• Dupe checking

    From Hans Beekhuizen@2:280/1018.1 to All on Wed Apr 7 17:33:43 2004
    Hallo All,

    I'm using the Linux version of Squish 1.11 and have a problem wioth dupe checking. According to me the following two messages should not be marked as dupes because they have different MsgIDs. At the moment I'm using "DupeCheck MsgID", but I also tried "DupeCheck Header MsgID". And I have "DupeLongHeader"
    in my config, although that should, I believe, not make any difference in this case.

    Could somebody please tell me what's wrong?

    Groetjes,

    Hans.

    === import ===
    = DUPEMAIL (2:280/1018.1) =====================================================
    Msg : 1 of 2
    From : Johannes D.H. Beekhuizen 2:280/1018 06 Apr 04 14:18:38
    To : Iedereen
    Subj : Kijk eens aan! =============================================================================== AREA:DUINHEKS.NEWS
    @MSGID: 2:280/1018@fidonet 3a7ef378
    @PID: MBSE-FIDO 0.51.3 (GNU/Linux-i386)
    @CHRS: LATIN-1 2
    @TZUTC: 0200
    @TID: MBSE-FIDO 0.51.3 (GNU/Linux-i386)
    Systeem naam duinheks.xs4all.nl
    Sysop Johannes D.H. Beekhuizen (jbeekhui@duinheks.xs4all.nl)

    >cut<

    Met vriendlijke groet: Johannes D.H. Beekhuizen.

    ... If it glows don't touch it!

    -+- MBSE BBS v0.51.3 (GNU/Linux-i386)
    + Origin: The Wizard's Linux box (2:280/1018)
    SEEN+BY: 280/1018 1042
    @PATH: 280/1018

    = DUPEMAIL (2:280/1018.1) =====================================================
    Msg : 2 of 2
    From : Johannes D.H. Beekhuizen 2:280/1018 07 Apr 04 00:36:46
    To : Iedereen
    Subj : Kijk eens aan! =============================================================================== AREA:DUINHEKS.NEWS
    @MSGID: 2:280/1018@fidonet 3a7ef472
    @PID: MBSE-FIDO 0.51.3 (GNU/Linux-i386)
    @CHRS: LATIN-1 2
    @TZUTC: 0200
    @TID: MBSE-FIDO 0.51.3 (GNU/Linux-i386)
    Systeem naam duinheks.xs4all.nl
    Sysop Johannes D.H. Beekhuizen (jbeekhui@duinheks.xs4all.nl)

    >cut<

    Met vriendlijke groet: Johannes D.H. Beekhuizen.

    ... We'll give you piece de resistance and a tour de force


    -+- MBSE BBS v0.51.3 (GNU/Linux-i386)
    + Origin: The Wizard's Linux box (2:280/1018)
    SEEN+BY: 280/1018 1042
    @PATH: 280/1018

    === tropmi ===

    --- GoldED+/LNX 1.1.5cvs031202
    * Origin: The Wizard uses MBSE/Linux (2:280/1018.1)
  • From Bo Simonsen@2:236/100 to Hans Beekhuizen on Thu Apr 8 01:20:51 2004
    Hello Hans.

    07 Apr 04 16:33, you wrote to all:

    I'm using the Linux version of Squish 1.11 and have a problem wioth
    dupe checking. According to me the following two messages should not
    be marked as dupes because they have different MsgIDs. At the moment
    I'm using "DupeCheck MsgID", but I also tried "DupeCheck Header
    MsgID". And I have "DupeLongHeader" in my config, although that
    should, I believe, not make any difference in this case.

    Could somebody please tell me what's wrong?

    We did have some major problems with dupechecking and the 3.03b release of maximus, they are fixed in the CVS version.

    So you can do.

    1. Use the CVS version

    cvs -d :pserver:anonymous@cvs.sourceforge.net:/cvsroot/maximus co maximus

    2. Only do dupechecking on Header. In squish.cfg

    DupeCheck Header

    instead of DupeCheck Header MSGID.

    Unfortionally can't I send you a patch because Bob Jones did correct the problem, and I can't remember what was wrong (properly some collition of 32- and 16 bit integers).

    If you have other problems with Squish feel free to report them.

    Bo

    --- GoldED+/LNX 1.1.5-31012
    * Origin: =====> The Night Express <===== (2:236/100)
  • From Mike Tripp@1:382/61 to Hans Beekhuizen on Thu Apr 8 08:35:52 2004
    Hello Hans!

