17,101 addresses have some current amount of maidsafecoin (ie not zero)
32,883 addresses have no maidsafecoin (ie received some then sent all of it)
I’ve been using the omniexplorer api to scrape this data. Since the rate limit is about 4 requests per minute it took many weeks to get the initial data. But now it will be pretty fast to update it, so I’ll aim to update the data at the start of each month and post a link to the new dataset here.
Some deeper analysis I’d like to try later on:
check for inconsistencies or errors to improve the scraping and parsing
Very tempting. My inner analyst suggested scraping forum posts this morning, to do time series analysis on who posts when, who responds, who seem to come and go in groups, time zones etc. So the itch is there already! I don’t have a framework for time-series though.
Its probably the way omni does it. I seem to remember they have a special way to do the distribution of the coin that is not a typical send on BTC transaction.
I’ve also got a script to do this direct from the blockchain rather than pulling it from omniexplorer.info, but the blockchain is still syncing so I’ll use it next time.
Out of curiosity, is anyone running an omnicore node? Would be neat to create this list from multiple sources and check they are identical. Would be a good test run for a decentralized snapshot (if that were to be an option for transition to the new tokens).
Yes I’m running 2, one is nearly completed with sync and other one is catching up.
I was thinking to upload all the chaindata and blks as immutable data.
The one I’m recently syncing taking now over 310 hours with an 10 vCPU Cores, 60 GB RAM, 1.6 TB SSD, 1 Gbit/s Port. I will try to tweak the cache settings a bit and add a huge amount of swap because those can affect speed a lot!
I really hope the newcomer safe network experience is less arduous than this. I’m sure it will be, but man, blockchain initial block download is no joke!
Yeah, the initial sync took me almost a month. I keep multiple backups of the block files now. If Core crashes or doesn’t shutdown cleanly for some reason it’s easier to restore the block backup and let it re-sync the past few weeks than it is to let it re-validate the entire chain. The re-validation takes days vs. just a few hours to re-sync from a backup. If you have any pointers to improve performance I would be very interested.
Not directly for Omnicore, but you could use a Copy-on-Write filesystem to make more snapshots in a given time, so that the intervals between them are smaller than “normal” backups. Each snapshot shouldn’t take much space (just the differences to other snapshots, like changes in files or directory structures) and are very fast to make.
But: CoW FSs are usually slower than normal FSs. This might be a problem for you if your bottleneck is your FS/Disk.
This is the first time the transaction data is parsed directly from the blockchain using omnicore. Previously it’s been taken from omniexplorer.info
There’s two keys removed from every transaction to make the data immutable: “ismine” which depends on the person running it, and “confirmations” which depends on when the script is run.
Wow it’s so small in file size! I will try to follow your format and parse one from my omnicore aswell so people can verify it from another source. Do you have a script for this on github?