From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jon Loeliger Subject: Notes for CVS Converts Date: Wed, 31 Aug 2005 14:08:36 -0500 Message-ID: <1125515315.13577.53.camel@cashmere.sps.mot.com> Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit X-From: git-owner@vger.kernel.org Wed Aug 31 21:10:11 2005 Return-path: Received: from vger.kernel.org ([209.132.176.167]) by ciao.gmane.org with esmtp (Exim 4.43) id 1EAXwv-0001rb-BL for gcvg-git@gmane.org; Wed, 31 Aug 2005 21:08:41 +0200 Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932522AbVHaTIi (ORCPT ); Wed, 31 Aug 2005 15:08:38 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932524AbVHaTIi (ORCPT ); Wed, 31 Aug 2005 15:08:38 -0400 Received: from az33egw01.freescale.net ([192.88.158.102]:15582 "EHLO az33egw01.freescale.net") by vger.kernel.org with ESMTP id S932522AbVHaTIi (ORCPT ); Wed, 31 Aug 2005 15:08:38 -0400 Received: from az33smr01.freescale.net (az33smr01.freescale.net [10.64.34.199]) by az33egw01.freescale.net (8.12.11/az33egw01) with ESMTP id j7VJHjWa012304 for ; Wed, 31 Aug 2005 12:17:45 -0700 (MST) Received: from [10.82.19.2] (cashmere.am.freescale.net [10.82.19.2]) by az33smr01.freescale.net (8.13.1/8.13.0) with ESMTP id j7VJCkZB013744 for ; Wed, 31 Aug 2005 14:12:46 -0500 (CDT) To: Git List X-Mailer: Ximian Evolution 1.4.6 (1.4.6-2.ydl.1) Sender: git-owner@vger.kernel.org Precedence: bulk X-Mailing-List: git@vger.kernel.org Archived-At: Guys, I was recently bitten by a cg-rm behavior where it silently Did The Wrong Thing, IMO. I think that other CVS Converts might stumble across the same pitfall, so it might warrant a comment in the documentation, perhaps in the "Notes For CVS Users" section. I am in the process of doing some file merging, resulting in some files being eliminated from my source tree. But I wasn't prepared to just out-right delete my original files until some testing had been completed first. So I moved my files out of the way so the build would proceed without them. Later, when I was satisfied the build was good, I executed a "cg-rm floof/somefile.h" and then cg-commit'ed them. However, I realized, they were not actually removed from the cache. Reading through "cg-rm", I figured out that it expected to actually rm the files from the filesystem after first proving they existed with a "find". That in turn also drove the "git-cache-update --force-remove". However, in my case I had already moved/removed the original files. The "find" didn't find them and the cache update didn't happen. And no output was generated either. But I thought nothing of it; after all, "rm" doesn't produce output when it works successfully. So two possible suggestions here: First, perhaps cg-rm should issue a warning if the generated temporary file results in an empty list of files to be removed (and then maybe with a --silent option subsequently?), Secondly, maybe the documentation needs a section with "Notes for CVS Converts" section that clearly states that cg-rm _expects_ to actually remove the file from the filesystem, directly opposite the normal CVS behavior. Thanks, jdl