    07 Apr 04 16:33, Hans Beekhuizen wrote to All:

    I'm using the Linux version of Squish 1.11 and have a problem wioth
    dupe checking. According to me the following two messages should not
    be marked as dupes because they have different MsgIDs.

    It appears that both of the messages you posted info for were the ones that Squish placed in the DUPE area. That means that each of these two has a twin that already lives in the target area, not that these two are twins of each other. You would need to have received triplets in order to end up with twins in the DUPE area.<g>

    .\\ike

    --- GoldED 2.50+
    * Origin: -=( The TechnoDrome )=- Austin,TX 512-327-8598 33.6k (1:382/61)
  • From Hans Beekhuizen@2:280/1018.1 to Mike Tripp on Fri Apr 9 17:37:08 2004
    Hallo Mike,

    Op 08 Apr 04 schreef Mike Tripp aan Hans Beekhuizen:

    I'm using the Linux version of Squish 1.11 and have a problem wioth
    dupe checking.
    It appears that both of the messages you posted info for were the ones that Squish placed in the DUPE area. That means that each of these two has a twin that already lives in the target area,

    That's what it *should* mean, but in this case it doesn't...

    not that these two are twins of each other. You would need to have received triplets in order to end up with twins in the DUPE
    area.<g>

    For me that's obvious :) When I set up this point, collecting the Fido packets
    with ftp and using squish to make them available for GoldED+, I of course first
    suspected something wrong in my setup when I saw this happening. But I soon found out these weren't real dupes.

    Groetjes,

    Hans.

    --- GoldED+/LNX 1.1.5cvs031202
    * Origin: The Wizard uses MBSE/Linux (2:280/1018.1)
  • From Hans Beekhuizen@2:280/1018.1 to Bo Simonsen on Fri Apr 9 17:55:39 2004
    Hallo Bo,

    Op 08 Apr 04 schreef Bo Simonsen aan Hans Beekhuizen:

    I'm using the Linux version of Squish 1.11 and have a problem wioth
    dupe checking.
    We did have some major problems with dupechecking and the 3.03b release of maximus, they are fixed in the CVS version.

    Aha :))

    1. Use the CVS version

    I just installed it, so will find out tomorrow morning if this problem has been
    solved.

    2. Only do dupechecking on Header. In squish.cfg

    Except for the date, the headers in these messages are all the same. Would it work then?

    Unfortionally can't I send you a patch because Bob Jones did correct the problem, and I can't remember what was wrong (properly some collition of 32- and 16 bit integers).

    That's no problem. I have ADSL so don't have to worry about those nasty ticks in the telephone :)

    If you have other problems with Squish feel free to report them.

    Of course I will :)) But don't expect too much from me: I only set up this point system to help one of my points who is interested in reading his Fido messages under Linux, instead of in a Dos-box under Windows. For myself I'm happy with mbse [a Linux-native all-in-one package] and GoldED+.

    Groetjes,

    Hans.

    --- GoldED+/LNX 1.1.5cvs031202
    * Origin: The Wizard uses MBSE/Linux (2:280/1018.1)
  • From Maurice Kinal@1:153/401.1 to Hans Beekhuizen on Fri Apr 9 10:51:58 2004
    Hey Hans!

    Apr 09 16:37 04, Hans Beekhuizen wrote to Mike Tripp:

    But I soon found out these weren't real dupes.

    Let me guess, duped MSGID's? I've seen quite a few of those from a hub who doesn't scan the MSGID's but don't see those exact same messages from the hub that does, although that hub might not see them either as they maybe getting filtered out before that hub gets them. If the MSGID worked then it would be a
    good, fast method but it obviously has flaws when dealing with software that thinks it's getting it right.

    Life is good,
    Maurice

    --- Msged/LNX 6.1.1
    * Origin: Little Mikey's Brain BBS - A work in progress (1:153/401.1)
  • From Bo Simonsen@2:236/100 to Mike Tripp on Thu Apr 8 19:34:51 2004
    Mike Tripp wrote to Hans Beekhuizen about ''Dupe checking'':

    I'm using the Linux version of Squish 1.11 and have a problem wioth
    dupe checking. According to me the following two messages should not
    be marked as dupes because they have different MsgIDs.

