# I am the Watcher. I am your guide through this vast new twtiverse.
#
# Usage:
# https://watcher.sour.is/api/plain/users View list of users and latest twt date.
# https://watcher.sour.is/api/plain/twt View all twts.
# https://watcher.sour.is/api/plain/mentions?uri=:uri View all mentions for uri.
# https://watcher.sour.is/api/plain/conv/:hash View all twts for a conversation subject.
#
# Options:
# uri Filter to show a specific users twts.
# offset Start index for quey.
# limit Count of items to return (going back in time).
#
# twt range = 1 20
# self = https://watcher.sour.is/conv/w7kpceq
Hey @prologic, I think feeds are being removed from my following list. I've certainly followed more than seven people, because I click "follow" on almost everyone involved in conversations here. Users I am very sure that I have followed here include @xjix and my main feed on mckinley.cc which I just followed for the third time. I am fairly sure I've followed @lazarus and @jlj as well.
@mckinley Hmmm
Can you refollow the feeds/users you think are missing now? I want to make sure there isn't a nasty bug.
I am aware that an abrupt (_think kill_) can cause data loss, But it shouldn't, something to work on.
@mckinley Hmmm
Can you refollow the feeds/users you think are missing now? I want to make sure there isn't a nasty bug.
I am aware that an abrupt (_think kill_) can cause data loss, But it shouldn't, something to work on.
@mckinley Hmmm\n\n \n\nCan you refollow the feeds/users you think are missing now? I want to make sure there isn't a nasty bug.\n\nI am aware that an abrupt (_think kill_) can cause data loss, But it shouldn't, something to work on.
Okay I saw the event and verified your Following list is updated.
I _think_ this was the result of a crash. Have you only noticed this one or maybe twice?
Okay I saw the event and verified your Following list is updated.\n\nI _think_ this was the result of a crash. Have you only noticed this one or maybe twice?
Okay I saw the event and verified your Following list is updated.
I _think_ this was the result of a crash. Have you only noticed this one or maybe twice?
I followed several feeds about a half an hour ago, and they all appeared in my following list as they should. I've had to re-follow my feed on mckinley.cc at least one time in the past, but I haven't consciously noticed other users missing before today.
I followed several feeds about a half an hour ago, and they all appeared in my following list as they should. I've had to re-follow my feed on mckinley.cc at least one time in the past, but I haven't consciously noticed other users missing before today.
My educated guess was the power outage I had a few weeks ago. Sorry about that.
My educated guess was the power outage I had a few weeks ago. Sorry about that.
I'll see if I can also prevent this from happening in the first place, the database is meant to have a "write-ahead-log" -- so as long as key/value pairs are written to the log, it should be possible to replay them on a crash or power outage.
I'll see if I can also prevent this from happening in the first place, the database is meant to have a "write-ahead-log" -- so as long as key/value pairs are written to the log, it should be possible to replay them on a crash or power outage.
No need to apologize, I was just curious. Thank you very much for looking into it. I know you're a busy guy.
No need to apologize, I was just curious. Thank you very much for looking into it. I know you're a busy guy.
Recent shutdown confirm that what I originally suspected:\n\n 
\n\nThat is, abnormal shutdown causes _some_ keys to not be correctly written and persisted to disk. This is _likely_ a but with prologic/bitcask -- Which I'll investigate separately as the idea of having a write-ahead-log is NOT to have this happen at all! 🤣 But I suspect it's a bug with indexing and not actual data.
Recent shutdown confirm that what I originally suspected:
That is, abnormal shutdown causes _some_ keys to not be correctly written and persisted to disk. This is _likely_ a but with prologic/bitcask -- Which I'll investigate separately as the idea of having a write-ahead-log is NOT to have this happen at all! 🤣 But I suspect it's a bug with indexing and not actual data.
Recent shutdown confirm that what I originally suspected:
That is, abnormal shutdown causes _some_ keys to not be correctly written and persisted to disk. This is _likely_ a but with prologic/bitcask -- Which I'll investigate separately as the idea of having a write-ahead-log is NOT to have this happen at all! 🤣 But I suspect it's a bug with indexing and not actual data.