dpc

``` (gdb) list 154 if (i % ARGON2_ADDRESSES_IN_BLOCK == 0) { 155 next_addresses(&address_block, &input_block, &zero_block); 156 } 157 pseudo_rand = address_block.v[i % ARGON2_ADDRESSES_IN_BLOCK]; 158 } else { 159 pseudo_rand =...

``` (gdb) p instance->lanes $4 = 0 (gdb) p *instance $5 = {memory = 0x7f326a7b12e0, version = 1786451528, passes = 32562, memory_blocks = 1786451360, segment_length = 32562, lane_length = 2,...

Can this be hmmm... easily optional or something? In some contexts the simpler the better. Making a new connection every time is easy, and has less chance of triggering any...

https://github.com/mtoohey31/cache-flake-attrs

My problem is `encfs` produces too long filenames, and due to this I can't sync it with other hosts using tools like `syncthing`. Wouldn't it be possible to have an...

> When a batched request comes, we first get all heights we want to scan and de-duplicate them. Then we fetch all those blocks from Core. Maybe it makes sense...

> Hmm probably yes if it's shared among all threads that are doing requests. Having it all go through one instance of `Arc` (from [my comment on caching](https://github.com/romanz/electrs/issues/530#issuecomment-944116370)) one could...

> I'm currently in mental state incapable of understanding what your proposal with `global_offset` achieves. I had to re-read it again, and my understanding is that it allows skipping decoding...

> However, I am not sure how the design above will handle reorgs correctly... I'm not sure what exactly the problem is, but one way or another since reorgs affect...

I did not fully understand last reorg handling comments, but just wanted to note that reorgs are so extremely rare, that any tradeoff should be at reorg handling expense, and...