OpenStreetMap is so cool

For some reason I love maps and aerial photos, so naturally I love OpenStreetMap!

If you don't know what OpenStreetMap is about, its essentially a project to create "Open Source" (actually a Creative Commons license) maps for the UK (because the real ones are copyrighted and cost loads to use!) and the rest of the world using a load of volunteers wandering around the country with GPS units. Sounds like it might be crap but check out this image of London generated a few months ago:


Pretty cool huh? Obviously there are big areas without any data, but the potential is there! At the moment one of the main guys in the project is doing a community-sponsored week of full time work on the project so there is a lot of activity going on, so go and check out the progress.

I liked the whole concept so much I even donated some money...lets just hope I don't regret giving the money away in a year or two if it ends up like the stupid mess that Wikipedia has become!

StumbleUpon for fun and profit - Part 3: Analysis

In my previews two posts about my run-in with StumbleUpon, I talked mainly about the number of visitors that ended up going to one of my sites.

I concentrated on the numbers of visitors as, according to the monetization (sic) community anyway, its all about visitors; the more you have the more money you get. Unfortunately for me, it turns out that this was the exact opposite. Instead of making a few extra dollars for a couple of days during the peak in traffic, I actually experienced a drop in click throughs for the next month!

To me that sounds kinda counter-intuitive. Lets take a look at the graphs:


The top graph is the real number of click throughs a day (not the percentage!), and the lower graph is impressions across all of my sites that run Adsense ads. The thick line is the line of best fit (showing a nice steady increase in traffic and money!).

As can be clearly seen, around about the 22nd September the StumbleUpon traffic arrived for one of my sites. It died away again really quickly (see the full write up about the traffic profile here: StumbleUpon for fun and profit - Part 2: Results).

But what is really curious here is the uncharacteristic drop in click throughs for the following month across all my sites. As you can see, click throughs are way below the trend up to ... are you waiting for this ... the 22nd October, i.e. exactly a month after the StumbleUpon traffic!

Does Google check referrers?

So this lead me to thinking - is Google checking the referrers before it decides which adverts to display for my pages. Maybe if it detects a lot of "low quality" traffic from sites like Reddit, Digg, StumbleUpon etc it puts your account on a black list, and then puts up some uninteresting, low paying adverts that no one is interested in to protect its advertiser's budgets?

That seems a bit far fetched maybe, perhaps it is the case that Google simply monitors the number of impressions and, upon detecting suspicious activity such as a huge spike, sends out the uninteresting, low paying adverts as above?

Whatever happened, exactly one month later my click throughs were back up to a healthy level. Unfortunately I didn't notice this happening at the time otherwise I might have disabled my adblockers and had a look at what adverts were actually showing - hopefully next time I'll catch it in time!

Windows Live Messenger/MSN8 loves your ram.

A while ago I switched over to Windows Live Messenger (aka MSN 8). I'd avoided the beta for months and months and months because I often have problems with Microsoft's beta stuff and its just not worth the hassle so I thought I'd wait until it was all dandy and fixed and ready for release.

Turns out that its still a huge memory pig. At the moment it is using about 25 meg of ram. Due to the quirks of Windows though, if I minimise it, it drops to using about 4 meg of ram - but if I send it to the tray, it still uses about 25meg! Windows unloads an app's working set from ram if its minimised, so it seems like the developers missed a trick when sending it to the tray ... I'm guessing they are just making the form invisible rather than minimising it. See my blog is both educational and informative! :)

Anyway that all pales into insignificance compared to what I experienced the other day. Came back to my computer and thought things were going a bit slow (click for bigger image)...

Thats right - its using over 260 meg of ram, and 470 meg of virtual memory! Thats more than even Firefox which is well known for its memory leaks!

From my experience, you need to completely shut Messenger down and restart it every few days, otherwise it tends to go mad and do things like this.

StumbleUpon for fun and profit - Part 2: Results

Ok so in my previous post I talked about StumbleUpon and a huge increase in hits I got from it having received a stumbling.

I was wondering how long the effect would last, and what the hits would be like afterwards. Turns out that what I blogged about before was pretty much it - nothing much else happened after:


As you can see I got about 3 days of high traffic, then it dropped off quicker than it picked up! One side effect though is that I am now getting about 3 or 4 times the average previous hits since the stumbling, and a slow trickle of referrals from StumbleUpon still. I appear to be getting more search engine traffic now too - this could be coincidence of course.

Its a bit of a strange profile - I can only presume that my site was temporarily on a "Whats New" type page for a little while, and has now slumped down to the usual listings.

Still, an interesting experience!

Update: I have had a closer look at what happened to my advertising revenue during and after this event. Check out the details at StumbleUpon for fun and profit - Part 3: Analysis

StumbleUpon for fun and profit

For the past few months, I've created a couple of websites made purely for putting adsense adverts on. My first site was a bit of an experiment, and has been pretty successful (for what I wanted anyway - I doubt others would class it as a success) by making a quid or two a day in advertising which is exactly what I wanted.

So I thought I'd start a new site to get some more money as its so easy. As is often the trouble though, getting it noticed and getting people to visit it can be a problem. With the first site I traded a link or two, sneaked a link into Wikipedia here and there etc. Its only getting between about 50-100 unique visitors a day which is pretty crappy really, but the click through rate is amazingly high (we're looking at anything in the region of 5-25%!) so it works.

