"Junk Mail" problem. help!
in [Eudora]
|
Prev: Data folder
Next: convert selected to uppercase
From: John H Meyers on 12 Nov 2009 15:46 On Thu, 12 Nov 2009 10:08:28 -0600, Jim Higgins interestingly wrote: > Eudora 7.1.0.9 > > This morning I un-junked two messages with no change seen to the > UserJunkDB.txt file !MessageCount values UNTIL EUDORA WAS SHUT DOWN. > After shutdown it could be seen that the first count had been > incremented by 2, no change to the second count. > > Now these same two messages were junked with no change seen to the > file until Eudora was again closed. After shutdown it could be seen > that the first count was DECREMENTED by two and the second count was > incremented by two. > > It appears that Eudora only updates the UserJunkDB.txt file at > shutdown, or at least at longer intervals than my testing lasted. Any > testing that doesn't include shutting down Eudora before checking for > changes to the UserJunkDB.txt file is potentially flawed. In the cases I reported earlier, I always first closed Eudora, before opening and inspecting the file content which I then posted, so I got "different mileage" on this, which apparently indicates that I didn't hit the same conditions -- perhaps the order of operations is significant, or doing several consecutively (I only did one operation at a time, with shutdowns in between), or who knows? What do you feel is responsible for your own counter (also in 7.1.0.9) not having "gone negative" after long use? Well, thanks for illustrating that something is up, whatever it may be. If this whets anyone's further curiosity, join the "group testing team," and add anything which seems to clarify (or muddle) the understanding of "when it does, vs. when it doesn't" Perhaps the junk scorer would have been even more accurate, had it acted consistently? (people have reported getting better results with "mailwasher," for example, which is another "probabilistic" system, whose internal details I do not know). --
From: Jim Higgins on 13 Nov 2009 12:35
On Thu, 12 Nov 2009 14:46:02 -0600, "John H Meyers" <jhmeyers(a)nomail.invalid> wrote: >On Thu, 12 Nov 2009 10:08:28 -0600, Jim Higgins interestingly wrote: > >> Eudora 7.1.0.9 >> >> This morning I un-junked two messages with no change seen to the >> UserJunkDB.txt file !MessageCount values UNTIL EUDORA WAS SHUT DOWN. >> After shutdown it could be seen that the first count had been >> incremented by 2, no change to the second count. >> >> Now these same two messages were junked with no change seen to the >> file until Eudora was again closed. After shutdown it could be seen >> that the first count was DECREMENTED by two and the second count was >> incremented by two. >> >> It appears that Eudora only updates the UserJunkDB.txt file at >> shutdown, or at least at longer intervals than my testing lasted. Any >> testing that doesn't include shutting down Eudora before checking for >> changes to the UserJunkDB.txt file is potentially flawed. > >In the cases I reported earlier, I always first closed Eudora, >before opening and inspecting the file content which I then posted, >so I got "different mileage" on this, which apparently indicates >that I didn't hit the same conditions -- perhaps the order of >operations is significant, or doing several consecutively >(I only did one operation at a time, with shutdowns in between), >or who knows? I manually un-junked two messages at the same time, observed no change to the file date in Explorer, shut down, observed a change to the file date, opened file and observed change to counters. Restarted Eudora, manually junked those same two messages at the same time, observed no change to the file date... etc same as above. >What do you feel is responsible for your own counter >(also in 7.1.0.9) not having "gone negative" after long use? I can't explain, but I'll ramble on a bit about my situation here. Dell Optiplex 330 new 11/27/2007, fresh install of Eudora 7.1.0.9 (registered version with X1 message search engine). Eudora has done an outstanding job of catching spam using only the default criteria that come with a fresh installation of Ver 7.1.0.9 so I have manually junked and manually unjunked relatively few messages over the last two years. Manually unjunked is small in comparison to manually junked and manually junked is very small in comparison to automatically junked. Perhaps 5 - 7 spam a week on average over those two years with Eudora handling the vast majority of it automatically. >Well, thanks for illustrating that something is up, >whatever it may be. Glad to help. I usually don't dig deeper into these things because explaining something that can't/won't be fixed/updated is low among my priorities. >If this whets anyone's further curiosity, join the >"group testing team," and add anything which seems to clarify >(or muddle) the understanding of "when it does, vs. when it doesn't" > >Perhaps the junk scorer would have been even more accurate, >had it acted consistently? I don't know that the spam scorer hasn't acted consistently since we clearly don't know all the factors that control it... IMHO. The difference in the results of our own recent testing illustrates this. > (people have reported getting better results with > "mailwasher," for example, which is another "probabilistic" > system, whose internal details I do not know). I'm inclined to ignore (or at least take with a grain of salt) self-selected testimonials about anti-spam software (or pretty much anything else). They just aren't statistically meaningful. Everyone's own crying baby sounds fine while the one next door is a pain in the butt. To at least put numbers on a personal testimonial I was going to quote spam counts, percentages, false positive and false negative results, etc., from my Eudora statistics file, but see that if anything is flawed in Eudora, statistics is it. Yearly stats say I get 7% spam. That would be 15 a day and I don't get that in a week. It also says 98% effective, meaning 2% inaccurate. That would be 4 messages a day improperly handled by the spam filter. Utterly ridiculous. I can't begin to give a true count for the year, but I recall one in the past couple of months, not 4 per day. I only mention this before folks split this thread and start ranking Eudora performance vs Mailwasher (or whatever) based on flawed Eudora statistics. -- Please don't be a "Help Vampire" http://slash7.com/pages/vampires |