I'm back! I got a bit of a chill when I first came back due to needing to wash and the weather was like 15`C and then back down to 10`C so I am a bit slow and headachey at the moment, but I have a series of minor adjustments. I published them on the facebook page for parallelcoin, here it is:

------

ParallelCoin
Published by Loki Verloren · February 1 at 3:10 PM ·
Our lead developer, Loki, is imminently returning to work.

Among the news for the project, the GUI is now in early prototyping, the primary developer being Djordje, and confirmed to build on Windows, Mac, Linux and FreeBSD, with confirmation of such coming soon for Android and iOS. In the plan is to form the beginnings of a truly cross platform GUI framework, built on the principles of pure Functional and Concurrent programming paradigms.

While in hiatus, Loki has not been idle and has several planned changes to the post Plan 9 hard fork consensus:

First being a novel multi-interval parallel block timing and reward scheme that uses prime number based dithering in order to produce a more regular clearance pattern;

Secondly all transaction signattures after the fork will be Schnorr signatures, that produce a consistent 64 bit size for every transaction no matter how many inputs are involved;

Third is the use of Blake2b hashes for all post hard fork block hashes, which have a very fast processing time (blake did not win the SHA3 standard - the less prettily named keccak won but it is slower and about equally secure from collisions)

Last is the integration of the required recognition and validation required to become Cosmos IBC compatible, in preparation for and to avoid a second fork to attaching a Tendermint Protocol sidechain, which will provide 3-5 second clearance time and a throughput capacity that will not be so dwarfed by interbank and international payment clearinghouses, and enabling further integration with the rapidly growing Cosmos internet of blockchains.

This is the minimum target and paves the way for future deployment and readies Parallelcoin to ascend the ladder of market capitalisation and transaction volume.

We should be starting beta testing in early March with all these features in operation, and have hard fork height set in stone by May.

-----

I am starting to work on the first thing in my part of the work now, just getting comfortable and recovering from my exile from Serbia, making a more precise specification. It's essentially going to be somewhat a return to the origins of Parallelcoin's difficulty regime in that there will now be 9 different algorithms, but they all run at a different timing, specifically:

9, 15, 33, 51, 123, 177, 201, 249

This is, first, take the first 9 primes, then select the primes indicated by each of these first 9 primes, then multiplied by 3 (as 3 seconds is unreasonably short)

This creates an intentionally non-repeating pattern which I estimate will take about 39,000 years to repeat. My hypothesis is that since finding blocks is random, a randomised timing pattern will yield a more regular pattern of blocks. Chaos is fickle like that, when you do things randomly you can leverage capabilities that are otherwise impossible, and you can stretch the bounds of probability. Besides all that, everyone would be familiar with the concept of 'Dithering', used to increase the precision of a mathematical system by adding noise to it.

The new scheme is considerably simpler than the old one, which was pretty nice already, but it should produce a more even clearance time and be resistant to any of the known ways to perform a timewarp attack, since to do so you have to craft your mining output to fit with nasty indivisible prime numbers that have a very long period of repetition, the same thing will work at one time but maybe even 10 seconds later conditions will not be ideal.

The other things I am doing are about improving performance. I have been interacting with the upstream btcd repository relating to the initial sync process and improving the throughput. In the last couple of months of my work before I had to stop at the end of november, I already improved the transaction clearance rate to around an average of 800tx/s for validation, peaking around 2000, and this is quite improved compared to the reference implementation. The post-hard fork chain will use Schnorr signatures which are faster to compute and allow you to create multiple signatures all within the same 64 byte signature size, no matter how many inputs (probably there is some practical limit but beyond what would be typical, in the hundreds or thousands of inputs.

I am not yet aware of any attempts to implement Cosmos IBC on a non-Tendermint chain, so this last task is probably the most challenging. However, the purpose of making Parallelcoin IBC enabled is that without a hard fork, we can roll out new tendermint based sidechains, for fast transactions, for other types of data transactions (IPFS, onion routing), and for that latter other type, implementing sidechains running Mimblewimble privacy protecting transaction mixing based on new technologies related to cryptographic recognition filters.

I will go back to my work now, I am just getting back to being comfortable and back in the programming mindset, circumstances literally stopped me from doing any significant amount of code, laptops died, phones had several functions fail... it has been a really difficult time for me. But I am confident I can roll out all these features and be in beta testing soon, and @marcetin is now significantly progressing towards having a usable GUI interface for users.