Dashboard Initiative

I started working with Drupal 15 years ago. I had previous experience contributing to open-source projects, but PHP wasn't my strength, and to be fair, it took me a bit to understand Drupal and most of my work could be done via site building. Some months later, I decided it had to change and attended the Drupal Developer Days 2012 at Barcelona (wow! the site is online! 🤩🤩). There were coding sprints planned on the Drupal 8 Multilingual Initiative, which actually was the beginning of this blog. I blogged about my own progress every day. 

Keeping track of what was going on in an initiative was affordable, thanks to the great job Gábor Hojtsy did leading it. We had IRC, our weekly meetings... I followed the D8MI tag on Drupal.org issues, but that was a manual process. I tried every single thing I could think of for tracking the issues I actively worked on, those that I had an interest in, and those where I would like to help if I had the time. Nothing really worked well for me.

On Jul 1st, 2013 (almost 13 years ago!) I decided to subscribe by email to ALL Drupal Core issues. All, because you cannot subscribe by email to a tag. So it was all, or missing information that I cared about. This was a huge game changer.

First, the filters + labels system I developed — using different tags with associated colors in Gmail — helped me keep track of every issue I was involved in or that had certain tags.

Second, because search is amazingly fast (and works even when there are Drupal.org outages).

And third, because I have a backup of Drupal.org core issues 😜

Image

But the real game changer is seeing things passing by. As you can see, I learnt quickly that infoxication was stressful, and assumed that Inbox-Zero wasn't for me, so I learnt to ignore the number of unread mails. But just by scanning those labels I could have a big picture of what was going on because of the email subject. If something catches my attention, I open the email, learn more about the issue, and there's a high chance I will follow it.

This, plus being on drupal.org and Slack, kept me pretty well informed. So good that this has worked for me for 13 years.

But priorities change, and that an issue exists doesn't always mean that it will land. Tracking what's actually making it into the project is a different beast. I was missing this piece of information between the noise of every single issue. Reading every commit for me was out of the question.

I think it was in 2023 when I met Marco Villegas (marvil07). What a great and passionate guy. So passionate that he shared that he would go through the last commits to Drupal Core, read them, and create and share a journal summarizing what was every commit about. He spoke about it in DrupalCon Barcelona 2024, and you can check his session Following Drupal core development: Is it possible to understand every added change? on YouTube.

But what's a blogpost these days without AI? Dries himself shared recently about his new process for following Drupal projects, and how AI is helping him to summarize the information for the increasing number of projects he needs to track. This is a huge resource. 

For the last month, I often visited his Drupal Digests, and looked around the repo to read the digests, but issue numbers are not super-helpful to keep track. Dries built this to scratch his own itch, so it's highly focused on RSS feeds. And my problem is that I don't really use RSS feeds anymore. Turns out, that wasn't entirely true.

I helped build the Dashboard module for Drupal CMS. There's an aggregator module (used to be, but no longer in Core) that I had installed in this blog, and where I already had some of the feeds I care about. And I visit this blog a lot to create drafts that go nowhere. 1+1 = 2.

So my email process is now complemented by a Dashboard, using the aggregator module provided blocks.
It didn't look pretty, so I had to include some custom CSS. This is how it looks now:

Image

If you want to replicate this, you'll need to inject a CSS similar to

/*
* Aggregator block theming
*/
.block-aggregator-feed-block {
  margin-block-start: var(--space-s);

  & .item-list ul {
    margin-block-start: 0;
    margin-inline-start: 0.875em;
    list-style-position: outside;

    li {
      list-style-type: none;
      margin: 0 0 0.25em 0.25em;
      padding: 0;
      padding-block-end: var(--space-m);
      padding-inline-end: var(--space-s);
    }
  }

  & a {
    text-decoration: none;
    color: var(--color-absolutezero);

    &:hover {
      text-decoration: underline;
      text-underline-offset: 2px;
    }
  }

  .more-link {
    margin-block-end: var(--space-m);
    margin-inline-end: var(--space-l);

    a {
      text-decoration: underline;

      &:hover {
        text-decoration: none;
      }
    }
  }
}

I considered including this in the Dashboard module styles itself, but might be too much opinionated. What do you think? Edited: created issue to discuss.

And how do you keep track of what's going on in Drupal?

I haven't kept up with updating this blog, so this is a short summary of what happened, and where we are with the Dashboard Initiative.

Our initial goals

When the Dashboard Initiative was announced, the goal was to get this into Drupal Core. 

As with most initiatives, we started in a sandbox, then moved to a contrib module.

Based on the above, our idea was

• providing a very simple Dashboard framework,
• according to Drupal core standards,
• based on layout builder,
• providing a set of blocks that could cover the requirements for 80% of sites
• that users would see first thing when they log in to their site,
• and with nice defaults for the standard profile in Drupal core.

We had weekly meetings, where we worked on defining what the defaults for the Standard profile would be (with Aaron, Cristina and Megh leading the way), while the module was steadily progressing. We attended many events, where we (Cristina, Pablo and I) spoke about the initiative, gathered lots of interesting feedback, and could work on sprints where we got dozens of new contributors to help with the initiative.

Fast-forward to the Drupal Starshot initiative

And that was our plan for DrupalCon Portland 2024. But we were surprised with the Drupal Starshot announcement. Our session about the dashboard, scheduled just after the keynote, was quite an experience for Pablo and me.

The Drupal Starshot initiative has evolved, but its foundation was packaging a new version of Drupal providing a great out-of-the-box experience. So we pivoted and the Dashboard initiative became one of the Starshot Tracks that would become Drupal CMS 1.0.

