ifelse - blog tag:ifelse.org,2009:/blog//21 2009-08-11T15:31:21Z Movable Type 3.31 Star Rating System and the Amazon Kindle tag:ifelse.org,2009:/blog//21.3698 2009-08-11T14:57:33Z 2009-08-11T15:31:21Z This post has been removed becuase I'm an idiot and interpretation of basic math eludes me without the second cup of coffee. Normal service will resume shortly. I've been looking at Amazon's Kindle product page since their whole debacle on... Michael I've been looking at Amazon's Kindle product page since their whole debacle on removing books that users bought and downloaded to the Kindle without permission. It brings up issues of rights, ownership versus subscription and DRM. As a result there has been an organized campaign to inform would be Kindle owners about this through Review on the product page. As of writing, there are 5579 reviews, or which 3265 give it 5 Stars and and 795 have rated it 1 Star. There's plenty in-between but despite all this the average Star rating is: 4 Stars out of 5. To me this didn't seem very accurate reflection of views, not off by much, but not quite accurate either. Its not that Amazon is doing anything deceitful or playing with the numbers. The Star Ranking system is used almost everywhere, but I wonder how honest a system it really is. To that end, I did a quick look at what the Amazon Kindle would be rated had it a Grade, A being highest, E being lowest. Using such a system the Kindle does poorly: (Link) Using plain Grades, Amazon Kindle scores 1 out of 5. a very significant difference? I don't think a Grade system is the answer to giving consumers a good impression of a products worth. However I think the Star system, after a point becomes useless. I've been told, by people much smarter than I, that with enough reviews all ratings average out at 4 eventually (out of 5) - just browse Yelp long enough to see what they mean. Or maybe I screwed up the basic math... ]]> The Presentation-Layer Arms Race? tag:ifelse.org,2009:/blog//21.3695 2009-03-03T18:13:32Z 2009-03-03T18:36:41Z Oh dear. It seems that a new sort or arms race may break out shortly. The nice people at Arc90 Labs work hard to improve the user experience on the internet and one of their targets is the online-ad. One... Michael Arc90 Labs work hard to improve the user experience on the internet and one of their targets is the online-ad. One line ads are often badly implemented and more often than not end up annoying the very people they're supposedly trying to entice. Its also a sad reality that they also tend to drive the design of the page as you have to consider where to put that Big Ad or Banner Ad position when you're re-working layout... *sigh* So back to Arc90 Labs... what did they do...? they created a simple and wonderful way to remove that clutter. Fantastic. I love it. Its called Readability. But... I have a problem with this. Two problems in fact. The first is I wouldn't want anyone to change the layout of a page on my site without a little bit of discussion. I've worked hard to create a page they way it is, why can't you respect that. Of course thats kinda impractical so thats more a pet peeve. I'm a big fan of GM scripts that make minor tweaks here and there, like extra buttons in Gmail, or a new theme etc. However Arc90 Labs kicks off a radical, unrecognizable change to the page to achieve its aim. The second problem is that right now this ad revenue, as much as you hate it, is what keeps many sites afloat. Most sites hate to be dependent upon ad revenue as their bread and butter (some aren't) but its a reality they face - and you do too by proxy. Ads are necessary at this moment in time to keep the content there. All is well as long as the vast majority of users do the normal site navigation and the minority use these new toys. However I feel its a matter of time before tweaker and publisher get locked in a small arms race over enabling and disabling some site modifications just like this. In this case, if a blogger or site owner wanted stop Arc90 Labs from working they could embed this code in their web pages (or in a referenced JS file thats on each page): 

setInterval(function() { 
if (document.styleSheets) {
	var d = document.domain.split(/\./);
	var base = (d.length > 1) ? (d[d.length-2] + "." +  d[d.length-1]) : document.domain;
	for (var i=0; i<document.styleSheets.length; i++) {
		try { if (document.styleSheets[i].href.indexOf(base)==-1) { location.reload(true) } }catch(err){}
	}
}
}, 5000);
The script is simple and was done in a rush (hence the try{})... you can also view it here. It checks every 5 seconds for any external stylesheets that are on the page that do not originate from the base domain. If a non-base domain stylesheet is found then it reloads the page so as to restore the original. Its simple (but not fully tested) and I'm sure Arc can find away around it, at which point I (or someone else) will update this and the back and forth will go on and on and on... I'd rather not have to do that.]]> Financial Meltdown in Real-Time tag:ifelse.org,2008:/blog//21.3677 2008-10-10T19:41:22Z 2008-10-10T19:57:55Z I created a quick Greasemonkey script on my lunch break so I could watch the markets collapse/rebound in (almost) real-time. It updates the business chart on the New York Times homepage every 10 seconds (anything less is uninformative and unnecessary).... Michael I created a quick Greasemonkey script on my lunch break so I could watch the markets collapse/rebound in (almost) real-time. It updates the business chart on the New York Times homepage every 10 seconds (anything less is uninformative and unnecessary).

You can download the script here, or here.

]]> TimesPeople tag:ifelse.org,2008:/blog//21.3675 2008-09-29T17:58:23Z 2008-09-29T21:31:03Z After 9 months in various stages of development (starting as a prototype within a Firefox plugin) TimesPeople was launched on Monday 22nd across the nytimes.com web site. I've the distinct impression that I shouldn't go into the specifics on this... Michael TimesPeople was launched on Monday 22nd across the nytimes.com web site. I've the distinct impression that I shouldn't go into the specifics on this project just yet. I will say that it was the most rewarding and challenging project I have been involved in. Watching the number of existing and new Times users beginning to use it in real-time took my breath away. So far the feedback is mostly positive which is always nice. As of writing the next push of tweaks and changes should address some minor issues that won't matter to 99% of people. Among them are corrections to existing microformats and fixes to ensure full W3C compliance (current at: 59 Errors, 4 warnings, soon to be: 0). One thing I will say, based off my experience - I will be very very happy when IE6 is no longer a white-listed browser.]]> Google Maps Case Study tag:ifelse.org,2008:/blog//21.3662 2008-09-21T23:08:58Z 2008-09-29T18:53:27Z Back in 2006 I worked as part of the team for the Travel redesign project. It was intense and fun and I got to work with a lot of good people. It was also my first opportunity to use third... Michael I don't think I will ever be able to articulate how awkward it feels to be in a video... its the whole 'that doesn't sound like me' thing plus bad hair day in one. Still, Google Maps was a big factor in my moving to a web app based direction. The actual page for this is here.]]> Avoiding Interstitials tag:ifelse.org,2008:/blog//21.3653 2008-07-29T20:48:29Z 2008-08-22T15:37:41Z For many reasons I've found the need to scrape article pages of The New York Times - either to prototype something or gather sample data (though this should become obsolete once the API's become public). Automated scraping of anything is... Michael nytimes.com any automated is going to hit an interstitial advert at some point. The easy way to avoid this is to append "no_interstitial" to the URL arguments. http://www.nytimes.com/2008/07/30/business/30bags.html?no_interstitial That was easy. One less thing to worry about.]]>