sudevsu asked:
Yes this is the difference that makes it easy. But when starting a project completely new without any Team this is real tough for me. Only one developer is everything and is entirely new. So what do you suggest? Best is to choose web application or web site?
At first, this question looked weird to me, after answering you in Solution 1, but then I realized that this is, essentially, a different question (still weird in formulation, but understandable).
My first reply would be: "Well, you should know your ultimate goals, something I have no idea about, so how can you ask me to make a choice for you?" That would be fair, and I hope you don't understand that I won't give you one choice of the two. The logic is apparent: if there were two different solutions, and one would be always better than another one, don't you think that the inferior solution would be pretty much forgotten with time. But both Web sites and application still co-exit for a long time. You need to check the points discussed in the MSDN article referenced against your goals and requirements. But this is too trivial.
So, I guess I have to answer on different question: what to do if the choice is still not clear for you at the moment. I can admit it would be possible.
The idea is simple:
create a simple prototype in the form of a Web site and another one in the form of Web application. The prototypes should be simple (apparently, one goal is to reduce development time), but never trivial; it should address your goals and most characteristic specific requirements, say, questionable or non-obvious features. If the choice is still not apparent, make the prototype more complex.
Generally, don't be afraid of making more prototype. Don't throw them out, always store them in the same revision control system (I hope you use one, if not, your team is just wasting time, not really developing). One of the biggest fallacy of project management is very usual opposition to prototyping. Stupid manager afraid they waste developer's time. It's hard to think of something as wrong. Prototypes save from major disasters.
On next step, you can make a wrong decision, even when on the "real" project. First of all the change of the way is possible:
http://msdn.microsoft.com/en-us/library/aa983476%28v=vs.140%29.aspx[
^].
(I did not find a manual for the migration in the opposite direction, but this is possible, too.)
But better settled it more confidently on the prototypes. Your know, people asked me several times: "what is good architecture?"
I answered:
Better architecture is not the one which has no mistakes. Better architecture is the one where fixing those mistakes costs less.
Good luck,
—SA