💾 Archived View for thebackupbox.net › ~epoch › blog › veilid-2 captured on 2024-12-17 at 10:11:36. Gemini links have been rewritten to link to archived content

-=-=-=-=-=-=-

more pokes

yep. an update so soon.

update

earlier post

I was looking for anything that might hint at what went wrong in the commits.

I first found a closed issue about windows computers not being able to connect to the bootstrap.

sounded like the right place to look.

eventually noticed the changelog say:

BREAKING CHANGE: ALL MUST UPDATE

so, I figured what was breaking was that I had old data stored in /var/db/veilid-server

moved it to a backup dir, and started it and got connected to the network.

then the fun part of trying to keep as much data as I could.

I went through and found that if I completely deleted protected_store or table_store it wouldn't work.

so I kept the protected_store and then opened __veilid_all_tables in sqlitebrowser to see what the randomly named files meant.

then went through and deleted one table at a time that sounded not important.

after a few of those still causing the bootstrap to not work, I got impatient and deleted all the ones that didn't sound important...

which are these iirc.

__veilid_config
crypto_caches
RouteSpecStore
routing_table
remote_subkeys

those 5 seemed like the least important and the most likely to be causing the bootstrap to fuck up.

might've been just one of those. maybe remote_subkeys, but I figure all of that could get regenerated from the network.