Is the browser on life support?


I've said in the past that I haven't alot of faith in today's browsers going forward as I think they have outlived their need and what I'd dare say is we need to realistically think about what can be done better in this space. It won't happen over night and it make take years to wind down from the browsers we see today, to the next "agent" we use for the web - but - I have a theory that it will happen eventually.

It's not a popular theory and a lot of people think it's a looney one, but I stand by it as I think software has radically changed in the last 5 years. We are seeing more and more applications becoming context specific in their approach, solving very small problems pieces at a time and doing so is becoming more and more relevant to our day to day lives (i.e why does concepts like SOA or SaaS exist?).

An example would be Flickr, although its currently adopted a browser focused approach to its distribution, it however needs to evolve and grow further in its empowerment to its consumer(s) overall. Yet, I'd wager they (Flickr) are smart enough to know that to move forward in this manner will require less bleeding edge technology but a reliance on popular "agents" in opening their reach/growth pontential.

NY Times reader built in WPF That being said, say Adobe Apollo or Windows Foundation Everywhere (WPF/e) gain massive amounts of popularity in the next 1-2 years, this could open up more possibilities for companies like Flickr as they could do more things with the content and provide a richer user experience then what the current web browser today an offer.

Yet, they have shifted gear - in that one could argue they are keeping fingers in both barrels if this scenario above were to play out (it's a reasonable scenario at that). Again, what role is the browser playing it today's application and how much usage of the browser is being used. If I were to integrate Flickr with say Windows Media Centre, is this not serving the same purpose but providing more depth in terms of reach - yet where is the browser going in this space?

An important thing to note is why has technologies like Flash grown in widespread adoption and use and more and more interactive content is showing up in it.

The browser plays no role in this equation, as Flash Player typically is doing the bulk of the work and it can be housed in any type of browser overall.

There is the AJAX perspective, in that it's growing with popularity and concepts like WPF/e are playing a role in this space by complimenting these skills going forward. I can see more potential around the browser still in this light - yet - I can't but help think that the bulk of the "goodness" is being shifted towards WPF/e's runtime and less about what the browser itself - standalone can do.

I'm not being popular when I say "The browser has a terminal illness and is dying." as it depends on what you define a browser as and more importantly the path forward in making it still relevant especially when you start to see more and more large companies shift gears and try out new pieces.

eBay Prototype built in Adobe Apollo We've seen this already today, with companies like eBay adopting Apollo as a point of difference. SAP have done the same with Project Muse (Apollo), ORACLE are rumored to be shopping around for its next "UX" approach to things. A lot of companies are looking deeper into Windows Presentation Foundation (WPF) and WPF/e because they can see the roadmap forward in this space.

AJAX still plays a role and at the same time, other companies are still seeing what can be done better by using this renewed faith in HTML + JavaScript - yet - they are aware there will be limitations in terms of interaction and more importantly latency mixed with User Experience (will innovation push the browser into fixing these? is the browser the right platform? will this then kick of yet another browser war of the late 90's?)

A great perspective on the matter can be found at this site (Will browsers ever deliver applications instead of documents? )

It's not about XHTML and complying with standards its more about points of difference - how can I do the same thing my competitor is doing in a web driven world, but better.

I say this as I think that's probably the one piece that the XHTML argument gets ignored by and I'm all for standards, I embrace them whenever I can but at the same time they have to be relevant, they have to appeal to the points of difference argument and if they can't, then they will pushed aside.

It's a radical theory per say and I can appreciate others are against it with every fibre of their body which I'd probably say to them this. If you could re-create Firefox/Internet Explorer for example what would you do differently.


Comments (7)
  1. MSDN Archive says:

    (forgive the scattered thoughts.. distracted at the moment, but this has been a peeve of mine for years)

    Honestly?

    I’d ditch HTML.  It’s too crufty and non-expressive for the markup needs of today (hence PDF, XPS, Flash, WPF/e, AJAX, etc).  It was a good first few experiments, but, in the end, it’s almost useless.  We spend countless man-years working on forcing programming models into the HTML model, navigating the rubble of the javascript/dom wars which means writing too many things twice or more, and it never.. quite.. works.

    Now that .NET and Mono are both widely available, I’d propose we go the managed route — the browser as the host window and sandbox for applications.  A few standard applications — document browsing, etc — could be brought in for rapid document publishing.  We could have much richer sites, easier to program, and arguably more secureable (CAS on the web, signed documents, strong-named pages?) than we have today.

    WPF/e was written as it is on the presumption that nobody wants to download and install the managed stack (Microsoft’s or Novell’s), so it’s a minimalist approach.  Why?!?  Vista already includes all of .NET, it would be nothing to distribute the framework on credit-card sized CDs for low-bandwidth users, and the Linux types can certainly get a non-Microsoft implementation unless they’re just being political weenies.

    The web taught us a few things we can carry on — the possibilities of describing data in xml and style in xslt + css.  We can achieve the same using modern techniques, with less of the mid-90s cruft, and get a more satisfying experience.

    I imagine plug-ins would be easier as well.

    So what do we need?

    Browser:

    * Hosted CLR under LUA

    * System.AddIns (pages as addins to the browser)

    * WCF (just because HTML is limited doesn’t mean HTTP must go away, but that doesn’t mean HTTP can’t happen over other channels…)

    * WPF for hosting pages

    * etc

    Pages:

    * strict, universal DOM — none of this proprietary crap from the NS vs MS vs everyone-else wars

    * XAML for the behavior and presentation

    * XML for the content

    * some safe subset of WPF

    * some safe subset of WCF

    * Workflow

    * LINQ for querying data

  2. I’m addicted to the New York Times Reader(beta) released last Fall (even if it is only for Vista and XP). Why? Because I can "read the paper" without having to surf to the NY Times page. Can I get the…

  3. Jessica says:

    Don’t know how the 2nd comment relates (missed spam perhaps).

