Here are a few stories that were in the news recently. They have two things in common — see if you can you guess what they are:

• Money laundering fears as universities accept £52m in cash (Times)
• Almost 1,000 UK homeless deaths recorded in 2020 (ITV)
• Covid bike and walking schemes do not delay ambulances, trusts say (Guardian)
• Councils fail to pay £1.3bn of emergency Covid business grants (Times)
• Covid-19: NHS trusts deny restricting PPE during pandemic (BMJ)
• Half of London boroughs found using Chinese surveillance tech linked to Uighur abuses (Japan Times)

If you’ve been keeping up with mySociety’s posts, it’s probably no surprise that the first thing these stories have in common is that they are all based on Freedom of Information requests — in fact, multiple requests made across many bodies.

We often mention how useful Freedom of Information can be in helping campaigns, journalists or individuals to gather information from a variety of sources, to create a dataset that didn’t exist in one place before.

Naturally we are all in favour of such stories — but we think the organisations and media behind these requests are missing an extra trick, and that’s the second thing they have in common.

In every case, it seems the journalist or organisation has submitted their requests, and gathered the data, then written the story — and that’s the end of it. That data is hidden away, and no-one else can access it to verify the story, dig further or to find more interesting leads.

Journalists understandably gather information for their stories in private so that they aren’t ‘scooped’: this is one factor that led us to develop WhatDoTheyKnow Pro, which allows users to embargo requests and responses until their story has been published. But, once it has, the tool features strong encouragements to put the underlying data live, so that everyone can access it.

After all, at this stage there is often little benefit to the journalist from keeping the data all to themselves — and lots of potential public good from putting it out in the open. This is also a great way of providing extra credibility for a news item, showing that the facts back it up.

Here are those stories again, together with details of the requests that informed them:

• University money laundering fears: The Times surveyed multiple British universities to break this front-page story.
• Homeless deaths: The Museum of Homelessness put in over 300 FOI requests to gain one part of the information backing up their Dying Homeless project.
• Bike and walking schemes not delaying ambulances The charity Cycling UK asked 10 ambulance trusts for their data.
• Councils fail to pay grants  400 FOI requests were issued by the Event Supplier and Services Association to local authorities across England.
• NHS trusts deny restricting PPE: The BMJ sent Freedom of Information requests to 130 acute, community, integrated, and ambulance trusts.
• London boroughs using Chinese surveillance tech FOI requests were submitted to all 32 London councils and the next 20 largest UK city councils.

If you’re a journalist or campaigner yourself, we’d like to suggest that you consider making your data public next time you use FOI like this. Do it via WhatDoTheyKnow Pro, or, if you prefer, do it elsewhere: naturally, the choice is yours, though it’s worth noting that data on WhatDoTheyKnow is easy for people to find, thanks to our excellent search engine positions.

Pro also has other features that aid journalists in their investigations, including the ability to send batch requests to multiple authorities.

With our citations tool, you can even link directly to your story, giving it a boost in visibility that is also accelerated by our good standing with Google et al (or other users can link to it in an annotation).

On the other hand, if you’re just an interested citizen who would love to know more about one or more of those news stories, don’t forget that you could use WhatDoTheyKnow to request the same information, and it will then be public for all to see.

For example, if the homelessness or the PPE story is of interest to you, you could make an FOI request to your own council or NHS Trust to get the local picture. Once you have the facts, you might take informed action on them: perhaps lobbying your local representatives for change, or contacting the local media if there’s a story to be told.

And, to help us in our attempts to get more journalists thinking about opening up their data, you could keep your eyes open for stories like these in the future.

If you see one, perhaps give the writer a friendly nudge to publish their data. After all, they’re using transparency to get their scoop — why not also practice transparency for the good of all?

In 2019, mySociety was involved in several projects working with local councils around using participatory or deliberative democracy to address a local issue (Public Square and the Innovations in Democracy Programme). Something that kept coming up at the fringes of these projects were the political considerations that led councils to find the idea of alternate forms of democracy appealing in the first place.

