Andrew,

Have you looked into the Zend PHP Framework?

If you have nosql data but want a relational db you can do HSTORE in postresql.
FYI if you weren't aware yet.
http://railscasts.com/episodes/345-hstore

I don't do many non-Domino projects, but when I do I prefer the Apache / mySQL
/ PHP (backend) / JavaScript with jQuery (front end) approach. In all honesty,
I feel a real sense of freedom when I do those projects, but also very out of
my element.

Have you considered cloud services and the stacks that they support? Perhaps
your vision for how you want it hosted/deployed/managed should take precedence
over picking a stack.

If you like simplicity, but want a powerful web platform choose "Play
Framework".

If you like the database part of Domino but want more look at "OrientDB".
It is document and graph database. It has master <-> master replication, ACID,
SQL queries and much much more. AND it is super fast!

/Jesper Kiaer

Since you said you might be interested in selling, I would think that the
preferred path would be IIS/.Net.

I say that because most of the vendors I can think of that you might need to
integrate with are .Net/SQL based. Windows resources are also available in
your target markets, and that always makes the decision makers more comfortable
with a solution.

This might be one of those times you put Mr. Developer off to the side and let
Mr. Business Man make the decision.

Good Luck with your decision to move forward with the migration, I'm sure
whatever platform you pick it will be a rock solid solution.

Let me put in a good word for Ruby on Rails.

First, as you well know from the Notes/Domino (hereafter Notes) industry,
perhaps the most important factor in selecting a technical platform is
community.

Both commercial platforms, like Notes and .Net, and open source platforms like
Rails, Django, Drupal, have ecosystems of users, consultants and vendors that
form around them. But in open source, the platform developers themselves, by
definition, form a strong community in which anyone with the appropriate skills
and inclinations can participate. In other words, the notion of community is
baked into the very development of the platform, not added in for marketing
purposes. This is a very powerful concept.

Second, both Ruby and Rails share the mission of developer happiness (see, for
instance: http://gettingreal.37signals.com/ch10_Optimize_for_Happiness.php).
This largely contributes to the strength of the community, but also to the
happiness of customers who are dealing with developers who truly love what they
do.

Third, while it is often criticized on these grounds, Rails resembles Notes in
being highly "opinionated" in that it makes many important architectural
decisions for you, leaving you the bits that actually add business value. While
this does involve a learning curve, and some uncomfortable periods of trying to
coax the platform into conforming to your preconceptions, when you take the
time to appreciate and master "the Rails way", I think you'll find, as I have,
that the decisions that are being made for you are very good ones, and they
save you a great deal of trouble and complexity of experimenting with different
alternatives. The thought leadership for Rails
(http://david.heinemeierhansson.com/) is indeed, in very good hands.

Fourth, don't make the mistake of lumping Rails, an application development
framework, in with Drupal, an extensible application (i.e. content management
system). In this sense, Drupal resembles Notes more than Rails does, but in my
opinion the Rails approach clearly wins. You know how, much as you try, a Notes
app will always look likeu2026well, a Notes app? Same with Drupal. But not
Rails. While its UI development approach is extremely powerful
(http://railscasts.com/episodes/279-understanding-the-asset-pipeline), it's a
framework for building user interfaces, and not a set of lego blocks for you to
assemble and fight with or give up when you want to do something slightly
different.

This last point may seem to contradict the previous one, but on the contrary,
it illustrates to me its uncanny knack, like good government, for making the
right decisions, but staying out of the ones it has no business making.

There's much more to say, and feel free to follow up with me if you like.

One caveat, though. Rails is almost 10 years old now, and many in the web
development industry are already beginning to treat it as yesterday's news. All
things node.js (javascript everywhere) is the rage these days, and if what's
cool and attractive to VC's forms an important part of your selection criteria,
then this would be the way to go. This may sound cynical, but I actually have
nothing against it (although I would find it painful to give up Ruby, my true
love, for JavaScript, yet again). The trade-off, for now anyway, is very much
one of stability and maturity vs. life on the bleeding edge. Depends, I guess
on your appetite for risk. But I've been working with Rails since its beta, and
I will say this about it: In its 10 years it has shown a remarkable resilience
to change, and has managed to remain very relevant, and in many ways the
measure against which the other frameworks are judged.

Depends where your target market (for selling) is. In the startup world you
will find OpenSource dominating with LAMP at the front, but Ruby and Node.js
coming quite close. Twitter is good ol' Java by now (after its start as a Ruby
app) and What'sApp is Erlang (as is CouchDB).
.net/MS-SQL is a no-no in a startup environment, but well entrenched in
corporate. You might look at CouchDB or MongoDB which are closer to the Domino
way of doing things. Node.js/Mongo is a "no-explanation-needed" combination in
the startup scene, but might not be well known outside.