May 24

Chrome Web Store; the opportunity for breaking out of silos with experience

Tech, Web Browsing with tags: , 2 Comments »


I am sure that we are going to see an explosion of people learning how you can create Chrome apps simply by creating a manifest and building a .crx as they rush to get their web apps into the Chrome Web Store.

This is a huge advantage of adding an “app” model on top of Web technology. We have been creating Web apps all along, and now we are getting to the packaging layer of the stack as we turn the Web inside out. Instead of hosting our apps on our own URLs, we are now packaging them up and putting them in external app catalogs for extra distribution. Kinda weird in many ways huh? ;)

Moving past simply wrapping up your Website so throw it in a store, I think we have an opportunity to do much more here.

With Greasemonkey user scripts, I am able to both personalize my experience across the Web, and pull in experiences.

I was just chatting with Phil Windley about Kynetx, his company that is doing interesting things in the “professional Greasemonkey” / enterprise client side mashup space. He showed me some of this in action. What if you are a member of AAA. By installing an app/add-on you get to carry that information around the Web with you. AAA has a ton of discounts that you never really know about. What if when you went to purchase something at it told you that you can get 10% off with a given AAA code? could choose to display all of the affiliates that it has, but that doesn’t scale. Instead, show me the affiliates that I have a relationship with!

Here Phil walks through a Twitter mashup example:

I want to break out of the app silo. I was drawn to webOS synergy because it allowed me to do just that. Our Facebook app is only part of the Facebook story. A bulk of the Facebook functionality is spread across the entire device. If a user enters their Facebook account, that follows them. Want to share a photo? Now you have a Facebook option. If users choose another social network, that can follow them instead. I don’t want to have to open an app to do everything.

This is where the trojan opportunity of the Chrome Web Store comes into play. Being able to package the Web application with extension like ability is key. Like Gmail? Download the app and:

  • Works offline
  • Nice big icon to click on
  • Permissions problem “solved”
  • Takes the Gmail experience throughout your browsing (mailto: goes to gmail, input type="email" auto completes on Gmail)

One great advantage of the Web is the mashup potential. There are a huge number of Web experiences that are vertical (go to a Web page to experience) as well as horizontal (having a favourite mapping service). The browser itself can add hooks (e.g. map: hooks into GMaps or Ovi Maps or MapQuest or …), but the browser is always behind the times. We can allow extensions to do this work. We can install a local library “web app” that means that wherever you see book information (e.g. Amazon, BN) you also get info about whether the book is available at the library. Right in the page.

I hope that the Chrome Web App store doesn’t become just wrapped Web pages, but we move the state of the art and break out of the silo.

May 19

What should the future of Web App Stores be?

Google, Tech with tags: 2 Comments »

NOTE: These are my random thoughts on Web App Stores and do not necessarily reflect the thoughts of Palm. Don’t read anything into this :)

App Stores. Catalogs. Markets. Gardens. They have been an incredibly hot topic ever since the success of the iPhone App Store. Everyone has wanted to own and control an app store ever since, for many different reasons. Not all of them are good. Not all of them can extract from the App Store experience.

With that, today Google unveiled a Chrome Web Store. Many folks have talked about how the Web, and the Open Web, could benefit from an app store. At a high level, the Web should give developers great options to distribute and monetize (if they so wish) their products, projects, and abilities.

There was a phase where people assumed that Internet meant “everything would be driven to be ‘free’”, and the model to make money is ads. Ads can potentially be a great avenue depending on your product, but people have always been willing to part with money to get value (or perceived value) back… and it shouldn’t be too shocking to have see that play out again via app stores. Deliver great experiences and they may come. There is a lot to be said about “expectations” though, and I feel for some of the business models that are in trouble (media etc).

So, it would be nice if a developer could build something useful, stick a price tag on it, and sell it… using great distribution channels. One model for this is to package the application in a way that works in the distribution channels (e.g. PhoneGap, Titanium, or for us on webOS…. natively) but surely we can do better?

When you look at what an app store is, it consists of many pieces indeed.

For example,


Application Platform

This is the set of APIs and technology that you can use to build your applications. We feel strongly that having many fragmented devices with their own set of APIs, and thus their own platform, is a path to madness. To fix that, we need a unifying platform. We could unify on a proprietary platform (e.g. Flash, Cocoa, Silverlight/.NET, or what have you) or we could learn from history and honour the very lucky place that are in as an industry, and choose Open standards… and in concrete, the Web.

The Chrome Web Store very much honours this. Their apps are just Web apps. This is fantastic and a big step forward. The fact that you could take the same codebase and ship it in a store that runs in Chrome and other browsers (and web OS’s :) is great news.

But this is only one level of the stack.

Application Packaging

How do you package up your application to place in a store? Chrome has a prelim crack at this with their crx format that we have seen via Chrome Extensions. There are also others:

W3C widgets, Android, webOS, Nokia WRT, and more. Libraries like PhoneGap already have to deal with this world. We need scripts and tools to navigate this world. In the short term lets create and use those tools, but how about all getting together and making a format that we can all live with and extend in the right ways?


How do you distribute your apps? Right now you either: a) put up a website and folks find you (via search {and thus SEO}, the social web, and links), b) put an application into a store itself.

