Tech Off Thread

10 posts

Smart Clients vs. Web Clients

Back to Forum: Tech Off
  • User profile image
    DadofEight

    Hey Channel 9'ers-

    I recently attended a conference where Billy Hollis (forgive me Billy if I have the wrong guy) quoted a credible industry source that asserted by the end of 2005, 60% of applications being developed will have Smart Client front-ends vs. Web front-ends.  This is exciting news for people like me who have been spending years trying to cobble together a reasonably user-friendly front-end in a browser. 

    I think that as a programmer, I see the 60% statistic as reasonable.  After all, we're leaving .dll hell beyind us and deployment technologies in the current .net platform and those coming in Whidby make the installation and automatic updates of smart client technology reasonably painless.

    It seems that many businesses and corporate IT shops that I speak to are still resistent to Smart Client solutions.  I have one very large client here in Vermont whose IT shop is digging their heals in insisting that a web application is the way to go for their solution.  This client has volumous data-entry tasks with many intricate dependencies.  A web client simply will not cut it.  But, they are still in the mentality that installs across desktops are simply too time consuming and painful.

    So are you all experiencing the same things out there?  Are you talking Smart Clients to businesses?  How has the reception been?  Do you think the pendulum is truly swinging away from the browser back to Smart/Rich Client front-ends?

    Al

  • User profile image
    bitmask

    The healthcare industry, in my opinion, hasn't been receptive. Thin client solutions still seem to be the "in" thing even while PDAs and tablets make some inroads. Technologies like Citrix and Terminal Services are popular. Many facilities still run client server applications and upgrading to a web solution seems like a hot idea. The IT staffs are usually overworked and view anything touching the client as running the risk of requiring more support. Maybe in another 3 years.

  • User profile image
    Lwatson

    In Heathcare I have found that their mentality is typically 5 years or so behind the curve. Understand that they are all saddled with the mistakes of the last 5 years and many of them would not want to trade off anything that know about now for something that they don't know about.

    I often get the feeling that they are corfortable with the pain they now have and are even on some level happy with that.

    As for the we client vs smart client if a shop is already going web client it might be next to impossible to get them to switch. If they have not dont either then a smart client is definately a better way to go but they might just have been sold on the thin approach. Of course when they get thin and dont like the results they wont have the money to do anything different, but they will feel like they have seen it before and in 5 years will likely repeat their mistakes on the next paradigm...

  • User profile image
    MarkThe​Viking

    Personally, I've found the market has been very receptive to rich clients, and every .NET solution I've produced has used a rich client in some way. It must be said that so far the target audience has been a max of 50 desktops, but having said that one customer is actually seriously looking at producing a rich client version of one of their extranets (albeit leaving the web version in place for those who can't or won't install the .NET framework).

    On the other hand, if I was in charge of a large desktop base, I must admit it would take some serious evangelising before I would be convinced that there wouldn't be serious extra support work involved in running a rich client rather than a website.

    But yes, I think the tide is turning.

  • User profile image
    JasonBunting

    I would say that in respect to theoretical ideas, I see things turning towards smart client apps as well, and like others don't see it taking hold for a while. Many are hesitant to try anything uncommon, though smart client apps have been extant for a couple of years now. But if you look at the reason why we went to web apps in the first place, and, as the first posted mentioned, you see why those reasons are becoming invalid, it is hard not to come to the conclusion that web apps are not as compelling as they used to be.

    Nearly 3 years ago Billy wrote an MSDN article, "Death of the Browser?" - I remember when it first came out, reading it and thinking at first that he was being a little dramatic. But as I read this I became more convinced that he had something . . .

  • User profile image
    DadofEight

    I think that it's helpful to try to quantify the amount of time and money it takes to build a reasonable user experience for a web client.  The amount of time and energy my company has put into building javascript/DHTML solutions in the past to achieve what you mostly get for free in a rich client is pretty amazing.  Yet some companies still clamour for the thin client (but moan of course as they're waiting for server round trips or the next version of IE breaks their client-side scripting code). 

    I also think as has been mentioned that it's just going to take some time and education to make folks understand why so many sacrificed user experience to a greater or lesser degree to embrace the thin client model.  WHen they really understand those reasons and see that the majority of them are no longer valid due to advances in technology, there should be many that seriously reconsider the Smart Client option where it makes sense.

  • User profile image
    DadofEight

    BTW: I just tripped over the new smart client development center on MSDN:

    http://msdn.microsoft.com/smartclient/

  • User profile image
    figuerres

    DadofEight wrote:
    I think that it's helpful to try to quantify the amount of time and money it takes to build a reasonable user experience for a web client.  The amount of time and energy my company has put into building javascript/DHTML solutions in the past to achieve what you mostly get for free in a rich client is pretty amazing.  Yet some companies still clamour for the thin client (but moan of course as they're waiting for server round trips or the next version of IE breaks their client-side scripting code). 

    I also think as has been mentioned that it's just going to take some time and education to make folks understand why so many sacrificed user experience to a greater or lesser degree to embrace the thin client model.  WHen they really understand those reasons and see that the majority of them are no longer valid due to advances in technology, there should be many that seriously reconsider the Smart Client option where it makes sense.


    Very well put!

    one of the things I say to folks is almost the same as what you are saying...

    that there are limits to what you can shove onto one web page and still make it work right...

    and at that point you have two paths to chose from:

    1)   break the page into a set of pages....
    possibly spending weeks / months re-working logic to fit.

    2)   move from a web app to some form of client app.


    it is not so much a # of controls on the page as a function of how much the app needs say

    "dropdown N is populated based on the selection in control n-5 and n-7"

    and much the same for the validation rules and work flow for the next page or tab....

    when it starts to be that the client side logic is more complex than the server side....

    you are in toruble....

    if its a web page...

  • User profile image
    scottmace20​02

    Why doesn't Billy Hollis have a weblog? Where we could learn more about the industry source he was quoting? For someone billing himself as the .Net authority, it's odd these days that he doesn't blog.

    Scott Mace

  • User profile image
    AQ

    For Windows development, I think smart clients are a natural choice. There is simply too much power on the client PC side which goes untapped using only IE. 

    Also XHTML / DOM / JScript is a nightmare for developers trying to create dynamic UI, based on dynamic content and data. It breaks, constantly.

    Really Flash is the only option for creating the rich user experiences people expect these days and at its core you have static, frame based-animation, or ActionScript which can't interop with your logic.

    I am currently developing an online game system utilizing DirectX Smart Clients. One-click launch from a web page, they run on the client's PC in windowed or full-screen mode. 

    The system requires the .NET 1.1 Runtime, DirectX 9, and a medium level of trust for the assembly. Its a 100% .NET implementation using ASP.NET for the web interface, WinForms for the smart Clients, managed DirectX for multimedia and XML Web Services for data storage and processing. Really, .NET makes this a very elegant solution, completely object-based. Now if only someone would solve the anonymous micropayment ecommerce problem, I would have a perfect cyberarcade! 

    Still in development but some demo games should be available, including full source code, shortly. I'll post it when it becomes available.

    aq

Comments closed

Comments have been closed since this content was published more than 30 days ago, but if you'd like to continue the conversation, please create a new thread in our Forums, or Contact Us and let us know.