Understanding more about this seemed important to the future spread of these ideas, and so as part of the Public Square project, we set out to find out how local councillors viewed ‘new’ forms of democracy, and how these views varied by the political situation of the councils and of the councillors themselves.

Using a survey of local councillors, we tried to learn about different awareness and attitudes towards deliberative or participative exercises. We found that partisan and structural factors shape the perceptions of local representatives of citizen participation, and a wide-range of views among local councillors. Some were supportive of more weight being put on citizen participation, while others argued that if decisions are made by elected councillors there is someone to hold accountable. Both awareness and support for participatory methods increased if there was local experience of an exercise. Even opposition councillors tended to be quite supportive (76%) of participatory processes when run by the current leadership of the council.

Where there is more disagreement was in how the outcome of processes should be handled. Very few councillors favour approaches where the result is authoritative or binding. Councillors in councils where there is no one party with an overall majority are more likely to give greater weight to participatory exercises (59%) than those where there is a single party majority (38%).  Every policy area except Children’s Social Care had over 50% acceptance that a participatory exercise could be appropriate. Programmes related to environment and cultural programmes rated highly, while programmes concerning social care scored lower. For all categories except planning and public health, councillors rated these activities as more appropriate if their council had previously engaged in such an exercise.

Overall, this survey told us that councillors make personal evaluations of participatory exercises based on a mix of political and practical factors. While there is a tension between participatory and representative democratic structures, in practice this tension can lead to a variety of outcomes. The success or failure of future participation requires understanding about how this tension affects not just the form of deliberative exercises, but how results will be interpreted and implemented.

The full report can be read online or downloaded as a PDF.

—

Image: Lucas Benjamin

Artificial Intelligence, innovative use of data and the arms industry – now there’s a bunch of areas you’d want oversight on. And yet, a high-profile new government research agency appears to have been absolved from the obligations of public scrutiny before it even begins work.

News broke this week that the treasury has authorised £800 million of funding over the next four years for the Advanced Research and Invention Agency (ARIA). This research agency was originally conceived by Dominic Cummings, and, according to the 2019 Conservative manifesto, will be producing “high-risk, high-payoff research, at arm’s length from government”.

More explicitly, The Guardian sees the agency very much working in the area of defence, while also noting that many technologies developed in this area have gone on to benefit society more widely. The BBC says ARIA has been inspired by US agency DARPA, which is “credited with funding the development of the internet and GPS”.

All well and good, perhaps, until you see the government’s assertion that “the new body is being set up so it can take fast, agile decisions without bureaucracy.”

Judging by multiple press reports and a comment from Ed Milliband, although the agency is to be funded by taxpayers’ money it will be exempt from Freedom of Information law. While we very much hope this is not the case, this aspect has been reported by several sources.

FOI is explicitly for the purpose of allowing citizens to demand transparency from the institutions which we fund. The Times, reporting this story, also takes a moment to remind readers that it, along with other major news outlets  — not to mention organisations including mySociety —  is calling for urgent action on declining levels of governmental transparency, and we can see from the ICO’s many notices to Whitehall departments that the current administration are not complying with their obligations.

Our friends at the Campaign for FOI point out that DARPA, the blueprint for ARIA, is in fact subject to the US FOI Act, so removing those obligations would be something that has been built in as part of ARIA’s conception:

This new taxpayer funded body must not be exempt from #FOI. The Act has an exemption to protect commercial interests. The US agency, DARPA, on which it is modelled is subject to FOIA. So ministers have no basis for claiming FOI would give rivals an unfair competitive advantage. https://t.co/Ch4Spnp91S

— Campaign for Freedom of Information (@CampaignFoI) February 17, 2021

Additionally, the WhatDoTheyKnow team point out that any authority wholly owned by the public sector is subject to FOI unless specific provision is made to exclude it — and so, dodging the obligations of the Act would require either setting up an opaque operating structure for that purpose, or a new exemption to be passed into law under the FoIA.

