http://bugzilla.spamassassin.org/show_bug.cgi?id=3331
------- Additional Comments From [EMAIL PROTECTED] 2004-04-29 10:09 ------- Subject: Re: Bayes option to keep original token as db data (not key). On Thu, Apr 29, 2004 at 09:58:19AM -0700, [EMAIL PROTECTED] wrote: > > > 1) How do you handle backup/restore? > > I don't see the issue, other than having to add code to write or read > the original token. > So when you run backup do you get the hashed token or the original token? What if not all original tokens are available? > > > > 2) Where are you going to put this data? In a seperate db file/table > > or alongside the current data? Be careful, your answer may require > > a ton more work. > > I'd put the original token in the same record used to hold the other > information (num ham, num spam, atime). I'm assuming db_file and SQL > can handle a variable-length string field and that their > implementations do so in a reasonable manner. (That is, there would be > no big performance difference between a table without variable-length > string fields and one with a string field that only stores nulls.) You just changed the database format. It's easy enough to just add onto a value in DBDm but since we pack several items into a single value it does require a database format change. SQL does not use variable length columns for data storage anymore. So you would have to add a column to the bayes_token table to place this value in, so it would change the database format. > > > > Patched welcome of course. > > I don't have SQL installed so there is no way I could test any changes > for that back end. If no one else starts working on the changes in a > few days I'll make them for dbfile (DBM). > Like I said, I welcome any patches, I think you'll find the storage code slightly more complicated once you start digging into it. After thinking on this a little I have a possible proposal that involves a seperate db_file/table but I haven't worked everything out yet. I'll think on it some more and reply back to see if it interests folks. ------- You are receiving this mail because: ------- You are the assignee for the bug, or are watching the assignee.
