C-Command Software Forum

Apple Mail 'Mark as Good' not working

Running Apple Mail 6.0 with 10.8 and SpamSieve 2.9.3

I had a message in today which SpamSieve sent incorrectly* to my Spam folder and so, with relevant message highlighted, I went to the ‘Message’ menu in Mail, and then clicked on ‘SpamSieve - Train as Good’. This action hung Spamsieve - I saw in marked as ‘Application not responding’ having pressed the ⌘+⌥+⎋ keys. Although Mail.app itself was NOT marked as ‘not responding’, it had also hung and I had to force quit that too as well as SpamSieve.

I tried installing the Apple Mail Plug-In from the SpamSieve menu again - no difference.

  • Interestingly, I had had two previous messages from this email address today, which went correctly to my Inbox, and had replied to one of them which in turn had been replied to and it was this latter message which was incorrectly delivered to my Spam folder. Weird.

Any ideas, Michael?

Best,
Tim.

First I would suggest that you check the log to see whether SpamSieve actually predicted this message to be spam. Normally, SpamSieve will whitelist an address that you’ve received a good message from (or two, in your case).

Please contact me via e-mail so that I can send you a pre-release version of SpamSieve that may fix this.

I’ve checked the log - the third incoming message was indeed “predicted” as spam.

I’ll send a copy of the log to you by email, Michael. Thanks for your help with this.

Thanks for sending the log. It looks like you had unchecked the Auto-train with incoming mail option. This is why SpamSieve didn’t auto-whitelist based on the good messages you had previously received.

Also, you had set the Spam-catching Strategy to make SpamSieve extra aggressive about thinking messages are spam.

Strange - please see first screen shot below showing that in fact ‘auto-training’ was switched on and has been since I’ve been using SpamSieve. If for some reason the setting has not actually been made, is there some amendment I could make to a .plist file or whatever?

My strategy slider is set to one above centre (and has been for ages as the auto-training) - would you recommend it being set at the centre position? - see second screen shot below.

Preferences-2.jpg Preferences.jpg

Same problem
I am having the same problem. SpamSieve is not catching messages it used to catch and when I try to train it again it freezes up mail, forcing me to force quit.

As to the above solutions, they only explain the settings problem, but why the freezing?

It’s possible that the setting is correct but that auto-training didn’t take place because a message from that sender had previously been trained as spam. Please send me your whole log file (from after you manually trained the messages) so that I can take a closer look.

That’s what I recommend in general, but I don’t think you necessarily need to make any changes while we’re still figuring out what happened.

The freezing problem is likely caused by some changes to SpamSieve’s architecture that were necessary for compatibility with Mountain Lion. Please contact me via e-mail so that I can send you a pre-release version of SpamSieve that fixes this.

Thanks Michael - log file sent by email.

It seems like you must have been receiving messages from this sender for a while, so some of the records are in older log files. Auto-training is working for other messages, though. So my guess is that, some time ago, you trained messages from Keith has both spam and good, and that’s why the address is neither blocklisted nor whitelisted.

The log shows that SpamSieve thought the message from August 6 was spam. You say it was good, but there’s no evidence that you told SpamSieve that. It is essential that you correct all the mistakes or SpamSieve will be working from false information.

There was also another message (from a different sender) that was incorrectly classified as spam due to the Encoded HTML mail is spam preference.

OK - thanks Michael.

Can you please identify the message to which you are referring to me by email please, Michael and then I can see what the log file entry looks like for that incorrect classification and so correct this sort of mistake for myself in the future. I’d appreciate it as I do for all your help - thank you.

This is now fixed in SpamSieve 2.9.4.