conduition

+1 for this issue. Happens for many servers, not just one special guy. Most servers I tested exhibit this bug. Including: ``` ssl://fulcrum.grey.pw:51002 ssl://electrum.jochen-hoenicke.de:50099 tcp://v7gtzf7nua6hdmb2wtqaqioqmesdb4xrlly4zwr7bvayxv2bpg665pqd.onion:50001 tcp://venmrle3xuwkgkd42wg7f735l6cghst3sdfa3w3ryib2rochfhld6lid.onion:50001 ssl://blackie.c3-soft.com:57002 ssl://e.keff.org:50002 tcp://udfpzbte2hommnvag5f3qlouqkhvp3xybhlus2yvfeqdwlhjroe4bbyd.onion:60001 tcp://kittycp2gatrqhlwpmbczk5rblw62enrpo2rzwtkfrrr27hq435d4vid.onion:50001...

I can confirm that this problem is not limited to batched transaction fetching. It also affects `batch_script_get_history`. If I had to guess, i suspect afflicted servers just don't respect JSON-RPC...

I'd be happy to take a look at fixing this. @romanz could you please clarify, is blocking on mempool sync is a desired behavior? My suggestion would be to block...

PR open to provide mempool scanning in a separate thread: https://github.com/romanz/electrs/pull/970

Requested changes force pushed, amending the existing commits. @romanz [you can view the diffs here](https://github.com/romanz/electrs/compare/0bf77378ced2a91dc5b7fc45ba69f16a43dade40..0162c5fdf405e3c8d65af72844ee4bed05da11cb). If you feel comfortable with that, i'll submit https://github.com/romanz/electrs/pull/970/commits/9b5f925934de1986c803881cb866f3770a872175 and https://github.com/romanz/electrs/pull/970/commits/ed42680992994488961bbf5aa3028a7a3d4b5dcc as a new PR.

[PR open here](https://github.com/romanz/electrs/pull/971), this PR is now dependent on that one

Rebase complete :+1:

Rebased :white_check_mark: I'm now looking into the mempool missing TX error. My guess is that async mempool scanning results in a delayed reaction where `electrs` only receives the newly broadcasted...

@romanz sorry for the delay. My suspicions were correct, and I applied a fix which automatically appends the newly broadcasted TX to the mempool immediately upon broadcast.

@romanz I rebased this now that https://github.com/romanz/electrs/pull/979 has been merged. I think this PR still has merit, because even if mempool scans are now much faster, there's still no need...