Network Based Replacement for Server-Side Processing?

[drehb]

@dirvine

The link is then verification and possibly another incentive to run nodes, but I do note (and it’s important) like skype the nodes running these are not paid, they do it because it’s right or helps them. However, I think incentives can help verification and allow for more critical consensus measurements

I like TrueBit’s approach of hashing the state of the program execution at each step and having validators being able to challenge results people have submitted. TrueBit. A verification and storage solution for… | by Christian | Medium

[dirvine]

With group consensus each member will have to provide the same result, so similar in a few ways. Groups get to consensus this way. I think it’s key to remove reward and bad nodes, but also allow nodes to grow in stature, so longer doing good node is better than a new node etc.

Nice to see so many projects tinkering about this area.

[intrz]

There are already multiple distributed computing project for ethereum right now and once SAFE is out and starts to get traction, I would expect someone would want to do ICOs for distributed compute projects running on SAFE. There’s many ways this could be done on the app side and it could be complimentary to the built-in compute on SAFE that will arrive eventually. There’s also a market for special purpose ones, for example there’s Render Token on ethereum, for rendering light-fields for AR and VR.

[beermenow]

@dirvine

I am not sure about Golem, I looked for code but found some python scripts (I need to check again though) and their first app seemed to be very high speed video rendering, but the mailing list had a load of messages suggesting is was incredibly slow. I could not see why video rendering was a good choice, so listened to their explanation video and was more lost, there they said it was an AI that would write itself based on user feedback. I would love a project to do what Golem says, i.e. allow decentralised AI, but I cannot find the thread that pulls it all together.

I’ll be careful and try not to let my own views come into this one and say I am invested in both projects.
Golem (what we can see of it anyway) is years behind SAFE development wise (irrespective of the funding behind them) and specifically with respect to your above question/s you may not see a sufficient answer to for quite some time yet.

I think the fact someone as inquisitive as yourself objectively looked at Golem and balked should be a little heads up. Fwiw I have asked some similar questions on the different forums and weren’t shown or provided anything that leads me to believe they won’t be anything but a pretty useful application which will most likely run most efficiently on SAFE. I just see a lot of good ideas but no concrete explanation as to how they are going to be linked together.

So in short, save someone adding something more beneficial, just look at what code is sitting there in front of you and trust your instincts, respectfully smoke and mirrors have served Eth (and their projects) well (I’m invested in many of them) but they still from technical perspective leave ALOT to be desired which any diligent cookie knows will only cause issues as they move along their development scale.

Keep doing what you do best

[Traktion]

I think a key difference with maidsafe is that they are building foundations for the future. They are creating a framework and a platform.

I suspect many of these other projects are interesting and some may even find a niche. However, something needs to draw them together to be truly useful. For them.to be a marketable product, they need a stable platform to operate from.

This reminds me a bit of web technology evolution. At first there were no tools other than basic editors and static content. CGI scripts bolted on to servers to provide basic dynamic content. Then came the simple scripting languages (asp, php) and they helped to create decent interactive sites. Then came the template engines, followed by the frameworks and now we have whole applications (.net, java, php) using sophisticated MVC frameworks (spring, symfony) running on cloud infrastructure (IaaS). We even have whole platforms integrating IaaS with software to create PaaS.

I see these various one off techs as being on the early part of this journey, but I maidsafe extending the end part of this journey. Different ambitions and implications, for sure. However, I see a market for integrating clever new functionality in a leading edge platform. I am not so convinced the other new crypto techs will progress beyond curiosities.

Of course, the challenge for maidsafe is the scope of what is trying to be achieved. Creating such a platform - which replaces big chunks of what Amazon, Google, Microsoft, any many other do - ia anhuge undertaking. Some realise this and are scared off, while others moan that is wasn’t finished yesterday like the other crypto curiosities. On the bright side, I believe maidsafe are doing an admiral job of handling this huge undertaking.