    It appears that both of the messages you posted info for were the
    ones that Squish placed in the DUPE area. That means that each of
    these two has a twin that already lives in the target area, not
    that these two are twins of each other. You would need to have
    received triplets in order to end up with twins in the DUPE
    area.<g>

    They got different MSGID so squish wouldn't juge it as dupe normally, but there
    is a bug with dupe detection in the current release.

    Bo

    ---
    * Origin: <=== The Night Express ===> (2:236/100)
  • From Bo Simonsen@2:236/100 to Hans Beekhuizen on Sat Apr 10 01:42:34 2004
    Hans Beekhuizen wrote to Bo Simonsen about ''Re: Dupe checking'':

    Hallo Bo,

    Hi Hans,

    I'm using the Linux version of Squish 1.11 and have a problem wioth
    dupe checking.
    We did have some major problems with dupechecking and the 3.03b
    release of
    maximus, they are fixed in the CVS version.

    Aha :))

    We are about to release a more stable version of maximus/squish/sqafix.

    1. Use the CVS version

    I just installed it, so will find out tomorrow morning if this
    problem has been solved.

    Okay, I'm 99% sure it's. :) Squish is so stable, so I've fixed all compiler warnings in it! ;)

    2. Only do dupechecking on Header. In squish.cfg

    Except for the date, the headers in these messages are all the
    same. Would it work then?

    I'm not sure, I didn't get mutch dupe by running Squish without msgid check, and right now there are still people which is using Software which seems to be buyed at antique-market, which doesn't set @MSGID. So Squish will only do dupechecking on Header, and it's not caused any dupes to me.

    Unfortionally can't I send you a patch because Bob Jones did
    correct the
    problem, and I can't remember what was wrong (properly some
    collition of
    32- and 16 bit integers).

    That's no problem. I have ADSL so don't have to worry about those
    nasty ticks in the telephone :)

    Okay. :) I've cable but it's the same princip.

    If you have other problems with Squish feel free to report them.

    Of course I will :)) But don't expect too much from me: I only set
    up this point system to help one of my points who is interested in
    reading his Fido messages under Linux, instead of in a Dos-box
    under Windows.

    Okay. :)

    For myself I'm happy with mbse [a Linux-native
    all-in-one package] and GoldED+.

    I see. :) I know MBSE it's good software, but there is some things I doesn't like about it. And if I ran software made by others, I can't figure out what to
    program on. ;)

    Bo

    ---
    * Origin: <=== The Night Express ===> (2:236/100)
  • From Mike Tripp@1:382/61 to Bo Simonsen on Fri Apr 9 20:52:38 2004
    Hello Bo!

    08 Apr 04 18:34, Bo Simonsen wrote to Mike Tripp:

    They got different MSGID so squish wouldn't juge it as dupe
    normally, but there is a bug with dupe detection in the current
    release.

    Just trying to make sure he was looking for a total of 4 messages across at least 2 areas and not just comparing the info from 2 that he posted from his dup area with each other. It would be rare for 2 messages that ended up in the
    dup area to match, even when both are legitimate dups (and dup-checking is working perfectly).

    .\\ike

    --- GoldED 2.50+
    * Origin: -=( The TechnoDrome )=- Austin,TX 512-327-8598 33.6k (1:382/61)
  • From Dale Shipp@1:261/1466 to Bo Simonsen on Sat Apr 10 00:35:00 2004
    On 04-08-04 18:34, Bo Simonsen <=-
    spoke to Mike Tripp about Dupe checking <=-


    They got different MSGID so squish wouldn't juge it as dupe
    normally, but there is a bug with dupe detection in the
    current release.

    Not so. Squish has the ability to use two different sorts of
    duplicate checking. One is based on MSGID. The other is based on
    the header info (TO, FROM, SUBJ and DATE). If *either* of these two
    things is a match, then Squish calls the message a dupe. Hence if
    someone happens to send a bunch of messages to the same person, all
    with the same subject and the same date/time stamp then squish will
    call all but the first a dupe if that option is set. Since some
    systems do Y2K protection by inserting the date/time of upload, this
    can cause a problem.

    dale (at) min (dot) net
    (1:261/1466)



    ... Shipwrecked on Hesperus in Columbia, Maryland. 23:16:45, 09 Apr 2004
    ___ Blue Wave/DOS v2.30

    --- Maximus/NT 3.01
    * Origin: Owl's Anchor (1:261/1466)
  • From Hans Beekhuizen@2:280/1018.1 to Bo Simonsen on Sat Apr 10 08:33:22 2004
    Hallo Bo,

    Op 10 Apr 04 schreef Bo Simonsen aan Hans Beekhuizen:

    We did have some major problems with dupechecking and the 3.03b
    release of maximus, they are fixed in the CVS version.
    Aha :))
    We are about to release a more stable version of maximus/squish/sqafix.

