Mailman 3 May 2004 - SpamBayes

this might seem fickle...
by Nathan Sudds Sept. 25, 2004

Sept. 25, 2004

But I can't seem to get the icons back for my spambayes toolbar... I installed a toolbar for putting paypal payment buttons into my emails... it's another outlook addon.. ever since then the icons for Spambayes show up as the paypal icons... it's a minor thing but it's pretty annoying for a visual person like myself... any advice? I'm using the Windows version of the Outlook Addon of spambayes and I am running windows XP Pro with Outlook 2000. SpamBayes version .81 to start with but now I'm running the new .9 version.. and I'm glad it's much faster on my machine than .81 was... but I wouldn't mind having those icons back... Thanks, Nathan Sudds | CreativeDirector nsudds(a)arkitekture.com ::: ARKITEKTURE:::::::::::::: focusing our sites on building yours -------------------------------------------------------------------- [launch] // www.arkitekture.com

1 1

[slightly OT] spammer 'random' words that are addressee-related?
by Denis Haskin Aug. 18, 2004

Aug. 18, 2004

(This isn't *quite* a spambayes question but I think there are knowledgable people here <grin> who can point me in the right direction to get more info. Feel free to declare this too off-topic and send me someplace else.) Like many of you (I'm sure), I'm sysadmin/network ops/hardware tech/customer support/etc for my family's home network. My wife showed me some spam she got this morning that included the usual 'random' words in the subject, but she was wondering if the words maybe weren't so random. The words seemed related to 2 email lists she's on. Now maybe this is just complete coincidence, but is it possible in fact that her email address was harvested somewhere and words were picked up from the harvesting context to be used when constructing spam to her? This seems like a lot of effort for not a lot of return... Also, we're *pretty* sure that both of these email lists are pretty well protected (e.g. they're not archived in any public place as far was we know). Also, they're obscure and small enough that I find it hard to believe it would be worth spammers' while to 'infiltrate' these lists (as they do with yahoo groups). It also occurred to me that perhaps someone else on these lists has a virus which is harvesting this from a local PC and sending it off to a spammer? Anyone have any thoughts? Is there any web site that catalogs known spamming tactics? Thanks, dwh

3 2

RE: [Spambayes] Spambayes Configuration with Norton Antivirus HelpPlease
by Tony Meyer June 14, 2004

June 14, 2004

> Problem is I am now without email virus > scanning protection. > I saw one user was instructed to put "pop3.norton.antivirus" > or something like that into to the spambayes server config > to set up a Server->Norton->SpamBayes->email client chaining... > I know above is not exactly what I tried, but I do know > that I tried the exact text I saw in the discussion list > posting but this didnt work for me. It's not working for him, either, yet :) (I don't have Norton, and don't use OE, so am answering blind, to a certain extent. I have read that others have successfully done this, though, so it can be done. (If only they were speaking up...) Keep watching the "RE: [Spambayes] Spambayes on Outlook Express" thread - the instructions should apply equally to you. If I understand things rightly now, Norton requires that it be connected to by SpamBayes, which means that you need to chain like: OE <-> Norton <-> SpamBayes <-> Mail Server So set SpamBayes to connect to the mail server, but proxy on 8110. Leave OE connecting to Norton (which I gather it does as soon as you enable Norton). Then tell Norton that your mail server is localhost 8110. =Tony Meyer --- Please always include the list (spambayes(a)python.org) in your replies (reply-all), and please don't send me personal mail about SpamBayes. This way, you get everyone's help, and avoid a lack of replies when I'm busy.

