From skip at pobox.com  Fri Jul  6 13:22:57 2007
From: skip at pobox.com (skip at pobox.com)
Date: Fri, 6 Jul 2007 06:22:57 -0500
Subject: [spambayes-dev] 1.1b1?
Message-ID: <18062.9745.201378.379989@montanaro.dyndns.org>

Anyone think it's about time for a beta release?  I'll be out of town this
weekend, but will try to put one together when I return.  Maybe a release
schedule like this:

    1.1b1     July 11
    1.1b2     July 25 (if necessary)
    1.1rc1    August 1
    1.1final  August 8

?  Given that the bulk of the users are Outlook users I think we should try
and recruit a few extra people to put it through its paces.  In particular,
getting some Outlook 2007 and Vista users to use it would be good I think.
That means Mark's or Tony's availability to create Outlook installers is key
to all these dates.

Ideally, I'd like to get the new core_server.py and sb_server.py merged back
together, but that will require extracting the POP3 bits from sb_server.py
into a plugin for core_server.py then renaming core_server.py to
sb_server.py.  I'm not sure I'll have the time to do the POP3 plugin, so
that will probably have to wait until 1.2 if we are to ever get a 1.1
release out.

Skip

From dave at boost-consulting.com  Fri Jul  6 17:03:59 2007
From: dave at boost-consulting.com (David Abrahams)
Date: Fri, 06 Jul 2007 11:03:59 -0400
Subject: [spambayes-dev] spoof detector
Message-ID: <87fy411sm8.fsf@grogan.peloton>


Something that comes up over and over in spam is a link of the form:

    <a href="http://url/of/spammers/site">
       http://url/of/some/legit/site
    </a>

Does SpamBayes have a token that represents that information and an
option I can set that will use it?

-- 
Dave Abrahams
Boost Consulting
http://www.boost-consulting.com

The Astoria Seminar ==> http://www.astoriaseminar.com


From skip at pobox.com  Fri Jul  6 17:46:50 2007
From: skip at pobox.com (skip at pobox.com)
Date: Fri, 6 Jul 2007 10:46:50 -0500
Subject: [spambayes-dev] spoof detector
In-Reply-To: <87fy411sm8.fsf@grogan.peloton>
References: <87fy411sm8.fsf@grogan.peloton>
Message-ID: <18062.25578.308007.229411@montanaro.dyndns.org>


    David> Something that comes up over and over in spam is a link of the
    David> form:

    David>     <a href="http://url/of/spammers/site">
    David>        http://url/of/some/legit/site
    David>     </a>

    David> Does SpamBayes have a token that represents that information and
    David> an option I can set that will use it?

The SpamBayes tokenizer essentially splits the message at word boundaries,
so the two urls are considered separately.  Their physical and structural
proximity is not noted.  Synthetic tokens based on hostname or IP address in
the urls will be generated if you add x-pick_apart_urls:True to the
Tokenizer section of your config file.  For completeness here is my current
set of tokenizer settings (haven't changed them in a long while):

    [Tokenizer]
    record_header_absence:True
    summarize_email_prefixes:True
    summarize_email_suffixes:True
    mine_received_headers:True
    x-pick_apart_urls:True
    x-fancy_url_recognition:False
    x-lookup_ip:True
    lookup_ip_cache:~/tmp/dnscache.pck
    x-image_size:True
    x-crack_images:True
    x-ocr_engine:gocr
    max_image_size:100000
    crack_image_cache:~/tmp/imagecache.pck

Skip


From dave at boost-consulting.com  Fri Jul  6 21:20:27 2007
From: dave at boost-consulting.com (David Abrahams)
Date: Fri, 06 Jul 2007 15:20:27 -0400
Subject: [spambayes-dev] spoof detector
In-Reply-To: <18062.25578.308007.229411@montanaro.dyndns.org>
	(skip@pobox.com's message of "Fri\, 6 Jul 2007 10\:46\:50 -0500")
References: <87fy411sm8.fsf@grogan.peloton>
	<18062.25578.308007.229411@montanaro.dyndns.org>
Message-ID: <87644xz6dg.fsf@grogan.peloton>


on Fri Jul 06 2007, skip-AT-pobox.com wrote:

>     David> Something that comes up over and over in spam is a link of the
>     David> form:
>
>     David>     <a href="http://url/of/spammers/site">
>     David>        http://url/of/some/legit/site
>     David>     </a>
>
>     David> Does SpamBayes have a token that represents that information and
>     David> an option I can set that will use it?
>
> The SpamBayes tokenizer essentially splits the message at word boundaries,
> so the two urls are considered separately.  

Yeah, I know that's the default behavior.

> Their physical and structural proximity is not noted.  Synthetic
> tokens based on hostname or IP address in the urls will be generated
> if you add x-pick_apart_urls:True to the Tokenizer section of your
> config file.  For completeness here is my current set of tokenizer
> settings (haven't changed them in a long while):
>
>     [Tokenizer]
>     record_header_absence:True
>     summarize_email_prefixes:True
>     summarize_email_suffixes:True
>     mine_received_headers:True
>     x-pick_apart_urls:True
>     x-fancy_url_recognition:False
>     x-lookup_ip:True
>     lookup_ip_cache:~/tmp/dnscache.pck
>     x-image_size:True
>     x-crack_images:True
>     x-ocr_engine:gocr
>     max_image_size:100000
>     crack_image_cache:~/tmp/imagecache.pck

That doesn't sound like it's doing what I'm asking about.  I want a
special token that is generated each time a link's text is just a URL
and the link and the URL text don't point to the same place.  Messages
with this property are always spam and account for a large percentage
of my unsures.  No matter how much I train on them, they keep falling
into unsure, so I thought if Spambayes could actually recognize their
distinguishing feature I could easily train it to consider them spam.

>From what you say above it looks like pick_apart_urls will generate
tokens describing different parts of a given URL, but will do nothing
to help capture this particular spammy relationship between enclosed
text and actual link.

Or did I misunderstand you?

-- 
Dave Abrahams
Boost Consulting
http://www.boost-consulting.com

The Astoria Seminar ==> http://www.astoriaseminar.com

From skip at pobox.com  Fri Jul  6 22:55:41 2007
From: skip at pobox.com (skip at pobox.com)
Date: Fri, 6 Jul 2007 15:55:41 -0500
Subject: [spambayes-dev] spoof detector
In-Reply-To: <87644xz6dg.fsf@grogan.peloton>
References: <87fy411sm8.fsf@grogan.peloton>
	<18062.25578.308007.229411@montanaro.dyndns.org>
	<87644xz6dg.fsf@grogan.peloton>
Message-ID: <18062.44109.162303.926899@montanaro.dyndns.org>


    >> Their physical and structural proximity is not noted.  Synthetic
    >> tokens based on hostname or IP address in the urls will be generated
    >> if you add x-pick_apart_urls:True to the Tokenizer section of your
    >> config file.

