Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Openrsync: An implementation of rsync, by the OpenBSD team

github.com

Post reply on HN
Print view View original

181–190 of 193 posts

Re: Openrsync: An implementation of rsync, by the OpenBSD team

#181

Post by fsckboy »

m463 wrote: post #174
SoftTalker wrote: post #126

Earlier quoted context omitted.

It's not, for example cp -R doesn't change behavior on the basis of a trailing slash on directory names.

I was implementing something recently and stumbled across that cp difference. ugh. the trailing slash is pretty convenient.

that's not a cp difference, cp is the granddaddy here

I think trailing / could be a nice way to indicate some meaningful difference, but since autocomplete always sticks it in, just feels like a bad idea to me. I might like it if directory names always had to have a trailing /, but I am less motivated by "convenience of common cases" and much more by "absolute precision/specificity/unambiguity" belt and suspenders.

(kind of unrelated but along the same lines, I toy with the idea of getting rid of . and .. visible in the filesystem, and make them only part of the syntax of paths. then you could have unambiguous multiple links to a directory: ".. is where you came from" and .. in the root is still the root, so chroot works too)

Re: Openrsync: An implementation of rsync, by the OpenBSD team

#182

Post by m463 »

fsckboy wrote: post #181
m463 wrote: post #174

Earlier quoted context omitted.

I was implementing something recently and stumbled across that cp difference. ugh. the trailing slash is pretty convenient.

that's not a cp difference, cp is the granddaddy here I think trailing / could be a nice way to indicate some meaningful difference, but since autocomplete always sticks it in, just feels like a bad idea to me. I might like it if directory names always had to have a trailing /, but I am less motivated by "convenience of common cases" and much more by "absolute precision/specificity/unambiguity" belt and suspenders. (…

I think . and .. are useful. there's also perforce's ... which I find interesting (for example foo/... means everything below directory "foo")

Re: Openrsync: An implementation of rsync, by the OpenBSD team

#183

Post by maleldil »

isityettime wrote: post #180
maleldil wrote: post #167

Earlier quoted context omitted.

> favorite Interesting choice of word

She's excellent and her stuff has made it to the front page many times. I love seeing her work come up and I imagine many others here feel the same way.

If you're someone who can separate the work from the author, sure. She's a very intelligent person. Many of us can't.

Re: Openrsync: An implementation of rsync, by the OpenBSD team

#184

Post by alphabeta3r56 »

WD-42 wrote: post #6

What's the deal with the name? Openrsync implies to me that it's an open source alternative to a closed source program. But the original Rsync is GPL? Is this just the pushover license making it "more open"?

Also

> This system has been merged into OpenBSD base. If you'd like to contribute to openrsync, please mail your patches to tech@openbsd.org. This repository is simply the OpenBSD version plus some glue for portability.

Seems more cathedral than Bazaar to me.

Re: Openrsync: An implementation of rsync, by the OpenBSD team

#185

Post by dolmen »

denysvitali wrote: post #175
mtlynch wrote: post #143

Earlier quoted context omitted.

The gokrazy team being Michael Stapelberg : ) https://github.com/gokrazy/rsync/graphs/contributors

I mean, yes, mostly him (and he gave multiple presentations about this topic, very interesting!). - but there are other contributors too :)

  stapelberg: 329 commits
  jonas-fi: 15 commits
  others: 1 commit each
The bus factor is 1.

Re: Openrsync: An implementation of rsync, by the OpenBSD team

#186

Post by port11 »

brianwawok wrote: post #178
prmoustache wrote: post #161

Earlier quoted context omitted.

When LLM allows you to produce code n% faster, it also allows you to introduce bugs n% faster. I find it quite strange that people do not seem to be aware of that...I think many started worshipping the tool as if it was some kind of divinity and lost all objctivity. This doesn't bode well for the future if people aren't able to review code anymore.

Don’t need people to review code, need more LLMs. People aren’t that magical.

Maybe one day. Right now I can have Opus carefully draft a plan, Sonnet execute it, and have Opus review the code. It will complain about things that it had dismissed in the plan.

Example: it recently flagged ID derivation from a signing key. Opus itself suggested it! When reminded of that, it did this new 4.8-style “my bad, and here’s why I was right in the first place.”

Humans are still badass.

Re: Openrsync: An implementation of rsync, by the OpenBSD team

#187

Post by avadodin »

throwaway27448 wrote: post #153
somat wrote: post #118

Earlier quoted context omitted.

My understanding is that much of the point of openrsync is to create a second implementation of a protocol so the standards bodies don't balk at including it in their standards. Or to put it more concretely, people working on the rpki standard(who happened to also be openbsd devs) wanted to use rsync to transfer bulk data. The standards body was hesitant, while rsync is ostensibly a documented protocol, there was onl…

Why would we explicitly develop code legally available to capital?

Most people using open licenses over right–restricting GPL–style licenses do so for philosophical reasons.

They don't necessarily think they are better for getting reciprocal contributions from evil corporations even though many do argue that today, especially after GPLv3.

See the redis licensing debacle and why and how it was resolved. If I understand it correctly, the FSF could technically sell all of GNU tools to Microsoft which could in turn keep them closed source including all contributions over the past 40 years. Not without losing all of their remaining credibility, but still.

Re: Openrsync: An implementation of rsync, by the OpenBSD team

#188

Post by isityettime »

maleldil wrote: post #183
isityettime wrote: post #180

Earlier quoted context omitted.

She's excellent and her stuff has made it to the front page many times. I love seeing her work come up and I imagine many others here feel the same way.

If you're someone who can separate the work from the author, sure. She's a very intelligent person. Many of us can't.

Fair enough. I didn't know what was supposed to be objectionable about her personally until your last comment made me do some Googling. Bleh.

To be honest, I feel like I still don't really know much about who she really is or what real political work she's doing recently, if any. And I kinda don't wanna know anyway; I don't wanna play political blacklist enforcer.

But I understand your reaction now. :-\

Re: Openrsync: An implementation of rsync, by the OpenBSD team

#189

Post by trollbridge »

asdsg wrote: post #176
trollbridge wrote: post #137

Given the sudden spike in vibe-coded commits to the rsync codebase, and regressions that’s introduced, this is very good news.

Yes, I also don't know how license critical distributions like Debian can even ship rsync now because it contains a) laundered code and b) Tridgell cannot claim copyright over the Claude additions. So rsync should be pinned to an older version and just get security updates.

Yeah. Pretty discouraging particularly since I have always associated Tridgell with high-quality code.

Re: Openrsync: An implementation of rsync, by the OpenBSD team

#190

Post by NewJazz »

tptacek wrote: post #28

rsync has specific running modes for the super-user. It also pumps arbitrary data from the network onto your file-system. openrsync is about 10 000 lines of C code: do you trust me not to make mistakes? No, but that's why almost nobody runs it outside of strict trust boundaries. This security section would make more sense if rsync was like curl, which routinely deals with hostile counterparties. If the other side of…

With rpki, you sync from sources you may not trust AIUI.
Post reply on HN