Yeah, I tend to think just hardforking the emissions curve to lower the present reward so that rewards last longer would be more acceptable at large than a involuntary re-emissions program labeled as a softfork, because a hardfork of the emissions curve is really one’s goal, so it were better to plainly state that.
But, I also think it is better to keep the emissions as they are.
I have been thinking about a way to provide miners some extended use of their equipment and I think making Ergo dApps that perform GPU accelerated curve fitting and dataset prediction services using Ergo Oracles would be a great idea.
Like, have ErgoPredictor dApp that is essentially an auction house for returning extrapolations of different oracle pools, users could submit a bid for a prediction with a margin of error, expiration date, and price for prediction of course.
For instance, a bid would be like “Prediction for the pirce of Ergo within +/-10% for tomorrow for 1 ERG” and other users could take up this contract and submit a prediction, and when the time comes for the prediction to be validated the first user to submit a prediction within the margin of error gets the 1 ERG for making the prediction.
Then, miners could run gpufit software (open source Levenberg-Marquardt non-linear GPU curve fitter) that could monitor the ErgoPredictor market and possibly pause or partially pause some of their GPUs to attempt to be the first to make a robust prediction. This prediction market would also provide GPU owners a source of revenue after mining ended.
Of course gpufit predictor executables that were compatible with Ergo GPU miners would need to be programmed, but they could be developed similar to how miners are developed, maybe some miner programmers would integrate them directly and allow end users to tweak the model parameters or choose the fit model to use to make their predictions more accurate, that would be the job the miner to also learn about curve fitting specifics to make their GPUs more useful to them. But, the software could also track the likelihood of the prediction being correct based on the chosen fit model and do some calculations to determine whether pausing mining to predict is likely to be profitable or not based on the price of open market bids for predictions.
I had this idea rolling in my head all day, so I just wanted to mention it, I think it serves at least as a good example of how there may be better ways to extend the usefulness of miners than extending the emission curve.