Gnu: Net | Hacker News Reader

Gnu: Net

gnunet.org

262 pointstomrod50 comments

Open articleView on HN

Discussed in June: https://news.ycombinator.com/item?id=20319419

A related post from September: https://news.ycombinator.com/item?id=20884463

Previous discussions include 2017: https://news.ycombinator.com/item?id=15877908

2013: https://news.ycombinator.com/item?id=6961018

https://news.ycombinator.com/item?id=6194553

https://news.ycombinator.com/item?id=5845435

2011: https://news.ycombinator.com/item?id=2515519

I love the principle behind this initiative. One suggestion to the people behind this work. Give this page a Stripe like ease of getting started with the first use case. I'm sure developers would love to get behind something like this. We just need to make it easy for them to get started.

Question: How do I get involved in contributing to these types of projects, I wish there were mentorship programs focused on helping software professionals get involved in open source/libre contributions. Any thoughts?

send them email.

The GnuNet people have an entire page detailing ways to get involved: https://gnunet.org/en/engage.html

But in general, just ask! Whenever someone posts a thoughtful comment, question or patch to one of my open source projects, I'm very appreciative.

Do you have a particular project you'd like to contribute to? If not, is there a particular area that's interesting to you?

Truthfully, at this point anything goes. Browser, mail, instant messenger, bookmarks manager, a message [bus] that can be used to open a link on another (authorised) device, etc. etc.

Do you guys have any project that is more contribution-friendly to people not deeply invested in the development of GNU Net so far?

> But in general, just ask!

You mean like in...

> Question: How do I get involved in contributing to these types of projects[...]?

That's a fair point.

Ask the _maintainers_ of the project, not a random person on the Internet.

There are lots of open source mentorship programs, some of them accept the kind of folks you are talking about:

https://github.com/fossjobs/fossjobs/wiki/resources#internsh... https://github.com/tapaswenipathak/Open-Source-Programs

I found the best way to contribute to open source is to dive in. Contributing to the open source things you use by fixing issues affecting you is a common way to get involved. If the software you use doesn't have any downsides that affect you, usually they will have an issue tracker, sometimes with "newcomer" tags. In the worse case you can always find some spelling errors using codespell or similar, get those fixed and proceed from there to other static analysis tools and then to reviewing the code and fixing any issues you find.

Although it's not sexy, lots of projects are in dire need of documentation/documentation maintenance.

How can I build port an application that depends on reliable, unordered packet delivery to GnuNet? Does GnuNet have its own web browser? Are there "WebSockets for GnuNet?"

From the browser perspective: https://gnunet.org/en/use.html#gns_browser

Are there any example application that makes use of GNUnet?

There are literally five+ applications listed and described further down the page.

The scope of ambition for this project is breathtaking. Best of luck to them!

A more cynical take:

The scope of this project was ambitious when it was launched, eighteen years ago. Today, it's abundantly clear that it was too ambitious (and/or the project was poorly managed); they still don't have anything usable, whereas other projects like Tor and I2P have produced usable tools in much less time.

ReactOS has been in development since 1996. Hurd since 1990. Some things take time.

ReactOS is entirely usable, if niche. GNUnet isn't usable. The home page describes it as "not ready for usage beyond developers", "still only suitable for early adopters with some reasonable pain tolerance", and "unlikely to provide good anonymity" -- it's unfit for purpose.

Hurd is an abject, embarrassing failure. Individual developers have produced more stable, complete operating systems as a joke (like PonyOS) than the entire GNU project has managed in 29 years.

Does contributing to Linux not count? Some things are hard and ultimately fail. It's impossible to have any success without toleration of the possibility of failing.

I think they should have backed out of it a little bit earlier. Being able to recognize your own failing might be as important as success.

Well, you have to fail anyway, so failing quicker might save you some time and get you to a good solution quicker as well. Understanding that should be a core part of any HNer at least. Probably wouldn't hurd the free software world either. (pun intentended)

Some of the GNU userspace projects surrounding Hurd have become important parts of the Linux userspace. Hurd itself, however, is wildly different from the Linux kernel, and has had little to no influence on Linux.

