin was born there and holds an Irish passport, which makes everything a whole lot easier. No visas, work permits, paperwork etc to deal with. Aside from that, a good Irish accent always puts a smile on my face, and our experiences of Irish hospitality so far have been excellent.

We leave New Zealand on the 21st February 2009 and will probably live in Dublin, but that isn't set in stone. We aren't sure exactly how long we will be gone for, but it's likely to be 6-9 months in Ireland, plus a bit of time in France and possibly Germany on the way back if finances allow.

How will the business run?

RagePank and Capsaicin will carry on running with surprisingly few changes. Because of the wonders of modern technology (VoIP), you will still be able to reach me on the same Auckland phone numbers, but most of the business is run by email anyway. I'll still be carrying on with all current projects, except I might actually have a half-decent internet connection in Dublin.

Obviously I won't be available for any meetings - but those who know me will know that I have been avoiding meetings like the plague for some time now - most things can be resolved over the phone or via email, which is almost always quicker. The other significant change is that office hours will be reduced to 7am - 12pm (NZ time), so if you need to call me, call me in the morning. I'm toying with the idea of redirecting calls outside these hours to someone local, but I'm not sure it is really needed or would help.

I'm currently trying to cut down as much paper mail as possible as this will be going to a 3rd party and will be checked fairly irregularly. I'd really appreciate it if customers could try to pay their bills electronically rather than by cheque, if possible.

If anyone wants to catch up before we go, just send me an email. I'll be available for face-face meetings during January, if one is needed, but February is going to be pretty busy.

I'll likely start a travel blog of some sort, and I'll post details if this goes ahead. I'm also giving serious thought to writing a book about potatoes - I hear the Irish like their potatoes.

Source: Ireland beckons...

Hey Google, check out my 301s

Thu, 18 Dec 2008 00:00:00 +1300

Every so often, you need to change the domain name for a website. This isn't something to be taken lightly as there are several implications.

One of these implications is having stale pages lingering around in Google's index from the old domain. The issue here is that due to dupe content issues, Google may not be willing to index your new content until the old content is gone. Sure, you have setup a 301 from the old pages to the new pages, but maybe your site just doesn't have much authority and Google doesn't care much about your redirects.

Without any links to those old redirected pages, Google won't be revisiting those pages anytime soon. Meaning Google won't get to see your 301, the old page stays indexed, and the new page is ignored because it's dupe content.

My solution to this problem is simple - write an open letter to Google. I like to keep it light hearted, it's not like a person is going to read it. Something like this...

Hey Googlebot, long time no see mate.
Hey, I was wondering if I could bother you for a bit - I have this site I'm in the middle of optimising for a client, and he has the .com domain and the .co.nz domain.

Thing is, the site has had dupe content issues in the past and we are only now doing something about it. You assumed that the 'master' copy was the .com but actually we want the .co.nz to be the main copy because the client mostly deals in New Zealand so to help out we have put a 301 in place.

So Googlebot, could you be so kind as to check out our 301s on the following pages, and do the decent thing and update your index?

Cheers mate. If you want to come round for dinner one night we would love to have you - get your wife to bring some of that lovely link juice that she makes :)

Cheers,

Harvey.

You don't post the letter - you just place it on a page somewhere that Googlebot will find.

So basically I'm just recommending that if you want to speed up the reindexing of your site, you need to get some links. Wow - aren't I just the leading authority on new SEO information!</sarchasm>

Thing is, the pages I mentioned above will never generate any direct links under normal circumstances - anyone linking to a little site like this one will be linking to the homepage. Just adding one link to each of the pages that Google has indexed means that there is a much better chance of Google seeing the 301 and updating the index quicker. If the links are placed on an external site that has been around for a while, then things should happen even quicker.

XML sitemaps

Anoher thing I will sometimes do is include the old URLs in the XML sitemap - if the search engine attempts to follow the link they will see the redirect and (in theory) update the index.

Obviously this method doesn't scale particularly well - but for a 5 page site with almost no links, it's a good way to get things happening just a bit quicker than normal.

Source: Hey Google, check out my 301s

Announcing NewFirst Thunderbird add-on

Tue, 16 Dec 2008 00:00:00 +1300

Again, I stray from the path of SEO blogging and post more non-SEO content. However, seeing as my blog gets more Firefox traffic than IE traffic, there's a good chance that many of you are using Mozilla Thunderbird as an email client, so this post might be of some use.

