17 March 2012

WebSharper - Is it worth it?

So, I hate it when I comment on a software platform and then can't get it out of my mind. This is what happened with F# and WebSharper.

I hate been asking some questions that I think may have ticked off the WebSharper people. If any of you are reading this, I swear that I just wanted the information before I invest too much time into re-learning F# and learning WebSharper.

As I've gone further into GWT and znEve, I've found that the platform is becoming very cumbersome, as all large Java-based projects tend to become after a while. Luckily, based on the comments in the above threads, it seems that F# and WebSharper may be a viable alternative. I have decided to write a small test application to see whether this will work or not.

Here is the new structure for how znEve will work.

Server component:
  • written in GWT, PHP or Python, depending on what components it needs to hook into and where it'll run
  • expose RESTful interface
  • serve information based on tokens - most of the pages will not be accessible by the "public" (those not logged in)
  • perform OpenID authentication for clients
  • manage and protect user data
Client component:
  • written in either GWT/Java or WebSharper/F#
  • should be self-contained
  • will contain the algorithms needed to re-calculate statistics for ship fittings, skill plans, etc.
    • this may be an issue depending on how resource-intensive the algorithms are. will have to test on mobile platforms (tablets, phones) to find out more.
Hopefully, this will get me off to a decent start. The first step is to now write that test application and see how well it performs (or not).