    Dave> That doesn't sound like it's doing what I'm asking about.  

No, it's not, however, you might be surprised how helpful it is to generate
tokens for the /8, /16, /24 and /32 address blocks can be.  I what I was
implying is that maybe you don't need the spoof detection you were asking
for if the address tokens generated from the spammer's IP address are
spammy.

    Dave> I want a special token that is generated each time a link's text
    Dave> is just a URL and the link and the URL text don't point to the
    Dave> same place.

That will require actually parsing the HTML at some level.  SpamBayes just
sees a stream of tokens.  It doesn't really know much (if anything) about
compound structure.

    Dave> Messages with this property are always spam and account for a
    Dave> large percentage of my unsures.  

Try these two settings

    x-pick_apart_urls:True
    x-lookup_ip:True

and see if they help.

    Dave> From what you say above it looks like pick_apart_urls will
    Dave> generate tokens describing different parts of a given URL, but
    Dave> will do nothing to help capture this particular spammy
    Dave> relationship between enclosed text and actual link.

    Dave> Or did I misunderstand you?

No, I probably misunderstood myself.  The IP address hacker is the
x-lookup_ip option I believe.  They are both helpful though.

Skip

From dave at boost-consulting.com  Sat Jul  7 02:56:03 2007
From: dave at boost-consulting.com (David Abrahams)
Date: Fri, 06 Jul 2007 20:56:03 -0400
Subject: [spambayes-dev] spoof detector
In-Reply-To: <18062.44109.162303.926899@montanaro.dyndns.org>
	(skip@pobox.com's message of "Fri\, 6 Jul 2007 15\:55\:41 -0500")
References: <87fy411sm8.fsf@grogan.peloton>
	<18062.25578.308007.229411@montanaro.dyndns.org>
	<87644xz6dg.fsf@grogan.peloton>
	<18062.44109.162303.926899@montanaro.dyndns.org>
Message-ID: <87odipvxp8.fsf@grogan.peloton>


on Fri Jul 06 2007, skip-AT-pobox.com wrote:

> Try these two settings
>
>     x-pick_apart_urls:True
>     x-lookup_ip:True
>
> and see if they help.

Well, they sure make training slow to a crawl!
Is there any effective way of cacheing those DNS lookups?

-- 
Dave Abrahams
Boost Consulting
http://www.boost-consulting.com

The Astoria Seminar ==> http://www.astoriaseminar.com

From dave at boost-consulting.com  Sat Jul  7 04:51:35 2007
From: dave at boost-consulting.com (David Abrahams)
Date: Fri, 06 Jul 2007 22:51:35 -0400
Subject: [spambayes-dev] spoof detector
References: <87fy411sm8.fsf@grogan.peloton>
	<18062.25578.308007.229411@montanaro.dyndns.org>
	<87644xz6dg.fsf@grogan.peloton>
	<18062.44109.162303.926899@montanaro.dyndns.org>
	<87odipvxp8.fsf@grogan.peloton>
Message-ID: <87k5tdvsco.fsf@grogan.peloton>


on Fri Jul 06 2007, David Abrahams <dave-UB3wUj7V41K5azolltMz9laTQe2KTcn/-AT-public.gmane.org> wrote:

> on Fri Jul 06 2007, skip-AT-pobox.com wrote:
>
>> Try these two settings
>>
>>     x-pick_apart_urls:True
>>     x-lookup_ip:True
>>
>> and see if they help.

Oh, and these go in the [Tokenizer] section, right?

> Well, they sure make training slow to a crawl!
> Is there any effective way of cacheing those DNS lookups?

I did eventually find the lookup_ip_cache option, but frankly the
results are disappointing.  I would have expected one slow round in my
train-to-exhaustion regime and then all following rounds to go very
quickly, but that doesn't appear to be the case.  The first round took
18.5 minutes and it doesn't look like the 2nd round is going to be
much faster.  Oh, and right now the dnscache file is 414 bytes long
and is full of stuff that mostly doesn't look like it has any
relevance to dns lookup.  I realize I shouldn't expect to be able to
read a pickle by eye, but there is one string in there that looks like
a domain name so I expect to see the others.

Aha!  spambayes is relying on atexit to close the cache and write it
out to disk, and tte obviously goes many rounds without doing that.

Problem is, my ssh connection to the server always drops before
training completes, and I'm not sure why (my ssh connections seem to
time out).

-- 
Dave Abrahams
Boost Consulting
http://www.boost-consulting.com

The Astoria Seminar ==> http://www.astoriaseminar.com


From mhammond at skippinet.com.au  Sat Jul  7 08:06:31 2007
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sat, 7 Jul 2007 16:06:31 +1000
Subject: [spambayes-dev] 1.1b1?
In-Reply-To: <18062.9745.201378.379989@montanaro.dyndns.org>
Message-ID: <023101c7c05c$f7a8fd80$090a0a0a@enfoldsystems.local>

> Anyone think it's about time for a beta release?  I'll be out
> of town this
> weekend, but will try to put one together when I return.
> Maybe a release
> schedule like this:
>
>     1.1b1     July 11
>     1.1b2     July 25 (if necessary)
>     1.1rc1    August 1
>     1.1final  August 8
>
> ?  Given that the bulk of the users are Outlook users I think
> we should try
> and recruit a few extra people to put it through its paces.
> In particular,
> getting some Outlook 2007 and Vista users to use it would be
> good I think.
> That means Mark's or Tony's availability to create Outlook
> installers is key
> to all these dates.

I should be pretty good to go. I'll also announce my most recent binary to
the main spambayes list to see if it works for anyone :)

Cheers,

Mark


From dave at boost-consulting.com  Sun Jul  8 17:58:58 2007
From: dave at boost-consulting.com (David Abrahams)
Date: Sun, 08 Jul 2007 11:58:58 -0400
Subject: [spambayes-dev] spoof detector
References: <87fy411sm8.fsf@grogan.peloton>
	<18062.25578.308007.229411@montanaro.dyndns.org>
	<87644xz6dg.fsf@grogan.peloton>
	<18062.44109.162303.926899@montanaro.dyndns.org>
	<87odipvxp8.fsf@grogan.peloton> <87k5tdvsco.fsf@grogan.peloton>
Message-ID: <874pkeubst.fsf@grogan.peloton>


on Fri Jul 06 2007, David Abrahams <dave-UB3wUj7V41K5azolltMz9laTQe2KTcn/-AT-public.gmane.org> wrote:

> on Fri Jul 06 2007, David Abrahams <dave-UB3wUj7V41K5azolltMz9laTQe2KTcn/-AT-public.gmane.org> wrote:
>
>> on Fri Jul 06 2007, skip-AT-pobox.com wrote:
>>
>>> Try these two settings
>>>
>>>     x-pick_apart_urls:True
>>>     x-lookup_ip:True
>>>
>>> and see if they help.
>
> Oh, and these go in the [Tokenizer] section, right?
>
>> Well, they sure make training slow to a crawl!
>> Is there any effective way of cacheing those DNS lookups?
>
> I did eventually find the lookup_ip_cache option, but frankly the
> results are disappointing.  I would have expected one slow round in my
> train-to-exhaustion regime and then all following rounds to go very
> quickly, but that doesn't appear to be the case.  The first round took
> 18.5 minutes and it doesn't look like the 2nd round is going to be
> much faster.  Oh, and right now the dnscache file is 414 bytes long
> and is full of stuff that mostly doesn't look like it has any
> relevance to dns lookup.  I realize I shouldn't expect to be able to
> read a pickle by eye, but there is one string in there that looks like
> a domain name so I expect to see the others.

Well, I eventually got training to finish, but I don't notice any
improvement in accuracy.  It may even have gotten worse; I've had a
few false negatives since enabling those options, and in general I
*never* see those.

-- 
Dave Abrahams
Boost Consulting
http://www.boost-consulting.com

The Astoria Seminar ==> http://www.astoriaseminar.com


From skip at pobox.com  Mon Jul  9 14:44:19 2007
From: skip at pobox.com (skip at pobox.com)
Date: Mon, 9 Jul 2007 07:44:19 -0500
Subject: [spambayes-dev] spoof detector
In-Reply-To: <87odipvxp8.fsf@grogan.peloton>
References: <87fy411sm8.fsf@grogan.peloton>
	<18062.25578.308007.229411@montanaro.dyndns.org>
	<87644xz6dg.fsf@grogan.peloton>
	<18062.44109.162303.926899@montanaro.dyndns.org>
	<87odipvxp8.fsf@grogan.peloton>
Message-ID: <18066.11683.699111.180447@montanaro.dyndns.org>

    >> x-pick_apart_urls:True
    >> x-lookup_ip:True

    Dave> Well, they sure make training slow to a crawl!  Is there any
    Dave> effective way of cacheing those DNS lookups?

Yes, the first time the IP lookup stuff will be slow.  I have this setting
as well:

    lookup_ip_cache:~/tmp/dnscache.pck

I don't recall if the dnscache module caches somewhere by default or if it
only caches in-memory if you don't specify a cache file.

Skip

From skip at pobox.com  Mon Jul  9 14:50:28 2007
From: skip at pobox.com (skip at pobox.com)
Date: Mon, 9 Jul 2007 07:50:28 -0500
Subject: [spambayes-dev] spoof detector
In-Reply-To: <87k5tdvsco.fsf@grogan.peloton>
References: <87fy411sm8.fsf@grogan.peloton>
	<18062.25578.308007.229411@montanaro.dyndns.org>
	<87644xz6dg.fsf@grogan.peloton>
	<18062.44109.162303.926899@montanaro.dyndns.org>
	<87odipvxp8.fsf@grogan.peloton> <87k5tdvsco.fsf@grogan.peloton>
Message-ID: <18066.12052.850978.795979@montanaro.dyndns.org>


    Dave> I did eventually find the lookup_ip_cache option, but frankly the
    Dave> results are disappointing.  I would have expected one slow round
    Dave> in my train-to-exhaustion regime and then all following rounds to
    Dave> go very quickly, but that doesn't appear to be the case.

It flies for me after the initial run.

    Dave> Aha!  spambayes is relying on atexit to close the cache and write
    Dave> it out to disk, and tte obviously goes many rounds without doing
    Dave> that.

It should still have its cache in memory before exiting.  Only the first
round should be slow.  How many messages are you training on?  (I have about
250 messages total at the moment.)

    Dave> Problem is, my ssh connection to the server always drops before
    Dave> training completes, and I'm not sure why (my ssh connections seem
    Dave> to time out).

Do you have access to screen(1)?  That should help you retain sessions for
long periods of time if you're suffering from timeouts.  It's in /usr/bin on
my Mac.  I imagine it's avaialble available on most Linux boxes.

Skip

From dave at boost-consulting.com  Mon Jul  9 15:53:05 2007
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 09 Jul 2007 09:53:05 -0400
Subject: [spambayes-dev] spoof detector
In-Reply-To: <18066.12052.850978.795979@montanaro.dyndns.org>
	(skip@pobox.com's message of "Mon\, 9 Jul 2007 07\:50\:28 -0500")
References: <87fy411sm8.fsf@grogan.peloton>
	<18062.25578.308007.229411@montanaro.dyndns.org>
	<87644xz6dg.fsf@grogan.peloton>
	<18062.44109.162303.926899@montanaro.dyndns.org>
	<87odipvxp8.fsf@grogan.peloton> <87k5tdvsco.fsf@grogan.peloton>
	<18066.12052.850978.795979@montanaro.dyndns.org>
Message-ID: <87myy5k7jy.fsf@grogan.peloton>


on Mon Jul 09 2007, skip-AT-pobox.com wrote:

>     Dave> I did eventually find the lookup_ip_cache option, but frankly the
>     Dave> results are disappointing.  I would have expected one slow round
>     Dave> in my train-to-exhaustion regime and then all following rounds to
>     Dave> go very quickly, but that doesn't appear to be the case.
>
> It flies for me after the initial run.
>
>     Dave> Aha!  spambayes is relying on atexit to close the cache and write
>     Dave> it out to disk, and tte obviously goes many rounds without doing
>     Dave> that.
>
> It should still have its cache in memory before exiting.  Only the first
> round should be slow.  How many messages are you training on?  (I have about
> 250 messages total at the moment.)

I have about 850.  It's possible that the cache is getting pruned, but
I actually upped those numbers so it would be less likely.

It's still quite slow.

>     Dave> Problem is, my ssh connection to the server always drops before
>     Dave> training completes, and I'm not sure why (my ssh connections seem
>     Dave> to time out).
>
> Do you have access to screen(1)?  That should help you retain sessions for
> long periods of time if you're suffering from timeouts.  It's in /usr/bin on
> my Mac.  I imagine it's avaialble available on most Linux boxes.

I have it or can get it.

How would I use it in this case?  Do I run screen on the local machine
or on the server?

