contrib cleaned up

Russ Allbery rra at stanford.edu
Sun Jan 14 14:25:51 UTC 2001


I've gone through and cleaned out contrib a bit, as well as adding a
couple of new things.  Various programs that don't work with the storage
API and look like they'd either be quite difficult to make work with the
storage API (amounting to a complete rewrite) or whose purpose is lost now
have been removed.

I also removed newsbot (a program to trigger on incoming articles and
perform various actions like issue cancels) because it's separately
maintained on www.frmug.org, it's nearly 100KB all by itself, and I don't
think the average user of INN is going to (or should be) running a cancel
bot.  Folks who want it can grab it from the maintainer's site.

The stuff that was in contrib as saved news articles or mail messages are
now executable or compilable directly, and there's a Makefile and a
complete README in that directory now.  I have ninpaths pending to go in
(see my other message).

fixscript is now run by the Makefile on all the Perl scripts to handle the
path to Perl (and in some cases to handle innshellvars).

I left a few things alone and didn't move them into the new system or make
Makefile entries for them because I think something else should be done
with them in the near term:

count_overview.pl -- Does anyone use this?  It counts occurrences of
groups in Xref data in the overview database, but would only work with
tradindexed.  I think it was intended for use with the old unified
overview.  If no one claims it, I'm going to remove it; it seems pretty
simple to me.

fixhist -- makedbz should just ignore malformed lines and have an
additional option to rewrite the history file to remove all malformed
lines while reporting them to stderr, and then this script would be
obsolete.  I'm thinking about working on the history API shortly, though,
and don't want to fiddle with makedbz too much before then since it would
be partly wasted effort.

mkbuf and reset-cnfs.c -- These are both doing essentially the same thing
for different reasons, namely writing zeroes to a file.  I think we should
introduce the idea of storage-specific utility programs and drop something
in storage/cnfs that does both of these.  (We could also move cnfsstat and
cnfsheadconf into storage/cnfs, where they'd fit a lot better, and maybe
down the road move the ovdb_* programs into storage/ovdb for similar
reasons, to keep all the code for particular backends together.)

-- 
Russ Allbery (rra at stanford.edu)             <http://www.eyrie.org/~eagle/>



More information about the inn-workers mailing list