From mboxrd@z Thu Jan 1 00:00:00 1970 X-Msuck: nntp://news.gmane.io/gmane.emacs.gnus.general/66377 Path: news.gmane.org!not-for-mail From: Ted Zlatanov Newsgroups: gmane.emacs.gnus.general Subject: Re: split-fancy and gnus-registry confusion Date: Thu, 28 Feb 2008 08:32:43 -0600 Organization: =?utf-8?B?0KLQtdC+0LTQvtGAINCX0LvQsNGC0LDQvdC+0LI=?= @ Cienfuegos Message-ID: <86ir09p2wk.fsf@lifelogs.com> References: <86y796owf6.fsf@lifelogs.com> <763ardp3zh.fsf@dev-d01.ppllc.com> NNTP-Posting-Host: lo.gmane.org Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Trace: ger.gmane.org 1204209114 9754 80.91.229.12 (28 Feb 2008 14:31:54 GMT) X-Complaints-To: usenet@ger.gmane.org NNTP-Posting-Date: Thu, 28 Feb 2008 14:31:54 +0000 (UTC) Cc: Norman Walsh , ding@gnus.org To: Jake Colman Original-X-From: ding-owner+M14868@lists.math.uh.edu Thu Feb 28 15:32:17 2008 Return-path: Envelope-to: ding-account@gmane.org Original-Received: from util0.math.uh.edu ([129.7.128.18]) by lo.gmane.org with esmtp (Exim 4.50) id 1JUjnp-0002AZ-5k for ding-account@gmane.org; Thu, 28 Feb 2008 15:32:05 +0100 Original-Received: from localhost ([127.0.0.1] helo=lists.math.uh.edu) by util0.math.uh.edu with smtp (Exim 4.63) (envelope-from ) id 1JUjn6-0007dy-9z; Thu, 28 Feb 2008 08:31:20 -0600 Original-Received: from mx1.math.uh.edu ([129.7.128.32]) by util0.math.uh.edu with esmtps (TLSv1:AES256-SHA:256) (Exim 4.63) (envelope-from ) id 1JUjn4-0007dc-BY for ding@lists.math.uh.edu; Thu, 28 Feb 2008 08:31:18 -0600 Original-Received: from quimby.gnus.org ([80.91.231.51]) by mx1.math.uh.edu with esmtp (Exim 4.67) (envelope-from ) id 1JUjmx-0005RY-SP for ding@lists.math.uh.edu; Thu, 28 Feb 2008 08:31:18 -0600 Original-Received: from mail.blockstar.com ([170.224.69.95]) by quimby.gnus.org with esmtp (Exim 3.35 #1 (Debian)) id 1JUjn3-0004VI-00 for ; Thu, 28 Feb 2008 15:31:17 +0100 Original-Received: from tzlatanov-ubuntu-desktop.jumptrading.com (unknown [38.98.147.130]) by mail.blockstar.com (Postfix) with ESMTP id 6A5733E8081; Thu, 28 Feb 2008 06:31:52 -0800 (PST) X-Face: bd.DQ~'29fIs`T_%O%C\g%6jW)yi[zuz6;d4V0`@y-~$#3P_Ng{@m+e4o<4P'#(_GJQ%TT= D}[Ep*b!\e,fBZ'j_+#"Ps?s2!4H2-Y"sx" X-Hashcash: 1:20:080228:ndw@nwalsh.com::0R1WHNbDXjuUHurm:0002kqh X-Hashcash: 1:20:080228:colman@ppllc.com::Yx10BVHur8AZt/Pi:03AY6 X-Hashcash: 1:20:080228:ding@gnus.org::2+QaHjjlMBl9PLeM:00005k7K In-Reply-To: <763ardp3zh.fsf@dev-d01.ppllc.com> (Jake Colman's message of "Thu, 28 Feb 2008 09:09:22 -0500") User-Agent: Gnus/5.110007 (No Gnus v0.7) Emacs/23.0.50 (gnu/linux) X-Spam-Score: -2.6 (--) List-ID: Precedence: bulk Xref: news.gmane.org gmane.emacs.gnus.general:66377 Archived-At: On Thu, 28 Feb 2008 09:09:22 -0500 Jake Colman wrote: >>>>>> "TZ" == Ted Zlatanov writes: TZ> The attached patch against today's CVS will give you proper TZ> logging, plus I reworked the function TZ> gnus-registry-split-fancy-with-parent to be much nicer. Make TZ> sure you have gnus-verbose turned up to 9, then try this out and TZ> let me know if the logging helped, and if things generally worked TZ> OK. TZ> I am reworking both logging and splitting, so comment on both TZ> please. Anyone else interested in testing, please give it a TZ> shot. TZ> Ted JC> I am up and running with gnus from 20080228 with your patch. It must be JC> working because the registry has got a heck of alot more to say! JC> Below are two examples: one tha(I guess) did as it should and the other JC> complaining about extra matches. What exactly is this telling me? How JC> do I read this? JC> gnus-registry-split-fancy-with-parent (extra tracking) traced sender "Rafiq Ahmed" to groups (INBOX/Principia/StateControl INBOX/Principia/SFP INBOX/Principia/Management INBOX/Principia/SFP INBOX/Principia/SFP INBOX/Principia/StateControl INBOX/Principia/SFP INBOX/Principia/SFP INBOX/Principia/SFP INBOX/Principia/Management INBOX/Principia/SFP INBOX/Principia/StateControl INBOX/Principia/StateControl INBOX/Principia/StateControl INBOX/Principia/StateControl INBOX/Principia/SFP INBOX/Principia/StateControl INBOX/Principia/SFP INBOX/Principia/StateControl INBOX/Principia/StateControl) (keys ()) JC> gnus-registry-post-process-group: too many extra matches ((INBOX/Principia/StateControl INBOX/Principia/SFP INBOX/Principia/Management INBOX/Principia/SFP INBOX/Principia/SFP INBOX/Principia/StateControl INBOX/Principia/SFP INBOX/Principia/SFP INBOX/Principia/SFP INBOX/Principia/Management INBOX/Principia/SFP INBOX/Principia/StateControl INBOX/Principia/StateControl INBOX/Principia/StateControl INBOX/Principia/StateControl INBOX/Principia/SFP INBOX/Principia/StateControl INBOX/Principia/SFP INBOX/Principia/StateControl INBOX/Principia/StateControl)) for sender "Rafiq Ahmed" . Returning none. I think this is one splitting example with two output lines. The registry found matches by sender, then in gnus-registry-post-process-group the matches were filtered. The post-processing says that the sender has multiple messages, whose registry entries point to several different groups. Currently the registry post-processing just gives up if that happens, so you don't end up with inconsistent splitting. I could change that logic, since I've reworked the splitting code anyhow. Maybe the user could set a strategy: - give up on multiples (current) - select first group alphabetically or some other stable sort method (predictable, easy to understand, easy to implement, not necessarily accurate) - select a group if it's "preferred" (configuration burden is on the user) - ask user (annoying if it happens a lot, but the "preferred" list of groups could be learned dynamically this way) - some mix of the above - other ideas? Ted