[nSLUG] Gnu "info" and other documentation (was: Stage Lighting)

Mike Spencer mspencer at tallships.ca
Sat Jan 31 01:01:16 AST 2009


Daniel Morrison <draker at gmail.com> quoth:

> Man, I _hate_ info.
>
> $ man gnuish_software
> 
> gnuish_software(1)
> 
> There is no documentation here.  If your system was installed
> correctly, you can get the full documentation with info.  What's that,
> you're not an emacs user?  Bwahahahahahahahaha good luck!!!!
> 
> See also: info(1), emacs(1), how_to_waste_the_sysadmins_time(1)

and David Payne <david at payneful.ca> added:

> I second that.  When ever I find documentation only in info I find
> it easier just to search with Google.

It's enough to make you want to invoke chainsaw(8) [1].

Actually, I'm a dedicated GNU emacs mainliner. The very first thing I
do in a new install is get my favored version of emacs running the way
I want it.  And I still hate info.

My workaround for info (and for information overload, an aging brain
and general befuddlement) is to keep a personal ~/man directory, a man
script called (for differentiation from man) "mann" and a personal
apropos-like flat database.  Anytime I have to puzzle something out, I
create a new page in ~/man or add to an existing one.  If there's only
(gnu-format) info available and I consult that doc often, I run the
whatever.info-* through a script and dump output to ~/mann/whatever.1
as concatenated ASCII text.  The mann script parses filename
extensions such that, say, mplayer.1 (the official manpage reformatted
to suit myself) and mplayer.mem (my notes and de-obfuscations) can live
in the same directory.

How to keep track of all this?  Well, for nearly everything, there are
words or terms that come first to mind, possibly the wrong ones, when
thinking about something.  Those associations are entirely
ideosyncratic so either (1) there is a file in ~/man named by that
word or (2) such words go into the apropos database along with
pointers to files in ~/man or to other docs or clues.  Some things
that I frequently forget (such as the word "jonquils") and which will
fit on one line are just inserted in the apropos file after the word I
*do* always remember but with no associated SEE ALSO reference.

Just for entertainment value, here's a few lines from the apropos
file:

daffodils     -- Did you mean jonquils? Narcissus?  Ho hum.
date          -- See sdate, phoon; dated (alias)
daylight      -- saving time: See DST; txt/misc/daylight-saving
daylight      -- See pkg/timezone; zic(8) (/sbin/zic)
dd            -- See box/pkg/sdd (Faster dd from author of CD burn pkg)
dd            -- Dennis Ritchie days it's from "define data" OS 360 jargon
dental        -- See dental-drill; dobbs
derusting     -- See electroclean
devices       -- See gpib (Instrument control boards and keyboards)
display       -- ~/bin/disp is ImageMagick's display-3.8.8. 5.3.5 is buggy
display       -- Saves 2-byte PGMs: see mann image, pnmdepth
display       -- ftp://ftp.imagemagick.net/pub/ImageMagick/binaries/ 6.9MEG
display       -- Sat 16 Feb 2008 V. 6.3.8; See display-6.html
dns           -- See resolv.conf, whois, rootservers; rfc dns; mann resolv
dobbs         -- Dr. Dobbs Jurnal of Computer Calesthenics & Orthodontia,
dobbs         -- Running Light WithOut OverByte; DDJOCCAORLWOOB
drill         -- See dental-drill
dst           -- pkg/timezone; zic(8) (/sbin/zic); mann daylight

``mann foo'' results in ``grep -i ^foo'' in this file if it can't find
~/man/foo.{1,2,3,4,8,doc,mem}.  If nothing there matches, it does
``alias |grep foo'' and some other things.  As a last resort, when
nothing else is found, it does ``locate foo |
<elide-anything-not-in-~>'' which will turn up the odd .pdf, .pod,
.html or .txt that I've squirreled away in places other than ~/man.

This also captures non-computer lookups, such as wiring diagrams for
electric motors I have, specs for machinery (e.g. my obsolete
belt-driven dental drills), pharmaceutical data sheets, MSDS docs,
parts and materials supplier info, obscure acronyms and more.

Doesn't work well for stuff that's new to me (witness my recent
ignorance of /usr/bin/isoinfo) but stuff I've seen before and long
since forgotten usually turns up. 

Sorry if this is a ho-hum digression but it's an indispensible tool
for me.  Probably wouldn't work all that well in a shared environment
but for a home box, it exploits, rather than suffers from,
idiosyncrasy.  Similarly wouldn't work well with a very large or
complex document set but, given the speed of current computers, I
won't live long enough for my humble, digital-carboard-box-full of
miscellany to bog down.


FWIW,
- Mike


[1] chainsaw(8): http://home.tallships.ca/mspencer/alien/chainsaw.html


-- 
Michael Spencer                  Nova Scotia, Canada       .~. 
                                                           /V\ 
mspencer at tallships.ca                                     /( )\
http://home.tallships.ca/mspencer/                        ^^-^^



More information about the nSLUG mailing list