my comments based on using DDN since v2.0.

My Vote: VB.NET

VB.NET Project
I seem to remember that DNN was the "largest and most successful" open source project based on VB.NET.  That was with v2.0.  Seems like that statement is even more true today.  This would somewhat dilute the C# arguments being discussed.

More Availability of C# Programmers
Maybe there are more unemployed C# programmers than VB.NET programmers...grin
It is also well accepted that VB.NET developers have a better sense of humor.  Hey, Google it if you don't believe me.

The Influx of C# Developers to DNN
This somewhat confuses me.  Over the years there seems to be a steady increase of C# developers writing modules for DNN and using the framework for consulting work and other projects.

1. Where are these developers coming from?
2. Since there are dozens and dozens of C# Portal/CMS projects out there why do they migrate to a VB.NET based project?
3. If C# will represent a better core, why are the other C# projects not able to recruit and keep all these developers?

VB.NET as the Core
It would seem that, at least in the early years, since DNN was the only major project written in VB.NET, that all non C# developers were drawn.  As it grew, I guess the feature set is what causes others of all platforms to continue to adopt DNN.  This then makes me wonder if the other C# based projects draw in VB.NET developers in any substantial numbers.  Hey, the only difference is the syntax.

Developing Well with VB & C#
I can somewhat read a snippet of C# code but cannot alternate productively between the two.  We use Infragistics NetAdvantage controls for our modules UI.  The learning curve for any similar product (Infragistics, Teleric, ComponentOne, etc) is huge.  If anyone can be productive in VB and C# along with using these 3rd party control suites then you are quite a valuable source.  We have standardized on VB "AND" Infragistics, iTextSharp and a few others. 

Summary
The only way to determine if migrating the core, all core modules and the forum module to C# will improve the project is to actually do it.  Since we just left Vegas, I would place a bet that in 5 years we will still be having this same argument.

Keeping DNN as the largest, most successful, most feature rich project written in VB.NET is a huge acheivement.  Becoming just another C# project will never be as significant.

Well, I could be wrong also.  That's why making the right decision is a skill only achieved by common sense, real world experience and a lot of luck.  I do not envy Shaun on this topic.  He got it more than right the first time.


mikez

You mean this community port: first C#-Port
Instead now this port is official: dnnc
Under people tab - is zyhfish (Ben) the same person as wso_angel?
What I think is really interesting is DNN running on mono and DNN running with postgresql - but first things first.
I had a lot of work to convince the people inside our team to use dotnetnuke. And there are still people who are not completely convinced of the choice because of the partially closed source stack. I don't know how much Microsoft is helping the DNN-team - I think enough. But with orchard project they are heading to a completely new direction. Additionally to this DNN is not really compatible with azure.

So the Cloud-Wave is rolling and DNN is missing this wave. But there are other options - not the microsoft way of cloud.
If DNN wants to be relevant inside of the "elastic cloud"-ecosystem it has to get rid of the windows-license-price burden.
I think Miguel de Icaza is giving you a hog more than once if you decide to port DNN to mono. I think DNN could be the biggest open-source-project running on the mono stack. This could bring a real kick in the open source community.

There is a new approach of doing load-balancing - google app engine is doing it like this and azure is doing it too.

The classic approach:
------http-load-balancer------
----/--------|---------\---
---/---------|----------\--
-Web1--------Web2--------Web3-

The new approach:
----------NGINX------
-------/-----|-----\--
-FastCGI----FCGI----FCGI-
---/---------|----------\--
-mono1------mono2-------mono3-

NGINX with two or more fcgi-machines
NGINX python configuration
DNN is constantly searching for ways to get attention. But in the last two years google trend of DNN is going down. I think that supporting mono will be the right direction. In the last couple of month I was searching for a good CMS for GAE or a good CMS for Azure. I tried Orchard - well very beta.
If it is possible to have a scaling stack with NGINX + some FCGI mono machines + postgresql (which is a great db with a really great license) it could be ground breaking. Developing modules with VS2010 and running it in a cost-free, open-source, scaling stack on Amazon EC2.

What about scaling the DB - really demanding jobs can be off loaded to a NoSQL-Engine - what do you think about MongoDB?
C# MongoDB-Driver
There are first modules for Drupal & MongoDB
MongoDB & Drupal Session at Drupalconf
A lot of stuff for one post  :)

Regards,
Thomas