Mailing-List: contact cygwin-apps-help AT sourceware DOT cygnus DOT com; run by ezmlm Sender: cygwin-apps-owner AT sourceware DOT cygnus DOT com List-Subscribe: List-Archive: List-Post: List-Help: , Delivered-To: mailing list cygwin-apps AT sources DOT redhat DOT com Date: Mon, 7 May 2001 11:07:42 -0400 From: Christopher Faylor To: Corinna Vinschen Subject: Re: Forcing SYSTEMROOT (opinions needed) Message-ID: <20010507110742.B26951@redhat.com> Reply-To: cygwin-apps AT cygwin DOT com Mail-Followup-To: Corinna Vinschen References: <20010502222849 DOT A1238 AT redhat DOT com> <20010503111926 DOT Y24200 AT cygbert DOT vinschen DOT de> <20010503133346 DOT A5353 AT redhat DOT com> <147585893249 DOT 20010507121105 AT logos-m DOT ru> <20010507114813 DOT V24200 AT cygbert DOT vinschen DOT de> <3AF69D43 DOT C5528BF0 AT yahoo DOT com> <20010507153134 DOT C24200 AT cygbert DOT vinschen DOT de> <15607524704 DOT 20010507181138 AT logos-m DOT ru> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.3.11i In-Reply-To: <15607524704.20010507181138@logos-m.ru>; from deo@logos-m.ru on Mon, May 07, 2001 at 06:11:38PM +0400 On Mon, May 07, 2001 at 06:11:38PM +0400, egor duda wrote: >Hi! > >Monday, 07 May, 2001 Corinna Vinschen cygwin-apps AT cygwin DOT com wrote: > >CV> On Mon, May 07, 2001 at 09:04:03AM -0400, Earnie Boyd wrote: >>> Corinna Vinschen wrote: >>> > Nevertheless I agree that the setting in an environment variable >>> > doesn't fit our needs in the future. Shouldn't we discuss creating >>> > a settings file which may override or supplement the CYGWIN environment >>> > variable settings? >>> > >>> >>> Perhaps doing away with the CYGWIN environment variable altogether. > >CV> Yes, personally I strongly agree but I'm somewhat reluctant to break >CV> backwards compatibility... even if it's ugly. > >i don't think it's ugly. on the contrary, configuring a program via >environment is much better and easier to administrate then using >files, etc. For example, i wrote some big script, that's supposed to >run on many machines across our company LAN. some machines in the LAN >do have installed cygwin, others don't. the script uses some cygwin >programs, so some cygwin configuration is needed. to work correctly, >this script is making some assumptions about cygwin configuration, for >example, it needs binmode, or codepage:oem, or something else. on the >other hand, i don't want to interfere with other CYGWIN settings >already set on those machines, where cygwin is already installed. >what's worse, some programs called from this script may want >nobinmode. with options passed via environment, i have no problems to >implement such script. without it it's certainly harder. I agree. The idea of a file has come up from time to time. I don't see any reason for it. It will just slow down cygwin startup further. If you want to store your settings somewhere, store them in the registry. This discussion has gotten far afield of its primary goal regarding SYSTEMROOT, etc. I guess I will make an executive decision. Lets always add SYSTEMROOT and SYSTEMDRIVE. We'll forgo an API to control this for now (since noone took the bait and did any research for UNIX-like ways to control this behavior). I'll check in a patch today. Just to put an end to this discussion. I don't like the idea of a cygwin initialization file. Nothing that has been discussed here has convinced me that it would be preferrable to the CYGWIN environment variable. I'm not going to introduce YA thing for users to tweak unless a see a clear benefit. cgf