reading standard digests

reading standard digests

[Dick Knowles]

I noticed a change of format from gnus to pgnus in the nndoc or
nnvirtual buffer when I read a digest.  For example, I marked 4
digests from the same mailing list, then typed M-C-d.  

In gnus v5.6.45 I got this nnvirtual summary buffer:

E -[  27(    0): nobody            ]Jan    0:  (none)
E -[  22(    0): nobody            ]Jan    0:  (none)
E -[  22(    0): nobody            ]Jan    0:  (none)
E -[  18(    0): nobody            ]Jan    0:  (none)
E  [  16(    0): Frank and Lois Ric]  23-Oct:  [PERRY-L] Address for Granny Peach, and a challenge
   [  10(    0): Matthews Family   ]  24-Oct:  [PERRY-L] the meaning of the name PERRY
       <  39(    0): Michael Perry     >  26-Oct:  
   [  57(    0): James Fremon      ]  24-Oct:  [PERRY-L] Perrys in the VT Chronical
   [  49(    0): Martha Veselka    ]  24-Oct:  [PERRY-L] Confederate Indigent Families Lists of Texas 1863-1865. 
   [  10(    0): Matthews Family   ]  25-Oct:  [PERRY-L] Does anyone have a scanned picture of the Perry Coat of Arms?
   [  15(    0): Matthews Family   ]  25-Oct:  [PERRY-L] update on my Perrys
   [  13(    0): Rita Ashley Fishma]  25-Oct:  Christinana PERRY b, 1805
   [  21(    0): Ron McCandless    ]  25-Oct:  [PERRY-L] Was your Perry in Jail?
   [  35(    0): Myline99@aol.com  ]  25-Oct:  [PERRY-L] Picture of Civil War ancestors
   [  60(    0): Frank and Lois Ric]  26-Oct:  [PERRY-L] Re: PERRY Help Please
   [  10(    0): Matthews Family   ]  26-Oct:  [PERRY-L] A big thankyou- re coat of arms and Perry meaning
   [  14(    0): Kenna Matthews    ]  26-Oct:  [PERRY-L] I have plenty of JPEGS re Coat of Arms if anyone interested!!!
   [  14(    0): David & Eunice Bou]  26-Oct:  [PERRY-L] Coats of Arms
       <  13(    0): David & Eunice Bou>  26-Oct:  
       <  16(    0): Debbi Geer        >  26-Oct:  
   [  49(    0): Ron McCandless    ]  26-Oct:  [PERRY-L] inherited Coat of Arms vs NOT inherited
   [  44(    0): James Fremon      ]  26-Oct:  [PERRY-L] Perry soldiers in Civil War photos
   [   6(    0): kjvernon          ]  27-Oct:  [PERRY-L] PERRY /GOULBURN AREA
   [   9(    0): Matthews Family   ]  27-Oct:  [PERRY-L] re Jpegs for coat of arms
   [   7(    0): Dan W. Perry      ]  27-Oct:  [PERRY-L] Perry Family in Maryland

In pgnus v0.97 I got this nnvirtual summary buffer:

   [  17(    0): Frank and Lois Ric]  23-Oct:  [PERRY-L] Address for Granny Peach, and a challenge
   [  10(    0): Matthews Family   ]  24-Oct:  [PERRY-L] the meaning of the name PERRY
       <  57(    0): James Fremon      >  24-Oct:  [PERRY-L] Perrys in the VT Chronical
       <  49(    0): Martha Veselka    >  24-Oct:  [PERRY-L] Confederate Indigent Families Lists of Texas 1863-1865. 
       <  10(    0): Matthews Family   >  24-Oct:  [PERRY-L] Does anyone have a scanned picture of the Perry Coat of Arms?
       <  16(    0): Matthews Family   >  24-Oct:  [PERRY-L] update on my Perrys
   [  13(    0): Rita Ashley Fishma]  25-Oct:  Christinana PERRY b, 1805
       <  21(    0): Ron McCandless    >  25-Oct:  [PERRY-L] Was your Perry in Jail?
       <  35(    0): Myline99@aol.com  >  25-Oct:  [PERRY-L] Picture of Civil War ancestors
       <  60(    0): Frank and Lois Ric>  26-Oct:  [PERRY-L] Re: PERRY Help Please
       <  10(    0): Matthews Family   >  26-Oct:  [PERRY-L] A big thankyou- re coat of arms and Perry meaning
       <  14(    0): Kenna Matthews    >  26-Oct:  [PERRY-L] I have plenty of JPEGS re Coat of Arms if anyone interested!!!
       <  13(    0): David & Eunice Bou>  26-Oct:  [PERRY-L] Coats of Arms
       <  16(    0): Debbi Geer        >  26-Oct:  
       <  39(    0): Michael Perry     >  26-Oct:  [PERRY-L] the meaning of the name PERRY
       <  50(    0): Ron McCandless    >  26-Oct:  [PERRY-L] inherited Coat of Arms vs NOT inherited
   [  14(    0): David & Eunice Bou]  26-Oct:  [PERRY-L] Coats of Arms
       <  44(    0): James Fremon      >  26-Oct:  [PERRY-L] Perry soldiers in Civil War photos
       <   6(    0): kjvernon          >  27-Oct:  [PERRY-L] PERRY /GOULBURN AREA
       <   9(    0): Matthews Family   >  27-Oct:  [PERRY-L] re Jpegs for coat of arms
       <   8(    0): Dan W. Perry      >  27-Oct:  [PERRY-L] Perry Family in Maryland

