Facebook was the key to the entire campaign, Wigmore explained. A Facebook ‘like’, he said, was their most “potent weapon”. “Because using artificial intelligence, as we did, tells you all sorts of things about that individual and how to convince them with what sort of advert. And you knew there would also be other people in their network who liked what they liked, so you could spread. And then you follow them. The computer never stops learning and it never stops monitoring.”

from: Robert Mercer: the big data billionaire waging war on mainstream media | Politics | The Guardian

Carole Cadwalladr’s article in today’s Observer, is both fascinating and frightening. The technology used by Cambridge Analytics is incredibly  powerful the use it has ben put too worrying. Andy Wigmore, Leave.EU’s comms director in the quote above doesn’t have a Facebook account quoted in the same article:

It is creepy! It’s really creepy! It’s why I’m not on Facebook! I tried it on myself to see what information it had on me and I was like, ‘Oh my God!’ What’s scary is that my kids had put things on Instagram and it picked that up. It knew where my kids went to school.

Featured image on this post created with a wee AppleScript Makes auto complete google search gifs.

For the last couple of weeks I’ve been playing with the Fridge I blogged about the other day.

I was feeling pretty please with myself until I got a DM for Eric earlier this (UK) evening. He noticed that the dolch list on the screenshot of the webpage I had made was pretty much the same one he had used on an activity he had made and blogged about: Control Alt Achieve: Wintertime Magnetic Poetry with Google Drawings. It was the same list. I remember seeing his site when I google dolch or wordlist. I can remember copying the list but I must have done so. I didn’t keep the list, replacing it on Sunday with a full (or almost full) dolch list. But I did take and use Eric’s curation of the list without attribution at the time.

Eric was gracious enough to DM me rather than call me out publicly, giving me a wee bit of breathing room to put things right for which I am grateful.

I am particularly embarrassed as I spend a fair bit of time talking about copyright and creative commons, the image search in the fridge provides automatic attribution. I’ve often blogged about it and the difficulty in getting pupils away from a straight google search. In the previous post, I credit Tom Woodward for the idea, and link to the code I’ve used. I just didn’t even think about the words and the work involved in curating a list. I am now presuming that my using the dolch list is ok?

Eric was also kind enough not to suggest I’d taken the whole idea from his template, which looks like a quite reasonable assumption. Here I can only plead I’ve been playing with virtual fridge magnets for a while, this effort: fridge is from 2002. Made with Flash and flamethrower (sic) 1.

I’ve added some attribution to The Fridge now, I’ll update any help for it that I produce and link to this post from Sundays shortly.

I can only repeat my apology to Eric and thank him for the lesson.

  1. The flash bit still works, Flamethrower is no more.

I’ve had a long term interest in digital ‘fridge’ poetry, making my first efforts with Flash around 15 years ago. A year or so ago I was excited by Fridge Poetry – Google Sheets as Database by Tom Woodward. There were a couple of goodies in that post, getting the word list from a google sheet and a nifty way to allow folk to easily make their own. I made a sheet and a poem and slotted the idea away.

I’ve revisited Tom’s post (and others) a few times, gathering tools 1 and wondering.

On the holiday weekend, given poor weather and a head cold, I revisited the idea and made my own Fridge.

This riffs & extends the idea a wee bit:

  1. You can add a background image to the poem, either from a built in flickr search or a local one.
  2. There is a standard common word list and a topical one from the google sheet.
  3. The words in the lists can be used more than once.
  4. I used JavaScript as opposed to php (except for proxying images to allow you to export).
  5. You can export the poem as an image.

I’ve edited Tom’s template a little, the new one:

  • Automatically generated a link to use. Tom got you to copy paste in the sheets own url and parsed that.
  • Adds a field for the image search.

Make a List this link should get you to create a copy of the list spreadsheet. You can edit the words (on the 2nd worksheet) and change the image search,  more info: Fridge Poetry.


I’ve gained a wee bit more JavaScript and jQuery. The idea of using Google sheets to populate a webpage or to display info from a sheet in a template is interesting. html2canvas is another tool that has interesting potential for storytelling on the web.

Using /copy at the end of a google sheet to allow anyone to make a copy is useful too.