    That's good :))

    1. Use the CVS version
    I just installed it, so will find out tomorrow morning if this
    problem has been solved.
    Okay, I'm 99% sure it's. :) Squish is so stable, so I've fixed all compiler warnings in it! ;)

    Of course, there were no new file announcements today... I'lllet you know as soom as I know something more, be it good or bad.

    2. Only do dupechecking on Header. In squish.cfg
    Except for the date, the headers in these messages are all the
    same. Would it work then?
    I'm not sure, I didn't get mutch dupe by running Squish without msgid check, and right now there are still people which is using Software which seems to be buyed at antique-market, which doesn't set @MSGID. So Squish will only do dupechecking on Header, and it's not caused any dupes to me.

    Well, I think yhat question is academic now that the problem is solved.

    Unfortionally can't I send you a patch
    That's no problem. I have ADSL so don't have to worry about those
    nasty ticks in the telephone :)
    Okay. :) I've cable but it's the same princip.

    ADSL came first in this village. I moved to it when my ISDN card died about 2.5 years ago.

    For myself I'm happy with mbse [a Linux-native
    all-in-one package] and GoldED+.
    I see. :) I know MBSE it's good software, but there is some things I doesn't like about it.

    It's still changing every day. Enhancements and bug fixes. The author is very
    keen. Usually when you report a bug it will be fixed within a couple of days. When I started playing with Linux and thinking about Fido software for it, I could not find anything I liked until I met mbse. There were a lot of problems
    with it then! I got a reward for reporting the most bugs :)) But I had been running Maximus and Squish under MS-Dos since March 1992!

    And if I ran software made by others, I can't figure out what to
    program on. ;)

    I Seem not to get to do much "private" programming these days. And anyway, I am anCbete, so would not be a great help. My main languages are Fortran, Pascal, COBOL and Abal.

    Groetjes,

    Hans.

    --- GoldED+/LNX 1.1.5cvs031202
    * Origin: The Wizard uses MBSE/Linux (2:280/1018.1)
  • From Hans Beekhuizen@2:280/1018.1 to Bo Simonsen on Sat Apr 10 10:16:53 2004
    Hallo Bo,

    Op 10 Apr 04 schreef Bo Simonsen aan Hans Beekhuizen:

    @PID: timEd/Linux 1.11.b1

    Ah, yes, Gerards programs have been ported to Linux as well... I was appointed
    as beta tester [three volunteers: you, you and you...] but could never get really used to it. Too much used to GoldED already :) But NetMgr and WIMM fully earned their place on my disk!

    Groetjes,

    Hans.

    --- GoldED+/LNX 1.1.5cvs031202
    * Origin: The Wizard uses MBSE/Linux (2:280/1018.1)
  • From Hans Beekhuizen@2:280/1018.1 to Mike Tripp on Sat Apr 10 10:17:42 2004
    Hallo Mike,

    Op 09 Apr 04 schreef Mike Tripp aan Bo Simonsen:

    They got different MSGID so squish wouldn't juge it as dupe
    Just trying to make sure he was looking for a total of 4 messages across at least 2 areas and not just comparing the info from 2 that he posted from his dup area with each other.

    I suppose some people would be that short sighted :) But I made certain that these messages were falsely marked as duplicates befoor I shouted boo :)) I only was not certain if this was a bug in the program or in my configuration. It was a long time ago when I last looked into the squish configuration: it was
    running nicely, so why would I bother...

    Groetjes,

    Hans.

    --- GoldED+/LNX 1.1.5cvs031202
    * Origin: The Wizard uses MBSE/Linux (2:280/1018.1)
  • From Maurice Kinal@1:153/401.1 to Dale Shipp on Sat Apr 10 09:02:24 2004
    Hey Dale!

    Apr 09 23:35 04, Dale Shipp wrote to Bo Simonsen:

    Not so. Squish has the ability to use two different sorts of
    duplicate checking. One is based on MSGID.

    Which can be duped without the actual messsage being a dupe.

