Remote Model Designs for Web Apps

Dec 17, 2009 18:06

I've been involved with writing back-ends and (limited) front-ends for multiple web-apps over the course of my career and one pattern that I observed at the last 3 of my 4 workplaces was that the application started as a single unit running on a webserver but the model sooner or later got ripped out and deployed in a separate layer over the network ( Read more... )

architecture, services, applications, webservices, ui, wsloader, web, front-end

Leave a comment

Comments 2

nileshg December 18 2009, 05:14:01 UTC
I haven't been involved in much FE. But on all the projects that I have worked on so far, I have observed that deployment of webservices were mostly driven by following two needs :
1. Keep the consumer app light-weight.
2. Webservices gives you better control over your code as
- You can control the Business Logic the way you want without depending on your clients/consumer much.
- They are easier to debug whenever there is a production issue, as you don't have to login into your client/consumer machines.
- The scaling is easy. As the scaling logic is completely oblivious from clients/consumer.
3. And webservices are platform independent.

I haven't worked on other decoupling models except web-services. But just by the amount of flexibility web-services provide, I will opt for decoupling the logic rather than a thick client app.

Reply


toast0 December 19 2009, 01:15:33 UTC
I would say that thick-servers aren't necessarily non-scalable, but they are harder to scale. Because all the layers are in one place, it's more difficult to reimplement a particular layer, if that's what you need to do to scale. You also don't get as good of insight into which layers are performing poorly ( ... )

Reply


Leave a comment

Up