The Web won on distribution. AOL couldn’t hold back the tide. Everything was Out There. Do we really want one or two companies in charge of what gets Out There? There is certainly value to curation, and finding applications via places you trust, but I personally want to see us solve the problem is a distributed open way. No entity should own the pipe. There should be no “right way” to censor or review applications. Different systems can choose values. Ideally they are transparent, and consumers can then choose where they want to look for the app content.


If we had a standard format for “installable web apps” a lot of good can happen. If we had a standard format, we could annotate the Web with it. Browsers could consume it (e.g. App Discover and the talk on just this.) and search engines could index it.

If I am on a Web site that has an application available, the browser should let me know. If I do a Google search, a one box should tell me about the matching apps for the platforms that I care about.

Anyone could create a merchandising experience and have access to apps to search for. We can go beyond that too of course. We can have web hooks that ping services with information about applications that have been created. A meta service would crop up that looks for the data and then pings the various Web app stores, just like we saw with trackbacks and the like.

There is still a ton of room to innovate in discoverability. Smarter searching that ties deeper than matching on the content will kick in.


Having access to the application data is important. The Palm ecosystem shares full feeds of the data from the catalog and this allows anyone to merchandize the content. They can be creative and present the right content to the user. There is a lot of innovation to be done here. Ideally, someone could create a great new social algorithm and have it run against all of the web apps out there. Right now, the stack is monolithic and we only have basic merchandising needs in most of the ecosystems. What about affiliate systems and the like which could enable innovation here?


How do you pay for a good. How does the developer get paid for the good. This is an obviously crucial layer. Apple has done very well because people already had iTunes accounts coming into the world of the iPhone. In a distributed model, users could choose to fulfill via Amazon, or Paypal, or Checkout, or SimpleBank (one day right al3x? :)

And more….

We haven’t even gotten into identity, single sign-on, reputation systems, and how reviews can be shared on the same applications in different stores. There is so much to be done here.

Looking forward, where does this next step take us? Is there a path where web applications themselves are back to just being websites that can live in various stores, and a pay gate can be setup in an easy way?

I can’t wait to see how this plays out. We could all go off and build ivory towered app stores, or we could come together and work out a better way. Are you in?

May 15

Palm Developer Day: Walking through the announcements in our keynote

Palm, Tech, webOS 2 Comments »

Ben and I gave a keynote at the Palm Developer Day where we set up the event, and also shared some of the technology that we are working on for developers. There are a lot of fun APIs, such as:

  • Accelerated CSS transforms and animations
  • JavaScript Services: You have always been able to publish messages to the Palm service bus (e.g. launch an application) but how about being able to subscribe and run code on the flip side? You will be able to do just that with JavaScript Services. You will get access to a slew of libraries in that world (your own process too!):
    • Background asynchronous services
    • File I/O
    • Binary JavaScript types
    • Low-level network I/O
    • “Medium-level” HTTP via CURL
    • Fast XML streaming parser
  • db8: what if you had access to a fantastic performant native JSON store? That is where db8 comes in, our new open source JSON datastore that includes:
    • Native JSON storage with query mechanism
    • Built-in primitives for easy cloud syncing (Easily query changed / deleted data, Designed to sync with CouchDB in the cloud)
    • Fine-grained access control for apps
    • Mobile-optimized and fast (especially for updates)
    • Pluggable back-end
  • Audio / Video Recording
  • Media Indexer Access
  • Secure Key Manager + Crypto Libs
  • Bluetooth SPP
  • Bonjour / Zeroconf

We just had fun doing a podcast on this and more. Join us as we go through the tech, and we also hope you enjoy the sales pitch at the 41 minute mark (not what you may think ;)

May 07

Watching you Mum use the iPad is fascinating :)

Mobile, Tech, UI / UX 1 Comment »

I got my Mum an iPad. I wanted to get one for her more than me (even though I got one too of cos ;) as it seemed like a perfect tool for her to consume content. Finally, I get to hand her something that won’t just break. She doesn’t need to be scared to use the device. One button. So much simpler for her than a normal computer.

She has loved it. My Dad has told me that she is always on it. This week she flew in and stayed with us, and I got to watch her use it first hand, and it was fascinating.

I had my brother setup the device for her first time around. He installed a bunch of apps, and must have shown her how to add a bookmark to the home screen. I say this, because when I asked her whether she had the Foo app, and she said “Yup, here it is.” However, it wasn’t that app at all, but rather saved onto the home screen. She had no idea that there was a difference! Now, I personally believe that the world of URLs on the device vs. apps should entirely blur…. but it already had to her! I then had to show her the “App Store” and how to install apps.

This is when I learned how poor the UX is for downloading and installing apps. I have thought this before, but didn’t realise how bad it truly is until I watched my Mum.

She had NO idea how to install an app. She could find them, but couldn’t install them. It turns out that the UI for doing an install is incredibly poor. You have to know to tap on a small [FREE] or [$1.99] button which then turns into the [INSTALL] button. It is small, and not obvious at all. I wonder why Apple hasn’t noticed this and made a more obvious [TAP HERE TO INSTALL THE APP] flashing green button?

This is bad on the iPhone:


But with large real estate on the iPad, it is really hard to find:


It is also hilarious when you see someone struggling on what to do with a device that has ONE button. I love you Mum. Thanks for letting me see a UI through different eyes this week. I hope you continue to enjoy the iPad!