[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Help-gnunet] GNS future

From: Christian Grothoff
Subject: Re: [Help-gnunet] GNS future
Date: Mon, 28 Jan 2019 18:41:21 +0100
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101 Thunderbird/60.4.0

On 1/28/19 12:17 PM, Alexandre Garreau wrote:
> Hello,
> I just found this article (in french, for those able to read it):
> I’m enjoyed to see GNS gains in interest, and think to keep supporting
> old ICANN DNS name for compatibility!
> However I have two questions currently:
> Is there really a rewritting going to happen in rust?

I'm not aware of any significant progress on that front. But in the end,
multiple implementations would be a good thing, but I'm still focusing
on getting the first one in C ready :-)

> Is it going to
> become the official implementation?

The best free implementation will be made official ;-).

> Will a Rust frontend be added to
> gcc? 

~6 months ago I was told that the gcc team considered that Rust was
still evolving too quickly to be supportable by GCC.  But that does not
mean "never", Rust just needs to become a stable standard and not
something different every 6 months (I was told this indirectly, don't
shoot the messenger).

> otherwise isn’t this a problem given the sortof rivalry between GNU
> and LLVM which purposedly allows, supports and now *has* contributing to
> create proprietary languages (such as Swift, iirc)?

I don't think that would keep GNU from trying to support popular
languages.  But it's obviously a matter of developer time. And indeed,
RMS himself counseled me against banking too much on Rust until there is
a 2nd (GCC-based) compiler available for it.  Still, experiments in that
direction are very welcome.

> What about graph- rather than tree(hierarchy)-based reference system?
> for instance if I want,,, to point to the
> same thing, how might that be used (I had difficulty formulating this
> question since the first time I learnt about GNS, in 2013)? Would there
> be some anti-redundance system to ensure for instance and
> point to the same thing (because otherwise keeping the distinction
> between and might be a problem)?

Why shouldn't you be able to point and to exactly the
same zone? Many companies do effectively this with all of their entries.  GNS is graph-based, not tree-based, you can have
cycles and whatever else, and there is no unique root (other than what
might be common consensus / default settings).  So maybe I'm not
understanding the question ;-)

reply via email to

[Prev in Thread] Current Thread [Next in Thread]