Lazycoder

Not necessarily Martin

Do we learn from history? Or when you started in the industry influences the solutions you propose

Martin makes some good points in his post. But I gues I’m an exception to the rule. I choose to write web based applications, and in some cases data mining/import/export applications that don’t even have a UI. I didn’t enter the programming pool during the dot-com era but a few years before it. My early apps were Form based applications written in VB5 and a little bit of 4. Why choose web based applications? I don’t like having to go around and upgrade my application everyones machine or build in complicated, and often easy to break, mechanisms for auto-updating my application.

I think what Martin is missing when he says:

However, what worries me is that some of the lessons of windows forms development have not been learned. There are still loads of web applicatons out there which have all there business and datalayer code built right into the web forms.

is that in the early WinForms applications (meaning pre-dotcom era) you didn’t really have any option for separating out the tiers. Most PCs weren’t networked together, so whether you built your tiers into separate components within the application or crammed all the logic into one big class largely didn’t matter. If you upgraded the application, you had to upgrade the WHOLE application. I’m sure Martin knows that though and his point is still valid. There’s no reason to flatten out your architecture now, although there are still people out there pissing in the wind over exactly how you should split up that architecture and what form the split should take.