nope - jl777 is not threatening to 'abandon' NXT, and he never has.
what he is doing is implementing some NXT functionality that SuperNET uses into SuperNET/BTCD.

it is a long story and much hyperbole, so please don't jump to hysterical conclusions.

What happens is that SuperNET/BTCD tech gets more streamlined.

There was no discussion with the SuperNET team about the 1.6.2 breaking all the installed base. The only "discussion" happened after I complained about it not being backward compatible. And this "discussion" was basically a confirmation that the promise of backward compatibility that was made last year is not being honored.

There have been many customer support issues due to the breaking of the API and a lot of unreported cases where the customers just stop using SuperNET (and NXT)

I have been assured that with each NXT release that there will likely NOT be any backward compatibility of the API, which means anything that uses it will likely break and need unknown amounts of testing, requalification and of course customer support and field update issues.

It is not possible to deploy a large scale system using a "platform" that isnt backward compatible due to the breaking of all the ALREADY installed and running systems. With decentralized systems it is not possible for centrally force an update on users. Just look at some of the versions NXT peers are still running. Therefore anything that is guaranteed to break is certainly not worth spending marketing dollars to acquire the installed base for.

The NXT devs lack a fundamental understanding about business reality and I am extremely disappointed in the breaking of the promise to keep things backward compatible. The ONLY exception is if there is an attack vector. Local node performance is not any excuse.

Due to this, I am rethinking my reliance on NXT for SuperNET. I have quite a lot of code that takes all my time to create and to be assured that I will have to keep reworking it forever due to backward compatibility breaking with each release. This is not practical for me.

So, NXT will have a more efficient API with much fewer users. That is the effect of this breaking of the backward compatibility promise.

Anyway, the first effect of all this is that ALL of my projects will be delayed. I will wait for NXT API to stabilize before I waste my time. The most likely scenario is that I will just remove all dependency on NXT and put whatever functions I need into BTCD.

James

tl:dr it is a MAJOR breach of trust as far as I am concerned that the API backward compatibility promise has been violated and I cannot rely on NXT for any important SuperNET functions.