NewFirst Thunderbird Add-on

This morning I decided to do something about a particular Thunderbird behaviour that has been bugging me for some time - the fact that when you click the date column header to sort mesages by date, they get sorted 'oldest at the top' by default. I want them to be 'newest at the top'.

Above: Clicking the date column header sorts in ascending order (oldest messages first). You need to click it a second time to get descending order.

I couldn't find any way to rectify this default behaviour, so I created a Thunderbird add-on called NewFirst to get things how I like them.

This is my first attempt at a Thunderbird Add-on, so I'd appreciate any feedback on this (either by email, or using the blog comments below). It's a very simple add-on, but solves a problem I'm sure others have noticed.

More info

You can either read the full spiel, or just download newfirst.xpi and get on your merry way.

Source: Announcing NewFirst Thunderbird add-on

Christmas hours 2008

Mon, 15 Dec 2008 00:00:00 +1300

RagePank will be closed for normal business from the 20th of December 2008 and will reopen 5th January 2009. I'm reachable by email, but won't be attending to or responding to anything non-urgent over this period.

As always, I'm available to fix anything genuinely urgent if such a need arises (such as a site outage). In this case, please email and be sure to include the word 'urgent' in the subject line. All such messages will be automatically forwarded to my mobile phone.

Have a safe and relaxing Christmas.

Harvey & the team.

Source: Christmas hours 2008

Trojan Horse for Sale

Wed, 10 Dec 2008 00:00:00 +1300

In a bid to clear up some space in the shed, I have decided to auction off my Trojan Horse. No, I'm not talking about a virus, I'm talking about a 2 metre wooden horse, custom made by yours truly and my father-in-law, Rob Tapper.

The trojan horse was custom designed to infiltrate border security at a NZ military base, and it succeeded in style. To be honest, I don't think the border guard at Hobsonville Airbase had heard the story about the trojan horse - we got straight through, no questions asked.
[[trojan-horse-video.tpl]]
A trojan horse is a great way to arrive at a party. The video above goes some way towards explaining why you need a trojan horse.

Perfect for parties

Load the horse onto a trailer
Park up outside the party
wait.
Soon enough, someone will notice a large wooden horse parked outside.
They will be thinking to themselves - "There's no way that fat bastard would fit into that thing" and conclude that it's safe to bring it inside. Just like .exe email attachments from unknown senders, your victims will fall for your clever trap.
An army is gathered to drag the horse inside the fortifications, using the generous length of rope attached to the horse.
Once inside, you can storm the gate and clear the path for your Spartan friends.

You can use it for...

Arriving late to a party used to be fashionable, but that's got nothing on arriving in a Trojan horse. That aside, the kids love playing on the horse, you can easily chuck a few toddlers in the inner compartment or under the body and take them for a ride. While I was testing out the wheel mechanism I took it for a quiet stroll down the street - you are guaranteed some great sideways looks from people driving past!

Technical specs:

It's approx 2 metres tall, and 2 metres long.
Spacious compartment holds a 120 kilo fat person.
Custom design, modelled to resemble the horse of Troy.
Solidly built - the plywood adds extra strength to the structure, and the back legs are solid wood (where the bulk of the weight is sitting).
Good, thick towrope included.
Is affectionately named 'Archippus', meaning 'Master of Horses'.

The horse has served me well, and I hope he may serve a new master just as proudly.

You can buy this unique piece on Trade Me - be quick.

Source: Trojan Horse for Sale

HP HDX18 vs HDX20 Notebook review

Tue, 09 Dec 2008 00:00:00 +1300

I have recently been under the spell of an unreliable notebook, and it's put the brakes on productivity, big time. Unfortunately, with top-of-the-line gear, you can't just rush out and grab a new one when things go wrong - you have to go through the proper support procedures, which takes time.

Sorry, this post has nothing to do with SEO, but I'm working on a little gem so do come back soon.

My old notebook, a Hewlett Packard HDX9108TX 20" notebook has been replaced with a HDX18-1013TX - the massive 20" models have been phased out in favour of smaller 16" and 18" units.

I'm in a unique position to be able to comment on both, so I thought I would share my thoughts.