Finally the ability of google sheets to get the id of the current sheet is really handy in simplifying the creation of links. This relies on a very simple script:

function getSheetID() {
  var r = SpreadsheetApp.getActiveSpreadsheet().getId();
return r;


you can then get the id by typing =getSheetID() in a cell.


There is more help on how to make and use a wordlist here: Fridge Poetry.

Hopefully someone will find this fun of useful, if you do and create new wordlists please let me know.

TL:DR I’ve found a link that leads to google image search for  images labeled for noncommercial reuse. This is handy on iOS where it is hard to get to the Usage Rights Filter, here is the link: https://www.google.co.uk/search?q=&lr=&safe=active&hl=en-GB&tbs=sur:f&tbm=isch.

Coming back to the classroom after 9 years I find I am still interested in searching for images and copyright. It still seems to be as hard to get young children to understand the problem and harder to understand and carry out attribution.

On my classes blog I link to various places to search for images with some advice on attribution. I include my FlickrCC Stampr tool which can simplify the attribution.

As well as the problem with attribution there is the ever present temptation just to search google. This is made worse by the fact that the Search Tools displayed on iOS lack the Usage Rights popup.

So I was interested in the link I saw today: How to find Google images with reuse licenses on an iPad iOS – Using Technology Better.

Unfortunately this method was described as a hack and took 6 steps to get to the advanced search and the usage rights pop up.

iOS google Search no Usages rights in Tools.

Google Image Search Mac, has Usage rights in Tools.

So I though I might have a search for the url parameters1 involved in a search with the Usage rights displayed.

There are quite a lot of parameters and although searching google for a list provides some these seem to be both undocumented and changeable:

You can expect that a lot of this will change. The reason why Google themselves do not provide any guidance or documentation on these parameters is probably that they want to retain full freedom to change how they work. You can expect that some will be removed, some will be added, and others will give a different result than before.

from: Google Search URL Request Parameters | DETECTED Which give a lot of details, but not the one I wanted.

So I went through the process in the Using Technology Better post and copied the url. I then started deleting the parameters until I found the ones that would produce the right kind of search:


The tbm=isch bit makes it an image search, the tbs=sur:f seems to set the usage rights to Labeled for noncommercial reuse.

So you can now give pupils on iOS a direct link to search for images that labeled for noncommercial reuse.

Featured Image: found searching Flickr for search with no known copyright restrictions: Image from page 211 of “Bulletin” (1961-1962) by Internet Archive Book Images No known copyright restrictions

1. The paramaters are the bits in the url after ? for example ?q=bus&safe=active, makes a search for a bus safe.

This is hopefully a useful tip for organising pupil created content on WordPress blogs like Glow Blogs.

Most of the content posed to blogs consists of posts or pages. Posts join the stream, pages are static generally accessed from menus.

This is another way to organise content that can be useful. Especially if several pupils are posting similar content at the same time. It helps organise that content and the early posters don’t get pushed away down the stream.

An Example Banton Birds

At school we are trying to pay attention to the local birdlife. We took part in the BIG BBC Bird watch with local Bird watchers, Mr & Mrs Carter. We started feeding the birds and trying to get some photos.

The pupils are doing a bit of research on the birds we saw (not too many it was a windy hour) and I hope they will continue over time adding other birds seen in the playground and around.

I want to organise these bits of research automatically collecting them together. To do this the pupils create, not posts or pages but a custom post type, project.

How it is done

This is just one way to do this but it is, at least on Glow Blogs, the most straightforward. It uses the JetPack Custom Content Types module and the unfortunately 1 named Portfolio Projects.

These projects are like pages in that they are out of the flow but like posts have Project Types (like categories) and Project Tags (like tags) that can be used to organise them. The also have a shortcode that can be used to create an index of the projects by type or tag2.

  1. Activate the Jetpack plugin.
  2. In Jetpack Settings Activate Custom Content Types.
  3. In Settings-> Writing Enable Portfolio Projects for this site
  4. Pupils create Projects for each item, add a project type (in this case Birds).

To create an index of these projects I add a page Banton Birds and add a shortcode to the page:

[portfolio include_type=birds display_types=false]

To make the index more visual the pupils make sure they have added a featured image to their posts. That automatically creates a thumbnail.