Meanwhile, our FOI site WhatDoTheyKnow does list authorities that are not subject to FOI if there is a good argument that they should be. If indeed it is officially exempted from the Act, we will also take this route with ARIA, just as soon as it formally comes into being.

EDIT: The official government press release is now here, and includes the statement: “Central to the agency will be its ability to deliver funding to the UK’s most pioneering researchers flexibly and at speed, in a way that best supports their work and avoids unnecessary bureaucracy.”

—

Image: Kevin Ku

Info Pro Vsechny (IPV) is the Freedom of Information site for the Czech Republic, run on our Alaveteli software.

Czech civil movement Million Moments for Democracy (Milion Chvilek Pro Demokracii) is currently using the platform to run a campaign, making for an interesting example of how such groups can leverage FOI sites to mobilise support, and to encourage citizens to engage in the democratic process.

Million Moments approached IPV, who were able to advise on the best way to allow their supporters to get involved, as the FOI site’s team explained when we chatted to them recently.

But first, to make sure we understood the context, we had a quick read of the Wikipedia page on the Czech Republic’s Prime Minister Andrej Babiš. It’s fair to say that Babiš is a contentious figure, as demonstrated by no fewer than eight entries in the ‘controversies’ section of that page.

Conflict of interest

Top of the agenda today, though, is a scandal currently under investigation by the European Commission. Babiš was instrumental in decisions to award EU grants to the massive Agrofert conglomerate, a holding company with over 250 subsidiaries across forestry, farming, food, construction and logistics industries, among others.

In doing so, he breached EU legislation. Why? Because he just happens to be the previous CEO of Agrofert.

While Babiš’ shares were subsequently transferred to a trust fund, as IPV told us, the European Commission has ruled that there is still a case to be answered: “They stated that the main fund beneficiary is still Babiš and the conflict of interest has not been resolved. And while they’ve asked the Czech government to act upon their recommendation, things are moving very slowly.”

This was the impetus behind Million Moments FOI campaign, which is currently encouraging their followers to use IPV to ask pertinent questions about this conflict of interest, and to potentially dig up others.

“They want to ensure that the Czech authorities are asking the right questions on behalf of the country’s citizens, rather than sweeping it under the carpet,” explain IPV. “So they’re encouraging people to ask all the institutions and semi-owned-state companies to what extent they deal with companies in the Agrofert holding.

“More questions, more people engaged, more institutions involved — it all puts greater pressure on the Prime Minister and owner of Agrofert.

“And one never knows, we might learn further things about how the state institutions co-operate with Agrofert companies.”

Providing a platform for a campaign

Million Moments provide example texts of the kind of requests their followers could make, pre-written on Google Docs, together with instructions on how to use IPV.

“This request is designed for state authorities, typically ministries, while this one is designed for state-owned companies, of which there are still quite a few”, explains the team.

“For example the state still owns a majority share in the globally famous Budvar brewery (brewers of Czech Budweiser, the real original according to many patent law victories around the world!)”

A site for everyone

At mySociety, our charitable status means that we must remain resolutely non-partisan, providing tools for anyone and everyone to use. This doesn’t mean that our partner organisations abroad have to stick to the same principles, though — they will be led by their country’s laws and their own funding structures.

Nonetheless we were interested to ask IPV whether it was a concern for them to be working with a campaign that has a clear political agenda.

They say, “We discussed at some length with Million Moments that the platform should only be seen as a technical facilitator of the campaign. As individuals we might or might not support their goals — but that is irrelevant, really. As an organisation, we’re only interested in providing a clear path for anyone who wants to use FOI to uncover information.

“That comes with some responsibilities. In particular we were concerned that the same few authorities would not be flooded with requests with exactly the same wording, which could incite the dangerous criticism that the platform facilitates spamming or politically motivated harassment.

