Discussion on javascript-syntax-checking

[Stephen McKamey]

I'm curious about people's feeling on the integrated JSLint component
of JsonFx.

On one hand having a gut check around syntax and common mistakes in
JavaScript has helped find many issues which would have otherwise
caused strange bugs in the future. Just having a general pass through
often is good for finding a missing semi-colon or curly brace.

On the other hand, needing to leverage the MSScript.ocx feels dirty.
It is the *only* non-.NET server side component. It complicates setup
and can cause issues in newer versions of Window, especially x64.

But perhaps more importantly due to the stubborn nature that JSLint is
created from, it doesn't allow some styles of perfectly valid
JavaScript to pass. For instance, jQuery which is arguably one of the
most quality JavaScript products out there immediately fails JSLint to
the point of it not completing the validation process.

Currently it only produces warnings rather than errors (since it isn't
always correct). Also if the MSScript.ocx fails to load then it
should disable itself listing a single warning during the build.

So my question: should the JSLint component be removed? Or does it
provide enough value in enough situations that it warrants its
existence? An in-between option might be to add an explicit method
for making it optional.

[Stephen McKamey]

In an effort to get closer to full Mono Framework 2.0 compatibility,
JSLint is now optionally compiled and self-disabling
(v1.0.0901.2513+). To compile without JSLint, remove the "JSLINT"
conditional compilation symbols from the JsonFx.BuildTools project.
In Mono Framework, this is automatically unset.