Core dev meeting #42

Hello,

I'm really sorry about the lack of video, we had recording problems, and couldn’t record the meeting.

Instead I wrote some notes to give you an idea of what was discussed, I'm sorry this is incomplete and we'll switch to @gtg's video solution next so that those issues won't happen again.

Developers sync:

On my end I mostly spent time updating my hivemind pull requests following the hardfork changes and supporting RC delegation questions (many heart attacks of people telling me it doesn’t work when in fact it’s something they did wrong).

Then I started working on implementing a feature request by @brianoflondon for recurrent transfers: https://gitlab.syncad.com/hive/hive/-/issues/382

On @blocktrades side (this is somewhat of a non-exhaustive list and will probably be in greater details in a post of his own):

There was some cleanup work following hard fork 26/27, doing some performance analysis on nodes, to see how things have changed.

Then they pushed a bunch of small changes:

• Fixed a problem with shutdowns that could happen sometimes when you did ctrl + c on a node.

• Fixes and improvements to the block statistics

• Added more RC statistics in the form of daily report, with a new API about it (this will happen in hf 27.1)

• New api calls for updated witness scheduling

• CLI wallet side, fixed a longstanding bug to allow controlling it from websockets.

• Updated our cryptography lib to increase performance (tests ongoing but so far the initial testing shows a 40% increase !)

• New block log utility to help manage compressed block logs:

  • Comparing compressed block logs to output a checksum even though they were compressed using different methods
  • Truncating a compressed block log (previously you had to uncompress, truncate, then re-compress)

• Improved get_block_range by forcing it to use http to improve performance due to a websocket issue. The current suspect is that websockets is doing compression which causes slowdowns

Adding a VOP to record successful RC delegation

We talked a bunch about the potential use cases and potential edge cases and issues.

The main discussion revolved around the fact that it’s hard to track if an RC delegation got executed or not (and what was executed). But after some back and forth, we concluded that since the node rejects incorrect transactions, it made sense to let the custom json denote of the change.

Hive version 1.27.1:

We want to ship version 1.27.1 very soon:tm: (aka in a few weeks) to provide improvements to HAF and hivemind. We’ll try to slot in a change to RC delegations to include an RC minimum to prevent spam and improve dev experience.

We’ll try to include at the bug that @engrave brought up: https://gitlab.syncad.com/hive/hive/-/issues/338

There's then a lot of discussions between @arcange and @blocktrades on his specific HAF setup that I don't think are really worth talking about.