I like the older style.  Each article of the digest is treated
individually and they are threaded based on the individual article
subjects.  In pgnus, each digest is treated as a single thread.  Also
the article buffers are different.  In gnus, I only see the actual
article headers, but in pgnus each article begins with the digest
headers, followed by the article.  Maybe this article splitting is why
the summary and threading work the way they do?  

So, are there variables I need to know to restore the previous
treatment of digests?  Thanks.

		       Dick

-- 
    Dick Knowles                  knowles@averstar.com
    AverStar, Inc.                http://www.averstar.com/~knowles
    23 Fourth Avenue              Phone: 781-221-6990
    Burlington, MA 01803-3303     FAX:   781-221-6991

Re: reading standard digests

[Lars Magne Ingebrigtsen]

Dick Knowles <knowles@averstar.com> writes:

> So, are there variables I need to know to restore the previous
> treatment of digests?  Thanks.

I'm guessing that these really are MIME digests, and not standard
digests.  In that case, you should force the standard-digest type on
the component nndoc groups.

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: reading standard digests

[Dick Knowles]

>>>>> On 07 Nov 1999 01:10:31 +0100, "Lars" == Lars Magne Ingebrigtsen  wrote:

  Lars> Dick Knowles <knowles@averstar.com> writes:
  >> So, are there variables I need to know to restore the previous
  >> treatment of digests?  Thanks.

  Lars> I'm guessing that these really are MIME digests, and not standard
  Lars> digests.  In that case, you should force the standard-digest type on
  Lars> the component nndoc groups.

MIME digest seems right, but according to some feedback from Shenghuo
Zhu (and nndoc.el itself), 'mime-digest' is gone from pgnus.  I stole
the 'mime-digest' code from gnus v5.6.45 for nndoc-type-alist and
nndoc-mime-digest-type-p, and then these digests work as I like.  In
the unmodified pgnus v0.97, neither mime-parts nor standard-digest did
the right thing.  Is there some other digest type in pgnus that will
work correctly?  Or should mime-digest be put back in?

                 Dick

-- 
    Dick Knowles                  knowles@averstar.com
    AverStar, Inc.                http://www.averstar.com/~knowles
    23 Fourth Avenue              Phone: 781-221-6990
    Burlington, MA 01803-3303     FAX:   781-221-6991

Re: reading standard digests

[Lars Magne Ingebrigtsen]

Dick Knowles <knowles@averstar.com> writes:

> In the unmodified pgnus v0.97, neither mime-parts nor
> standard-digest did the right thing.

Well, that depends on what you mean by "right thing".  Perhaps the
mime-parts interpretation of MIME digests is not the optimal.

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: reading standard digests

[Dick Knowles]

[Oops.  On previous send I replied directly to Lars rather than back
to the list.  Sorry Lars.]

>>>>> On 11 Nov 1999 07:26:07 +0100, "Lars" == Lars Magne Ingebrigtsen  wrote:

  Lars> Dick Knowles <knowles@averstar.com> writes:
  >> In the unmodified pgnus v0.97, neither mime-parts nor
  >> standard-digest did the right thing.

  Lars> Well, that depends on what you mean by "right thing".  Perhaps the
  Lars> mime-parts interpretation of MIME digests is not the optimal.

The "right thing" for me is just how gnus v5.6.45 mime-digest acts, at
least from my perspective as a user: each nndoc article from the
digest has only its own, original headers.  In pgnus, each nndoc
article has the main digest headers, followed by the original article
headers.  In gnus, the nndoc summary shows the original article date,
subject, and author; if threading is on, the threads work from the
original article headers- same subjects or references thread together.
In pgnus, the summary line is correct, but threading makes a single
digest one thread- all articles from that digest are children of the
digest.  If several actual articles are part of the same thread- in
one digest or across multiple digests- this is ignored.

