In this case A would be at an advantage to spend the 2 minutes looking for the next block. If they happen to find another block quickly they could release then in quick succession.
The benefit there is that if another miner released a block before that 3 minutes this miner still can release their first block and has already spent 2 minutes working on a block that could better validate their first block now that there are competing chains.
But the time spent by B is not wasted. If they find a block between minute 1 and 2, their block will be accepted, and A just lose the reward of the block they found.
When you reveal a block, it's not accepted instantaneously. When two competing blocks are revealed "roughly at the same time", it ends up in two competing chains.
If B finds a block between minute 1 and 2, they start working on their competing chain, but A is already working on theirs. And A had a headstart because it started working on it somewhere between minute 1. So it's more likely that A's fork wins the race in the end.
But the head start doesn't change anything. At this point A is mining on their block, B is mining on theirs. There's no advantage.
I'd even say that B is slightly more likely to keep their reward because they started propagating their block earlier, so it's more likely other miners are mining on this block.
If A finds a second block between minute 1 and 2, then they win, but it would be the same if the didn't withhold their block.
When A is mining on their hidden block, they mine for a potential height of 2 that would win against a miner only able to push a height of 1. But by doing that they put the block they found at risk of being abandoned because another miner found a block in the meantime.
So if you find a block, you get almost 100% chance it'll stay if you publish it immediately. If you withhold it and find another one you get 100% chance of keeping your 2 blocks. If you don't find that 2nd one, you get <50% chance of your block to be the main chain (depending on time of reaction of another block being published, and connectivity). On the other hand, if you don't withhold it and find 2 blocks in a row, you also get almost 100% chance of keeping your 2 blocks. I fail to see how withholding is profitable.
Because you keep ignoring the part where it is profitable :-).
> If A finds a second block between minute 1 and 2, then they win, but it would be the same if the didn't withhold their block.
Except that by withholding their block, they got a headstart so they are more likely to find the second block. So it's not the same.
And you keep ignoring the fact that they don't necessarily have to wait until someone else finds a competing block. Maybe a winning strategy is to always withhold the block for 5 seconds. If you slightly increase your likelihood to find the winning block, you increase your profit, and that's the whole point.
With the interesting consequence (and that's the game theory part) where if everybody starts withholding their block for 5 seconds, then it changes the winning strategy.
> Except that by withholding their block, they got a headstart so they are more likely to find the second block. So it's not the same.
Withholding their block (5s or whatever) doesn't make them more likely to find the second block. The probability of finding a block is always the same, given a hashrate.
They are the only ones mining on this particular chain, but that's not an advantage either. How mining on a hidden chain is an advantage?
On the other hand, withholding certainly makes them more likely to lose the reward of the block.
> They are the only ones mining on this particular chain, but that's not an advantage either. How mining on a hidden chain is an advantage?
It's easier to see the argument if you have a head start. Imagine you've somehow created a private chain that's 10 blocks ahead of the public chain. You could publish that now and earn 10 blocks of reward, or you could continue mining until the lead diminishes to 0 blocks, earning the same 10 blocks of reward plus however many blocks you've mined in the meantime.
If you have 50%+ε of the hash rate on the network, this argument would have you bully other miners out by almost always stranding their blocks, since in expectation you'll mine blocks faster than your competitors.
The insight is that this same situation can happen probabilistically with a finite but non-majority fraction of the hash rate on the network. With 49% of the hash rate you'll still be able to build a private chain some fraction of the time, so waiting a little bit to see if this occurs might have positive expected value.
the key to realize is that this strategy only makes sense if you have a considerable fraction of total hashrate. If you have 10% hashrate, delaying for 1 block period gives you a 10% chance of finding another block on top (that no one else can search for because you haven't published the first one).
But by withholding you also increase the risk that your first block will never end up in the main chain (if the remaining 90% find a block while you're withholding).
And you would sill have 10% chance of mining another block if you don't withhold.
It's still possible, you just need to declare which other apps you query for. Even then, there are loopholes that still let you query for all apps installed on the device.
> Apps that have a verifiable core purpose facilitating financial-transactions involving financially regulated instruments (for example, dedicated banking, dedicated digital wallets) may obtain broad visibility into installed apps solely for security-based purposes.
> Real-money gambling apps where the core purpose of the app is real money gambling and where the app requires broad package visibility in order to comply with technical standards mandated by applicable geofencing regulations.
I presume that's to allow the gambling apps to make sure you don't have a location spoofing app installed?
13 year olds can get groomed and addicted to gambling, be they at home, school, or a bus stop. But God forbid you install an app outside the approved™ app store®, citizen. What a world.
> I assume HSBC are using the "antivirus" use case.
There's an exception for banking apps
> Apps that have a verifiable core purpose facilitating financial-transactions involving financially regulated instruments (for example, dedicated banking, dedicated digital wallets) may obtain broad visibility into installed apps solely for security-based purposes.
> This may have a nontrivial memory cost, especially at high compression levels. (Don't set the compression window any larger than it needs to be!)
It sounds like these contexts should be cleared when they reach a certain memory limit, or maybe reset periodically, i.e every N messages. Is there another way to manage the memory cost?
LZ77 compression (a key part of gzip and zip compression) uses a 'sliding window' where the compressor can tell the decompressor 'repeat the n bytes that appeared in the output stream m bytes ago'. The most widely used implementation uses a 15 bit integer for m - so the decompressor never needs to look more than 32,768 bytes back in its output stream.
Many compression standards include memory limits, to guarantee compatibility, and the older the standard the lower that limit is likely to be. If the standards didn't dictate this stuff, DVD sellers could release a DVD that needed a 4MB decompression window, and it'd fail to play on players that only had 2MB of memory - setting a standard and following it avoids this happening.
That's a misunderstanding. Compression algorithms are typically designed with a tunable state size paramter. The issue is if you have a large transfer that might have one side crash and resume, you need to have some way to persist the state to be able to pick up where you left off.
> It's entirely dependent on the type of code being written. For verbose, straightforward code with clear cut test scenarios, one agent can easily 24/7 the work of 20 FT engineers. This is a best case scenario.
So the "verbose, straightforward code with clear cut test scenarios" is already written by a human?
I think you should avoid casting a wide net, i.e untargeted ads. Post or contribute to niche communities, reach out 1:1, ask friends for leads. If none of this works, you could try highly targeted ads.
Thanks for the detailed advice,
Not $200/Mo, Just need 200 one time paying users $49 per user, for my digital service [Turn your Face into 60+ animated doodles and get attention improve CTR]
> Turn your Face into 60+ animated doodles and get attention improve CTR
I don't understand the value prop, at least not to make me pay $49. I'm not the target audience clearly, but this seems like a mechanical, low sophistication task, why don't I just use gemini or whatever to do this for free? If I already pay $20/month for any LLM that's multi modal I can do this now already.
The asset (animated doodles) doesn't really connect me to how I'm going to get a better click through rate.
I think for this to work you need a specific niche that resonates with people more, like you have a "proven way to get a better CTR for $niche topic with this asset pack which bundles the animated doodles plus a playbook for how to use them" or something like that.
Thanks mate you cleared my view, i need to market it differently. I can understand without seeing actual product no one can predict sells. When you talk about generating with AI same as my output it will cost you around 80-120 days.
Thanks for the feedback KingKongjaffa. I'll work on hero section.
A finds a block after 1 minute, then powers off and waits for another minute. They reveal the block after 2 minutes.
B searches for the block for 2 minutes.
After 2 minutes, A has used 1 minute of their compute, and B has used 2.
reply