ReSpec: Evolution

Hi all,

if you use ReSpec, or participate in a group that uses ReSpec and think that your editors might not be on this
list, please take the time to read this email as it contains important information on the topic.

Before I delve in: if there is any part of this email that makes you worried, panicked, angry, fills you with
angst, or generally spoils your mood then please jump straight to the "WHAT THIS MEANS FOR YOU" section.
All is fine and dandy.

There has recently been a flurry of questions in various parts about ReSpec, mostly to do with what the
status of v2 is, how to load it, what to patch.

I started work on v2 not because it was fun to make a new version but because the v1 code was grown all the way up
from a 50 lines script into a tentacular monster. I recently had to fix a bug in the v1 WebIDL code and some
parts of my brain fried. Some of the configuration options interact in ways that I don't understand and
dare not touch. It is also constrained to a single type of document (W3C specifications) when many people
are asking for the same stuff for different documents; and even within the restricted scope of W3C
documents there are variants that were hard to accommodate without (even more of) a combinatorial
explosion in configuration options.

I didn't feel like I could fix any of that without taking all the pieces apart on the floor and finding a new,
better way of putting them all together.

That's what happened with v2. I figured out how to make this maintainable and flexible. But the problem was
that this was a rather large undertaking to carry out within bits and pieces of spare time. The v2 that's
available today is very roughly usable (some people do) but there are still a thousand paper cuts that need
fixing before it can replace v1. This would all be fine were it not that in the meantime, people still need
fixes and new features in what they use. So over time, v1 has kept moving and all of those changes need to be
integrated into v2, which takes time since it's a different code base. This has made it impossible for me to
reach the point where v2 could take over.