I use 20 paper wallets to keep the amount of BTC low in each wallet and store copies at two separate banks.
If BTC address are compromised through cryptography, BTC wil be worthless anyway. No point in spreading to a lot of address, just spread to a few different places.
Depends on the compromise. Bitcoin can adapt to certain compromises. And some compromises will affect some schemes but not others.
Still, similar to what DeathAndTaxes said, it depends in part on what counts as a wallet. If I keep a single text file with a bunch of private keys, independently generated with a true random number generator, on a USB drive in a safe-deposit box, is that one wallet, or many wallets (or no wallets)?
From a cryptographic standpoint, I'd say that counts as many wallets. Whereas if you're using a single BIP 32 chain based on a single parent key, I'd say from a cryptographic standpoint that counts as one wallet.
Um, no I wouldn't think so, because cryptography is the mainstay of the whole bitcoin system.
When the P=NP puzzle is solved or when quantum computing reduces P to NP, then basically bitcoin will become worthless. It's a main reason why so many of my mathematician and computer science friends won't invest in bitcoin.
That's why so many white hat hackers are stating that passwords, encrypting, hashing and salting are going to be dead in the future.
When the P=NP puzzle is solved or when quantum computing reduces P to NP, then we move to to a different signature system. If all your coins are in separate wallets, with the private keys independently generated with a true random number generator, then your public keys are unknown.
If SHA-256 and/or RIPEMD-160 are broken, especially if they are broken in certain ways, then maybe it won't matter, and maybe the whole system will come crashing down in an instant. But there are plenty of potential problems which might come up whereby your coins are safe for long enough for you to transfer them to a new signature scheme.
If the problem is with SHA-256 and/or RIPEMD-160, and it makes attacks easier but still require a lot of computing power, then you might be especially glad that you didn't protect all your coins with a single 128-bit seed (a la Electrum).
Based on the history of cryptographic breaks the later scenario is many magnitudes more likely. Flaws usually begin as academic breaks on simplified versions of the algorithm, then eventually evolve into faster than brute force attacks on the full algorithm. Those early weaknesses are usually impossible to implement. For example is a weakness reduces the security of SHA-256 from 256 bits to 102 bits well that is worrisome however a preimage attack would require more energy than the entire human race uses in a year. Eventually the combination of Moore's law and improved cryptanalysis reduces the cost/complexity of an attack until it is economical.
A good case in point is that SHA-1 has been considered "weakened" for almost a decade now. However the current estimated cost to perform a SHA-1 preimage attack is on the order of millions of USD (and months of time). Bitcoin could use SHA-1 for pubkeyhashes and unless your address holds tens of thousands of BTC you would be in no real danger. Remember the cost is NOW in the millions of USD after SHA-1 being considered degraded for almost a decade.
On a long enough timeline it is highly likely new stronger address types will be deployed. Bitcoin can survive cryptographic breaks just fine. Intelligence agencies use the same kind of risk model. If the Chinese get a hold of an encrypted classified document it is only a matter of time before they will gain access to the secrets so NIST sets encryption standards to ensure that so much time will have passed that the information will have very little value. For example say a document which had specs on the first stealth fighter was stolen (but remained encrypted). If cryptanalysis and Moore's law eventually break that protection 50 years from now well it has done its job. US stealth technology will be so far advanced from that first version the damage from the loss will be minimized due to the effect of time.
i hope you know what you are doing