MAID will be feeless and instant?

Features
23

I dunno, many times things are decided quick and need to be done quick. Also 1 second on a 32 core ryzen could be days or weeks or months on low end equipment. And 15 seconds on the high end equipment will be simply too long on low end.

But to do some work for the network where things even out a lot due to routing lags might be the area to look in rather than cpu usage. CPU usage is electricity waste and if we want 100K token transactions per hour/day over the network then multiply the electricity use for one machine by 100K and it then becomes significant. But if work is done for the network, like helping audit spending then its minimal work and takes enough time to minimise spamming, while doing work for the network that has to be done anyhow.

5 Likes
24

I agree 100% that if the ‘work’ is useful then that’d be great. Unless there is a clear and obvious path however, that may be too complicated to figure out pre-beta, so might be okay to have something not-so-great as spam prevention beforehand.

3 Likes
25

chunk auditing.

3 Likes
26

Actually the issue is nearly resolved anyhow because each transaction requires the client to do a reasonable amount of work across the network by obtaining signatures. The reason this is reasonable is that it involves communications lag which limits transaction rates any one client can achieve.

The topic was exploring the issue prior to DBCs and the requirement for clients to be doing the majority of the work.

13 Likes