From what I understand static typing makes programmers more productive and gives implicit documentation to the code base
- although explicit documentation is also very helpful. It also seems to make the tools we use much more powerful.
One of the problems of static typing is that your code becomes more verbose. With programming languages like F# the compiler will implicitly type as much of your program as possible. Where it fails, you need to type it yourself.
Not only that but many times static languages will provide tools to do the
typing for you. F# has Type
Providers
which will dynamically create types for your program on the fly, lessening the
burdens of a statically typed system. One of my favorite libraries that does
this is FSharp.Data.SqlClient - Not your grandfather’s
ORM which will
statically type your calls to SQL Server letting you even type raw SQL in your
code. Another useful one is the JSON Type
Provider.
With F#, WebSharper, and SQL Server Data Tools(SSDT) you can have static typing from your database all the way down to your client web app. You can literally keep everything in sync with each other from across your entire infrastructure. This seems really huge to me. Programming in any other way seems like programming in the past, by decades. Unfortunately, it seems like it takes ages for the programming community and business industry to get on board. It is understandable and mind boggling at the same time.
So, SSDT, can give you static typing across databases, it seems to work best
when they live on the same server when you track the databases in the same
SSDT solution. But you can also have this static typing across different
servers in different SSDT solutions by referencing the dacpac of the
different databases you are connecting to. It also makes testing much simpler
with the help of tSQLt testing framework.
I’ve never used WebSharper before but it looks like it is a wonderful way to write your back end and front end code. I’m not sure if other .NET frameworks can give you the same static typing that WebSharper can, but WebSharper can generate your end points for you in distributed applications, letting them be private or public end points.
So, with all this incredible technology that can make us so much more productive why do we not embrace this technology and stop doing the computer’s work for it? I’m sure there are many reasons, including psychology and business maintainability. Which I won’t enumerate some here, which I think could possibly be reasons. Let’s start embracing this technology and start slowly introducing them where we can in are work places. It might take another ten or twenty years to be able to really start letting the computer do its work rather than us doing its work, but the time to start planting the seeds is now.