Tell us your barriers and blockers to updating to GOV.UK Frontend v6.0.0

[calvin-lau-sig7]

About this work

The GOV.UK Design System team is working on our next breaking release, GOV.UK Frontend v6.0.0.

We want to improve the communication, guidance and user support we provide for this release.

What’s included in the upcoming release.

The release notes for our v6 beta releases (GOV.UK Frontend v6.0.0-beta.0, v6.0.0-beta.1 and later) show some of the things that will be in the breaking release.

Discussion pages in the Brand refresh category also show some of the design work going in as well.

How you can help

We’re looking for ways to inform and encourage service teams to update. This starts with making the update process and guidance as simple as possible — and we’re interested in finding out what else we can do.

In particular, we want to find out about the blockers and barriers that users in service teams face and ways we can help you overcome them.

For example, these could be issues like:

• understanding what actually needs to be done to update
• work needed due to breaking changes
• long-standing issues that need to be addressed first (sometimes known as technical debt)
• difficulty making updates to GOV.UK Frontend a priority on your team
• getting buy-in from leadership to work on this

If you’ve got more technical thoughts and comments, we’ve also started a GitHub discussion page for the GOV.UK Frontend 6.0.0-beta releases.

Other ways you can give feedback

If you’d rather give feedback another way, you can:

• email the Design System team
• join us on #govuk-design-system channel on UK Government Digital Slack.

When to give feedback by and our next steps

We’d appreciate hearing from you before 12 January 2026.

We’ll use your feedback to decide how we can support service teams with our next breaking release, GOV.UK Frontend v.6.0.0.

[querkmachine]

@JamesFNGibbons Please keep your comments genuinely constructive and remember that there is a code of conduct you must follow to contribute to any GDS repo or discussion.

It's rather difficult to parse from your post what difficulties you actually had given the majority of components, documentation, and the installation process did not significantly change between v4 and v5.

[querkmachine]

@JamesFNGibbons I wasn't the one who marked your comment as spam.

Regardless, your attitude and comments are not constructive. Continued trolling will not be tolerated.

[colinrotherham]

Have you got a Benefits of updating to v6 list, like the one for v5?

[colinrotherham]

Thanks @owenatgov

We learned for v5 that the "What's in it for me?" question was important for prioritisation

Keep up all your fabulous hard work—but Sass modules might swing it 😉

[owenatgov]

We'll get to it 😁

[owenatgov]

Being serious I do agree. There's a super interesting content project/challenge where we better enable both helping people understand what work is required to upgrade and why it's worthwhile, as well as why keeping up to date is worthwhile in the firstplace and journeys there. Keen for us to get onto it. This feedback is useful for that!

[colinrotherham]

Yeah 100%

Many teams won't be able to pitch "new ways to access colours" as a reason to prioritise v6 into the next sprint

[JamesFNGibbons]

I would also be interested in seeing this document when it's ready for release

[ch-ncolley]

One thing I have done for Companies House is build a bespoke upgrade guide for our applications specifically, I wonder if given that many users of GOV.UK Frontend will be using Express if a real world example of changes would be useful, but I suspect this may be better coming from the community maybe.

To put together that guide I did have to go through and parse what was related to what things, e.g. we use Nunjucks so I went through the CHANGELOG and disregarded anything that we'd get automatically because of Nunjucks usage.

Maybe there's something to explore around filtering and grouping changes based on types of usage? e.g. if you're not using Nunjucks, you need to do these steps.

As it relates to getting buy-in, it's hard, there has been buy-in for the rebrand so that's allowed for some upgrade work to happen but when no visual changes are mandated it becomes harder.

I think making the case around accessibility from a legal perspective might be useful in a similar way to security vulnerabilities e.g. making it really clear which versions support which version of WCAG, and that older versions are potentially inaccessible and therefore a risk. Much of our maintenance efforts outside of frontend stuff are backed by the amount of security vulnerabilities e.g. CVEs and then they get their dependencies updated as a result.
I havent thought about this very deeply though so just spitballing.

Final thing I've learnt is that it's common for services to stay at an old version and not get upgraded if there's no one working on them, so the guide I wrote for example can upgrade a 3.x service to 5.x. I'm not sure what the takeaway from that is but consider that some teams may be upgrading from 3.x -> 4.x -> 5.x -> 6.x when they do get around to doing it.

[ch-ncolley]

Ah one more thing, many people just dont know that new versions are released, so maybe some wider comms / advertising is worthwhile.

[JamesFNGibbons]

I thought that support for Nunjucks was dropped? I remember our department removing it from projects because of the well documented security concerns.

[ch-ncolley]

Nunjucks is still supported in the GOV.UK Design System.

The Nunjucks 3.x release has not had any active development for a while but this is mainly because the project is a port of Jinja and apart from a few features it's "done". The core maintainer started efforts on a 4.x release (https://github.com/nunjucks/nunjucks4).

While I think there is a valid worry of a vulnerability being found and the project not being updated quickly, the core maintainer has said he doesnt have time to do any feature work on the 3.x release, but last time there was a vulnerability he did put out a fix - I would imagine since he's still around doing work on 4.x that a fix could go out as it's not an abandoned project. See updating the test runner last week: mozilla/nunjucks#1516

Many of the dependencies we rely on are in a similar position, robust and stable and not updated for a long time, I don't think that's a high enough indicator of security concern to abandon that dependency.

Companies House is a smaller department so it saves us time being able to reuse the templates without having to port them, a lot of teams underestimate the effort require to keep a port up-to-date.

Coming back on topic the strength of Nunjucks is in it being stable and not changing, so that's always something to consider when changing public APIs.

[JamesFNGibbons]

Makes complete sense. I wonder if there is bandwidth to fork Nunjucks and fix some of the issues internally?

[jabley]

Might it be possible to use something like OpenRewrite to provide a smoother upgrade path for the community?

[36degrees]

Thanks for the suggestion, @jabley. I'm aware of CodeMod which looks like it operates in a similar space, and I have done some very rudimentary exploration of what's possible using that. We have not seen OpenRewrite before though.

Whether we have the opportunity to get something in place for v6 or not, it feels like something worth exploring at some point.

[JamesFNGibbons]

Will there be any support for WinForms in the upcoming release?