           Dick

Re: reading standard digests

[Lars Magne Ingebrigtsen]

Dick Knowles <knowles@averstar.com> writes:

> The "right thing" for me is just how gnus v5.6.45 mime-digest acts, at
> least from my perspective as a user: each nndoc article from the
> digest has only its own, original headers. 

I've reinstated the old mime-digest type in the CVS.

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: reading standard digests

[Dick Knowles]

>>>>> On 15 Nov 1999 21:44:37 +0100, "Lars" == Lars Magne Ingebrigtsen  wrote:


  Lars> I've reinstated the old mime-digest type in the CVS.

Thanks!!

                Dick

Re: Reading standard digests

[Shenghuo ZHU]

>>>>> "Dick" == Dick Knowles <knowles@averstar.com> writes:


[...]


Dick> I checked nndoc-article-type: in pgnus, it's mime-parts; in gnus
Dick> it's mime-digest.  As an experiment, I stole the mime-digest
Dick> piece of nndoc-type-alist from gnus and added it to my running
Dick> pgnus, and also stole and loaded nndoc-mime-digest-type-p.  When
Dick> I did this, the mime-digest (no longer mime-parts) worked as it
Dick> had in gnus.  Articles had their own headers only, not the
Dick> digest headers followed by article headers; and the summary and
Dick> threading worked as before.  So why was mime-digest removed?  It
Dick> appears that mime-digest and mime-parts dismantle the digest
Dick> differently.  What can I expect to break with my little
Dick> experiment, or can I safely make this change locally?

I don't know why Lars removed MIME-digest in February. 

I guess it should be safe.

I think a customizable preference variable should be added, since Gnus
users are always picky.

Dick> In case this helps explain anything, here are the headers and a bit
Dick> more from one of the digests I'm getting:

Dick>    X-From-Line: PERRY-D-request@rootsweb.com Wed Nov  3 00:48 EST 1999
Dick>    Received: from [...]
Dick>    From: PERRY-D-request@rootsweb.com
Dick>    Date: Tue, 2 Nov 1999 21:46:37 -0800 (PST)
Dick>    X-Gnus-Mail-Source: file:/var/mail/knowles
Dick>    Message-Id: <199911030546.VAA16738@bl-14.rootsweb.com>
Dick>    Subject: PERRY-D Digest V99 #235
Dick>    X-Loop: PERRY-D@rootsweb.com
Dick>    X-Mailing-List: <PERRY-D@rootsweb.com> archive/volume99/235
Dick>    Precedence: list
Dick>    MIME-Version: 1.0
Dick>    To: PERRY-D@rootsweb.com
Dick>    Reply-To: PERRY-L@rootsweb.com
Dick>    Content-Type: multipart/digest; boundary="----------------------------"
Dick>    X-Content-Length: 3367
Dick>    Lines: 113
Dick>    Xref: mesmer.i2ada mail.roots:1029

[...]

multipart/digest help a lot. While, in some digests
multipart/mixed+message/rfc822 is used instead, such as
comp.sys.mac.digest.

-- 
Shenghuo ZHU

Re: Reading standard digests

[Dick Knowles]

>>>>> On 03 Nov 1999 10:56:15 -0500, "Shenghuo ZHU" == Shenghuo ZHU  wrote:

>>>>> "Dick" == Dick Knowles <knowles@averstar.com> writes:
  [...]

  Shenghuo ZHU> Which type is the digest article? For standard-digest,
  Shenghuo ZHU> it should still work.  Lars removed the mime-digest
  Shenghuo ZHU> type, and mime-parts is used instead. The digest
  Shenghuo ZHU> article you list is not like mime-digest. You find out
  Shenghuo ZHU> the type by typing `C-h v nndoc-article-type' in the
  Shenghuo ZHU> summary.

  Dick> So, are there variables I need to know to restore the previous
  Dick> treatment of digests?  Thanks.

  Shenghuo ZHU> I guess no variable for the current version of pGnus.

I checked nndoc-article-type: in pgnus, it's mime-parts; in gnus it's
mime-digest.  As an experiment, I stole the mime-digest piece of
nndoc-type-alist from gnus and added it to my running pgnus, and also
stole and loaded nndoc-mime-digest-type-p.  When I did this, the
mime-digest (no longer mime-parts) worked as it had in gnus.  Articles
had their own headers only, not the digest headers followed by article
headers; and the summary and threading worked as before.  So why was
mime-digest removed?  It appears that mime-digest and mime-parts
dismantle the digest differently.  What can I expect to break with my
little experiment, or can I safely make this change locally?