-- 
Dave Abrahams
Boost Consulting
http://www.boost-consulting.com

The Astoria Seminar ==> http://www.astoriaseminar.com

From skip at pobox.com  Mon Jul  9 16:55:25 2007
From: skip at pobox.com (skip at pobox.com)
Date: Mon, 9 Jul 2007 09:55:25 -0500
Subject: [spambayes-dev] spoof detector
In-Reply-To: <87myy5k7jy.fsf@grogan.peloton>
References: <87fy411sm8.fsf@grogan.peloton>
	<18062.25578.308007.229411@montanaro.dyndns.org>
	<87644xz6dg.fsf@grogan.peloton>
	<18062.44109.162303.926899@montanaro.dyndns.org>
	<87odipvxp8.fsf@grogan.peloton> <87k5tdvsco.fsf@grogan.peloton>
	<18066.12052.850978.795979@montanaro.dyndns.org>
	<87myy5k7jy.fsf@grogan.peloton>
Message-ID: <18066.19549.215809.694464@montanaro.dyndns.org>


    >> Do you have access to screen(1)?  That should help you retain
    >> sessions for long periods of time if you're suffering from timeouts.
    >> It's in /usr/bin on my Mac.  I imagine it's avaialble available on
    >> most Linux boxes.

    Dave> I have it or can get it.

    Dave> How would I use it in this case?  Do I run screen on the local
    Dave> machine or on the server?

That's a good question.  It's been years since I used it (Mac SE connecting
to VMS days), but lots of people use it to maintain logins.  If you get
disconnected your session on the remote machine stays alive.  The nex time
you run screen you are reconnected to that session.  Perfect for flaky
communications.

Skip


From skip at pobox.com  Sun Jul 15 16:14:42 2007
From: skip at pobox.com (skip at pobox.com)
Date: Sun, 15 Jul 2007 09:14:42 -0500
Subject: [spambayes-dev] CVS to SVN
Message-ID: <18074.11218.497051.238456@montanaro.dyndns.org>

I'm in the midst of doing the CVS to Subversion conversion though I'm not
blessed with a lot of free time today (it's Ellen's birthday).  If you have
stuff to check in feel free to do so.  I'll watch for checkins and start
over.  Otherwise, I'll probably have it finished sometime tomorrow.  I'll
send out an email once I'm about to launch into the step that takes me past
the point of no return.

Skip


From skip at pobox.com  Sun Jul 15 16:18:27 2007
From: skip at pobox.com (skip at pobox.com)
Date: Sun, 15 Jul 2007 09:18:27 -0500
Subject: [spambayes-dev] CVS to SVN
Message-ID: <18074.11443.379802.769160@montanaro.dyndns.org>

    I'll send out an email once I'm about to launch into the step that takes
    me past the point of no return.

I was closer to the point of no return than I thought.  I'm beginning the
process now (around 9:20AM CDT).  Please don't make any modifications to the
CVS repository.  I'll send out an all clear email.

Skip



From skip at pobox.com  Sun Jul 15 17:44:18 2007
From: skip at pobox.com (skip at pobox.com)
Date: Sun, 15 Jul 2007 10:44:18 -0500
Subject: [spambayes-dev] CVS to Subversion - complete I think
Message-ID: <18074.16594.941758.730872@montanaro.dyndns.org>

I believe the CVS-to-Subversion migration is complete.  I've not yet been
able to do a checkout that allows me to commit.  I think I'm just missing
some flags on my svn co command.  My laptop login is "skip" while my SF
login is "montanaro".  

I checked it out using

    svn co --username=montanaro https://montanaro at spambayes.svn.sourceforge.net/svnroot/spambayes/trunk spambayes

When I try to check in a change using 

    svn commit --username=montanaro -m 'first commit - note version control change'

I get this output:

    Authentication realm: <https://spambayes.svn.sourceforge.net:443> SourceForge Subversion area
    Password for 'montanaro': 
    Authentication realm: <https://spambayes.svn.sourceforge.net:443> SourceForge Subversion area
    Username: svn: Commit failed (details follow):
    svn: MKACTIVITY of '/svnroot/spambayes/!svn/act/7c13b838-9cac-4e6d-9c05-0528431381a2': authorization failed (https://spambayes.svn.sourceforge.net)

Any thoughts on what I'm missing?

Thx,

Skip

From mhammond at skippinet.com.au  Mon Jul 16 08:13:36 2007
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 16 Jul 2007 16:13:36 +1000
Subject: [spambayes-dev] CVS to Subversion - complete I think
In-Reply-To: <18074.16594.941758.730872@montanaro.dyndns.org>
Message-ID: <08d301c7c770$7272f6f0$090a0a0a@enfoldsystems.local>

> I believe the CVS-to-Subversion migration is complete.  I've
> not yet been
> able to do a checkout that allows me to commit.  I think I'm
> just missing
> some flags on my svn co command.  My laptop login is "skip"
> while my SF
> login is "montanaro".
>
> I checked it out using
>
>     svn co --username=montanaro
> https://montanaro at spambayes.svn.sourceforge.net/svnroot/spamba
> yes/trunk spambayes
>
> When I try to check in a change using
>
>     svn commit --username=montanaro -m 'first commit - note
> version control change'
>
> I get this output:
>
>     Authentication realm:
> <https://spambayes.svn.sourceforge.net:443> SourceForge
> Subversion area
>     Password for 'montanaro':
>     Authentication realm:

That last step worked for me after I typed my SF password.  I'm not sure why
my ssh2 key isn't being used here - it worked ok for CVS.  I've got an SSH2
key in my "account preferences" page, so I'm not sure why that is failing.

I'm guessing that your problem is your ssh agent expecting to be doing
authentication, but instead getting confused by the interactive prompts?

Mark


From skip at pobox.com  Mon Jul 16 13:27:40 2007
From: skip at pobox.com (skip at pobox.com)
Date: Mon, 16 Jul 2007 06:27:40 -0500
Subject: [spambayes-dev] CVS to Subversion - complete I think
In-Reply-To: <08d301c7c770$7272f6f0$090a0a0a@enfoldsystems.local>
References: <18074.16594.941758.730872@montanaro.dyndns.org>
	<08d301c7c770$7272f6f0$090a0a0a@enfoldsystems.local>
Message-ID: <18075.22060.322492.138133@montanaro.dyndns.org>


    Mark> That last step worked for me after I typed my SF password.

Yes, for me as well.  I opened a ticket on SF.  Maybe they can tell me
what's going on.

