in reply to MonkReaper?!

Or perhaps you think there is a good reason for leaving usernames even if they haven't logged in for 5 years....?

Just being idle is not a good reason for removal.

For instance, take the example of paco, who has already been mentioned in this thread -- who would be associate paco's old post with? It wasn't written by the new paco, but the message needs to have an owner.

The only way that you can remove a record is if it's not referenced anywhere else. (a concept known as 'Referential Integrity' in database terms). Some things can be checked for -- if they have posted any messages, if they've earned any xp, etc.

But what about those who have been linked to? You'd have to go through every last message that has been posted, and look to see if someone has been linked to (eg, 'Barry' was mentioned in one of the other threads that has already been linked to -- so if you remove him, you have a dead link, which isn't so bad ... but if you allow someone else to use that name, then you have posts talking about a user that they didn't intend to talk about.

But ... what about those folks who were mentioned, but not actually linked to? Removing dead users is a much more difficult thing than just removing those items with no activity -- and if the benefits don't outweigh the cost of doing it (with the benefits being measured by the person who's doing the work, in volunteer situations), it's not going to be done.

Replies are listed 'Best First'.
Re^2: MonkReaper?!
by bofh_of_oz (Hermit) on Jun 29, 2005 at 17:36 UTC
    One way to deal with orphaned nodes left after a user has "left the monastery" is to anonymize them, i.e. assign Anonymous Monk as owner...

    As for other things, something surely can be done... And I will agree with you on one point: it could happen that the cost of deleting a user outweighs the benefits. Perhaps, the feasibility of this idea should be considered and approved/implemented or discarded according to the outcome...

    --------------------------------
    An idea is not responsible for the people who believe in it...

      I like the idea, however I would actually prefer that they be assigned to "Departed Monk" (or "Apostate Monk" or even "Deceased Monk").

      I'm not sure why I prefer Departed Monk over Anonymous Monk, repressed memories of bastardized databases floating to the surface perhaps. I think of Anonymous Monk has a specific definition of sorts and previously known monks don't really fit that definition, do they? So moving these monks to Anonymous Monk would tend to muddy the definition of Anonymous Monk. Having had to deal with some databases where folks muddied column definitions by reusing existing columns for new kinds of data rather than adding new columns as they should have (try extracting meaningful information from that kind of mess), I'm skiddish of changes that would tend to muddy the existing definition of something.

        I think i prefer "Apostate Monk" it has a nice ring to it and it implies they're off doing VB or something.

        Also don't confuse the username with the node id when it comes to referential integrity issues. The username attached to a node could be changed (I assume) without breaking any of the threads created by that node (user).

        Some 13 years down the line now this has indeed come to pass; the grand irony being that it is the OP of this very thread who has become the first to be removed at their own request. Unfortunately the name given to our departed brother is nothing so poetic, being merely (anonymized user).

Monastery graveyard (was: MonkReaper?!)
by Roy Johnson (Monsignor) on Jan 13, 2006 at 18:44 UTC
    My thought is that monks with no writeups and no logins in X years would be deleted. Those with writeups (but no logins in X years) would be renamed by a process that runs once a year. If "IdleMonk" met the criteria, he would be renamed (something like) "RIP2006:IdleMonk" at the end of this year.

    It seems fitting that a monastery have a graveyard for those monks who have passed on.


    Caution: Contents may have been coded under pressure.