Did you try to compare to mine with that node ? What's the difference of new job broadcasting relatively to any major pool ?

botcoinnotbot[m]: currently only sending transaction data over anonymity networks is supported, if you block other ports you won't receive new monero blocks anymore

I'm quite sure the same case is in wownero so I'm not sure what this dedicated I2P port is

so this isn't just about adding a command line option, this would require larger changes

<botcoinnotbot[m]> "for those who wish to have all..." <- What if you do `--p2p-bind-ip 127.0.0.1` and then you tunnel that through i2p/tor? Same as running a hidden service

Is there any other network activity by the node that doesn't go through that (ignoring rpc of course)

you can also use the --proxy option in the new release

yeah, you can always do that.

(p2p bind to localhost, tunnel that)

Hi! I noticed that it seems that all distributions (ArchLinux, Gentoo, Void Linux, Homebrew) download the sources from GitHub and ignore the tarball from the homepage. They all also check out a specific commit hash. Why is that?

I'm asking because I'm looking into packaging Monero for Fedora, but it's very suspicious to me that nobody wants to use the tarballs from the homepage.

the tarballs are useless

since there are several submodules in the build

Yeah, I noticed they all download several archives of repos. But Arch Linux for example goes through the trouble of cloning the entire repo just to check the GPG signature of a commit, rather than just downloading the tarball and check the signature of that.

But I guess since Fedora builds on offline-only machines, I should just check out the source via git, check out all submodules and then verify the signature of the commit I checked out?

that sounds workable

but you don't need all the submodules just to check one commit's signature do you?

nope, but I need all the submodules to build it, and if I verify the commit it should also check all submodules, as the hash at which a submodule is is part of the commit I checked out, right?

yes

This PR updates RPC calls on getmonero. Would be great if somebody from here could review it: erciccione/monero-site #27

Commented on get_miner_data RPC because this one was added by me :)

thanks sech1 🙂

What is wrong with make if it perfectly finds out which modified .cpp files to incrementally recompile but at the end does not link executables like `monerod` or `monero-wallet-cli` like it should?

If I delete the executables "by hand" it finally re-links

wrong dependencies declared probably

Where are those declared?

I haven't looked at Monero's Makefiles, but usually it's something like:

target: dep1 dep2

Strange thing is that I remember it working perfectly for a while, and a few days ago it stopped, without me messing anything with makefiles

Maybe it's a known thing that sometimes make makes a mess of its interal dependency info and stops to work incrementally in a correct way?

nope, make is extremely reliable, it is pretty much always a bug in the Makefile

Hmm, maybe in the case of Monero an omission rather that I somehow trigger now? Like a header file that you should import yourself but don't but which you always

indirectly get until one day you don't get anymore because of a change in another file?

*omission rather than bug

well, an omission of a dependency is a bug

what can happen is that make works fine without -j, and things happen to be in the right order by luck and then fail when using -j

but that's always a bug in the Makefile and not an issue with make

I see, thanks. Maybe I will stumble over where such a dependency might me missing. It's quite a big and complicated code base ...

of course the makefiles are all generated by CMake so it's the cmake files you need to analyze

ah, CMake. That doesn't even create real Makefiles.

All it does is create dummy Makefiles that call into CMake

uh huh

rbrunner isn’t it like sometimes cached symbols don’t get refreshed? Always annoying

Not sure which cached symbols you mean. You mean a failure of header compilation, so to say?

Yeah idk enough to say I know what I’m talking about :p

Yeah, the wonders of C++. As somebody said yesterday on reddit: "Why not Rust?" How about it, the two of us rewrite Monero in Rust!

Rust? Why not 6502 Assembly!

rbrunner: See my comment here:

If only there wasn't about a magnitude of difference in complexity between Monero the coin and BCH the coin ...

With corresponding effort needed to implement a node and a wallet

I don't think Monero needs six node implementations like BCH. Two would be nice, though. I am not a dev, so time costs would not be borne by me

Two would definitely be nice, but as soon as you set your eyes (as a dev) on the code, you will get quite strange feelings :) Been there, done that.

What do you mean?

Some parts of the code base are positively mind-boggling

all the more reason to rewrite it

In principle, yes. And if you take a really long-time view, maybe inevitable.

I am not a real programmer but I'd agree with hyc, FWIW.

I use to imagine what the talk of our grand-grand-children could be on the celebration of 100 years Monero, in 2114, and maybe the "great rewrite" may be a subject

mathwise there are complex calculations in here, but that's the only thing that is deservedly complex

everything else is just bog-standard internet server crap

selsta: would like 8070 / 8075 merged if possible

reviewed first of course

hyc: can sech1 review it?

I kind of reviewed it when I added support for it in p2pool

8070 fails 1 CI check, I added a comment there