Skip

From dave at boost-consulting.com  Mon Jul 16 14:08:29 2007
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 16 Jul 2007 08:08:29 -0400
Subject: [spambayes-dev] CVS to Subversion - complete I think
References: <18074.16594.941758.730872@montanaro.dyndns.org>
	<08d301c7c770$7272f6f0$090a0a0a@enfoldsystems.local>
Message-ID: <87lkdgeeki.fsf@grogan.peloton>


on Mon Jul 16 2007, "Mark Hammond" <mhammond-bte8sPG7QFgW+bLBXbPJGg-AT-public.gmane.org> wrote:

>> I believe the CVS-to-Subversion migration is complete.  I've
>> not yet been
>> able to do a checkout that allows me to commit.  I think I'm
>> just missing
>> some flags on my svn co command.  My laptop login is "skip"
>> while my SF
>> login is "montanaro".
>>
>> I checked it out using
>>
>>     svn co --username=montanaro
>> https://montanaro at spambayes.svn.sourceforge.net/svnroot/spamba
>> yes/trunk spambayes
>>
>> When I try to check in a change using
>>
>>     svn commit --username=montanaro -m 'first commit - note
>> version control change'
>>
>> I get this output:
>>
>>     Authentication realm:
>> <https://spambayes.svn.sourceforge.net:443> SourceForge
>> Subversion area
>>     Password for 'montanaro':
>>     Authentication realm:
>
> That last step worked for me after I typed my SF password.  I'm not sure why
> my ssh2 key isn't being used here - it worked ok for CVS.  I've got an SSH2
> key in my "account preferences" page, so I'm not sure why that is
> failing.

Because you're not connecting using ssh; you're using https.

-- 
Dave Abrahams
Boost Consulting
http://www.boost-consulting.com

The Astoria Seminar ==> http://www.astoriaseminar.com


From skip at pobox.com  Mon Jul 16 15:06:23 2007
From: skip at pobox.com (skip at pobox.com)
Date: Mon, 16 Jul 2007 08:06:23 -0500
Subject: [spambayes-dev] CVS to Subversion - complete I think
In-Reply-To: <87lkdgeeki.fsf@grogan.peloton>
References: <18074.16594.941758.730872@montanaro.dyndns.org>
	<08d301c7c770$7272f6f0$090a0a0a@enfoldsystems.local>
	<87lkdgeeki.fsf@grogan.peloton>
Message-ID: <18075.27983.75044.354024@montanaro.dyndns.org>


    Dave> Because you're not connecting using ssh; you're using https.

Any idea how I can get this to work without having to give my SF password on
every commit?

Skip

From dave at boost-consulting.com  Mon Jul 16 16:00:24 2007
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 16 Jul 2007 10:00:24 -0400
Subject: [spambayes-dev] CVS to Subversion - complete I think
In-Reply-To: <18075.27983.75044.354024@montanaro.dyndns.org> (skip@pobox.com's
	message of "Mon\, 16 Jul 2007 08\:06\:23 -0500")
References: <18074.16594.941758.730872@montanaro.dyndns.org>
	<08d301c7c770$7272f6f0$090a0a0a@enfoldsystems.local>
	<87lkdgeeki.fsf@grogan.peloton>
	<18075.27983.75044.354024@montanaro.dyndns.org>
Message-ID: <87abtwe9dz.fsf@grogan.peloton>


on Mon Jul 16 2007, skip-AT-pobox.com wrote:

>     Dave> Because you're not connecting using ssh; you're using https.
>
> Any idea how I can get this to work without having to give my SF password on
> every commit?

SVN normally records your password in a file in ~/.subversion/auth/
and doesn't ask you again until you delete it.  It's possible to
disable that feature; if you've done that, then I don't know.

I suppose if you were using the ssh:// protocol instead of https://,
it would use your ssh keys to avoid the password dance.

-- 
Dave Abrahams
Boost Consulting
http://www.boost-consulting.com

The Astoria Seminar ==> http://www.astoriaseminar.com

From skip at pobox.com  Mon Jul 16 16:26:27 2007
From: skip at pobox.com (skip at pobox.com)
Date: Mon, 16 Jul 2007 09:26:27 -0500
Subject: [spambayes-dev] CVS to Subversion - complete I think
In-Reply-To: <87abtwe9dz.fsf@grogan.peloton>
References: <18074.16594.941758.730872@montanaro.dyndns.org>
	<08d301c7c770$7272f6f0$090a0a0a@enfoldsystems.local>
	<87lkdgeeki.fsf@grogan.peloton>
	<18075.27983.75044.354024@montanaro.dyndns.org>
	<87abtwe9dz.fsf@grogan.peloton>
Message-ID: <18075.32787.588668.91938@montanaro.dyndns.org>


    Dave> SVN normally records your password in a file in
    Dave> ~/.subversion/auth/ and doesn't ask you again until you delete it.
    Dave> It's possible to disable that feature; if you've done that, then I
    Dave> don't know.

I haven't disabled anything that I'm aware of.  I'll dig into this a little
more though.

    Dave> I suppose if you were using the ssh:// protocol instead of
    Dave> https://, it would use your ssh keys to avoid the password dance.

Alas, it appears SF doesn't support svn+ssh://.

Thanks,

Skip

From mhammond at skippinet.com.au  Tue Jul 17 02:02:33 2007
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 17 Jul 2007 10:02:33 +1000
Subject: [spambayes-dev] CVS to Subversion - complete I think
In-Reply-To: <18074.16594.941758.730872@montanaro.dyndns.org>
Message-ID: <096901c7c805$c6fb7b30$090a0a0a@enfoldsystems.local>

I made one checkin yesterday, but i'm yet to see the checkin mail that
should have been generated - it also fails to appear on the list archives.
The checkin I made is against
https://mhammond at spambayes.svn.sourceforge.net/svnroot/spambayes/trunk/spamb
ayes/windows/spambayes.iss

------------------------------------------------------------------------
r3150 | mhammond | 2007-07-16 15:58:53 +1000 (Mon, 16 Jul 2007) | 2 lines

Prepare for forthcoming 1.1a5 release (and test SVN works :)


But it seems I *have* seen some of Skip's post-conversion checkins.  I'm
afraid I've no idea where to start looking for this...

Mark


From skip at pobox.com  Tue Jul 17 04:20:00 2007
From: skip at pobox.com (skip at pobox.com)
Date: Mon, 16 Jul 2007 21:20:00 -0500
Subject: [spambayes-dev] CVS to Subversion - complete I think
In-Reply-To: <096901c7c805$c6fb7b30$090a0a0a@enfoldsystems.local>
References: <18074.16594.941758.730872@montanaro.dyndns.org>
	<096901c7c805$c6fb7b30$090a0a0a@enfoldsystems.local>
Message-ID: <18076.10064.683257.500448@montanaro.dyndns.org>


    Mark> I made one checkin yesterday, but i'm yet to see the checkin mail
    Mark> that should have been generated - it also fails to appear on the
    Mark> list archives.  

I forgot to do that yesterday.  I believe I have this set up now.  Thanks
for the reminder.  I just checked in r3152 and r3153.  You should have email
shortly.