Emacs is also a more complete operating system.

To be fair, Hurd was also a joke... in the end.

Does anyone know what it’s tech stack is or whether it was architected/designed well enough to build off of? It’s a cool project but it has been running a long time

Ridiculously overcomplicated, and almost entirely unique to GNUnet:

https://gnunet.org/en/architecture.html

My uncharitable guess would be that most of it is completely unusable outside the context of this project.

I just needed one peek at that crazy diagram to agree with your point.

It might look complicated at first, but the parts seem reasonable too me. It might be simpler to understand when the building blocks are grouped and larger components only show top level (grouped) dependencies instead of the full graph.

Seems like the actual name is GNUnet, I don't know why this title of the submission but I guess it mirrors the logo... bad choice

I like the paper. It it is interesting to see other people that are interested in this.

When I try to explain why IPv6 is a privacy failure and consumers should turn it off I get modded to oblivion. https://privacylog.blogspot.com/2019/09/ipv6-hurts-your-priv...

But when other people are aware of passive attacks and see the economic incentives I think there might still be hope.

This has nothing to do with IPv6. Unless I am missing something, it is literally just a result of NAT, which using IPv4 does not guarantee, and which using IPv6 does not need to prevent.

Regarding the blog, yes it has nothing to do with the specification of IPv6. It has to do with the way that everybody (nearly everybody?) uses IPv6.

This distinction is already complicated enough that we should just tell everybody that "IPv6 is bad" because it is true enough.

How does it compares to Yggdrasil?

https://yggdrasil-network.github.io/

Although there are some minor similarities, it is quite a different beast. Yggdrasil aims to be a general purpose globally scalable routing scheme and the current implementations do this by providing end-to-end IPv6 transport. GNUnet seems to be much more tailored to specific applications at this stage and may not scale well either.

Is there a texting application built on GnuNet's SCTP-like technology?

How would one host a website over GNUnet?

Can browsers work on this network? Or is that not how this works?

GNU's own page on GNUnet https://www.gnu.org/software/gnunet/ redirects to gnunet.org, so I would think that it is associated with the GNU project.

GNUnet is AGPL as well.

As another comment[1] mentioned, the software is mentioned in gnu.org as a GNU package[2].

[1] https://news.ycombinator.com/item?id=21672653

[2] https://www.gnu.org/software/software.html

The connection is more clear in this sub-link imo: https://www.gnu.org/software/network/

It links to a Savannah page that links to gnunet.org

e.V. or eingetragener Verein, is a registered non-profit organization that must serve a good cause, imo. Usually used for sports, culture, science organisations etc. It's not a company

I wonder how the FSF feels about this, they seem to be using FSF trademarks without attribution.

It's a GNU package - see this list: https://www.gnu.org/software/software.html

From what I've seen, many GNU projects don't explicitly attribute the name.

> Even though transport encryption is increasingly being deployed on the Internet, it still reveals data that can threaten democracy: the identities of senders and receivers, the times, frequency and the volume of communication are all still revealed.

Can threaten democracy? What a weird wording... Typically where these informations are used to block people from accessing the internet, aren’t democratic countries to begin with.

>Typically where these informations are used to block people from accessing the internet, aren’t democratic countries to begin with.

Blocking people from accessing the internet is crude. Tracking people accessing the internet is even better, and that also happens in so-called "democratic" countries (that is, countries than once in 4 years allow the masses to vote for a party or candidate between 2-3 pre-selected options).

Minor nitpick: in the US, it’s two of the big parties, and a few of the small parties (so not 2-3 total). Last time (2016), there was the Libertarian Party and the Green Party on the ballot.

Yeah, but it's usually merely 2 parties that have any chances at all, by the way the voting system is setup, and how the whole game is played (for one, ruling parties having the money to advertise and secure their positions and playing with advantage beyond their program over any new party)

Reply on news.ycombinator.com