SLICE: The Web

There have been a lot of conversation about all the issues of the web over the last couple of weeks and they broadly group themselves into the following categories:

  • Performance
  • Lumpy inconsistent browsers
  • Rapidly expanding feature landscape.

I want to put these to the side for a couple of minutes to quickly talk about one of terms that we have used in Google to quickly describe the positive aspects of the web as a platform for users and developers: SLICE.

I can't find an original reference for it, but the underlying points that I will go into are well known. SLICE was mentioned at the first Chrome Dev Summit by Linus Upson in the 2013 keynote. When Linus talked about the properties of the web it wasn't in the correct order for naming but I do encourage you to watch this video. Note: Brett Cannon, a Microsofter (formerly a Googler) also recently mentioned it and it's a good read and has broadly similar conclusions to my post about Living with Web Apps

I think it covers a lot of good points:

  • Secure - All domains are sand-boxed from each other and sites are sand-boxed away from the users machine. The user can go to any site and know they are safe.
  • Linkable - You can point to any page or piece of content just by sharing a URL
  • Indexable - Because you can link to anything, if public it can be discovered by any person or machine that can index it to make it universally discoverable to everyone.
  • Composable - Iframes and JavaScript allow us to quickly compose and embed new sites, apps and services just by dropping in some JS and hooking things together.
  • Ephemeral - There is nothing to install, you go to the page and interact with it, leave the page and when you do it stops taking up resources.


As a set of abilities that the web encapsulates SLICE principles are well known yet frequently forgotten when considering the competition of native platforms.

As a term, I find that SLICE is a great way to quickly address benefits of the web today. It misses out a couple of major benefits of the web such as the ability to deploy updates instantly — SLUICE isn't a great acronym — but that is ok, SLICE as an acronym works well.

I use the SLICE model as a baseline for where we are going with the future of the web and the challenges we face and need to overcome to get there.

  • Secure - The web should remain sand-boxed and it should be end-to-end encrypted. We also need to work out what the model is for ensuring the user is in control and aware of how permissions to advanced API's.. For example, we recently started to ship a Bluetooth API how do you give users the confidence that it is safe and secure to use.
  • Linkable - I started this with Web Intents and although it ended early, I believe we are due for another generation of linking content, sites, apps and native experiences. Some of it requires new technology some of it requires education.
    • Linking in to Web Apps: I am going to dive in to this another time. TL;DR - product landing pages and log in pages don't help us link to web apps.
    • Deep linking into media: browsers for a long time have been able to link to any part of a file yet no one seems to do it.
    • Deeper linking in to text: The first time I saw this was Dave Winer's blog where you could link to any paragraph, more recently Medium are giving every paragraph a deep link.
    • Linking Real world Objects: The physical web for discovery of "things" around us, and new API's for talking to these "things" will reduce the friction in our every day lives.
  • Indexable - The headless web, i.e, parsers and indexers are getting more advanced allowing us to understand more about the content that is on the web, they will run JS and understand visually how the page renders but there are a lot of issues still:
    • Embedded can't describe semantically correctly (hence JSON+LD)
    • Media does not have a huge amount of meta data exposed in a public format.
    • Apps: Web Intent's tried to be a way that described what a web app could do. We don't have that any more and we are missing a massive way to describe the capabilities of what a web app can do. Take my airhorn app as an example, even though I expect no one to need horn functionality in their app, there is no way to find it other than to search for meta-data and it is one of the reasons why we have product landing pages on the web.
    • Internet Connected devices aren't being indexed and they don't describe what they can do. It's a missing piece for my in the Physical Web story, discovering capabilities. I feel like we need a Web Intents for IoT.
  • Composable - It would be easy to mention just only mention Web Components, but really we are talking about the broader ecosystem of reusable tools, libraries and frameworks:
    • There are massive issues of interop at the moment as frameworks try to own the entire stack.
    • We need to solve client-side delegated functionality. Web Intents tried this, but much of it is possible still on the web today but we don't do it. i.e, I have created a QR code snapper web app, why do you need to build one yourself to integrate it in to your own app, just use mine or any other pre-existing service.
  • Ephemeral - Two words: Service Worker.
    • Installability is the antithesis of Emphemerality. By it's very definition, when you install something it becomes a long running and integrated part of the device. Service Worker can give use the best of both worlds: a middle ground letting you choose how and when the site should be integrated deeper into the device. Combine this with the manifest and the user now has the choice of installing the "web app" or keeping it as an as-needed interaction.

So what are we missing? I am going to leave that for you to tell me, I suspect I am missing a lot. I do have a set of follow up posts where I will talk about how native platforms are taking a piece out of SLICE model for themselves as a way to entrench native apps further in to the everyday lives of users and how the web can differentiate even further.

Image Credit: Justus Hayes

About Me: Paul Kinlan

I lead the Chrome Developer Relations team at Google.

We want people to have the best experience possible on the web without having to install a native app or produce content in a walled garden.

Our team tries to make it easier for developers to build on the web by supporting every Chrome release, creating great content to support developers on, contributing to MDN, helping to improve browser compatibility, and some of the best developer tools like Lighthouse, Workbox, Squoosh to name just a few.