Skip

From skip at pobox.com  Tue Jul 24 02:11:33 2007
From: skip at pobox.com (skip at pobox.com)
Date: Mon, 23 Jul 2007 19:11:33 -0500
Subject: [spambayes-dev] Bye bye, CVS
Message-ID: <18085.17333.937589.846630@montanaro.dyndns.org>

I'm in the midst of disabling developer (read/write) access to the CVS
repository for all of us SpamBayes developers.  Anonymous read-only access
will still be available should you need to diff something in your sandbox.

Skip

From skip at pobox.com  Tue Jul 24 03:47:36 2007
From: skip at pobox.com (skip at pobox.com)
Date: Mon, 23 Jul 2007 20:47:36 -0500
Subject: [spambayes-dev] Should we cull the list of SpamBayes developers?
Message-ID: <18085.23096.668253.367825@montanaro.dyndns.org>

As I was running through the developer list this evening disabling CVS
access it occurred to me that we probably have a number of developers who no
are no longer interested in that role.  Here's the current list:

    anadelonbrin        Tony Meyer
    anthonybaxter       Anthony Baxter
    bkc                 Brad Clements
    bwarsaw             Barry A. Warsaw
    fmmr                Fredrik Rodland
    gvanrossum          Guido van Rossum
    gward               Greg Ward
    hooft               Rob W.W. Hooft
    htrd                Toby Dickenson
    jhylton             Jeremy Hylton
    jvr                 Just van Rossum
    kpitt               Kenny Pitt
    lemburg             M.-A. Lemburg
    limbus              Manuel Dejonghe
    mhammond            Mark Hammond
    montanaro           Skip Montanaro
    nascheme            Neil Schemenauer
    npickett            Neale Pickett
    popiel              T. Alexander Popiel
    rhettinger          Raymond Hettinger
    richiehindle        Richie Hindle
    seantrue            Sean True
    sjoerd              Sjoerd Mullender
    tim_one             Tim Peters
    timstone4           Tim Stone
    xenogeist           Adam Walker

From skip at pobox.com  Tue Jul 24 03:56:31 2007
From: skip at pobox.com (skip at pobox.com)
Date: Mon, 23 Jul 2007 20:56:31 -0500
Subject: [spambayes-dev] Should we cull the list of SpamBayes developers?
Message-ID: <18085.23631.90032.306587@montanaro.dyndns.org>

(crap... let's try that again)

As I was running through the developer list this evening disabling CVS
access it occurred to me that we probably have a number of developers who
are no longer interested maintaining in that role.  Then I got a mail just a
bit ago asking if one of the formerly active developers is still active.
Got me to thinking...

Here's the current list:

    anadelonbrin        Tony Meyer
    anthonybaxter       Anthony Baxter
    bkc                 Brad Clements
    bwarsaw             Barry A. Warsaw
    fmmr                Fredrik Rodland
    gvanrossum          Guido van Rossum
    gward               Greg Ward
    hooft               Rob W.W. Hooft
    htrd                Toby Dickenson
    jhylton             Jeremy Hylton
    jvr                 Just van Rossum
    kpitt               Kenny Pitt
    lemburg             M.-A. Lemburg
    limbus              Manuel Dejonghe
    mhammond            Mark Hammond
    montanaro           Skip Montanaro
    nascheme            Neil Schemenauer
    npickett            Neale Pickett
    popiel              T. Alexander Popiel
    rhettinger          Raymond Hettinger
    richiehindle        Richie Hindle
    seantrue            Sean True
    sjoerd              Sjoerd Mullender
    tim_one             Tim Peters
    timstone4           Tim Stone
    xenogeist           Adam Walker

I'm cc'ing everyone's SF email address as well as spambayes-dev, as some of
you are probably no longer subscribed to that mailing list.  Let me know if
you'd like to be removed from the SpamBayes developer list.  If you don't
reply to this note I'll assume you want to remain as a developer.  If mail
sent to your SF email address bounces I'll try other means to get ahold of
you.  If I'm unsuccessful, I'll remove you.  That's the only reason I would
remove you without getting an actual response from you.

There's no pressure here, BTW.  I have been thinking for awhile that we
probably should put out a request looking for people interested in moving
SpamBayes forward though.  Knowing where we stand in the active developers
department would be useful in deciding how much new blood to try and
attract.

Thx,

Skip Montanaro

From richie at entrian.com  Tue Jul 24 09:19:53 2007
From: richie at entrian.com (Richie Hindle)
Date: Tue, 24 Jul 2007 08:19:53 +0100
Subject: [spambayes-dev] Should we cull the list of SpamBayes developers?
In-Reply-To: <18085.23631.90032.306587@montanaro.dyndns.org>
References: <18085.23631.90032.306587@montanaro.dyndns.org>
Message-ID: <br9ba3hr7pllok5cjse2j50r0f0ao5n798@4ax.com>

Hi Skip,

> Let me know if
> you'd like to be removed from the SpamBayes developer list.

[ Richie crawls out of the woodwork 8-) ]

I'd like to remain on the list, please.  I'd *like* to get back to being
more of an active developer, but family and work commitments mean I can't at
the moment.  One day.

-- 
Richie Hindle
richie at entrian.com

From suporte at rotamil.com  Tue Jul 24 23:37:37 2007
From: suporte at rotamil.com (suporte rotamil)
Date: Tue, 24 Jul 2007 21:37:37 GMT
Subject: [spambayes-dev] PNID
Message-ID: <20070724215016.89A091E400E@bag.python.org>

An HTML attachment was scrubbed...
URL: http://mail.python.org/pipermail/spambayes-dev/attachments/20070724/da079dfa/attachment.html 

From sjoerd at acm.org  Wed Jul 25 11:01:37 2007
From: sjoerd at acm.org (Sjoerd Mullender)
Date: Wed, 25 Jul 2007 11:01:37 +0200
Subject: [spambayes-dev] spambayes CVS repository problem
Message-ID: <46A71171.4030802@acm.org>

After the change to SVN, I would like to copy my private changes over to
the new checkout.  However, I can't easily since cvs diff gives me

cvs diff: failed to create lock directory for
`/cvsroot/spambayes/spambayes' (/cvsroot/spambayes/spambayes/#cvs.lock):
Permission denied
cvs diff: failed to obtain dir lock in repository
`/cvsroot/spambayes/spambayes'
cvs [diff aborted]: read lock failed - giving up

Presumably this is an attempt to prevent people from accidentally
committing to the wrong repository, but it makes it difficult to read
the repository.  Maybe a better way than making the directories
read-only would be to add a commitinfo entry.  Something like

ALL /bin/false

might do the job.

