monero-update output on windows

Do you have GCC or CLANG installed ?

cl is the Microsoft compiler I believe. Not sure why it's both finding and not finding it.

And g++/clang++ too :)

moneromooo: Did someone test github.com/moneromooo-monero/monero…/blob/master/monero-update.qml#L174 on Windows?

Not to my knowledge.

Ok AFAIK Windows requires "file://"

you can simply use `QDesktopServices::openUrl(QUrl::fromLocalFile(path))` but you have to call it from C++

Are you implying the current one does not work on windows ?

The filename is from boost's temporary file API, so it should be an absolute filename.

Windows apparently requires this: "file:///C:/path/to/file"

:|

see here stackoverflow.com/questions/3156369…cesopenurl-with-file-url-containing

Alright, thanks.

and bugreports.qt.io/browse/QTBUG-48037

mooo. I am just using msys 64bot

bit*

also installed : packages.msys2.org/package/mingw-w64-x86_64-gpgme

And gcc/clang ?

whatever was on monero gui and cli

one liner

let me check

would have to think toolchain would have both

mingw-w64-x86_64-gcc-9.2.0-2

Does it have a gcc/g++ alias ?

Packages (2) mingw-w64-x86_64-llvm-9.0.0-1 mingw-w64-x86_64-clang-9.0.0-1

both were not installed with toolchain, GCC was

A name like that isn't likely to be found automatically.

If you run "gcc --version", what does it say ?

gcc.exe (Rev2, Built by MSYS2 project) 9.2.0

Then cmake seems not to like it. You can post CMakeOutput.txt and CMakeErrors.txt, maybe they'll have clues.

*windows update restarting.....

lol

wtf CMAKE_C_COMPILER:FILEPATH=cl

file path CL

brb, going to let it install clang

If yo have MSVC installed, it might be cmake chooses it. That'd be a mistake.

You can try removing it from the PATH before running cmake (with a clean build tree).

I dont have msvc installed on this guy

must delete monero blockchain to install mcvs :(

cant build monero on msvc

so... what *is* the deal with kovri.

there seems to be sentiment that its dead, even though the crypto media seems to think its still kicking

wasn't it some kind of scam and the author did an exit?

AFAIK, anonimal dropped off, says he's working alone in private on sekreta, a layer above kovri/tor/etc.

I dont see kovri as an scam

more like an half complete project

The GUI people are using i2p0 now instead.

gingeropolous: last statements from anonimal say kovri is still in deelopment and should be released soon. see twitter.com/whoisanonimal/status/1186908160134729729 and repo.getmonero.org/monero-project/monero-site/issues/990

reminiscent of fusion power

<learninandlurkin> I guess when kovri/sekreta drops there's going to be some debate about which direction to take going forward. Now that tini2p, i2pzero and tor integration have come along so far.

We are actually using fusion as an energy source now, more and more.

<learninandlurkin> I guess it depends on whay sekreta actually is

<learninandlurkin> Remote fusion power

AIUI, it's a layer that sits between a user program (eg, monero...) and a number of overlay networks (eg, tor, kovri...). Monero uses sekreta, which delegates to whatever overlay networks it's got available.

Yes, pretty remote :)

So it's got the potential to get the better of the networks it can worth with, but also the worst. The devil will be in the details.

So instead of having ext2 calls, you deal with libc, and libc delegates and your code works also with zfs, tmpfs, whatever.

wouldn't use of several networks, while providing decentralisation, require good network bridges to avoid long forks, make bug tracking far more complex and perhaps provide additional attack surfaces

is this normal? paste.debian.net/hidden/dbcfe757

Yes.

Some poor sod who's trying to fork but messed it up I think.

Or someone who finds amusement in spammings people's logs, you never know, some people can spend lots of time on pointless stuff.

In any case, harmless.

well thanks for choosing me i guess 😁

kinghat, the Choosen One!

Better than to be the Cheesy One.

i dont know man. hard to get paid on my single process from someone behind a version

kinghat: we are all chosen :)

Oh. Sad to see shapeshift was hit by legal crap and had to abandon their entire premise.

seen sideshift.ai ?

Nope. Thanks for the suggestion. I'm looking at alternatives now.

<azy "seen sideshift.ai ?"> Whats good about sideshift?

No kyc requirements unlike shapeshift.io (now). I'm just coming out of a crypto-hibernation and these events from 2018 are news to me.

Hey! Quick question: when mining can/should i make use of my node?

It'll be slower, but it can be used normally otherwise.

hey, 3900X owner here. i've seen xmrig rx/0 benchmarks from other 3900X owners showing >14KH/s with mild CPU overclocks, but they're saying their crazy high hash rates are due to memory timings. Has anyone A/B'd these personally?

> It'll be slower, but it can be used normally otherwise.

Thank you

i've done a manual all-core OC to 4.2GHz, set the IF to 1800MHz, and loaded the XMP C16 timings, and getting ~11.5KH/s, which appears fairly standard for a well-tuned 3900X. Not a lot of info out there, but i'm curious if anyone has tooled around to diff C14 3200MHz to C16 3600MHz ram.

Do you have huge pages enabled ?

I bet he has, or results would be worse

fortranny: check ryseb dram calculator. subtimings are very important.

Inge-: my DJR series (follows the hynix CJR recommendations) won't post even on those safe timings.

(it's g skill trident Z neo DDR4-3600)

I get 13.8KH with tuned 3200C14 ram at 4 - 4.1ghz all core

okay, that sounds more like it then.

good B die + tight timings

but others get similar rates around 3600 too

rev e and higher clocks

so i did the thaiphoon burner SPD dump into the ryzen calculator (1.6.2) V1 profile for 3600MHz single-ranked DDR on an x570 chipset, and neither fast nor safe will post

some of the settings the calculator throws out are insanely different than what's currently configured, like an order of magnitude larger

evening all would any body have some time to give me some advice regarding dev work thanks

ok, jammed away at those timings and DIMM voltage again, got to 14KH/s even on a 4.2GHz all-core OC

96C on stock cooling, way to win it

any devs about that give me a few mins of there time thanks

just ask!

fortranny: what voltage?

fortranny: clocking down to 4100 or even 3950 wont cost much hash and allows volting down and better efficiency

Inge-: it takes 1.34375 to hold 4.2GHz, can get it cooler and much less voltage to go down to 4.05 or 4.1

and yeah, not much loss of hash

i've downclocked them to 4100 @ 1.25 for now, only lost about 300H/s