[Nix-dev] Google Summer of Code 2015

[stewart mackenzie]

100% Reproducible Hydra builds

NixOS already has a very impressive track record in delivering quality
reproducible services via AWS and other cloud platforms. The secret
sauce is NixOps, Nix and most importantly Hydra. Hydra is the heart
and liver that keeps the packages cleanly building and circulating to
all our users. Due to the high demand, Hydra often becomes a
bottleneck. Our goal is to live off master branch, or at least live as
close to master branch as possible. By living on master our community
responds quickly to arising issues.

This project lays the foundations of a truly distributed build system.
It will be your job to ensure Hydra

Skills
* Perl
* C/C++
* Familiarity with Named Data Networking

On Thu, Feb 5, 2015 at 3:04 PM, stewart mackenzie <setori88 at gmail.com> wrote:
> I prefer the NDN approach for a number of reasons:
>
> * An Alan Kay quote: The Internet was done so well that most people
> think of it as a natural resource like the Pacific Ocean, rather than
> something that was man-made. When was the last time a technology with
> a scale like that was so error-free? The Web, in comparison, is a
> joke. The Web was done by amateurs.
> * IPFS.io tailors to patching flaws in HTTP and is aimed at the browser.
> * IPFS has many moving pieces, NDN is much simpler in comparison.
> * NDN has provenance built into the protocol, IPFS does not.
> * NDN isn't dependent on the browser and is meant to support other
> non-browser applications.
> * A web-of-trust implementation on NDN would be much simpler, as key
> dissemination is done via the NDN. Eelco's public key could be
> distributed with each Nix installation. As long as the NDN obtained
> list of trusted contributors signed by that key it's safe.... within
> the social limitations of a web-of-trust.
>
> So, no, I do not support a bittorrent P2P styled setup.
>
> On Thu, Feb 5, 2015 at 2:29 PM, Michael Raskin <7c6f434c at mail.ru> wrote:
>>>Possibly so, though maybe limiting the scope of this GSoC project to
>>>reproducible builds is a suitable approach? Thus laying the foundation for
>>>whatever dissemination strategy to be adopted in future.
>>
>> I hope the paragraph before the one you cited expresses my support for
>> that idea.
>>
>> Bittorrent/GnuNet PoC for P2P substituter could also be added (with
>> manual content hash list management).