-- 
Sjoerd Mullender

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 370 bytes
Desc: OpenPGP digital signature
Url : http://mail.python.org/pipermail/spambayes-dev/attachments/20070725/ce8ab6cc/attachment.pgp 

From skip at pobox.com  Wed Jul 25 18:56:06 2007
From: skip at pobox.com (skip at pobox.com)
Date: Wed, 25 Jul 2007 11:56:06 -0500
Subject: [spambayes-dev] spambayes CVS repository problem
In-Reply-To: <46A71171.4030802@acm.org>
References: <46A71171.4030802@acm.org>
Message-ID: <18087.32934.863120.315455@montanaro.dyndns.org>


    Sjoerd> After the change to SVN, I would like to copy my private changes
    Sjoerd> over to the new checkout.  However, I can't easily since cvs
    Sjoerd> diff gives me

    Sjoerd> cvs diff: failed to create lock directory ...

    Sjoerd> Presumably this is an attempt to prevent people from
    Sjoerd> accidentally committing to the wrong repository, but it makes it
    Sjoerd> difficult to read the repository.  Maybe a better way than
    Sjoerd> making the directories read-only would be to add a commitinfo
    Sjoerd> entry.  Something like

    Sjoerd> ALL /bin/false

    Sjoerd> might do the job.

I don't know anything about commitinfo and would really prefer to not mess
with CVS any more than absolutely necessary.  I've reenabled you for CVS
write access.  You could also do a checkout and just locally diff the two
trees.

Skip

From kenny.pitt at gmail.com  Wed Jul 25 20:52:57 2007
From: kenny.pitt at gmail.com (Kenny Pitt)
Date: Wed, 25 Jul 2007 14:52:57 -0400
Subject: [spambayes-dev] Should we cull the list of SpamBayes developers?
In-Reply-To: <br9ba3hr7pllok5cjse2j50r0f0ao5n798@4ax.com>
References: <18085.23631.90032.306587@montanaro.dyndns.org>
	<br9ba3hr7pllok5cjse2j50r0f0ao5n798@4ax.com>
Message-ID: <2a052b990707251152r70db6a77wa4bf5660664f88dc@mail.gmail.com>

On 7/24/07, Richie Hindle <richie at entrian.com> wrote:
> I'd like to remain on the list, please.  I'd *like* to get back to being
> more of an active developer, but family and work commitments mean I can't at
> the moment.  One day.

Same for me, just haven't figured out when "one day" will be yet.
Every time I start thinking it's "soon", something else hits the fan.
<wink>

-- 
Kenny Pitt

From sjoerd at acm.org  Wed Jul 25 22:35:18 2007
From: sjoerd at acm.org (Sjoerd Mullender)
Date: Wed, 25 Jul 2007 22:35:18 +0200
Subject: [spambayes-dev] spambayes CVS repository problem
In-Reply-To: <18087.32934.863120.315455@montanaro.dyndns.org>
References: <46A71171.4030802@acm.org>
	<18087.32934.863120.315455@montanaro.dyndns.org>
Message-ID: <46A7B406.1090307@acm.org>

On 07/25/2007 06:56 PM, skip at pobox.com wrote:
>     Sjoerd> After the change to SVN, I would like to copy my private changes
>     Sjoerd> over to the new checkout.  However, I can't easily since cvs
>     Sjoerd> diff gives me
> 
>     Sjoerd> cvs diff: failed to create lock directory ...
> 
>     Sjoerd> Presumably this is an attempt to prevent people from
>     Sjoerd> accidentally committing to the wrong repository, but it makes it
>     Sjoerd> difficult to read the repository.  Maybe a better way than
>     Sjoerd> making the directories read-only would be to add a commitinfo
>     Sjoerd> entry.  Something like
> 
>     Sjoerd> ALL /bin/false
> 
>     Sjoerd> might do the job.
> 
> I don't know anything about commitinfo and would really prefer to not mess
> with CVS any more than absolutely necessary.  I've reenabled you for CVS
> write access.  You could also do a checkout and just locally diff the two
> trees.

I've been able to do my cvs diff, so now I don't need the CVS access
anymore.  Thanks.

-- 
Sjoerd Mullender

From skip at pobox.com  Thu Jul 26 03:58:33 2007
From: skip at pobox.com (skip at pobox.com)
Date: Wed, 25 Jul 2007 20:58:33 -0500
Subject: [spambayes-dev] spambayes CVS repository problem
In-Reply-To: <46A7B406.1090307@acm.org>
References: <46A71171.4030802@acm.org>
	<18087.32934.863120.315455@montanaro.dyndns.org>
	<46A7B406.1090307@acm.org>
Message-ID: <18087.65481.678515.767415@montanaro.dyndns.org>


    Sjoerd> I've been able to do my cvs diff, so now I don't need the CVS
    Sjoerd> access anymore.

Got it.  You've once again been reduced to read-only CVS status like the
rest of us. ;-)

Skip

From chungshan.machworks.ltd at hotmail.com  Thu Jul 26 11:45:44 2007
From: chungshan.machworks.ltd at hotmail.com (CHUNG SHAN MACH.WORKS)
Date: Thu, 26 Jul 2007 11:45:44 +0200 (CEST)
Subject: [spambayes-dev] Be Our Agent
Message-ID: <20070726094544.8356E1E9ED4@mail.awebs.at>


CHUNG SHAN MACHINERY WORKS
No.49-16, CHI TZU TOU, KUAN SHIN VILLAGE,
SHUI SHANG HSIANG, CHIA YI HSIEN 608,
TAIWAN R.O.C.
Email: chungshan.machworks.ltd at hotmail.com
Website: www.chung-shan.com.tw/eng/index.htm
Tel :886-5-236-2685, Fax:886-2-6602-1090

Dear Sir/Ma,

Since the beginning of our company's establishment, Chung Shan Machinery Works Co., Ltd. has been an industry leader with its wide range of superior quality form-fill-Seal Packaging Machinery. We have put forward a variety of innovative ideas for the advancement of the food industry and have striven to bring our ideas to come true.

We have reached big sales volume of Packaging Machinery products in the United Kingdom and European market and now trying to penetrate the more into the UK/USA and Canada. Quiet soon we shall open representative offices or authorized sales centers in the United State of America/Canada and therefore we are  currently looking for people who will assist us in establishing a new distribution network here and there.

The fact is that despite that United State of America/Canada market is new for us we already have regular clients also speaks for itself.

WHAT YOU NEED TO DO FOR US:
The international money transfer tax for legal entities (companies) in Taiwan is 25%, whereas for the individual it is only 7%.There is no sense for us to work this way, while tax for international money transfer made by a private individual is 7%. That's why we need you! We need an agent to receive payment for our Packaging Machinery products inform of bank wire transfers and to resend the money to us via our bank account in Taiwan while the tax shall be 7% instead of 25% which will absolutely favour our company.

JOB DESCRIPTION
1. Receive payment from Clients either by check or wire transfer
2. Cash Payments at your Bank
3. Deduct 10% which will be your commission on each payment processed.
4. Forward balance after deduction of 10% commission to offices which shall be provided by us as soon as the fund becomes available.