In a few months we were able to redefine our goals and implement what was needed to launch Drupal CMS on January 2025, which included the Dashboard 2.0.0 first release.

Shortly after, I was appointed as one of the few Drupal CMS committers, which I consider a huge honor. 

Is becoming part of Drupal Core still the goal?

Drupal CMS is the promoted option for downloading on drupal.org. In DrupalCon Atlanta 2025, exploring a marketplace of site templates was announced, which would be based on Drupal CMS 2.0. Given that, having the dashboard as part of Drupal Core shouldn't be a priority, even if it would be satisfying.

In Dries' July 2025 update to the Drupal core strategy 2025-2028, you can find good arguments for both including it or not including it. If I leave my bias aside, I'm 50-50. The tie-breaker would be adoption.

Dashboard adoption

Before 2.0.0 was released, I worked on a proof-of-concept for updating Dashboards with Layout Builder module to be based on Dashboard, to ensure this was possible with an automatic upgrade path. That's feasible, but never really happened.

Of course, Dashboard adoption is great. Even if it's short-lived, since it's part of Drupal CMS and the future site-templates on top of it. So I'm looking at what distributions outside of that are adopting it.

The last releases of the Varbase distribution moved to Dashboard already (and they have been really collaborative on issues on the Dashboard queue, thanks!). 

Drupal Commerce introduced its own dashboard functionality in their 3.x releases ^[1]. I think adoption on high profile distributions/modules like Commerce would help make the point on having this as part of Core. 

Another case I'm looking at is drupal.org itself. With their upgrade to Drupal 10, it is quite possible some kind of dashboard is needed on drupal.org too for feature parity with what we have today. Opening an issue to see if they already have something in mind is on my to-do list.

What's next then?

If we look at our initial goals, providing a set of blocks that could cover the requirements for 80% of sites is not something I'm looking at right now. Thanks to Drupal's new recipes system, I don't think that belongs in the dashboard module itself anymore, but on dedicated recipes (as Drupal CMS proves) or modules.

We've been working on bugs, minor features (like our new coffee integration!), and keeping up with new core releases and PHP new features. We just released 2.1.0-beta1, which should be shortly followed by 2.1.0.

So to be honest, I don't know what's next. Feedback is welcome on the issue queue or at the #dashboard channel on Drupal Slack.
 

^[1] I was corrected that this happened with Commerce 2.37.
 

Photo by Paul Johnson, available for publishing under Creative Commons Licence with attribution via Flickr.

Last year, during DrupalCon Prague, Cristina approached me with an idea to include Dashboards in core. We met with Sascha, and since then a team emerged and we have been meeting more or less regularly defining how this would look like, and creating some proof of concepts.

What are we trying to solve?

When you log in to your Drupal site, you land on the /user page. There you can see how old your account on this site is. Useful, huh? Unless you have customized this behavior with contributed or custom modules, now you need to think about why you logged in, and go to wherever you can do what you wanted to do, which usually will take you several steps.

There's a big chance that this journey will start at the Content page, if you are a content manager or a content editor, but you might be a site builder, a site admin, or be responsible for other duties on your website. So we need a landing area where the content shown makes sense based on your recurring tasks. This is where customizable dashboards come into play.

Wearing multiple hats

Users often have multiple roles, or they might want to perform different tasks which might be unrelated to each other. The initial idea was to have a dedicated dashboard for each role.  That evolved into the idea that a user should have access to different dashboards, where groups of tasks and information might be grouped by their nature, not neccessarily by their role. You might wear your content manager hat one minute, and later you might need to perform some SEO duties.

So nothing should block you from having unlimited dashboards, and use the permissions system to give access to them to different roles.

Is this aligned with Drupal core goals?

Drupal Core Product Manager Lauri Eskola recently stated the three tracks were most efforts are concentrating in Drupal Core: 

• Reduce the time it takes for site builders to become proficient with Drupal.

• Empower site builders to deliver engaging editorial experiences.

• Reduce the cost of keeping Drupal applications secure.

Adding dashboards helps with the second track, enabling site builders to make editors´ lives easier by providing customized dashboards based on their editors´ needs. There would be benefits to the other two tracks as well. Having clear journeys on some common tasks for different roles will help site builders to become proficient in Drupal; and can reduce the cost of keeping Drupal applications secure if we surface the need for security updates on the dashboard and make it easy to get up to date with security updates or news from the Drupal project and the Drupal Association.

How to get this into Drupal core

When new ideas go into core, usually they are introduced as Experimental modules, and that's also our aim. Similar to other existing admin tools (like the content listing admin page) in Drupal core, our approach is based on progressive enhancement. This translates to:

• If you don't enable dashboard, you will be redirected to /user as of right now. Drupal core Standard profile cannot depend on an experimental module, so it won't be installed by default (for now).

• We will ship Dashboard module with default dashboards for common tasks in core. 
• If Layout Builder is also installed, you will be able to customize the layouts and blocks of those dashboards on the UI. 

How to get involved?

For now, we've been working in a sandbox, and we also have a GitHub project that contains an easy-to-evaluate site with some demo content, and triggers a Tugboat that we can use as demo or to manually test changes. Thanks to Tugboat.qa for providing that, and James for helping out setting it up!

If you want to help, we welcome you at the #dashboard channel on Drupal Slack.

Thanks to Gábor Hójtsy, Lauri Eskola and Cristina Chumillas for reviewing this blogpost.