2 1

Shelve database corruption?
by Lars Marius Garshol June 3, 2004

June 3, 2004

I recently got the problem that on some emails SpamBayes throws an exception, and the email does not get classified at all. The result is that SpamBayes catches some of my spam, but when this error happens the spam gets through, and it seems to happen for about half of the messages. I got this problem while running SpamBayes 1.0a7, and tried to solve it by upgrading to SpamBayes 1.0rc1, but that made no difference. The traceback I get is: Traceback (most recent call last): File "scripts/sb_server.py", line 472, in onRetr msg.setId(state.getNewMessageName()) File "/usr/lib/python2.2/site-packages/spambayes/message.py", line 258, in setId msginfoDB._getState(self) File "/usr/lib/python2.2/site-packages/spambayes/message.py", line 118, in _getState attributes = self.db[msg.getId()] File "/usr/lib/python2.2/shelve.py", line 70, in __getitem__ f = StringIO(self.dict[key]) error: (-30981, 'Unknown error 4294936315') What I'm looking for is some idea of what's wrong with the shelve database so that I can fix the corruption. I've trained SpamBayes on 37,000 emails, so the idea of starting again from scratch is not appealing... Help would be *much* appreciated, since I'm now back to drowning in spam again. :-( -- Lars Marius Garshol, Ontopian <URL: http://www.ontopia.net > GSM: +47 98 21 55 50 <URL: http://www.garshol.priv.no >

3 9

RE: [Spambayes] What performance is good?
by Tony Meyer June 2, 2004

June 2, 2004

> How much do you expect to see in Junk Suspects? Good results would be 2-5% of total incoming mail, IMO. > I get about 15% ham. All the ham goes in Inbox (I had one > ham in Junk Suspects last week), along with 1 or 2 spam, > which is beautiful. 25% to 30% of incoming mail goes in Junk > Suspects. Is that reasonable? Are you displaying the scores for these messages? If so, do they all tend to score over a certain value? You might find, for example, that you can simply reduce the spam threshold (say to 80%) and the problem goes away. What do you have the thresholds set for now? (SpamBayes->SpamBayes Manager->iltering) > My database is unbalanced (36 ham/84 spam), so I'm only > training on ham in Junk Suspects (i.e., hardly ever). Should > I be seeking ham to train on? IMO, that's not imbalanced enough to worry about. =Tony Meyer --- Please always include the list (spambayes(a)python.org) in your replies (reply-all), and please don't send me personal mail about SpamBayes. This way, you get everyone's help, and avoid a lack of replies when I'm busy.

2 1

500 server error
by Ben Foster May 31, 2004

May 31, 2004

Win XP home SB ver. 0.6 for Outlook Express 6 When I click on configure I get error message: 500 Server error Traceback (most recent call last): File "spambayes\Dibbler.pyc", line 461, in found_terminator File "spambayes\UserInterface.pyc", line 627, in onConfig File "spambayes\UserInterface.pyc", line 641, in _buildConfigPage File "spambayes\PyMeldLite.pyc", line 738, in __setattr__ File "spambayes\PyMeldLite.pyc", line 654, in _replaceNodeContent File "spambayes\PyMeldLite.pyc", line 640, in _nodeListFromSource File "spambayes\PyMeldLite.pyc", line 574, in _generateTree File "spambayes\PyMeldLite.pyc", line 499, in feed ExpatError: not well-formed (invalid token): line 1, column 34 Can you help? Thanks Ben Foster --------------------------------- Do you Yahoo!? Friends. Fun. Try the all-new Yahoo! Messenger

1 0

RE: [Spambayes] More Error Messages in log file?
by Tony Meyer May 31, 2004

May 31, 2004

> I have 10 Pst files with hundreds of folders in each of them. > You are suggesting that I set SpamBayes up to watch over > 1000 folders. It really doesn't sound like SpamBayes is the filter for you, so by all means google about for others. However, if you do want to try to use it, here are two suggestions: * You can use the "include subfolders" option when selecting folders. I presume that the hundreds of folders are in some sort of hierarchy, so this would mean only selecting a few. Still some work to setup, but that only gets done once. Of course, if there are subfolders that should be filtered, and subfolders that shouldn't, then this won't work. * If you're using POP3, then you could use the proxy rather than the plug-in. Not as good a UI, but not bad, either. This would guarantee that all messages were classified before Outlook got to them. =Tony Meyer --- Please always include the list (spambayes(a)python.org) in your replies (reply-all), and please don't send me personal mail about SpamBayes. This way, you get everyone's help, and avoid a lack of replies when I'm busy.