    The other is based on
    the header info (TO, FROM, SUBJ and DATE). If *either* of these
    two
    things is a match, then Squish calls the message a dupe.

    The problem with this is that often some of these fields are altered such as the SUBJ field, which is quite common I've noticed. Thus an actual dupe could slip by or nondupes end up a dupes. The DATE, containing time, is based on the
    originating computer's time which can malfunction without the computer malfunctioning (ie create unique messages all with the exact same time and date). The TO and FROM can be the same without the messages being dupes.

    The only true way is to check the actual message but I wouldn't do that for every one coming through but instead if it fails any of the preliminary tests then only those would get a more rigorous test ... maybe. :-)

    The thing is that there is too much broken, unsupported stuff out there and far
    too many kludges being tacked on along the way by everyone and their dog. Personally I have gotten to the stage where I am just stripping everything out of messages, minus the TO, FROM, DATE (<- ignoring it though), and getting the originating nodenumber from the Origin. Everything else is known without jumping through hoops once they are stripped. Without compression I found that
    I average a 50% reduction in archive size when compared to normal Fido archiving methods.

    Life is good,
    Maurice

    --- Msged/LNX 6.1.1
    * Origin: Little Mikey's Brain BBS - A work in progress (1:153/401.1)
  • From Bo Simonsen@2:236/100 to Dale Shipp on Sat Apr 10 23:35:54 2004
    Dale Shipp wrote to Bo Simonsen about ''Re: Dupe checking'':

    Hi Dale,

    They got different MSGID so squish wouldn't juge it as dupe
    normally, but there is a bug with dupe detection in the
    current release.

    Not so. Squish has the ability to use two different sorts of duplicate checking. One is based on MSGID. The other is based on
    the header info (TO, FROM, SUBJ and DATE). If *either* of these
    two things is a match, then Squish calls the message a dupe.

    Yes you may be right, I've not looked specifically in the code of that area.

    Hence if someone happens to send a bunch of messages to the same
    person, all with the same subject and the same date/time stamp
    then squish will call all but the first a dupe if that option is
    set. Since some systems do Y2K protection by inserting the
    date/time of upload, this can cause a problem.

    Yes but not in Hans's issue.

    Bo

    ---
    * Origin: <=== The Night Express ===> (2:236/100)
  • From Bo Simonsen@2:236/100 to Hans Beekhuizen on Sat Apr 10 23:40:34 2004
    Hans Beekhuizen wrote to Bo Simonsen about ''Re: Dupe checking'':

    We did have some major problems with dupechecking and the 3.03b
    release of maximus, they are fixed in the CVS version.
    Aha :))
    We are about to release a more stable version of maximus/squish/sqafix.

    That's good :))

    :-)

    1. Use the CVS version
    I just installed it, so will find out tomorrow morning if this
    problem has been solved.
    Okay, I'm 99% sure it's. :) Squish is so stable, so I've fixed all
    compiler warnings in it! ;)

    Of course, there were no new file announcements today... I'lllet
    you know as soom as I know something more, be it good or bad.

    Greate.

    2. Only do dupechecking on Header. In squish.cfg
    Except for the date, the headers in these messages are all the
    same. Would it work then?
    I'm not sure, I didn't get mutch dupe by running Squish without msgid
    check, and right now there are still people which is using
    Software which
    seems to be buyed at antique-market, which doesn't set @MSGID.
    So Squish
    will only do dupechecking on Header, and it's not caused any
    dupes to me.

    Well, I think yhat question is academic now that the problem is
    solved.

    Indeed it's what I'm trying to say is that Squish is not juging many dupes if you run without MSGID testing.

    Unfortionally can't I send you a patch
    That's no problem. I have ADSL so don't have to worry about those
    nasty ticks in the telephone :)
    Okay. :) I've cable but it's the same princip.

    ADSL came first in this village. I moved to it when my ISDN card
    died about 2.5 years ago.

    Aha.

    For myself I'm happy with mbse [a Linux-native
    all-in-one package] and GoldED+.
    I see. :) I know MBSE it's good software, but there is some things I
    doesn't like about it.

    It's still changing every day. Enhancements and bug fixes. The
    author is very keen. Usually when you report a bug it will be
    fixed within a couple of days. When I started playing with Linux
    and thinking about Fido software for it, I could not find anything
    I liked until I met mbse.

