Hi Norwid, At 2025-12-16T16:41:08+0100, Norwid Behrnd wrote: > Recently I noticed groff's bug tracker on savannah.gnu.org lists > groff's preprocessors in alphabetic order. The man page of groff (as > provided by with package 1.23.0-10 in Debian 14/forky, testing), > section `Preprocessors` (about 40% in to the document) does not follow > this sequence. > > Is this discrepancy because the former is a sort by a computer,
Well, a lexicographic sort, at any rate, and thus easily automated--yes. > the later by priority how frequently a typical user of groff likely > engages one of them? It's hard for me to say. I'm not responsible for the ordering that appears and no motivation for the order chosen is stated in comments in the source document, which would be the best place to motivate it (if not in the text itself). It certainly isn't the order in which groff(1) itself _arranges_ the preprocessors in the pipeline it creates. https://cgit.git.savannah.gnu.org/cgit/groff.git/tree/src/roff/groff/groff.cpp?h=1.23.0#n54 It appears to me that the list in question first arrived in this commit. commit 6f4bee460e1c8764ad19d1cb70571db0b012e5f2 Author: Werner LEMBERG <[email protected]> Date: Tue Jan 1 12:40:16 2002 +0000 * src/roff/groff/groff.man: Completely rewritten. ...however this was before we were actually using Git, but rather CVS. The comments in the file disclose further detail. diff --git a/src/roff/groff/groff.man b/src/roff/groff/groff.man index 805c6a173..da298c96c 100644 --- a/src/roff/groff/groff.man +++ b/src/roff/groff/groff.man @@ -1,5 +1,10 @@ .ig -Copyright (C) 1989-2000, 2001 Free Software Foundation, Inc. +groff.man + +Last update : 1 Jan 2002 + +Copyright (C) 1989, 2002 Free Software Foundation, Inc. +Rewritten in 2002 by Bernd Warken <[email protected]> [...] So in modern parlance, we'd probably say that Bernd was the author and Werner the committer, analogously to how we handled your fix to the chem example's ATP molecule.[1] > If not -- would it matter much if the two were in sync, i.e. both in > alphabetic order? I can't imagine how it could hurt. > Second, since groff initially was / is more about technical > communication / documentation, could `chem` and again explicitly > appear among the preprocessors in the pull-down menu, too? Maybe an > entry for each of the preprocessors the man page mentions, to the > result the entry `Preprocessors others/general` on the menu to become > `Preprocessors, general` alone? Web controls (often) have to be scrolled through, so it's an ergonomic win if the list of "Categories" for tickets is as short as possible. We balance that consideration against narrowly drawn categories that help users and developers judge the impact and breadth of expertise required to work around or resolve a defect or to implement a feature. The compromise that the project seemed to have reached before I started contributing is to have "other/general" categories to house tickets against parts of the groff software system that don't accumulate many reports. As much as I'd like to think that this is because those parts are unusually robust, defect-free, and have feature sets that are perfectly satisfactory to our users, the more likely explanation is that those parts aren't used much. The three preprocessors Bernd Warken originated--gperl, gpinyin, and glilypond--are cases in point. During my tenure, I've created a category for "Macro package me", specifically because Dave Kemper was notably productive in creating tickets for it. (After a while, I got into the act myself.) A stable set of categories also makes for a less noisy information channel when undertaking statistical measurement of ticket properties. See the groff 1.23.0 release announcement for an example.[2] > I don't consider either one as a bug and maybe there are reasons for > the current state I'm not (yet) aware. The foregoing is my understanding of how we arrived at the status quo. Regards, Branden [1] https://cgit.git.savannah.gnu.org/cgit/groff.git/commit/?id=86f816985f3998caba53c6a5bafd7ee6d03b4e98 [2] https://lists.gnu.org/archive/html/info-gnu/2023-07/msg00001.html
signature.asc
Description: PGP signature