1 0

RE: [Spambayes] More Error Messages in log file?
by VBCoder May 31, 2004

May 31, 2004

Tim, I have 10 Pst files with hundreds of folders in each of them. You are suggesting that I set SpamBayes up to watch over 1000 folders, surly you jest. You should have left it alone when you where ahead. I have no use for a tool that requires I chase after all of the destination folders. I don't know Python, nor do I have any desire to learn it. I have absolutely no intentions of downloading the source code to get descent documentation. This implies that only Python programmers are your intended audience for this application. I hope not. Paul -----Original Message----- From: Tim Peters [mailto:tim.one@comcast.net] Sent: Monday, May 31, 2004 4:52 PM To: 'VBCoder'; Spambayes(a)python.org Subject: RE: [Spambayes] More Error Messages in log file? [VBCoder] > It appears that all of the issues I am seeing are related to Spambayes > not seeing a message before my Outlook rules. This brings me to the sad > realization that SpamBayes is unusable in my environment. I'm not sure why. You can tell SpamBayes to watch any number of folders, in any number of .pst files. It's not limited to scanning messages *only* in the Inbox. I have Outlook rules too (although not many), which move messages from the Inbox to other folders, some in other .pst files. I configure SpamBayes to watch those destination folders too (in addition to the Inbox). The Outlook rules run first (with background filtering enabled, if necessary), and SpamBayes rips the spam out of all the folders. > I had asked a question about the purpose of the Filter timer and you > responded with "read the code". At that point, I had already responded in great detail about the *cause* of missing scores, in terms of the Outlook object model. If you want to know even more than that, then yes, read the code. You say you're a VB programmer, so the ItemAdd event shouldn't be a mystery to you. > I don't have the code The source code can be downloaded by anyone. > and I don't have the time or inclination t read the code. That's fine, and I gave you as much answer as I had time for. > If I did, I would write the code instead. The timer code is a very small part of the system. It would take anyone months to write all of this code. > I don't know what language it is written in, I am guessing > it is not one that I am conversant in like assembly or VB. It's written in Python, but uses MS's MAPI and Outlook interfaces. Those look pretty much the same no matter which language you call them from, although VB and Python are much easier to use for this than C++. > A simple answer would have sufficed. Indeed, no good deed goes unpunished <0.6 wink>. > Don't bother responding as it would make little difference as I am > already looking for a product that will work in my environment. As at the start, it's unclear why SpamBayes would not work in your environment. > Hint: If you wrote the Outlook piece as a custom rule, I would have been > able to put it first to insure it would always fire before my rules moved > it. But I am sure you already thought of that. Yes, but nobody has dreamt up a practical way of making that actually work. > You last comment about turning on background filtering is not the answer. > When turned on, almost all messages are missed (blank percent). This > setting seems to insure that it is always fired after the Outlook rules, > at least in OL2K. Yes, that's the intent. AFAICT, the only piece of this puzzle you're missing is configuring SpamBayes to watch your rules' destination folders too. --- Incoming mail is certified Virus Free(1). Checked by AVG anti-virus system (http://www.grisoft.com). Version: 6.0.692 / Virus Database: 453 - Release Date: 5/28/2004

2 1

RE: migrating SpamBayes from Windows to Linux
by Tony Meyer May 31, 2004

May 31, 2004

[Note that you're more likely to have success asking questions via the spambayes@python mailing list than you are with c.l.p/python-list(a)python.org. You're also more likely to receive a response if your reply address works.] > I have been trying Linux on-and-off for a couple of years but > keep going back to Windows out of frustration. I have been > using the Outlook plug-in for over a year and now that I am > playing around with Linux > again the first thing I did was install SpamBayes. I am > looking for some easy way to migrate the existing database > from Windows to Linux so I don't have to retrain from scratch. You can possibly just copy the database file over (default_bayes_database.db, which you can rename to hammie.db if you like). If that doesn't work (depends on the dbm versions installed), then you can use the sb_dbexpimp.py script. You'll probably have to run it on the Windows machine (which means installing Python if it's not already there, and getting the source archive) to get a successful conversion to either csv or a pickle. Then copy that across and convert back. =Tony Meyer --- Please always include the list (spambayes(a)python.org) in your replies (reply-all), and please don't send me personal mail about SpamBayes. This way, you get everyone's help, and avoid a lack of replies when I'm busy.

