From mboxrd@z Thu Jan 1 00:00:00 1970 X-Msuck: nntp://news.gmane.io/gmane.emacs.gnus.general/34932 Path: main.gmane.org!not-for-mail From: prj@po.cwru.edu (Paul Jarc) Newsgroups: gmane.emacs.gnus.general Subject: Re: regexp-based group parameters Date: 23 Feb 2001 15:54:03 -0500 Sender: owner-ding@hpc.uh.edu Message-ID: References: <873dfg9z1x.fsf@splinter.inka.de> <87wvcs8awc.fsf@splinter.inka.de> <878zp71ef0.fsf@lovi.inf.elte.hu> NNTP-Posting-Host: coloc-standby.netfonds.no Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Trace: main.gmane.org 1035170763 383 80.91.224.250 (21 Oct 2002 03:26:03 GMT) X-Complaints-To: usenet@main.gmane.org NNTP-Posting-Date: Mon, 21 Oct 2002 03:26:03 +0000 (UTC) Return-Path: Original-Received: from karazm.math.uh.edu (karazm.math.uh.edu [129.7.128.1]) by mailhost.sclp.com (Postfix) with ESMTP id BD870D049F for ; Fri, 23 Feb 2001 15:55:00 -0500 (EST) Original-Received: from sina.hpc.uh.edu (lists@Sina.HPC.UH.EDU [129.7.3.5]) by karazm.math.uh.edu (8.9.3/8.9.3) with ESMTP id OAC07161; Fri, 23 Feb 2001 14:54:43 -0600 (CST) Original-Received: by sina.hpc.uh.edu (TLB v0.09a (1.20 tibbs 1996/10/09 22:03:07)); Fri, 23 Feb 2001 14:53:46 -0600 (CST) Original-Received: from mailhost.sclp.com (postfix@66-209.196.61.interliant.com [209.196.61.66] (may be forged)) by sina.hpc.uh.edu (8.9.3/8.9.3) with ESMTP id OAA13280 for ; Fri, 23 Feb 2001 14:53:36 -0600 (CST) Original-Received: from multivac.cwru.edu (multivac.STUDENT.CWRU.Edu [129.22.96.25]) by mailhost.sclp.com (Postfix) with SMTP id 95B48D049F for ; Fri, 23 Feb 2001 15:54:05 -0500 (EST) Original-Received: (qmail 21016 invoked by uid 500); 23 Feb 2001 20:54:25 -0000 Mail-Followup-To: ding@gnus.org Original-To: ding@gnus.org In-Reply-To: (NAGY Andras's message of "23 Feb 2001 17:41:19 +0100") User-Agent: Gnus/5.090001 (Oort Gnus v0.01) Emacs/20.7 Precedence: list X-Majordomo: 1.94.jlt7 Original-Lines: 53 Xref: main.gmane.org gmane.emacs.gnus.general:34932 X-Report-Spam: http://spam.gmane.org/gmane.emacs.gnus.general:34932 NAGY Andras writes: > * Gnus functions could use a single interface to fetch group > parameters based on both methods, without having to deal with regexp > matching and stuff. This is an important goal. > * Users would be able to define every group parameter based on regexps > as well, in addition to the group/topic method. I'm not sure what you mean here. But see below for my suggestion. > * Group/topic parameters allow setting arbitrary variables, not only > `strict' parameters. I'm not sure this is a problem. Your gnus-parameters doesn't seem to address it, either. > To demonstrate this: I cannot (but would like to) write the following > in my .gnus: ... > I think the example speaks for itself. (Setting everything in one > place, independetly of newsrc.eld and topic topology.) Not bad, but I'm not especially fond of the regexp matching; my naming scheme for my groups should not be determined by whether they have similar parameters, any more than my topic hierarchy should be so determined. I'd like it if parameters were (or could be) stored in backends. Backends can already use -request-update-info to correct Gnus's parameters, so we just need a way for Gnus to notify the backend when the user modifies the parameters. (Then once all backends do this, Gnus could stop redundantly storing parameters in newsrc.eld.) Also, to provide for hierarchical inheritance, we could have a parameter called, say, parameter-parent, whose value is a group. Then when trying to find the value of parameter foo, if the group doesn't have a value for foo, you check the parent group, recursively, stopping when you find a group that has foo, or one that has no parent. Dummy groups could be used to hold sets of parameters; real groups would use the dummy as their parent. Your suggestion (like mine) fails to cover the capability we have now of putting a group into multiple topics, and using the parameters of whichever topic the user entered through, but this seems like more trouble than it's worth, especially in situations where there is no "current" topic. Thoughts? paul