    I read your post this morning and intend to do a bit more background research so that I can fully understand where you’re coming from in the posts you make about browsers being on their way out.

    I’ve now just read <a href="http://www.webstandards.org/2007/03/12/which-is-better-for-the-web-single-vendor-homogeneity-or-ossweb-20-style-innovation/">a review post</a> on the Web Standards Project site and the linked SXSW presentation/blog post which asks how we can improve and move forward with browsers, which lit a spark reminding me of your call for comments about how we would re-create them.

    If at all possible I’d be interested in your thoughts on the issues brought up over there (particularly the blog post), and whether or not I’m on the right track thinking that this is perhaps the equal and opposite reaction of the direction that you are promoting.

    This is by no means a flame or argumentitive post, but as a result of my previous misunderstandings I’m trying to understand what you do and where you’re coming from because I quite enjoy your blog and have a genuine interest in what your role at MS involves.

  4. Garry Trinder says:

    Hi Jess,

    Firstly, flame if you want to 🙂 I love the internet and that includes flames and what not hehe.

    Secondly here is what I posted on Brendans site as a comment so maybe this will fill you in on my response to Open Web.

    Hi, first time commenter long-time reader 🙂

    Being open can mean many things to many people, and I would argue that the only realistic reason for HTML’s success is the fact it was transparent, it slowly evolved over time (going from 1.0 to 4.0 was a great experience to be part of) but it’s reached a point where its now mutated into something that was realistically never even intended.

    I’d argue that the Rich formats proposed are an extension of this existance, in that it’s sole purpose is to give depth back to the web but in a more business orientated focus and less about page-like content.

    If you look at the mainstream use of Flash/Flex based applications they are projected as being interactive or a thin-ware approach to solving line of business problems.

    If you then look at mainstream websites built in HTML, they generall are “document” style formatted with a hint of interaction when Forms are presented to the end user.

    The problem over the last decade I guess is that we the developer/designer generation have bent and manipulated the browsers and HTML+JS into shaping up to be something that just seems to constantly go against the grain.

    Flash although it has limitations in some parts can offer the extension path, the path where you want to combine video/audio and rich vibrant forms doing on the fly validation with server-side binding (all those things that appeals to the average AJAX punter in terms of server-side to client-side connectivity) is the right path per say – in that it should never of been flash’s job to do that.

    It should of been the browsers, yet it didn’t and so that’s why I think that the open web is flawed going forward in some parts. As it had its chance and has somehow stumbled its way forward.

    I had great hopes a few years ago when Mozilla was hinting with XUL in that you had a rich client approach to things, bolted back onto the browser were I guess it should of been while at the same time you had transparency in terms of code base. So, the dotComs were going to get nervous about IP being stolen – which is fair call but overall it would of hopefully kicked off the “Internet Next Boom”.

    It didn’t.

    Flash has been sitting around for some time now, evolving every year I guess into something new and better – kudos to Macromedia/Adobe but it’s shaping up to solve a problem that the browsers can’t.

    Now comes the ask, on how do we semantically identify Rich media to enable indexing and search specific agents? – sorry no. As you ask that now for Desktop Applications and you get confused looks. What we do have as a model is the output these applicatons generate, in that a .doc file for example – that’s fair game, and that should be searchable.

    A Mortgage calculator shouldn’t be.

    The danger with Flash based initiatives is they are becoming the “brower” and the “content” all in one – that’s kind of what both appeals to me about Apollo and also why It makes me nervous. As you’re now forking away from the mainstream browser agent itself.

    WPF/e is kind of bringing some new mixed with old, where it will enable you to extend beyond the browsers current capabilities in different platforms while at the same time providing a transparent code-base for others to poke around in (which may appeal to the concerned folks about Microsoft and security).

    It’s an extension of WPF and while Flash and WPF/e will compete on some parts they serve individual needs and will provide a point of difference each to their own. One is more open then the other but will that hurt or benefit it.

    If i were someone homing in on the Open Web initiative, i’d use these two pieces as my “To watch” as on one hand you have a compiled binary (SWF) that has at present no “transparency” once packed together. On the flipside you will have one that does similiar job but has transparency.

    I look back on all the evolving technologies and can see that the next “browser wars” will be the “agent wars” – who out there can blur the line between internet and desktop will be the outside winner.

    Developers have upped the stakes, they want more then what the browser’s today can offer and are demanding it.

    Open Web has pontential but not in the current existance it is now.

    I’d further argue with technologies like Apollo, WPF etc – you the developer have an easier reach in terms of providing power to create your own Browser and implement what you feel right.

    Scott Barnes

    Developer Evangelist,

    Microsoft.

  5. Jessica says:

    Thanks for the response, if anything it was not so much the comment above but your more recent post that "clicked" for me. I think I understand what your vision is now. I’ve grown up with browsers so it’s hard to twist my head around it but I’ll look forward to your future posts with an open mind.

    Evangelist – 1 pt.

  6. Jack.Bee.Nimble says:

    there is a flickr wpf app called nostalgia. its not that bad but still new n buggy…

  7. There’s been talk of the browser being a dying breed when RIAs come along, and while I agree partially,

Comments are closed.

Skip to main content