Other reviews

It's worth noting that I read extensive reviews on both notebooks - and pretty much all of them missed the mark. Reviewers get sidetracked by flashing lights and impressive specs, but miss those everyday finer touches that make or destroy the experience.

The 20" Beast

Let's talk about the 20" monster. The HDX series is all about 'entertainment' - the big screen, big sound, and vast array of plugs that connect it to pretty much anything. I loved the double-hinged screen, the 5.1 output jacks, the dual hard drives, and the fairly decent graphics card. Overall though, I felt there were too many annoyances that should not have been there, and have largely been fixed in the newer model.

The 18" little sister

A bit smaller, a lot more portable, a bit more user-friendly about sums up this model. It lacks the impact of the big beast, but is better in basically every other aspect.

The Screen

The 20" model has a double hinged screen - everyone is impressed by this, and they should be. The screen is big, bright, and you can pull it closer and tilt it to watever angle you like. There is so much to love about the screen. While the screen isn't full HD1080 like the 18" model, it's well decent at 1680x1050, and I'd take this screen anyday.

The 18" model is 16:9 ratio rather than 16:10 like most notebooks. On the plus side, you don't get black bands across the screen when watching 16:9 movies - but that's not really a biggie, and I'd prefer to have that extra screen real estate. Having full 1080p res is cool, especially considering the notebook incorporates a Blu-ray drive.

Sound

The sound on the big beast was both wonderful, and terrible at the same time. The laptop features 4 speakers plus a sub - which means 4.1 sound. While this is cool, every review I read failed to note that 4.1 sound is not the same as 5.1 sound. When you use the default DVD software to watch a 5.1 movie, you lose the center audio channel - the one that contains all the speech from the movie.
I worked around this by loading VLC media player onto the machine, and adjusting the settings to do plain-ol' stereo sound instead. How nobody noticed this at design time, production time, or review time is beyond me.

On the plus side, the notebook has 5.1 output jacks - meaning I can plug in my lovely 5.1 Logitech speakers, and the sound is a treat. Something I have been dreadfully missing since I traded my desktop for a notebook a few years back.

Notably missing from the 18" model is the 5.1 analogue outputs - meaning my fancy Logitechs are idly stuck on 2.1 output. The notebook itself has 2.1 speakers, but to be honest, this is no less useful than the 4.1 on the 20" HDX. It's not that this is bad, jsut that I was a bit spoiled having the full 5.1 output on the old notebook.

Keyboard, touchpad

We all love a full-sized keyboard, and you would expect one on any notebook 17" or larger. The 20" keyboard was good, but marred by a half-sized right hand shift key - the key was made smaller to accommodate the arrow keys. As a programmer, I use the shift key more often than the space bar, and found myself constantly hitting the up arrow and messing up my code. Forgiveable on a compact notebook, not on this one. There was also something hot underneath where your left hand rests - I found this a little annoying as well.
The 18" HDX doesn't have either of these issues. Even better, the 18" doesn't have a remote control sitting on the left hand side taking up space and forcing your hand into a slightly uncomfortable position. The action is nice, though some other reviewers have commented that the silver keys can be a bit reflective in bright light. I really like this keyboard.

One-touch media keys

These are fucking annoying on both models, period. Other reviews with use words such as 'innovative' or 'creative' when describing these one-touch keys, and I thought that too until I actually tried using them.

To start with, on the 20", model there was an unresolved bug with them - every time you hibernated Windows, the one-touch keys would cease to work. To resolve, you had to shut down, remove the AC power and battery, then restart. Which basically means these keys were responsible for not being able to hibernate. Tech support seemed to think this was a perfectly reasonable solution, and didn't release a fix in the 12 months I owned the notebook. Thankfully, this seems to be resolved on the 18" model.

The volume control exhibits even more annoying functionality on both models - if (heaven forbid) you want to adjust the volume in a full-screen game, pressing the one-touch volume control will alt-tab you back to the Windows desktop to show you the little on-screen volume popup, and you have to manually alt-tab back into your game. Again, how did someone not notice this?

Further to that, the volume controls don't work properly until Windows is fully loaded, so if you had your music loud last night you can't turn it down a bit while all your taskbar stuff loads and beeps at you.

I'll take a plain old hardware-based volume control anyday. Just like on my old no-frills HP notebook.