    There are fundemental things I doesn't like about MBSE:

    1. It's using JAM, anyhow I could live with that.
    2. It's using UNIX accounts for users (I can't live with that).
    3. It's uses Binary configuration files (I can't live with that).

    - Maximus is more smart it's "compiles" a plain ASCII configuration file
    and take it from there.

    There were a lot of problems with it
    then! I got a reward for reporting the most bugs :)) But I had
    been running Maximus and Squish under MS-Dos since March 1992!

    Okay. :) It's good and very stable software even on UNIX! :)

    And if I ran software made by others, I can't figure out what to
    program on. ;)

    I Seem not to get to do much "private" programming these days. And anyway, I am anCbete, so would not be a great help. My main
    languages are Fortran, Pascal, COBOL and Abal.

    Aha, I do Pascal too, but not mutch latest, mostly C. :)

    I allways wanted to learn Fortran but I'm not studying a place where they are using that.

    Bo

    ---
    * Origin: <=== The Night Express ===> (2:236/100)
  • From Bo Simonsen@2:236/100 to Hans Beekhuizen on Sat Apr 10 23:41:40 2004
    Hans Beekhuizen wrote to Bo Simonsen about ''Re: Dupe checking'':

    @PID: timEd/Linux 1.11.b1

    Ah, yes, Gerards programs have been ported to Linux as well... I
    was appointed as beta tester [three volunteers: you, you and
    you...] but could never get really used to it.

    Yes timEd is very good for Fidonet purpose, but there is some missings then it gets to using Gated Newsgroups.

    Too much used to
    GoldED already :) But NetMgr and WIMM fully earned their place on
    my disk!

    I did port NetMgr to UNIX no big deal, Gerard know how to write portable code! :)

    Bo

    ---
    * Origin: <=== The Night Express ===> (2:236/100)
  • From Dale Shipp@1:261/1466 to Bo Simonsen on Sun Apr 11 02:30:00 2004
    On 04-10-04 22:35, Bo Simonsen <=-
    spoke to Dale Shipp about Dupe checking <=-

    duplicate checking. One is based on MSGID. The other is based on
    the header info (TO, FROM, SUBJ and DATE). If *either* of these
    two things is a match, then Squish calls the message a dupe.

    Yes you may be right, I've not looked specifically in the code of that area.

    Nor have I -- that was quoting from the manual.

    Because of the fact that the header checking can lead to false
    duplicate detection, I have long ago turned that off and only use
    MSGID checking. Of course, there are a few BBSes out there that do
    not use MSGIDs.


    dale (at) min (dot) net
    (1:261/1466)


    ... Shipwrecked on Hesperus in Columbia, Maryland. 00:55:06, 11 Apr 2004
    ___ Blue Wave/DOS v2.30

    --- Maximus/NT 3.01
    * Origin: Owl's Anchor (1:261/1466)
  • From Johannes Beekhuizen@2:280/1018 to Dale Shipp on Tue Apr 13 09:52:33 2004
    Hallo Dale,

    Op 09 Apr 04 schreef Dale Shipp aan Bo Simonsen:

    One is based on MSGID. The other is based on the header info (TO,
    FROM, SUBJ and DATE). If *either* of these two things is a match,
    then Squish calls the message a dupe.

    That's not how I understand it. Allow me to quote from the manual:

    === import ===

    DupeCheck [<type>...]

    The DupeCheck keyword controls the dupe-checking algorithm
    used by Squish:

    <type> can be either or both of "Header" or "MSGID".

    "Header" instructs Squish to check the message header to
    determine whether or not a message is a dupe. Squish will
    hash the "To", "From" and "Subject" fields into a 32-bit
    identifier. It will append the message date to this,
    resulting in a 64-bit duplicate identifier.

    "MSGID" instructs Squish to check the MSGID kludge to
    determine whether or not a message is a dupe. Squish will
    hash the text of the MSGID "address" field into a 32-bit
    identifier. It will append the MSGID serial number to this,
    resulting in a 64-bit duplicate identifier.

    If only one of the above settings is enabled, Squish will
    only use that method when determining whether or not a
    message is a dupe.

    However, if both MSGID and Header are specified, Squish will
    perform both checks. If EITHER the MSGID or the header is
    duplicated, Squish will declare the message to be a dupe.

    === tropmi ===

    I read is as: If and only if both Header and MsgID are the same, the message is a dupe. Which seems sensible to me. That, at least in the last Linux release, it didn't work is a different matter.