“We initially suggested the possibility that one “master question” could be put to each authority, and all the other followers could just sign up to follow the requests. However, Million Moments wanted to let people feel they were actively participating, so the compromise is that some  examples are offered as suggestions for questions, but in the end  individuals decide for themselves.”

You can see the campaign page here (in Czech – here’s the same page on Google Translate).

A swell in users

The campaign started with a mailout to Million Moments’ 400,000 followers, and this alone has brought a great result for IPV, a site which was operating with a fairly small userbase. When we spoke to them, it had been live for six days.

“We’ve already got over 400 new users”, they say, “which means we’ve increased our total userbase by nearly 25%, and many of these will likely use the site in the future as they are obviously active citizens. Between them, they look to have placed around 200 questions already.

“We’ll be looking to use this campaign as a platform to build up interest from journalists, who are one of the categories of people who can really benefit from using FOI.

“The Million Moments campaign has definitely given us some momentum! The next burst of interest will probably come when we see how the questions are answered…or not.

“But we have to be pleased with such an increase in our userbase in the space of a single week, especially as we’d expect many of these people to return.

“They are the type of citizens we believe the site is made for.”

We share IPV’s interest in this campaign, and will watch with interest to find out how it develops, and what it might uncover. Thanks to the team for keeping us informed — we always love to hear stories from our many Alaveteli partners about how their sites are making change.

—

Image: Anthony Delanoix

Open Democracy’s recent Art of Darkness report highlighted the worsening state of Freedom of Information request-handling in central government, with concerns over a gravely dwindling response rate, stonewalled responses and a disregard for the ‘applicant blind’ principle.

In combination, these deficiencies have served to erode government transparency at a time when scrutiny is vital. That’s we’ve signed Open Democracy’s open letter calling for an urgent investigation.

In The Art of Darkness, report write Lucas Amin states, “Central government granted fewer and rejected more FOI requests in 2019 than ever before, according to official statistics collected by the Cabinet Office. The percentage of requests granted in full has declined every year since 2010 – from a high of 62% in 2010 to 44% in 2019. The percentage of requests withheld in full has steadily increased from 21% in 2010 to 35% in 2019.”

The report also notes the the government’s increased use of a central ‘clearing house’ through which sensitive requests must be passed. Open Democracy have uncovered evidence that, contrary to the FOI Act’s principle of ‘applicant blindness’ (ie, information is accessible to all, with no consideration of who is making the request), this clearing house, which has been functioning since 2007, is in the practice of identifying which requests are made by journalists and exercising increased caution in their handling.

With this report also picking up many fundamental procedural errors in the way in which requests are being handled, it seems particularly timely that at mySociety we’re working on a tool to help request-makers to understand the reasons for refused requests and guide them in seeking an internal review as part of wider updates within our own WhatDoTheyKnow service.

But perhaps more importantly: as an organisation that campaigns for transparency from our authorities, and works closely with journalists, we recognise the danger of such practices going unquestioned.

That’s why we’ve added our voice to those of the many editors, journalists, campaigners and citizens who call for an inquiry. You can do the same here.

—

Image: Gianluca D’Intino

Here at mySociety, before pressing ‘post’, we sometimes pass the wording of a tweet or blog entry past a couple of colleagues, just to make sure it strikes the right tone. 

So when we saw emails from the Department of Health and Social Care (DHSC) and the Cabinet Office’s ‘Rapid Response unit’ going back and forth to get the wording of a tweet absolutely correct, we sympathised.

“The quote would need to be shrunk down to fit, what should the main focus be?”

“Have added some bits, not sure what the highlighted section was meant to be?”

“[Redacted] wants us to delete the tweet for relationship management purposes and replace with the below.”

“I’ll check at this end, but isn’t doing that just going to reignite?”

“It could potentially reignite it, yes. But the Mail Online did not approach us for a comment and their headline is very misleading so feel we should rebut with less confrontational language.”

