Happy 7th Birthday Blog!
No, I’m not seven years old (even though I may act like it sometimes)…but this blog is! My first blog post was on the 11th November 2008 – a rather philosophical post entitled Wisdom and Knowledge – and so on this anniversary I thought I’d look at my blog over the years, the number of visits I’ve got, my most popular posts, and so on.
The earliest snapshot of my website available at www.archive.org is shown below, with my second ever post on the home page (an article that was later republished in the Headington School magazine).
Traffic
Unfortunately it’s a little difficult to look at statistics across the whole of this blog’s life – as I used to use Piwik for website analytics, but switched to Google Analytics part-way through 2014 (when Piwik just got too slow to use effectively). Cleverly, I didn’t export my Piwik data first – so I can’t look at any statistics prior to 2014.
However, I can say that since August 2014, around 240,000 people have viewed my blog, with around 300,000 individual page views – a fairly impressive total, averaging at around 15,000-20,000 visitors per month, but increasing in recent months to around 25,000 per month.
There’s a very noticeable pattern in page views, with significantly fewer views at weekends – probably reflecting the work-like nature of my blog.
The vast majority of these views come from so-called ‘organic searches’ – basically just people searching online for something. The referrals I get tend to be either from social networks (mainly Twitter, but sometimes Facebook too), or from ‘blog aggregators’ that some of my posts are contributed to (my R posts can be found on R Bloggers and my Python posts on Planet Python).
Most popular posts
So, what are these people coming to read? Well, a lot of them are coming to find out how to fix a network printer suddenly showing as offline in Windows – in fact, 51% of all of my page views have been for that post! It’s actually become quite popular – being linked from Microsoft support pages, Microsoft TechNet forums, SpiceWorks forums and more.
My second most popular post is also a how to post – in this case, how to solve Ctrl-Space autocomplete not working in Eclipse (10% of page views), and my third most popular ‘post’ is actually the second page of comments for the network printer post…obviously people are very keen to fix their network printers!
The next most popular posts are:
- John Snow’s Cholera data in more formats
- How to: Set raster values to NoData easily in ArcGIS 10
- How to: Fix ‘WARNING: terminal is not fully functional’ error on Windows with Cygwin/Msysgit
- Producing polar contour plots with matplotlib
- How to: Reset the Software Update URL in OS X
- John Snow’s famous cholera analysis data in modern GIS formats (this gets a lot of referrals from the Guardian article which uses my dataset)
- How to: Fix problem where EndNote only shows the EndNote Web toolbar in Word
- How to: Set up a simple service to run in the background on a Linux machine (using daemontools)
- Free Julian Day calendar poster download
- Introducing recipy: effortless provenance tracking with Python (impressive that this is already in this list, given that it was only published at the end of August this year)
These posts make sense as the most frequently viewed posts: they are either posts that help solve very frustrating problems, posts which have been linked from high-traffic websites (like the Guardian), posts which show you how to do something useful, or posts that have received a lot of attention on Twitter.
My favourite posts
It’s hard to choose some favourites from the 130-odd (some very odd!) posts that I’ve published – but I’m going to try and choose some from various categories:
- Really interesting:
- Default Deny: A signal of academic maturity? – an interesting article discussing how not accepting things other people say is an essential part of academia
- Leaky abstractions in science – another more philosophical post trying to generalise Joel Spolsky’s ‘Law of Leaky Abstractions’ to science
- The best debugger is a good compiler – this is something I miss slightly, now I spend most of my time programming in Python
- Programming << Pointer Programming << Parallel Programming – thoughts on the complexity of parallel programming
- You’ve done your literature review – what about a data review as well – if I ever supervise a MSc/PhD student, I will probably force them to do this!
- Really useful
- How to choose a co-ordinate transformation in ArcGIS – so few people seem to know about this document, and just pick a transformation at random, but there is a far better way!
- Want to write some code? Get away from your computer – I reference this post a lot when teaching programming, as it is far too tempting (particularly for newbies) to start fiddling around with the code, rather than actually thinking!
- Configuring legends in ArcGIS – a slightly less horrible way to do it – it’s a sign of how poor the ArcGIS interface is that I have to read this post every time I want to make a legend that doesn’t look awful!
- How to fix a sudden VPN problem in Vista – I had a similar problem in Windows 7 a few months ago, Googled for a solution and found my post – and it worked too!
- Encouraging citation of software: introducing CITATION files – this was an idea I had ‘all by myself’, and people have actually started doing it!
- Things I never really took forward
- (These posts had good ideas in them, but I never really followed them up…maybe in the future)
- Rules of thumb in remote sensing – a list of common ‘rules of thumb’ in my field and roughly where they come from. I think this sort of ‘common sense’ should be documented more in most fields.
- Standard test images for remote sensing – I wish I’d got round to putting together a set of images like this; I think it would be very useful.
- Historically important (for me!)
- John Snow’s famous Cholera analysis in modern GIS formats – this is the post that got me in the Guardian!
- My programming journey – a nice summary of my development as a programmer
- Review: Image Analysis, Classification and Change Detection in Remote Sensing – my first book review, which led to many more
- Behind the paper: Spatial variability of the atmosphere over southern England and its effect on scene-based atmospheric corrections – my first behind the paper post (hopefully the first of many)
- Introducing recipy: effortless provenance tracking with Python – the blog post that introduced the world to recipy, automatically posted during my talk at EuroSciPy 2015
How and why do I do this?
Well, partly because it’s fun, partly because it’s useful and partly because I feel I ought to. Let’s take those separately:
- It’s fun to write interesting articles about things. I enjoy sharing my knowledge about topics that I know a lot about, and I enjoy writing the more philosophical articles.
- It’s useful because it records information that I tend to forget (I can never remember why I get this error message without looking it up – and I’ve found my own blog from Google searches many times!), it gets me known in the community, it publicises the things that I do, and it shares useful information with other people.
- I feel that I have an obligation to share things, from two perspectives: 1) As a scientist, who is funded mostly by taxpayers, I feel I need to share my research work with the public and 2) As a user of lots of open-source software, I feel a need to share some of my code, and share my knowledge about programming.
(For another useful guide on reasons to blog as an academic, see Matt Might’s article – in fact, read his whole blog – it’s great!)
Conclusion
I started this blog as an experiment: I had no idea what I would write, and no idea whether anyone would read it. I posted very little in the first few years, but have been posting semi-regularly since 2011, and people actually seem to be reading it. Hopefully, I will continue for another seven years…
If you found this post useful, please consider buying me a coffee.
This post originally appeared on Robin's Blog.
Categorised as: Meta
Leave a Reply