The touchpads are different on both models - however both have a button for disabling the touchpad, which is the best feature ever. I hate touchpads and will carry a mouse around with me, so having the touchpad disabled when you don't need it is very cool.
On the 20" model, the touchpad was made from the same stuff as the handrest - they melded together almost seamlessly. While this looked nice, I found there to be a little more resistance on the touchpad than I was used to, and would have preferred a normal touchpad.
On the 18", they have gone with a super smooth finish that incorporates design elements from the rest of the casing. Again, it's so smooth that if there is even a tiny bit of moisture on your finger, the resistance goes up. I'd prefer a normal touchpad over this as well, but then I use a mouse 95% of the time anyway so it's not a show-stopper for me.

Grunt

Both have the same CPU. The 20" arrived with 2Gb of RAM, but I upgraded to 4Gb pretty quickly because Vista was running like a tortoise. The 18" has 4Gb as standard, good to see.
The 20" was using Vista Ultimate x64 whereas all new HP models are using 32 bit Vista - in this case Home Premium. I liked the 64 bit Windows - I had only one or 2 apps that wouldn't work for me so in my case, the problem people were complaining about wasn't really a problem.
The 18" has a better video card, and it is noticeable when gaming.

Plugs

Both models have lots of plugs. E-SATA, HDMI, Firewire, RJ-45, lots of USB, media card reader, jacks for 2x headphones, internal webcam and mic, TV tuner etc. Notably missing is a DVI output, but the HDMI plug will do this with a special adaptor. Both have a plug for a proper docking station, which I'd like to get at some stage.

Portability

Hey, guess what - the 18" model is better here! I'm big and strong, but I have to admit moving the big 20" thing around was getting a bit much. You can't buy bags for them either, and the power pack was 1.5kg by itself.
I think this is the reason why HP moved to more practical sizes - the 20" is basically a desktop without cables and I was actually finding myself moving it around more than I should have been.

I'm still having trouble finding a good bag for the 18" model, but at least I have options - I could go grab an average bag any time I liked, but I'm holding out for a good one (I like the ones where you drop the notebook in from the top rather than having to open the case right up).

Overall

All in all, I think I bought the 20" too early - HP just hadn't done enough testing to iron out the bugs, and it just didn't fulfill the promise of a top-end notebook. The 18" is more refined, and much more portable. Most of my original complaints are now gone, and in the 3 weeks I have had it, it's been nice and productive.

I'm happy to recommend the 18" model, it's a good machine for the mix of web development and light gaming that I use it for. But I won't be buying a 'first of the series' notebook again.

Source: HP HDX18 vs HDX20 Notebook review

Hiding ugly affiliate links using Javascript

Sun, 16 Nov 2008 00:00:00 +1300

At some point in your affiliate career (usually earlier rather than later), you will come across affiliate links that look a little bit too much like affiliate links.

I'm talking about www.example.com/?AFFILIATE_ID=1234 and similar format links. Nothing screams "I AM AN AFFILIATE LINK" like having the word "affiliate" in the URL.

I'm a big fan of disclosing your interests, but the amount of disclosure does depend on the context. For example - if the affiliate link is in the sidebar and looks like an ad, you don't want to reinforce that it's an ad by drawing attention to the affiliate ID in the URL. While it's good to make it obvious that a link is financially motivated, you don't want to make it *too* obvious. Generally speaking, I really don't like big obvious affiliate links.

However, as an affiliate you have to use whatever format link you are given. If the link format contains a big obvious affiliate ID, here's one effective way to make it go away.

Javascript

Javascript is an interesting demon. It receives a lot of flack, but it's actually brilliant. Anything done using javascript does require a little extra thought - you need to design the script to work well for users with javascript, but also ensure that the page doesn't fall to bits for users without javascript.

What we are going to do is use the affiliate link on the web page as per normal. Our javascript will modify the link to remove the ugly affiliate ID, then when the user clicks on the link, the affiliate ID is re-attached. If the user doesn't have Javascript enabled, none of this trickery happens and they just see the normal affiliate link.

Let's consider you want to link to www.example.com?prod_id=765 (a normal page) but you want the affiliate commission for any clicks. So you actually need to link to www.example.com?prod_id=765&AFF_ID=1234 - but you don't want to show the "AFF_ID=1234" to the user.