“But you can’t replace a tweet? You can only delete and then go back on the original article with a new comment, so you’re rebutting twice, only the second time around admitting that you went too hard first time? Which just creates another story. Isn’t it better to just leave it?”

Admittedly the DHSC’s predicament was higher stakes than ours generally is — they were responding to a piece in the Mail Online and tackling disinformation about coronavirus statistics. The email that kicks off all this discussion reads:

“Flagging growing engagement with a Daily Mail article claiming that Covid-19 statistics around fatalities and hospitalisation have been twisted to create fear among the public (6.6k interactions).

“Although not very high engagement, the article has now been picked up by several high profile lockdown sceptics such as Simon Dolan and Adam Books.

“Given these damaging claims could affect compliance, we recommend that the press office contact the Daily Mail to make them aware of the public health impact, and if possible, include a government line in the article.”

For those who work in communications, and perhaps everyone else too, it is quite interesting to see the authority’s rebuttal process roll into action, with each statement requiring sign-off by a named person, presumably for accountability purposes (these names have, though, been redacted before release).

Available thanks to FOI

How did we come to see these internal memos? Because a WhatDoTheyKnow user requested them under the Freedom of Information Act. 

We can’t know this user’s intent*: were they hoping to reveal evidence that there is indeed a governmental coverup over lockdown, or perhaps to argue the case that there is none? Either way, it seems pretty clear from the released email threads that if there is a conspiracy at play, the staff frantically scrambling to get the right message out to the public don’t know about it.

This request is also notable because the user, spotting that the authority had not provided everything they had asked for, requested an internal review, as is anyone’s right if they believe their FOI request has not been handled correctly.

To DHSC’s credit, they did go on to provide the missing data, and also went out of their way to give some background information “outside of the scope of the FOI Act, and on a discretionary basis”.

It’s worth noting that, for all the effort put in by DHSC’s communications team, however, Mail Online does not appear to have amended the article.

FOI as fact-checker

As we’re in the midst of a fast-changing pandemic situation, it’s perhaps inevitable that there’s lots of misinformation and confusion flying around at the moment — and thanks to social media, it spreads fast.

Freedom of Information requests can play one small part in countering ‘fake news’, by bringing background information into the public domain, helping us understand the full picture a bit better.

Is it always useful for such data to be public? That’s a matter for debate, and a question that is baked into the ICO’s FOI guidance for authorities.

We’ve been doing some in-depth work around exemptions recently, so it is interesting to see COVID-related requests like this one about ill-effects of vaccines in the light of Section 22 exemptions, which cover ‘information intended for future publication and research information’. We suspect a Section 22 exemption may be applied here.

That request is for Adverse Drug Reaction (ADR) data, and includes the instruction, ‘This information should be made available now as raw data, not held back to be accompanied by any analysis.’ 

Would that be desirable, or is the release of raw data just opening the door wide for potential misinterpretation and the drawing of erroneous conclusions? 

When applying a Section 22 exemption, the ICO says that the authority must perform a public interest test to assess whether there is more public good in releasing the requested information than there would be in withholding it. 

Their guidance specifically notes that “In most instances public authorities will not be able to argue that information is too technical, complex or misleading to disclose, or that it may be misunderstood or is incomplete, because they can explain it or set it into context.”

And so, the ideal scenario is that the data is released, with robust explanations and in a way that can be understood by all. That would be a great outcome made possible by FOI**.

* UPDATE: The request-maker has now added an annotation which explains their intent.

** UPDATE: Vaccine adverse reaction data is available, with context, on the GOV.uk website.

—

Image: Garry Butterfield

The climate emergency is, of course, a massive concern, and that’s why we often urge you to contact your MPs and councillors to demand faster, better, greener progress.

And that’s important — but also, we really should take the time to give positive feedback, thanking those councils and politicians who are doing the right thing.

