HDegroote

Below I outline my experience in trying to get beaker/hyperdrive to work on a Raspberry Pi 4. It does not help in resolving this issue, but it details a work-around...

Can confirm that all three remote cores from the tutorial are no longer being seeded (at least at the moment). I'll open another issue to clarify this for future users.

While looking into this I also noticed that this line: https://github.com/holepunchto/hyperdrive-next/blob/973bc95cdaa630a40c523d57f8f1443756bc9940/index.js#L119 is probably not relevant (I don't see how it could be true). However, assuming it is relevant, then I...

> still relevant @HDegroote ? This test now passes in main, so the underlying issue was already fixed in the mean time. I updated the PR to only include the...

Fixed a pending issue of accessing private `corestore._closing` (thnx to just-published corestore version which uses ready resource), so adding this test should now be good to merge

I wasn't aware of the fact that a writable core loaded in another namespace is still writable. It doesn't seem like a good pattern to do this, which makes this...

Converted to draft until I updated it. The clean solution is to throw a clearer error when the block is not locally available

Added a similar partial fix for createReadStream and createHistoryStream, but I'm not sure they are desired. They seem to solve the race condition (without those fixes the corresponding tests fail),...

The `--private` flag uses a different approach. In particular, the 'allowed' list is incompatible with `--private` mode in the server, and the `-i` option is incompatible with the `--private` option...

Converted to draft while we're doing an iteration on secure-key