3 ms·

When I briefly paid attention to Monero/XMR and ran a mining rig it was glaringly obvious that every time they changed the algorithm, something Monero deliberat

When I briefly paid attention to Monero/XMR and ran a mining rig it was glaringly obvious that every time they changed the algorithm, something Monero deliberately did(does?) regularly for anti-ASIC reasons, those participating in the development would have their miners active effectively across the changeover with zero down time. It seemed pretty close to insider trading to me. While the network was getting back to its previous aggregate hash rate the difficulty would be exceptionally low, so everyone closely involved had a window of easy mining every time the algo changed. It would take days for the aggregate hash rate to recover 100%...

gruez 4y ago
I don't buy it. I'm not intimately familiar with how the monero project is run, but shouldn't you able to pick this up by looking at the PRs? Or are the repository owners just keeping their PRs under wraps and merging them immediately after they're created? Even then, the bigger question is how these changes get deployed. If you're going to change the hashing algorithm, you'll probably have to get most clients on the network to update in advance, otherwise all the people running outdated clients would be on a forked chain. Therefore it seems unlikely that monero insiders would have much of an advantage compared to an observant outsider.
- xur17 4y ago
  Agreed. The network would grind to a hault if the mining algorithm was changed without notifying everyone first. I have to imagine a new version of the node software is released (fork) that changes the hashing algorithm starting at X block. Anything else would be.. stupid.
  - pengaru 4y ago
    When your currency intends to keep changing the algorithm, the protocol itself should describe the algorithm to use and automate the changeover process. It's not the manual aspect that is defending against ASIC miners, it's the non-determinism of what the algorithm is going to be. If you keep it that way for years, it starts looking very deliberate. In this scenario I'd describe the "insiders" as those participating in and following monero's development closely enough to be continuously aware of these changes at the source level. We're not talking about SEC regulated crap where insider trading is well defined, this is all unregulated wild west software land. Insiders are the nerds cloning git repos and building from source, lurking in monero chat rooms all day, etc.
- pengaru 4y ago
  It just comes down to how we define "insider" in this scenario. This was years ago now, but at the time at least this wasn't an automated mechanism. So yes, you could follow the repository and pay very close attention, and if you're savvy enough maybe even automate that process to deploy the new miner whenever changes landed. (mine often broke by amdgpu opencl incompatibilities though, it often took some hacking to make the new algo work) But from where I was sitting it was very clear that the manual nature of it created tons of opportunity for the developers and those savvy and close enough to practically be in the same privileged set as developers. Whenever they changed the algorithm the network reset and the hash rate was a tiny fraction of its previous quiescent state, with a low difficulty, for days. I was paying semi close attention at the time, and even being purely manual with a single eGPU mining rig was able to somewhat capitalize on the opportunity created. Practically all my pittance of shares were mined in post-reset low difficulty periods. Those with actual large mining farms normally competing with the full rate difficulty surely were quite productive in those windows. There's no question that the switchovers would kill the hash rate, we can infer from that most the miners were being manually maintained by folks not closely monitoring the upstream repo for every change.
Blahah 4y ago
It's not insider information because the information is accessible and discoverable. Anyone who wants to can follow the development plans, run their own nodes, pull and build the new branches, etc. There's certainly a lot of insider trading (adjacent) bad behaviour in crypto generally, but the example you've given is the opposite.

gruez 4y ago

I don't buy it. I'm not intimately familiar with how the monero project is run, but shouldn't you able to pick this up by looking at the PRs? Or are the repository owners just keeping their PRs under wraps and merging them immediately after they're created? Even then, the bigger question is how these changes get deployed. If you're going to change the hashing algorithm, you'll probably have to get most clients on the network to update in advance, otherwise all the people running outdated clients would be on a forked chain. Therefore it seems unlikely that monero insiders would have much of an advantage compared to an observant outsider.

xur17 4y ago
Agreed. The network would grind to a hault if the mining algorithm was changed without notifying everyone first. I have to imagine a new version of the node software is released (fork) that changes the hashing algorithm starting at X block. Anything else would be.. stupid.
- pengaru 4y ago
  When your currency intends to keep changing the algorithm, the protocol itself should describe the algorithm to use and automate the changeover process. It's not the manual aspect that is defending against ASIC miners, it's the non-determinism of what the algorithm is going to be. If you keep it that way for years, it starts looking very deliberate. In this scenario I'd describe the "insiders" as those participating in and following monero's development closely enough to be continuously aware of these changes at the source level. We're not talking about SEC regulated crap where insider trading is well defined, this is all unregulated wild west software land. Insiders are the nerds cloning git repos and building from source, lurking in monero chat rooms all day, etc.
pengaru 4y ago
It just comes down to how we define "insider" in this scenario. This was years ago now, but at the time at least this wasn't an automated mechanism. So yes, you could follow the repository and pay very close attention, and if you're savvy enough maybe even automate that process to deploy the new miner whenever changes landed. (mine often broke by amdgpu opencl incompatibilities though, it often took some hacking to make the new algo work) But from where I was sitting it was very clear that the manual nature of it created tons of opportunity for the developers and those savvy and close enough to practically be in the same privileged set as developers. Whenever they changed the algorithm the network reset and the hash rate was a tiny fraction of its previous quiescent state, with a low difficulty, for days. I was paying semi close attention at the time, and even being purely manual with a single eGPU mining rig was able to somewhat capitalize on the opportunity created. Practically all my pittance of shares were mined in post-reset low difficulty periods. Those with actual large mining farms normally competing with the full rate difficulty surely were quite productive in those windows. There's no question that the switchovers would kill the hash rate, we can infer from that most the miners were being manually maintained by folks not closely monitoring the upstream repo for every change.

Blahah 4y ago

It's not insider information because the information is accessible and discoverable. Anyone who wants to can follow the development plans, run their own nodes, pull and build the new branches, etc. There's certainly a lot of insider trading (adjacent) bad behaviour in crypto generally, but the example you've given is the opposite.