This year, we’re taking part in the Climate Coalition’s Good News Day which, since 2015, has asked “organisations, institutions, household names and millions of people to use the power of green hearts to join together and ask politicians to put aside their differences and tackle the climate crisis.”

Here’s how you can get involved

1. On Friday February 12, use our Climate Action Plans database to search for your local council and see if they have a plan in place.
2. If they have, drop your councillors a line on our WriteToThem service to let them know you appreciate it.
   Local authorities and councillors who are taking action need to know they’re supported in their actions, some of which may be radical or taking them into new territories — so let’s thank them for everything they’ve done so far, and maybe give them the support to go further, too.
3. If they haven’t? Let them know you care about any climate-related action the council have taken, and urge them to get a wider plan approved.
4. Maximise the power of your action by shouting about it on social media. Use the hashtag #ShowTheLove, and use a picture of a green heart (we’ve added links to some royalty-free images below you can download or copy and paste) to join in with the national Good News Day movement. Or, if you want to go all out, make your own crafty green heart: there are some ideas on the Climate Coalition’s worksheet and on cafod.org.uk.
5. If you’d like to do more, see the Climate Coalition’s collection of downloadable resources.

If you’re on a roll…

There are other ways you can #showthelove, too.

We think the prompt to ‘ask politicians to put aside their differences and tackle the climate crisis‘ is a particularly important one, so:

• You could also use WriteToThem to email your MP with this message…
• …or go public and tweet them!

And finally, there is encouragement to share everything your own organisation is doing to help the climate. With that in mind:

• Here’s mySociety’s own environment policy — which any other organisation is welcome to copy and adapt for your own needs.
• And a reminder of our post showing the various ways in which our services can be used to help the climate.
• If you’re one of our many friend and partner organisations in the UK and around the world, please consider joining Good News Day and getting the word out to your own followers.

Green heart pictures

Pictures on Unsplash are free to use and you don’t even have to credit the photographer, although if we’re talking about showing the love, we should of course do the same for the creative people whose work we benefit from!

Top row L-R: Ronak Valobobhai, Siora Photography, Adithya Vinod.
Bottom row: Volodymyr Hryshchenko, Patrick Fore, Bekky Bekks.

In this technical deep-dive, our designer Zarino Zappia explores how we used the latest web performance techniques to make our 2020 annual report quick and easy to download — no matter what kind of connection or device you’re using.

Each year, mySociety produces an online annual report sharing a round-up of our achievements, a thank you to our supporters, funders, and volunteers, and our vision for the immediate future.

For the last two years, that annual report has taken the form of a single, multi-section web page.

This year, with more news than ever to cover across both mySociety and our commercial subsidiary SocietyWorks, we knew it was super important to get that single page running as smoothly and efficiently as possible in visitors’ web browsers.

Our annual report is often shared as an introduction to the work mySociety does. It’s crucial that it sets a good first impression, and performance is a big part of that. It’s also an embodiment of what sets us apart from the crowd – we sweat the small stuff, to give users the best possible experience.

What’s a PageSpeed score anyway?

Google has done a lot of work championing fast loading websites, and their PageSpeed Insights tool is often used as a benchmark of web performance. Developers aim to achieve a mobile performance score of 90 or more – but often this is easier said than done!

In this fairly technical blog post, I’m going to go through a few of ways we really optimised the performance of our 2020 annual report, to achieve a consistent 90-plus PageSpeed score. Through a combination of common sense, technological restraint, and a little extra effort, we manage to serve 6,500 words (a 33% increase over last year), 90 images, and two videos, over nine sections of a single page, in the blink of an eye.

Here’s how.

Keep it simple, stupid

Like most mySociety products, the annual report starts with a very boring, very unfashionable, but very fast base – plain old HTML – in this case generated by Jekyll. Jekyll helps to keep our work maintainable by letting us refactor common elements out into their own includes, but since everything compiles to plain HTML, we get lightning fast page renders from our Nginx web server. We use static HTML sites a lot at mySociety – for example, for our documentation, our prototypes, and even some of our user-facing products. Sometimes simple is just better!