But with the new site I've not really done what I did before. I was getting a small dribble of search engine traffic, but it was only about 10 uiques a day. So I decided to submit it StumbleUpon to try and get some more traffic. That was a day or two ago now - today I checked on the site's webalizer stats and was pretty damn surprised by what I saw:


As you can see, thats some pretty mental growth in hits. Yesterday I got 1277 uniques; so far in the first half an hour of today I've got about 40 uniques which works out at about 1900 uniques multiplied up. But is this ethical? Well who knows - I'd argue that its more of a "grey hat" method - I'm not gaming the system, the site is genuine and there is no hidden tricks or anything.

I've got no idea how long this will last - will it just dribble off into nothing after a day or two, or will it carry on? I'll post an update on this blog at some point in the future when something interesting happens!

Update: I have blogged about the outcome and side effects of a StumbleUpon "stumbling".

RSS Sucks. Do we need a new alternative?

Near enough every blog you visit, every web site you visit, pretty much anything you visit on the net these days is advertising their own RSS feed.

Anyone would think people actually use RSS!

Ok so there are the hardcore bloggers, maybe a few professionals who are super-keen on keeping up to date with the goings on in their industry, but how many common or garden internet users actually care about RSS? According to a Yahoo! whitepaper, only 12% of internet users are even aware of it, and only 4% actually having knowingly used it - i.e. have actually subscribed to a feed.

I will be the first to admit that RSS is damn great for what it was designed for - sites like My Yahoo! or My MSN where news content is "syndicated" - i.e. news sources are pulled together from various sources onto one page. According to that whitepaper 27% of internet users take advantage of such sites. Thats great - RSS is doing its job and people don't even know about it.

But why are people still saying that RSS is "about to go mainstream" and how it is going to "revolutionise the net"? I am totally unconvinced it ever will, and here is why:

1. Crap software. There are loads of RSS readers out there and 99% of them are shite copies of the basic "three pane" email client view. What is the benefit of using them? What compelling reason do they offer other than maybe sorting, searching etc? To be fair there are some interesting things like RSS screensavers, but then you can actually use them or the screensaver turns off!
   
2. Information Overload. People are already complaining that their lives are plagued by too many emails that they will never get the chance to read. How will having 50, hell even just 10, RSS feeds all sitting there full of even more cruft for you to read through help with the constant nagging voice in the back of your head telling you to "check your emails", other than changing it to tell you to check your feeds now too?
3. Its dull! So you can sit there and read about things from your favourite author or website in a plain, boring three pane email client ripoff like we used to do (and maybe still do) with usenet and ancient BBSs in the "old days". Or maybe like me, you like going to the bloody websites! I like to visit the site and really "get" the whole experience of the visit as intended by the author - they spent a lot of money or time (or both) creating the site to best reflect their personality, their style, their content, their ethos - damn it I want to experience it! We are at a point in the internet where a genuine "multimedia" (I hate that word!) experience is coming of age with embedded music, videos, interactive animations, and stunningly attractive (yet still accessible!) web pages, but people think that we would rather just ignore all this and read a boring text RSS feed? Madness!
4. If its not dull, its full of HTML! So I said that sometimes its dull - some people have addressed that problem by using HTML in their feeds. For obvious reasons sending the HTML along with your content sucks. So you can either have dull and boring text in your feeds, or you can have partially formatted feeds with limited HTML capabilities that breaks a lot of software and devices trying to use the content!
   

Clearly there is no sense in boring RSS readers. However, one of the main arguments I have heard from RSS users (some of those 4%!) is "It [RSS] saves me time instead of visiting all those sites!" - fair enough, but with Firefox you can open all of the sites simultaneously into separate tabs with a single click if you have book marked them...

I think what they really mean is, "I get notification if there is a new post at a website without having to visit it." - OK, now we are talking!

Maybe what we really need is a new, lightweight alternative to RSS - something where the entire article is not pumped into a RSS file, and where there is specific and strict consistent use of attributes to better allow for machine reading and where it is actually viable to use this sort of thing on a mobile phone or a PDA? All this new format would need to do is provide details of new content (i.e. title, short description, date etc), and where it came from/is available using a URI, so a user can still subscribe to a feed and simply get small, efficient "pings" from a website when it has been updated. It doesn't even need to be XML!

A simple solution to "Domain Kiting"?

Bob Parsons - president of GoDaddy.com - recently posted about what he has called "Domain Kiting". I personally wasn't familiar with this term prior to reading this post, or "Check Kiting" either but I guess its probably just an Americanism that hasn't made it over to the UK.

Anyway - this article outlines what domain kiting actually is in a fairly simple and easy to understand. I'm not going to go over it again here, if you don't know just go and read the article. Parsons claims in his article that domain kiting is "out of control" and "must be stopped", but he doesn't go into any detail about. From what I can infer from his post, it looks like he wants to lean on ICANN to try and get them to shut down particular registrars who appear to be responsible for - or are at least facilitating - this process. That's a fair enough point, but then you remember that Parsons runs his own registrar and is maybe just looking to take out some competition? :) Maybe I am being a bit harsh!

But anyway it got me thinking - surely there is a really simple way to solve this problem without resorting to oppressive actions from ICANN, who appear to be pretty unresponsive anyway.

Perhaps a better way to prevent this is to simply set a "block" on all cancelled domains, so they cant be reused for a short period?

For example, say a domain kiter registered the domain "domainkiting.com" with the intention of "kiting" it over and over and over:

1. Kiter registers "domainkiting.com"
   
2. Kiter cancels domain after 5 days and gets money back
3. Registry puts a 30 day "block" on "domainkiting.com" preventing further registrations.
4. Kiter cannot re-register domain for a month.
5. Domain Kiting solved?

Maybe I am looking at this in a really really simplistic way - but that seems like such an obvious solution to me.