We create a HTML page containing the original affiliate link amongst the content.
We make up a list of all the affiliate links we want to use across the site - we record the affiliate version of the link (eg www.example.com?prod_id=765&AFF_ID=1234) and the non-affiliate version (www.example.com?prod_id=765) in our Javascript file.
Once the page loads, our Javascript scans all the links on the page. If any link on the page is in our list of affiliate links, it gets replaced with the non-affiliate version of the link.
When a user clicks on a link, it gets replaced with the affiliate version of the link just before they leave the page. They arrive at the merchant's site with the affiliate ID intact.

All this happens rather seamlessly in the background. And thanks to jQuery, the code required is fairly minimal.

A word of caution

It's worth pointing out that users don't like being tricked. And with the purpose of this script being to show you one link when you mouseover, and send you to a slightly different link when you click, there's definitely the potential for this to be seen as misleading.

I would suggest that this script is only used where there is some form of disclosure already happening on the page. It's easy for a visitor to look at the source code and see what it really going on, and people don't like to be mislead.

Good examples:

I have used company X and found them to be well priced - www.example.com (aff)
This site is sponsored by www.example.com
Browse top quality X from www.example.com (with the text being situated where it obviously looks like an ad, not content)

Bad examples:

Yeah, you should check out www.example.com
I recommend www.example.com as they are good to deal with

With the good examples, the user can reasonably conclude that the link is paid, so (in my opinion) there is no harm in masking the affiliate ID in the URL. With the bad examples, the user would reasonably assume that an unbiased recommendation was being made, and are likely to get angry if they find out that is not the case.

Google adwords is a classic example of the link not truly representing the destination. But because the ad is disclosed as being paid, and the destination URL isn't too far different from what was advertised, people don't get upset. Follow these guidelines, and there is nothing to worry about.

See this script in action

You can see a stripped-down example of this technique here. Mouse-over the example link - If the link looks normal, this means the script is working. If you see an affiliate ID in the link when you mouseover, either you have Javascript disabled, or there is an error in the script. In either case, the link will still work and you still get your commission.

Download

Download a copy of hide affiliate links. Code is LGPL licensed. If you like it, please consider placing a link to www.ragepank.com or www.capsaicin.co.nz, which always puts a smile on my face.

Installation

Upload the included jQuery Javascript file, and link to it in the <head> of your HTML, eg:

<script type="text/javascript" src="jquery-1.2.3.pack.js"></script>

Alternatively, link to Google's copy of jQuery:

<script type="text/javascript" src="http://ajax.googleapis.com/ajax/libs/jquery/1.2.3/jquery.min.js"></script>

Note that jQuery is approx 20kb - that's a lot of extra weight to add to the page if you aren't already using jQuery for something else. Luckily jQuery is awesome, and easily justified on most sites.
Edit aff.js - add all your affiliate links into this file in the following format:

aff['http://www.domain.com/AFFILIATE_LINK/'] = 'http://www.domain.com/REGULAR_LINK/';
Link to aff.js in your HTML - this must appear after the jQuery link:

<script type="text/javascript" src="aff.js?v=1"></script>
Link to affiliate URLs as you normally would throughout the site content, using the Affiliate URL.
If you need to modify aff.js for any reason you will need to update the version number in the code used in step 3 - otherwise aff.js will be cached by visitor's browsers, and they won't see your updated code. Simply increment the version number each time you make a change:

<script type="text/javascript" src="aff.js?v=2"></script>

Enjoy.

Source: Hiding ugly affiliate links using Javascript

Would you pay $30,000 per Gb for data?

Mon, 06 Oct 2008 00:00:00 +1300

I was chatting with a friend recently, who was complaining about the data costs he incurred while travelling to the USA on a 6 month trip.
He said it cost 30 cents per 10kb of data.

I gave this some thought and realised it equated to a whopping $30,000 per gigabyte. I can understand that Blackberry users would never need a whole gig of data, but it's easy to go through that if you use your phone as a modem like I do.

Might I recommend using extra caution before considering data roaming - 30 cents doesn't sound like a lot of money, until you realise how little you can do with 10kb of data. Vodafone do advise caution on their data roaming page, but they don't mention the cost in gigabytes, for obvious reasons.

My friend is employed by the Government, so I'm sure his complaints would have been louder if it was coming from his own pocket (thanks again taxpayer).

Source: Would you pay $30,000 per Gb for data?