Where we want to improve the user experience through interaction – things like hiding and showing the menu, and updating the URL as you scroll down the page – then we add small amounts of progressive enhancement via JavaScript. Again, like most mySociety products, the annual report treats JavaScript as a nice-to-have – everything would work without it, which means the browser can render the page quickly and our users spend less time looking at a blank white loading screen.

On the styling side, we use Bootstrap to help us quickly pull together the frontend design work on the annual report in about a week at the beginning of December. We build our own slimmed down version of Bootstrap from the Sass source, with just the modules we need. Libraries like Bootstrap can sometimes be a bit of a sledgehammer to crack a nut, but by cherry-picking just the modules you need, you can dramatically reduce the size of your compiled CSS, without losing the flexibility and fast prototyping that a library like Bootstrap provides. The annual report makes heavy use of Bootstrap’s built-in components and utility functions, and then anything truly bespoke is handled by a few of our own custom styles laid on top.

Use the tools at your disposal

Performance tweaks that were cutting edge a few years ago have quickly gained widespread browser support, so if you’re not using them, you’re missing out. Most of them take pretty much no effort to set up, and will dramatically improve perceived loading time for your users.

As a single, big, long page with almost 100 image tags, it’s imperative that we use lazy loading, to prevent the browser requesting all of those images at once on initial page load. Doing this is as easy as adding loading="lazy" to the image tags, but we also add a JavaScript fallback for browsers that don’t support the loading attribute, just to be safe. We even wrap it up in a nice Jekyll include, to make it super easy when managing the page content.

We minify our CSS and JavaScript files, and the server is set to gzip them in transit. These things are really easy to forget, but can make a big difference to load time of a page full of static files.

We also make an effort to efficiently encode our images – serving images at roughly the size they’re actually displayed, to reduce wasted bandwidth, and using WEBP compression with JPEG fallbacks, for the best balance of quality and filesize.

Reduce run-time dependencies

A big part of webpage performance is all the stuff that happens after the browser starts rendering the page. Loading external JavaScript libraries, tracking services, logging services – it all imposes a cost on your users. For some larger projects, you might decide the benefits for your team and the health of the product outweigh the network and performance cost for your users. But for the annual report we did the typically mySociety thing of reducing, reducing, reducing, to the simplest possible outcome. For example…

We don’t load any social media JavaScript. Data is patchy on whether people actually use social sharing buttons anyway, but just in case, where we do display sharing buttons, we use old-fashioned sharing URLs that just open in a new tab. Rather than embedding tweets using Twitter’s JavaScript embed API, we build our own tweets out of HTML and CSS. Not only does this mean we’re avoiding having to load megabytes of third-party JavaScript on our page, but it also helps protect the privacy of our visitors.

Sometimes you can’t avoid third-party JavaScript, though. For example, YouTube embeds. What you can do is defer loading the third-party JavaScript until right before it’s needed. In the case of YouTube embeds on the annual report, we use a little bit of our own JavaScript to delay loading any YouTube code until you click on a video thumbnail. To the end user, the entire thing happens instantaneously. But we know we’ve saved a whole lot of time not loading those iframes until they were really needed.

Unusually for a mySociety project, we don’t even load jQuery on the annual report. While jQuery still makes sense for our bigger products, on something as simple as our annual report, it’s just not worth the filesize overhead. Instead we write modern, efficient JavaScript, and include small polyfills for features that might not be present in older browsers. It makes development slightly slower, but it’s worth it for a faster, smoother experience for our visitors.

Nothing’s perfect!

Overall, some sensible tech decisions early on, and a lot of hard work from the design team (including our new hire, Clive) resulted in a 2020 annual report I think we can be really proud of. I hope it’s also a taste of things to come, as we start introducing the same techniques into our other, bigger products.

That said, there’s still a few things we could improve.

