Pif01010 wrote:Okay sorry I've been gone for a bit, I was a bit overwhelmed and didn't know where to begin, but let's start by me fulfilling your request.
I talked to one of my tech support members and we believe this is probably what you are looking for...
Okay, so, on the computer where Namecoin Core is showing the "ERROR: ReadBlockOrHeader: Deserialize or I/O error - ReadCompactSize(): size too large at CBlockDiskPos" message in debug.log, first make sure that Namecoin-Qt and Namecoin Core aren't running, then move the following files/folders to any other empty directory:
- blocks
- chainstate
- database
- blk0001.dat
- blk0002.dat
- blkindex.dat
- nameindex.dat
Make sure that you remember what empty directory you moved them to, so that you can move them back if I recommend doing so later. I'd suggest creating an empty directory as a subdirectory of your Documents folder and moving them there.
Then start Namecoin Core (not Namecoin-Qt). It will re-download the blockchain, and hopefully this will resolve the corruption. It will probably take a few hours. Your coins and names will stay intact, but you might not be able to make new transactions until the blockchain has downloaded. Once the blockchain has downloaded, the issue should be resolved (and you can keep using Namecoin Core on that computer; don't go back to Namecoin-Qt 0.3.80 after that).
Pif01010 wrote:1. when I try to register a domain it asks for a JSON string or something like that?
Does that ring a bell?
This is the JSON value that's attached to the name, which would include information such as the IP address that you want the domain name to point to. If you're just holding the name and not pointing it to a website, you can enter whatever you want. It looks like there's a bug right now that prevents leaving it completely blank (see
https://github.com/namecoin/namecoin-core/issues/118 ), but you can enter a single character (e.g. a letter or number or space or question mark) and it will work fine. (The longer the data you enter, the higher the fee will be, so keep it at a single character unless you actually want the domain name to point to a website.)
Our NLnet funding does include some funding to make this particular aspect more user-friendly, so hopefully that'll be improved soon.
Pif01010 wrote:1.75 now it keeps bugging me to go through with the transaction, even though it's already registered on my wallet
If I recall correctly, when Namecoin Core asks you whether you want to finish the transaction, say Yes. It will then ask you for your wallet password. It will then probably show an error saying that the name is already registered (or something to that effect), and it will ask you if you want to cancel the transaction. At that point, tell it to cancel the transaction, and it won't bother you anymore about it. The relevant GitHub issue is
https://github.com/namecoin/namecoin-core/issues/119 .
Pif01010 wrote:1.99 Which brings me to when I have these sort of transactions it doesn't let me cancel the transaction an error that says
The transaction has to be canceled the way I described above. The technical reason for this is that the usual method of canceling transactions is used to cancel transactions that are already signed and present in your memory pool; the name_firstupdate transaction isn't signed or added to your memory pool until you click Yes when prompted to finish the transaction and enter your wallet password.
Pif01010 wrote:2. Perhaps more concerning though is on my core it doesn't show which transactions are which other than a long string of transaction Id's that is very difficult to tell which transaction is which. Is this right or is this a bit of a glitch or something?
I think this is only the case for name_new transactions; the name_firstupdate and name_update transactions should list the name. The technical reason for this is that name_new transactions don't actually contain the name being registered, they only contain a salted commitment to the name (this is to prevent someone else from seeing names that you register and then trying to register the same name themselves before your registration gets mined). I think the old 0.3.80 client had some custom logic in place to label the name_new transactions with the name, even though the transactions themselves don't have a name. I'll look into whether we can add a similar feature to Namecoin Core; the relevant GitHub issue is
https://github.com/namecoin/namecoin-core/issues/180 .
Pif01010 wrote:Since I can get transactions through on my other computer which works fine without resorting to such extreme measures
in theory could I up the reward for minors to get transactions through?
Transactions generated by Namecoin 0.3.80 sometimes (50% of the time) violate something called the low-S rule. Most miners will reject transactions that violate the low-S rule regardless of fee paid, because transactions that violate the low-S rule are sometimes used by attackers to harm the network (allegedly this kind of transaction was used as a component of the MtGox theft). So basically that rule is there for security reasons, and miners probably won't allow transactions that violate it. (Technically it's possible for an irresponsible miner to mine them anyway, and such a miner might be more willing to do so for a high fee, but I don't recommend it.)
That said, if you create a transaction with 0.3.80 and it hasn't been mined within a few blocks, that's usually a sign that it was rejected due to the low-S rule, and the straightforward solution is to simply cancel the transaction and then try again (at which point it will succeed with 50% probability, if it doesn't get mined within a few blocks, keep canceling the transaction and trying again until it goes through).
Pif01010 wrote:I'd really like to at least move another 500 NMC to my other qt, that feels like a good move as that one seems to have domains getting closer to expiring everyday now.
And that would give me some relief on that front.
In theory you can send coins to that wallet from any other wallet (but if you're sending from an 0.3.80 wallet, there's a 50% chance that it won't go through, see my above comment).