[CRITICAL] Nu Network fork experienced at block 407,740

[Sabreiib]

@sigmike How is going about cold mint of Nu even B&C ?

[Sentinelrv]

@sigmike How is going about cold mint of Nu even B&C ?

Last time I talked to Sunny about this he said that he still needed to think about the implications of Sigmike’s solution before he decided to include it in Peercoin or start over from scratch. He’s not interested in rushing cold minting. The earliest it would be included is probably v0.6. I’m not sure about Nu and B&C though.

[Sabreiib]

SunnyKing… He is just thinking and thinking, don’t know what he is coding now, anyway, he is a volunteer, otherwise will be fired by his/her boss. Just a Joke.

[mhps]

SunnyKing… He is just thinking and thinking, don’t know what he is coding now, anyway, he is a volunteer, otherwise will be fired by his/her boss. Just a Joke.

You get more than what you paid (which is zero)

[Ben]

Update:

The OS X Nu 1.2.0 application package (nu-1.2.0-osx.dmg) is now available for download at https://nubits.com/download. Please let me know immediately if you have any problems.

[desrever]

Please let me know immediately if you have any problems.

Like the new dmg installer, well done!

[woodstockmerkle]

A big thank you to sigmike, Ben, tomjoad, JordanLee, Coingame, Pennybreaker and the rest of the team to get in front of the situation, get the word out to the community, discover the issue, and get a fix out there.

The team has shown its true colors once again.

[seki]

so, if we are in the correct chain, we don’t need to update anything?

[mhps]

My Raspberry Pi has been unable to find any block since the fork even though it is on the same branch. My laptop running in the same LAN with the same S wallet has been able to find blocks.
The R-pi is minting as shown by nud getinfo and load level by top.

I have compared debug.log of the r-pi when when minting is on or off on the r-pi. When the laptop finds blocks, the log shows the same activity regardless minting being on or off.

I think this is the same problem happened here and here.

[Hyena]

My Pi is minting blocks but since I currently don’t have physical access to the node I can only guess that it’s on the main chain. If http://blockexplorer.nu shows blocks being minted to the address of my raspberry pi, then does that mean I am certainly on the right chain and don’t need to do anything?

[Sentinelrv]

so, if we are in the correct chain, we don’t need to update anything?

Yes, can somebody answer this? Do we need to update if we didn’t have the problem? Will updating prevent this problem from happening in the future?

[tomjoad]

Version 1.2.0 is listed as a mandatory upgrade for all users on https://nubits.com/download. We should all upgrade as soon as possible to ensure our network remains protected.

[Sentinelrv]

Can we edit the title and body of this thread to ensure people know this is a mandatory upgrade? I don’t think it’s clear.

[Hyena]

The source code still points to the old version:
https://nubits.com/about/source-code
->
https://bitbucket.org/JordanLeePeershares/nubit/branch/1.0.1-Stable-Release

However, the latest should now be
https://bitbucket.org/JordanLeePeershares/nubit/branch/1.2.0-Stable-Release
???

I need to know which source code to download on my Raspberry Pi 2 because I need to build it from source to get the update.

[Ben]

Thanks. I’ll update that page to point to the repository rather than include a specific version note.

I need to know which source code to download on my Raspberry Pi 2 because I need to build it from source to get the update.

You should be able to use the 1.2.0-Stable-Release branch to build for your RPi. As far as I know there’s nothing special you need to do other than compile it directly on the Pi. @CoinGame and @mhps may have suggestions.

[masterOfDisaster]

I have tried that on my RaPi.

• downloaded https://bitbucket.org/JordanLeePeershares/nubit/get/1.2.0-Stable-Release.tar.bz2 to ~/downloads

wget https://bitbucket.org/JordanLeePeershares/nubit/get/1.2.0-Stable-Release.tar.bz2 ~/downloads/1.2.0-Stable-Release.tar.bz2

• unzipped “1.2.0-Stable-Release.tar.bz”

bunzip ~/downloads/1.2.0-Stable-Release.tar.bz2
...wait...
tar xvf ~/downloads/1.2.0-Stable-Release.tar

• adjusted /src/makefile.unix to not support UPNP (USE_UPNP:=-):

• installed dependencies according to ~/downloads/JordanLeePeershares-nubit-2bcdba8652aa/doc/build.txt
• compiled it

make -f makefile.unix

• found that libcurl was missing
• fixed that

sudo apt-get install libcurl4-openssl-dev

Et voilá:

nud getinfo
{
    "version" : "v1.2.0-unk-beta",
    "protocolversion" : 50000,
[...]

This is by no means a perfect guide to get your own compilation of nud on your RaPi running Raspbian as I tried to remember the steps I took. But it may help guiding you.

[Hyena]

And you didn’t run out of memory while compiling? The problem with Pi2 is that it only has 1 GiB of RAM but the compilation of nud consumes approximately 1.5 GiB of memory. Since Pi utilizes SSD it does not have a swap file by default. Adding a swap file temporarily allows the compilation to finish but I don’t think it’s normal for a compiler to consume more than 1 GiB of RAM. That’s insane, there’s something intrinsically wrong with the source code of nud.

[Ben]

@Hyena, what image are you using for your RPi2 (Raspian, Debian, etc.)? I’d need to check but I’m pretty sure that by default the Nu build process has a lot of the debug flags turned on. Turning them off would reduce the build process memory usage.

I’ve never built for RPi but I know that compiling Bitcoin and Peercoin (along with Nu) require me to enable a swap any time I’m spinning up a small DigitalOcean droplet or it will run out of memory. It could be that the Satoshi-client derived wallets aren’t optimized (maybe because of the libraries used, like Boost – I’m just speculating at this point).

Anyways, ultimately it would be nice if we could cross-compile the ARM builds on beefier machines and then shift the binaries over to an RPi, primarily to avoid the ultra-long build time even if memory wasn’t a problem.

[Hyena]

I’m using Ubuntu MATE ( https://www.raspberrypi.org/downloads/ ).

[masterOfDisaster]

And you didn’t run out of memory while compiling?

Nope. And I have only 100 MB swap (in addition to the 500 MB RAM (RaPi1)):

swapon -s
Filename                                Type            Size    Used    Priority
/var/swap                               file            102396  0       -1

...

top - [...]
KiB Mem:    496592 total,   481240 used,    15352 free,    13764 buffers
KiB Swap:   524284 total,    11260 used,   513024 free,    55272 cached
[...]

I’m running Raspbian on my RaPi.

there’s something intrinsically wrong with the source code of nud.

I can’t confim that - on the contrary it worked well.

Anyways, ultimately it would be nice if we could cross-compile the ARM builds on beefier machines and then shift the binaries over to an RPi, primarily to avoid the ultra-long build time even if memory wasn’t a problem.

I was trying to get a cross-compiling environment for ARMv6/v7 ready on my Ubuntu machine, but finally gave up. That seems to be somrthing for real savvy people