We could achieve almost instant rendering on first pageload by extracting the CSS required to display the first few hundred pixels of the page, and inlining it, so it renders before the rest of the styles have been downloaded. We already do this on fixmystreet.com, for example, using Penthouse.

There’s also an external dependency we could drop – Google Fonts. We’re currently using it to serve our corporate typeface, Source Sans Pro. The advantage of loading the typeface off Google Fonts is that most of our other sites also load the same typeface, so chances are the user will already have the font files in their browser cache. Self-hosting the fonts is unlikely to be much faster than loading from Google, but it would be a win for user privacy, since it’s one less external resource being hit for each pageload. Something to consider for future!

Photo by Indira Tjokorda on Unsplash

We have the opportunity to help one organisation in Europe set up and run their own Freedom of Information website. Could you be that organisation?

The background

Thanks to ongoing funding from Adessium, we’ve been working with a number of partners right across Europe to set up new Alaveteli websites, and upgrade existing ones with the Pro functionality. The ultimate aim is to increase the quality, quantity and simplicity of European and cross-border Freedom of Information based investigations.

So far we’ve helped organisations in France and Netherlands to launch their own sites, and we’ve added Pro to AskTheEU, Belgium, Sweden and Czechia.

Now we have space to provide technical help and support for one more organisation who would like to launch their own brand new Alaveteli site.

What would that involve?

Running an Alaveteli website is no light undertaking, we’ll be the first to admit it. While we can help you with all the technicalities of getting the site up and launched, there is an ongoing commitment for the recipient organisation, who will need to factor in significant time to administer it, moderate content and help users.

On the plus side, we have masses of experience that will get you set off on the right footing; we’ll do most of the technical stuff for you; and there’s a global community of other people running Alaveteli sites who are always quick to offer friendly advice when you need it.

OK, sounds good – can we apply?

There’s just one important detail: we’re looking for organisations in European countries or jurisdictions where there isn’t already an existing Alaveteli site. Take a quick look at our deployments page to see whether your country is already on the list.

That’s the main requirement — but there are also a few details that the ideal organisation would fulfil.

• So that you understand the service you’d be offering to citizens, you’d already have transparency or freedom of information as a remit or strand of your work
• You might include some people with at least some basic technical or coding skills amongst your workforce;
• You’ll have a source of income (or plans for how to secure one) that will allow you to keep running the site after we’ve got you all set up.

We’re looking to start work in April, with a probable build phase that would take us to December 2021. All work is conducted remotely, and we’d have regular check-ins with you via video call to keep you updated.

We’d then give you all the support you needed in the first few months after your site’s launch, then from March 2022, you’d be all set to take the training wheels off — although, as we say, we and the rest of the Alaveteli community would be around to offer help and advice on an ongoing basis.

Right, that’s everything — so it only remains to say that if you’re still interested, please get in touch to have an initial chat. Or, if you know any organisations that might be a good fit for this opportunity, please send them the link to this post.

—

Banner image: Gia Oris

On 18 January we received a letter from Robert Largan MP regarding our parliamentary site TheyWorkForYou. He requested that we ‘correct a misrepresentation’ in the way that the site displays how he and his fellow MPs have voted on measures to prevent climate change:

The letter was co-signed by around 50 members of his party, and identified three votes not currently included in our climate change vote calculations, with the request that they be taken into account on their voting records pages.

This was not an unusual message: we often receive emails from MPs to TheyWorkForYou, asking us to explain or reconsider the data we publish on them — and the most common subject is the voting records pages.

The only differences with this letter were that Mr Largan had gathered the support of so many other MPs; and that it was covered in the press and shared on Twitter quite a few days before we actually had receipt of it. 

So we’ve treated it in the same way that we would any other, but given the amount of exposure the issue has already had, we thought we’d also share our considered response here. We’re glad to have this opportunity to illustrate how we run the site, and the judgements that we have to make in order to run the fairest, most factual service we can.

—

Image: UK Parliament