I’ve used this technique several times on blogs myself, including the Projects page on this blog. I’ve used it too to provide and organise photos and links on the Biggies blog for the pupils.

1. For Glow Blogs where we have a lot of pupil e-portfolios through a custom plugin.

2. You could do the same with pages using the hierarchy and the Display Posts Shortcode plugin, but not as easily.

On Wednesday this week I was talking into the mic. Not too unusual given I often broadcast on Radio #EDUtalk on Wednesday evenings.
This week however I was talking to [Ian Guest](‪@IaninSheffield ‬) for his research on Twitter for CPD.



Being even less prepared than usual I missed a few things. I was probably more negative about some aspects of twitter than some other folk.

What I did not talk about enough was the positive communities that build up on Twitter. For me #ds106 is a particularly good example, but there are lots more.

This morning when I looked at my notification I see this:

After breakfast with the radio news, a lovely contrast. Sometimes tweets are not just passing leaves in the stream and are worth grabbing.

A few years ago I posed about an interesting use of Google reverse image search:


And over the years I’ve read a good few posts about the tool on Alan’s blog (e.g. The Hidden Complexity of Attribution, Reverse Image Search ).

This week I’ve been reading more fascinating uses of the tool: Cleanup Time and Road Trip

I am even more strongly minded that we should be starting to teach these skills from a young age. How easy that will be I don’t know.

A few weeks ago, during the scary clown storm I was hearing about clown stories every day. One pupil was most insistent that there was a clown plague. The pupil presented me with ‘evidence’ from his iPad. This was a photo of a dead clown stretched out shot on a New York street. I took a look with the idea of demonstrating a wee bit of fact checking. On scrolling down below the picture I found the headline explaining that this was a fake photo! No detective work needed.

I am not quite sure where to start with this teaching. Perhaps using the reverse image search to identify things or creatures combined with some work on The Pacific Northwest Tree Octopus.

The problem is that the fake stuff is catchy, fun and enables us to grab a quick stance.

Featured image Reversed | Pekka Nikrus | Flickr BY-NC-SA

A little blue sky thinking, or borrowing some ideas from @davewiner & @mrkrndvs

I wonder if a WordPress plugin could ape pngWriter and create images for twitter cards with the text from the excerpt on the image used by twitter.
Like this, but with less fireworks.

I’ve been thinking about twitter again.

I think I’d have preferred twitter to be just text rather than being expanded to include all the media and ‘twitter cards’. I’d rather the reading flow would be twitter for short stuff and link to more visual or longer material out on your blog. That would perhaps make reading a little deeper and avoid the problem of folk just seeing the main point of an article and reacting to that 1.

But that cat left the bag a while ago. Now when I look at my twitter stream it is full of images. I occasionally use OneShot to grab and crop out interesting bits of text to share and surmount the 140 char limitation too.

I am not sure if this is a great idea

Recently I’ve watched Dave Winer experimenting with pngWriter (see the about page: About pngWriter). This creates images of blocks of text and sends them out in a tweet. It also creates an RSS feed of the text (pngWriter is not open for use at the moment).

This reminded me of how Aaron uses featured images in his blog: Creative Commons Starts with Making – A Reflection on Creating and Sharing – Read Write Respond

So I am kicking round a couple of ideas.

  1. If you used pngWriter you could pull the rss into your blog. You could do it in WordPress with the FeedWordPress plugin.
  2. A plugin could be made that would do the same sort of thing in WordPress, take the except text from a post and make an image for twitter from that.

Here is how I imagine that working

  1. Using an html5 canvas to automatically create an image of the post’s excerpt.
  2. Auto upload that to WordPress media library.
  3. Make it the featured image of the post

Alternatively (better?) make it the content of the twitter:image metatag for the post, that means that the featured image on the post would not have the text, visitors could read the post.

Or make a copy of the posts featured image, add the text and make that the twitter card image (I can imagine that might be different to get the text readable.).

For a couple of minutes I imagined that my babysteps php & JavaScript could put something like this together. Then reality kicked in. So I though I’d post it out here maybe someone else would think it would be a good idea.