2 1

RE: [Spambayes] More Error Messages in log file?
by VBCoder May 31, 2004

May 31, 2004

Hi Tim, It appears that all of the issues I am seeing are related to Spambayes not seeing a message before my Outlook rules. This brings me to the sad realization that SpamBayes is unusable in my environment. I had asked a question about the purpose of the Filter timer and you responded with "read the code". I don't have the code and I don't have the time or inclination to read the code. If I did, I would write the code instead. I don't know what language it is written in, I am guessing it is not one that I am conversant in like assembly or VB. A simple answer would have sufficed. Don't bother responding as it would make little difference as I am already looking for a product that will work in my environment. Hint: If you wrote the Outlook piece as a custom rule, I would have been able to put it first to insure it would always fire before my rules moved it. But I am sure you already thought of that. You last comment about turning on background filtering is not the answer. When turned on, almost all messages are missed (blank percent). This setting seems to insure that it is always fired after the Outlook rules, at least in OL2K. It is too bad that Spambayes will not work for me in my environment as it really looked promising. I did not want to use a proxy for one and I am not willing to change the other software I am currently using. Thanks for your time and input, Paul -----Original Message----- From: Tim Peters [mailto:tim.one@comcast.net] Sent: Monday, May 31, 2004 2:42 PM To: 'VBCoder'; Spambayes(a)python.org Subject: RE: [Spambayes] More Error Messages in log file? [VBCoder, using Outlook 2K] ... > 1). Does SpamBayes get control of the messages before any of my Rules? This appears impossible to predict, varying by such things as system load, processor speed, how fast email is coming in, Outlook version, and phase of the moon. It's possible that Outlook will tell SpamBayes about a new message before your rules or after your rules, it's possible that this will change from message to message, and, if you have a rule that moves a msg from one watched (by SpamBayes) folder to another watched folder, it's possible that Outlook will tell SpamBayes about the message twice -- or not at all! The good news is that enabling "background filtering" (on the SpamBayes Manager's Advanced tab) pretty reliably results in Outlook rules running first. There's no known way to reliably get SpamBayes to run first. If you are a VB coder, SpamBayes learns about new messages via ItemAdd events on the watched folders. So the answer to all questions is "easy" at that level: SpamBayes sees a new message when, and only when, Outlook fires ItemAdd. Unfortunately, ItemAdd isn't reliable: http://support.microsoft.com/?kbid=249156 Outlook 2003 introduced a NewMailEx event which *may* be a reliable way (and a more efficient way) to learn about new messages, but it looks like that fires before rules (and, of course, isn't available before OL2003). > 2). Is it normal for a message to be processed by my Outlook rules and > arrive in it's final destination (supposedly after being processed by > SpamBayes and my Outlook rules) with the Spam field blank (no > percentage)? If so, under what conditions could this happen? You really have to ask Microsoft, and they won't answer you <wink>: the spam field is left blank if ItemAdd never told us about the new message, and only MS can say exactly when Outlook fails to fire ItemAdd. > 3). What is the purpose of the Filter Timer? How does it work? Under > what conditions should I use it? Read the code. You should use it if anything flaky is happening (like messages not get scored). > I am finding many messages that have no percentage in the Spam field yet > a clues display will show a percent that is anywhere in the spectrum. A clues display scores the message from scratch. It doesn't matter to Show Clues whether the message has been scored before, or, if the message was scored before, what soore it got at that time. Show Clues always recomputes everything from the current state of training. > Many are over the %90 Spam limit and should have been sent to the Spam > folder. "no percentage" means ItemAdd never told SpamBayes about the message. > The messages that do arrive in the Spam and Unsure folders have a > percent value and appear to be processed correctly. I am trying to find > out if this is normal, if there is a setting that will change this > behavior if not, and if it is a bug that needs to be reported. Turn on background filtering. ItemAdd may still miss some messages, but far fewer. --- Incoming mail is certified Virus Free(1). Checked by AVG anti-virus system (http://www.grisoft.com). Version: 6.0.692 / Virus Database: 453 - Release Date: 5/28/2004

2 1