In case this helps explain anything, here are the headers and a bit
more from one of the digests I'm getting:

   X-From-Line: PERRY-D-request@rootsweb.com Wed Nov  3 00:48 EST 1999
   Received: from [...]
   From: PERRY-D-request@rootsweb.com
   Date: Tue, 2 Nov 1999 21:46:37 -0800 (PST)
   X-Gnus-Mail-Source: file:/var/mail/knowles
   Message-Id: <199911030546.VAA16738@bl-14.rootsweb.com>
   Subject: PERRY-D Digest V99 #235
   X-Loop: PERRY-D@rootsweb.com
   X-Mailing-List: <PERRY-D@rootsweb.com> archive/volume99/235
   Precedence: list
   MIME-Version: 1.0
   To: PERRY-D@rootsweb.com
   Reply-To: PERRY-L@rootsweb.com
   Content-Type: multipart/digest; boundary="----------------------------"
   X-Content-Length: 3367
   Lines: 113
   Xref: mesmer.i2ada mail.roots:1029

   ------------------------------

   Content-Type: text/plain

   PERRY-D Digest				Volume 99 : Issue 235

   Today's Topics:
     #1 [PERRY-L] Re: PERRY-D Digest V99 #   [L3udGen@aol.com]
     #2 [PERRY-L] David and James Perry      [fiedlerro@juno.com]
     #3 [PERRY-L] Harrison County Texas PE   [billtodd@pipeline.com]


                   Dick

-- 
    Dick Knowles                  knowles@averstar.com
    AverStar, Inc.                http://www.averstar.com/~knowles
    23 Fourth Avenue              Phone: 781-221-6990
    Burlington, MA 01803-3303     FAX:   781-221-6991

Re: Reading standard digests

[Shenghuo ZHU]

>>>>> "Dick" == Dick Knowles <knowles@averstar.com> writes:


[...]


Dick> I noticed a change of format from gnus to pgnus in the nndoc or
Dick> nnvirtual buffer when I read a digest.  For example, I marked 4
Dick> digests from the same mailing list, then typed M-C-d.  

[...]


Dick> I like the older style.  Each article of the digest is treated
Dick> individually and they are threaded based on the individual
Dick> article subjects.  In pgnus, each digest is treated as a single
Dick> thread.  Also the article buffers are different.  In gnus, I
Dick> only see the actual article headers, but in pgnus each article
Dick> begins with the digest headers, followed by the article.  Maybe
Dick> this article splitting is why the summary and threading work the
Dick> way they do?

Which type is the digest article? For standard-digest, it should still
work.  Lars removed the mime-digest type, and mime-parts is used
instead. The digest article you list is not like mime-digest. You find
out the type by typing `C-h v nndoc-article-type' in the summary.

Dick> So, are there variables I need to know to restore the previous
Dick> treatment of digests?  Thanks.

I guess no variable for the current version of pGnus.

[...]

-- 
Shenghuo ZHU

Reading standard digests

[Dick Knowles]

[Sorry if this appears more than once.  I tried posting before I was
subscribed to the list, and I don't think it made it.  But, maybe it
did! ]

I noticed a change of format from gnus to pgnus in the nndoc or
nnvirtual buffer when I read a digest.  For example, I marked 4
digests from the same mailing list, then typed M-C-d.  

In gnus v5.6.45 I got this nnvirtual summary buffer:

E -[  27(    0): nobody            ]Jan    0:  (none)
E -[  22(    0): nobody            ]Jan    0:  (none)
E -[  22(    0): nobody            ]Jan    0:  (none)
E -[  18(    0): nobody            ]Jan    0:  (none)
E  [  16(    0): Frank and Lois Ric]  23-Oct:  [PERRY-L] Address for Granny Peach, and a challenge
   [  10(    0): Matthews Family   ]  24-Oct:  [PERRY-L] the meaning of the name PERRY
       <  39(    0): Michael Perry     >  26-Oct:  
   [  57(    0): James Fremon      ]  24-Oct:  [PERRY-L] Perrys in the VT Chronical
   [  49(    0): Martha Veselka    ]  24-Oct:  [PERRY-L] Confederate Indigent Families Lists of Texas 1863-1865. 
   [  10(    0): Matthews Family   ]  25-Oct:  [PERRY-L] Does anyone have a scanned picture of the Perry Coat of Arms?
   [  15(    0): Matthews Family   ]  25-Oct:  [PERRY-L] update on my Perrys
   [  13(    0): Rita Ashley Fishma]  25-Oct:  Christinana PERRY b, 1805
   [  21(    0): Ron McCandless    ]  25-Oct:  [PERRY-L] Was your Perry in Jail?
   [  35(    0): Myline99@aol.com  ]  25-Oct:  [PERRY-L] Picture of Civil War ancestors
   [  60(    0): Frank and Lois Ric]  26-Oct:  [PERRY-L] Re: PERRY Help Please
   [  10(    0): Matthews Family   ]  26-Oct:  [PERRY-L] A big thankyou- re coat of arms and Perry meaning
   [  14(    0): Kenna Matthews    ]  26-Oct:  [PERRY-L] I have plenty of JPEGS re Coat of Arms if anyone interested!!!
   [  14(    0): David & Eunice Bou]  26-Oct:  [PERRY-L] Coats of Arms
       <  13(    0): David & Eunice Bou>  26-Oct:  
       <  16(    0): Debbi Geer        >  26-Oct:  
   [  49(    0): Ron McCandless    ]  26-Oct:  [PERRY-L] inherited Coat of Arms vs NOT inherited
   [  44(    0): James Fremon      ]  26-Oct:  [PERRY-L] Perry soldiers in Civil War photos
   [   6(    0): kjvernon          ]  27-Oct:  [PERRY-L] PERRY /GOULBURN AREA
   [   9(    0): Matthews Family   ]  27-Oct:  [PERRY-L] re Jpegs for coat of arms
   [   7(    0): Dan W. Perry      ]  27-Oct:  [PERRY-L] Perry Family in Maryland

In pgnus v0.97 I got this nnvirtual summary buffer:

   [  17(    0): Frank and Lois Ric]  23-Oct:  [PERRY-L] Address for Granny Peach, and a challenge
   [  10(    0): Matthews Family   ]  24-Oct:  [PERRY-L] the meaning of the name PERRY
       <  57(    0): James Fremon      >  24-Oct:  [PERRY-L] Perrys in the VT Chronical
       <  49(    0): Martha Veselka    >  24-Oct:  [PERRY-L] Confederate Indigent Families Lists of Texas 1863-1865. 
       <  10(    0): Matthews Family   >  24-Oct:  [PERRY-L] Does anyone have a scanned picture of the Perry Coat of Arms?
       <  16(    0): Matthews Family   >  24-Oct:  [PERRY-L] update on my Perrys
   [  13(    0): Rita Ashley Fishma]  25-Oct:  Christinana PERRY b, 1805
       <  21(    0): Ron McCandless    >  25-Oct:  [PERRY-L] Was your Perry in Jail?
       <  35(    0): Myline99@aol.com  >  25-Oct:  [PERRY-L] Picture of Civil War ancestors
       <  60(    0): Frank and Lois Ric>  26-Oct:  [PERRY-L] Re: PERRY Help Please
       <  10(    0): Matthews Family   >  26-Oct:  [PERRY-L] A big thankyou- re coat of arms and Perry meaning
       <  14(    0): Kenna Matthews    >  26-Oct:  [PERRY-L] I have plenty of JPEGS re Coat of Arms if anyone interested!!!
       <  13(    0): David & Eunice Bou>  26-Oct:  [PERRY-L] Coats of Arms
       <  16(    0): Debbi Geer        >  26-Oct:  
       <  39(    0): Michael Perry     >  26-Oct:  [PERRY-L] the meaning of the name PERRY
       <  50(    0): Ron McCandless    >  26-Oct:  [PERRY-L] inherited Coat of Arms vs NOT inherited
   [  14(    0): David & Eunice Bou]  26-Oct:  [PERRY-L] Coats of Arms
       <  44(    0): James Fremon      >  26-Oct:  [PERRY-L] Perry soldiers in Civil War photos
       <   6(    0): kjvernon          >  27-Oct:  [PERRY-L] PERRY /GOULBURN AREA
       <   9(    0): Matthews Family   >  27-Oct:  [PERRY-L] re Jpegs for coat of arms
       <   8(    0): Dan W. Perry      >  27-Oct:  [PERRY-L] Perry Family in Maryland

I like the older style.  Each article of the digest is treated
individually and they are threaded based on the individual article
subjects.  In pgnus, each digest is treated as a single thread.  Also
the article buffers are different.  In gnus, I only see the actual
article headers, but in pgnus each article begins with the digest
headers, followed by the article.  Maybe this article splitting is why
the summary and threading work the way they do?  

So, are there variables I need to know to restore the previous
treatment of digests?  Thanks.

		       Dick

-- 
    Dick Knowles                  knowles@averstar.com
    AverStar, Inc.                http://www.averstar.com/~knowles
    23 Fourth Avenue              Phone: 781-221-6990
    Burlington, MA 01803-3303     FAX:   781-221-6991