    Groetjes,

    Hans.

    --- GoldED+/LNX 1.1.5cvs031202
    * Origin: The Wizard uses MBSE/Linux (2:280/1018)
  • From Bo Simonsen@2:236/100 to Dale Shipp on Tue Apr 13 20:30:10 2004
    Hello Dale.

    11 Apr 04 01:30, you wrote to me:

    duplicate checking. One is based on MSGID. The other is based
    on
    the header info (TO, FROM, SUBJ and DATE). If *either* of
    these two things is a match, then Squish calls the message a
    dupe.

    Yes you may be right, I've not looked specifically in the code of
    that area.

    Nor have I -- that was quoting from the manual.

    Okay! :)

    Because of the fact that the header checking can lead to false
    duplicate detection, I have long ago turned that off and only use
    MSGID checking. Of course, there are a few BBSes out there that do
    not use MSGIDs.

    Or fidonet mail systems like D'Bridge! :) Which is only used by two persons
    on Fidonet...

    Bo

    --- GoldED+/LNX 1.1.5
    * Origin: The Night Express - Roennede, Dk (2:236/100)
  • From Dale Shipp@1:261/1466 to Johannes Beekhuizen on Wed Apr 14 02:37:00 2004
    On 04-13-04 08:52, Johannes Beekhuizen <=-
    spoke to Dale Shipp about Re: Dupe checking <=-

    Hallo Dale,

    Op 09 Apr 04 schreef Dale Shipp aan Bo Simonsen:

    One is based on MSGID. The other is based on the header info (TO,
    FROM, SUBJ and DATE). If *either* of these two things is a match,
    then Squish calls the message a dupe.

    That's not how I understand it. Allow me to quote from the manual:

    Then you understand wrong. Read the very two last lines in what you
    quoted. It is quite explicit.

    If EITHER the MSGID or the
    ^^^^^^
    header is duplicated, Squish will declare the message to
    be a dupe.

    That means that you do *not* need both, either one will cause the
    message to be declared a dupe.

    IMO, it would be better if it declared the message to be a dupe only
    if both items said so, but that is not the case.

    === import ===

    DupeCheck [<type>...]

    The DupeCheck keyword controls the dupe-checking
    algorithm used by Squish:

    <type> can be either or both of "Header" or "MSGID".

    "Header" instructs Squish to check the message header
    to determine whether or not a message is a dupe. Squish
    will hash the "To", "From" and "Subject" fields into a
    32-bit identifier. It will append the message date to
    this, resulting in a 64-bit duplicate identifier.

    "MSGID" instructs Squish to check the MSGID kludge to
    determine whether or not a message is a dupe. Squish
    will hash the text of the MSGID "address" field into a
    32-bit identifier. It will append the MSGID serial
    number to this, resulting in a 64-bit duplicate
    identifier.
    If only one of the above settings is enabled, Squish
    will only use that method when determining whether or
    not a message is a dupe.

    However, if both MSGID and Header are specified, Squish
    will perform both checks. If EITHER the MSGID or the
    header is duplicated, Squish will declare the message to
    be a dupe.
    === tropmi ===

    I read is as: If and only if both Header and MsgID are the same, the message is a dupe. Which seems sensible to me. That, at least in the last Linux release, it didn't work is a different matter.

    Groetjes,

    Hans.

    -!- GoldED+/LNX 1.1.5cvs031202
    ! Origin: The Wizard uses MBSE/Linux (2:280/1018)



    ... Shipwrecked on Hesperus in Columbia, Maryland. 01:03:40, 14 Apr 2004
    ___ Blue Wave/DOS v2.30

    --- Maximus/NT 3.01
    * Origin: Owl's Anchor (1:261/1466)
  • From Mike Tripp@1:382/61 to Johannes Beekhuizen on Wed Apr 14 08:55:08 2004
    Hello Johannes!

    13 Apr 04 08:52, Johannes Beekhuizen wrote to Dale Shipp:


    perform both checks. If EITHER the MSGID or the header is
    duplicated, Squish will declare the message to be a dupe.

    === tropmi ===

    I read is as: If and only if both Header and MsgID are the same, the message is a dupe.

    Check again: it is a dupe if it fails either test, not both.

    .\\ike

    --- GoldED 2.50+
    * Origin: -=( The TechnoDrome )=- Austin,TX 512-327-8598 33.6k (1:382/61)