The logic showing that warning tries to avoid false positives. I don’t know what burntsushi saw it - he disowned all his comments when he deleted his account, so I don’t know what his last few comments were to see what the flags were or who made them. In the last three years no mod had a private conversation with him, either about his posting or anything to do with the site (the mod notes were added three years ago, so I can’t speak confidently about before then). I don’t have more insight into why he deleted his account and I’m sorry to see him go.
Is there a way we can get his name back on his comments? Most of his comments were excellent and helpful, and it seems a shame to not have his name on those (not to mention making this whole scenario look a lot worse for Lobsters).
So it just runs an update comments set user_id=-1 where user_id=42 query. Unless you’re going to restore from a backup, I don’t think this can just be corrected.
Well, if there are database back-ups, it would certainly be possible to make a query against a back-up to find all the comments made by burntsushi, then run the queries to change those comments’ owner to burntsushi’s deleted account against the live database.
If this is to be done, it should probably be done for not just burntsushi, but also other users since december 2018 (which is when the bug was introduced). And if that commit contains other bugs as well, not just disowning comments, should it be done for it as well?
You would need the data. The statement itself won’t tell you what specific rows were affected by it at the time. If it somehow logged the ID of the stories and comments it was updating it would be trivial but that would be somewhat unusual to log.
haha yeah. I guess it’s kind of advertising sinklog.com, which is just a free utility that uses the methods in the blog post. I mostly just thought that it was somewhat novel and folks might be interested in how it’s built.
The catch-all is “content marketing”, offer your potential audience something useful or interesting so they’ll see your name, think well of you, and maybe eventually buy whatever you’re offering. The term covers both businesses posting to attract users/employees and employees writing personal blogs to build their own reputations.
The logic showing that warning tries to avoid false positives. I don’t know what burntsushi saw it - he disowned all his comments when he deleted his account, so I don’t know what his last few comments were to see what the flags were or who made them. In the last three years no mod had a private conversation with him, either about his posting or anything to do with the site (the mod notes were added three years ago, so I can’t speak confidently about before then). I don’t have more insight into why he deleted his account and I’m sorry to see him go.
The disowned comments might be a bug: https://twitter.com/burntsushi5/status/1399716212028985351
I found the line in the logs, it’s a very unfortunate bug:
I’ll take a look at the threads linked there, thanks for the reference.
EDIT: filed the bug
Is there a way we can get his name back on his comments? Most of his comments were excellent and helpful, and it seems a shame to not have his name on those (not to mention making this whole scenario look a lot worse for Lobsters).
The code is:
So it just runs an
update comments set user_id=-1 where user_id=42query. Unless you’re going to restore from a backup, I don’t think this can just be corrected.Unfortunate indeed :-(
Well, if there are database back-ups, it would certainly be possible to make a query against a back-up to find all the comments made by burntsushi, then run the queries to change those comments’ owner to burntsushi’s deleted account against the live database.
If this is to be done, it should probably be done for not just burntsushi, but also other users since december 2018 (which is when the bug was introduced). And if that commit contains other bugs as well, not just disowning comments, should it be done for it as well?
If the generated SQL statement is in the logs, it might not be too difficult to reverse.
You would need the data. The statement itself won’t tell you what specific rows were affected by it at the time. If it somehow logged the ID of the stories and comments it was updating it would be trivial but that would be somewhat unusual to log.
I guess I’m wondering why this is logging people’s access tokens and passwords?
I think that’s a literal
[FILTERED]in the logs, see here.I love ST
Dev-vertising? What’s the name for this sort of thing? :-\
haha yeah. I guess it’s kind of advertising sinklog.com, which is just a free utility that uses the methods in the blog post. I mostly just thought that it was somewhat novel and folks might be interested in how it’s built.
The catch-all is “content marketing”, offer your potential audience something useful or interesting so they’ll see your name, think well of you, and maybe eventually buy whatever you’re offering. The term covers both businesses posting to attract users/employees and employees writing personal blogs to build their own reputations.
I’m sure it was meant innocently but this really doesn’t read well. Never attack your readers.
Indeed, and I think learning how to break this would be a very rewarding and educational experience. That - to me - is a life.
yep that definitely came across too harsh.