Akasha - decentralized social media platform


Good timing! Just as I leave Facebook, a decentralized alternative is announced:

No doubt there will be many of the same aims as Heartbeat, and some different ones, and some different technology choices.

Tech stack seems rational enough:

  • Ethereum
  • IPFS
  • Electron, React with Redux and Node.js

Hopefully they can build something cool. Actually, hopefully WE ALL can build something cool together.

Will be open source and fully decentralized. Very interesting to see what they have up-and-running.


Sounds like worth a try, anything that keeps personal stuff personal is welcome, personally :slight_smile: I like Node but it’s single thread error catching fickleness is the reason why I would not use it for complex stuff. In that sense I hope swift will do better, ( still have to get it running on linux, currently i am busy getting to grips with it on a minimac ) IPFS seems a nice concept as does Ethereum, How do you see Maidsafe ?


I don’t know much about the current state of MaidSafe.

I was quite enthused about it in early 2014, but I think they have largely withered on the vine.

The reaction from my colleagues whenever MaidSafe is mentioned is “MaidSafe? Is that still going? Oh, right, they’re going to 'have something soon”, aren’t they? Snicker".

So while their approach is a little different to other projects, I think they’re missed their window of opportunity, and that they may be obsolete before they actually launch.

StorJ also seems to be in much the same place, though perhaps a little more real.

IPFS is certainly the leading solution right now. There is also a content-addressable storage solution within the Ethereum banner called Swarm, which has been hibernating a little, but which is coming back to life as we speak.

See https://gist.github.com/zelig/954937b915768b8a8d2c


PS. See also:


Done some more digging and wound up installing the VM for Ethereum. ( had to upgrade my VMware workstation 8 to a player 12 and for that needed to upgrade my Debian 7 to an 8.3 and in the same effort swapped my 1TB regular disk for a Hybrid in a Compaq 15-s130nb that is not designed for changing disks. Then reinstalled everything and pumped 18 VM’s over to my upgraded system )

Now everything looks fine again, shortly I will start to explore the Ether … found an interesting tutorial, hope it puts me in the right direction. (mc2-umd.github.io/ethereumlab/docs/serpent_tutorial.pdf)

any useful pointers you can give me ?


Glad to hear it!

Here is a list of some starting resources: https://www.reddit.com/r/ethereum/comments/3vxvlx/starter_guide_almost_all_the_links_youll_need_to/

The tutorials listed there are usually well received :slight_smile:

See also:

Also, if you want your mind blown at the madness of crowds, check out “The DAO” which has now raised $106M in a few weeks, and looks to be #1 crowdfunding of all time before it finishes.

Aral will hate what “The DAO” has become, which is essentially decentralized VC. The mob lives at:



The more I read and see about Ethereum, Blockchain and more specifically Compositionality the more it resonates with what I have been trying to do for the last 30 years in the circumstances and with the tools available to me throughout this period.

The significance of it for me is that this seems to be the first time I hear and see people who are talking about what was always at the tip of my tongue but never really took form other than in some hazy, undefined structuring of something that did not really fit with the materials it was built with. ( not unlike trying to build a car-engine with sticks and stones ) Not to mention that nobody every really got what I was trying to do.

Also the baseline philosophy to try to keep it all at an understandable level in the FOS environment instead of building yet another intellectual ivory tower only accessible to friends and elite-ists charms the bejesus out of me.

Thank you for pointing me in this direction,

I suddenly have a very clear view of what I want to do with my time for the next couple of months.


Glad to hear it!

Yes, I think there is huge potential in blockchain and related decentralization technologies. It’s still very early days and there is still plenty for us to work through before we know the best ways to apply this new technology.

We’ll all be a bit dumb and naive for a few years yet. I think the way we will know that we have succeeded is when blockchain and ledger technologies are seen as being as boring as databases. Just more tools in our armory. Great for particular uses, crap for others, but something which you would always consider.

It will be lots of fun on the journey, I think.

Another class of technology which I only heard of recently which is also fascinating and relevant for decentralization and collaboration is - https://en.wikipedia.org/wiki/Conflict-free_replicated_data_type.


I always avoided Operation-based synchronization because 1 wrong step could have totally unpredictable consequences for data or systems that might only become visible when it was way too late to do something about it.

Got to that conclusion very early on in my career seeing how people memorized procedures and then occasionally started executing them blindly in the wrong menu or on the wrong system.

The checking systems required to make sure everything was duly processed were often more complex ( or downright too complex ) than setting a basic state-based synchronization which of course could not capture how things had changed or why. Hence lots of time spent puzzling with outdated log-files and unexplainable state-changes but given the possibilities we managed.

Looking around now I often wind up wondering how to use stuff in ways it might not have been originally built tor but could have added value.

With the concept of a block-chain type mechanism in the middle between synchronizing systems as a recorder with programmable abilities one could not only guarantee delivery but even have a rewind and repeat function too, Or build sequencing and control of execution plus correctness into the block-chain contracts ( The contracts could be dynamically created like generating sql )

I’d love to get that into simulation scenarios. Like having a chain of events ( not limited to sql statements but any type action ) being locally recorded and stored - and here is the twist - why not try to add their ‘undo’ counterparts - so any simulation or test system can scroll backward and forward to replay from any desired point of the status and take another route from there. That in it’s turn would be recorded …

No doubt it already exists for specific systems but in my environment definitely not as a generalized concept, In this environment ( classic DB&SQL etc… in industrial settings ) something like that would have been very valuable to prevent mistakes and errors to be made over and over again.

Most of the managers I came across in my career don’t even have a full grasp on what a simulation is let alone what it can do so … I’ll need to play around with this in my own time.

Question: What do you think is a good tool/language to work with big multidimensional arrays of mainly numeric and ascii based data really fast ( for simulations ) ? I Came across Rasdaman but that is mainly for image processing as far as I can see.


Indeed. All of the possibilities in front of us in the next few years are just fascinating.

I’m afraid I have no recommendations for a tool or language for you. I’ve not done much of that kind of work myself.

All the best!