From owner-freebsd-commit  Wed Oct  4 22:28:09 1995
Return-Path: owner-commit
Received: (from root@localhost)
          by freefall.freebsd.org (8.6.12/8.6.6) id WAA19981
          for freebsd-commit-outgoing; Wed, 4 Oct 1995 22:28:09 -0700
Received: (from root@localhost)
          by freefall.freebsd.org (8.6.12/8.6.6) id WAA19968
          for cvs-all-outgoing; Wed, 4 Oct 1995 22:28:07 -0700
Received: (from root@localhost)
          by freefall.freebsd.org (8.6.12/8.6.6) id WAA19956
          for cvs-ports-outgoing; Wed, 4 Oct 1995 22:28:05 -0700
Received: from jhome.DIALix.COM (root@jhome.DIALix.COM [192.203.228.69])
          by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id WAA19948
          ; Wed, 4 Oct 1995 22:27:50 -0700
Received: (from peter@localhost) by jhome.DIALix.COM (8.6.12/8.6.9) id NAA12554; Thu, 5 Oct 1995 13:26:20 +0800
Date: Thu, 5 Oct 1995 13:26:19 +0800 (WST)
From: Peter Wemm <peter@jhome.DIALix.COM>
To: Satoshi Asami <asami@cs.berkeley.edu>
cc: joerg_wunsch@uriah.heep.sax.de, CVS-commiters@freefall.freebsd.org,
        cvs-ports@freefall.freebsd.org
Subject: Re: cvs commit: ports/devel/cflow/patches patch-aa
In-Reply-To: <199510050455.VAA07443@silvia.HIP.Berkeley.EDU>
Message-ID: <Pine.BSF.3.91.951005132159.509F-100000@jhome.DIALix.COM>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: owner-commit@FreeBSD.org
Precedence: bulk

On Wed, 4 Oct 1995, Satoshi Asami wrote:
>  * >   "gzip --best" -> "gzip -9nf" for manpage compression.  Otherwise it
>  * >   will ask if you want to overwrite it if the compressed manpage already
>  * >   exists.
>  * 
>  * I thought we won't use "gzip -9" any more?
> 
> Joerg, there are so many ports out there with gzip -9 in there, and
> now is not the time to go fix all of them. :>
> 
> Besides, I don't think any conclusion came out of the discussion a
> while ago.  However, if you ask me, I'd rather keep the -9 in there
> (or whatever the "best" compression is, in terms of compression ratio,
> regardless of speed).  I think most of our users out there use binary
> packages, so the extra time we pay, will save space for thousands of
> them. :)

My sentiments exactly.  For "once compressed, many, many times 
decompressed" type arrangements, gzip -9 is probably fine.

For things like the wuftpd's ftpconversions file, it most definately 
should be gzip -1 - who out there is running a busy ftp server and can 
afford to have their gzip's using 6 times more cpu for a couple of extra 
percent compression? :-)

Also, In the man page case, the overhead of actually invoking the "gzip" 
executable from the makefile is probably greater than the time to 
actually compress it. :-)  -1 or -9 for the average small manpage is 
probably not going to make much of a measurable difference time-wise.

Cheers,
-Peter

> Satoshi