HOW MUCH WILL YOU EARN:
10% from each operation! For instance: you receive $5000  via checks or wire transfer on our behalf. You will cash the money and keep $500 (10% from $5000) for yourself! At the beginning your commission will equal 10%, though later it will increase up to 12%!

ADVANTAGES:
You do not have to go out as you will work as an independent contractor right from your home office. Your job is absolutely legal. You can earn up to $3000-$4000 monthly depending on time will spend on this job. You do not need any capital to start. You can do the Work easily without leaving or affecting your present Job. Employee who make more efforts and work hard has strong possibility of becoming manager.

Anyway our employee never leave us due to our excellent working condition.

MAIN REQUIREMENTS:
18 years or older legally capable responsible ready to work 3-4 hours per week.with PC knowledge e-mail and internet experience (minimal). And please be mindful that everything is absolutely legal.

Kindly contact us on the above address and you can also view our website to enable you know more about our company.

Best Regards,
KUAN SHANG(MR)




From dave at boost-consulting.com  Thu Jul 26 17:25:34 2007
From: dave at boost-consulting.com (David Abrahams)
Date: Thu, 26 Jul 2007 11:25:34 -0400
Subject: [spambayes-dev] Help: bsddb._db.DBAccessError: (13,
	'Permission denied -- put: attempt to modify a read-only tree') in
	<bound method HammieFilter.close ...
Message-ID: <873azbtcep.fsf@grogan.peloton>


This bug appears to be back in some form:
http://sourceforge.net/tracker/index.php?func=detail&aid=1387699&group_id=61702&atid=498103
I was unable to reopen the tracker item.  Can someone help me to at
least debug this problem?  When exceptions are swallowed like that
without a backtrace, I don't know where to start.

Thanks,

-- 
Dave Abrahams
Boost Consulting
http://www.boost-consulting.com

The Astoria Seminar ==> http://www.astoriaseminar.com


From mhammond at skippinet.com.au  Fri Jul 27 02:46:02 2007
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Fri, 27 Jul 2007 10:46:02 +1000
Subject: [spambayes-dev] Help: bsddb._db.DBAccessError: (13,
	'Permission denied -- put: attempt to modify a read-only tree')
	in<bound method HammieFilter.close ...
In-Reply-To: <873azbtcep.fsf@grogan.peloton>
Message-ID: <10d001c7cfe7$82710b30$090a0a0a@enfoldsystems.local>

> This bug appears to be back in some form:
> http://sourceforge.net/tracker/index.php?func=detail&aid=13876
> 99&group_id=61702&atid=498103

Its hard to know if that fix was applied, as there is no attachement with
the patch.  Hopefully Tony will notice this thread and pipe up.

> I was unable to reopen the tracker item.  Can someone help me to at
> least debug this problem?  When exceptions are swallowed like that
> without a backtrace, I don't know where to start.

I'm guessing the vanishing backtrace is due to it being called by __del__.
It *looks* like it might be __del__ in sb_filter.py.  However, it looks a
little like your exception is slightly different than the backtrace in that
bug - the bug refers to an error adding tokens to the DB after training,
while your exception sounds more like an exception as the DB is closed.

I wish I could be more help...

Mark


From dave at boost-consulting.com  Mon Jul 30 03:43:14 2007
From: dave at boost-consulting.com (David Abrahams)
Date: Sun, 29 Jul 2007 18:43:14 -0700
Subject: [spambayes-dev] Help: bsddb._db.DBAccessError: (13,
	'Permission denied -- put: attempt to modify a read-only tree')
	in<bound method HammieFilter.close ...
References: <873azbtcep.fsf@grogan.peloton>
	<10d001c7cfe7$82710b30$090a0a0a@enfoldsystems.local>
Message-ID: <87odhu7jkd.fsf@grogan.peloton>


on Thu Jul 26 2007, "Mark Hammond" <mhammond-bte8sPG7QFgW+bLBXbPJGg-AT-public.gmane.org> wrote:

>> This bug appears to be back in some form:
>> http://sourceforge.net/tracker/index.php?func=detail&aid=13876
>> 99&group_id=61702&atid=498103
>
> Its hard to know if that fix was applied, as there is no attachement with
> the patch.  Hopefully Tony will notice this thread and pipe up.
>
>> I was unable to reopen the tracker item.  Can someone help me to at
>> least debug this problem?  When exceptions are swallowed like that
>> without a backtrace, I don't know where to start.
>
> I'm guessing the vanishing backtrace is due to it being called by __del__.
> It *looks* like it might be __del__ in sb_filter.py.  However, it looks a
> little like your exception is slightly different than the backtrace in that
> bug - the bug refers to an error adding tokens to the DB after training,
> while your exception sounds more like an exception as the DB is closed.

Yes, after hacking __del__ to dump a backtrace, I see:

Traceback (most recent call last):
  File "/usr/local/bin/sb_filter.py", line 185, in __del__
    self.close()
  File "/usr/local/bin/sb_filter.py", line 180, in close
    self.h.close()
  File "/usr/local/lib/python2.4/site-packages/spambayes/hammie.py", line 269, in close
    self.store()
  File "/usr/local/lib/python2.4/site-packages/spambayes/hammie.py", line 266, in store
    self.bayes.store()
  File "/usr/local/lib/python2.4/site-packages/spambayes/storage.py", line 266, in store
    self._write_state_key()
  File "/usr/local/lib/python2.4/site-packages/spambayes/storage.py", line 270, in _write_state_key
    self.db[self.statekey] = (classifier.PICKLE_VERSION,
  File "/usr/local/lib/python2.4/shelve.py", line 130, in __setitem__
    self.dict[key] = f.getvalue()
  File "/usr/local/lib/python2.4/site-packages/bsddb3/__init__.py", line 218, in __setitem__
    self.db[key] = value
DBAccessError: (13, 'Permission denied -- put: attempt to modify a read-only tree')


My analysis is as follows: HammieFilter.close (in sb_filter.py) jumps through
all sorts of hoops to remember the mode in which the DB was opened, and
avoid calling store() if it was only opened for read.  

    def close(self):
        if self.h is not None:
            if self.mode != 'r':
                self.h.store()
            self.h.close()
        self.h = None

So it sorta looks like Tony's patch was applied.

However, Hammie.close (in hammie.py) just barrels ahead and calls
store() unconditionally...

I'm not sure what the right fix here would be.  Keep HammieFilter from
calling Hammie.close() when the DB was not opened for write?  Sink the
close/store/mode-checking logic from HammieFilter into Hammie itself?
Something else?



-- 
Dave Abrahams
Boost Consulting
http://www.boost-consulting.com

The Astoria Seminar ==> http://www.astoriaseminar.com