From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 5914 invoked by alias); 21 Feb 2003 17:14:38 -0000 Mailing-List: contact gdb-patches-help@sources.redhat.com; run by ezmlm Precedence: bulk List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-patches-owner@sources.redhat.com Received: (qmail 5901 invoked from network); 21 Feb 2003 17:14:38 -0000 Received: from unknown (HELO crack.them.org) (65.125.64.184) by 172.16.49.205 with SMTP; 21 Feb 2003 17:14:38 -0000 Received: from nevyn.them.org ([66.93.61.169] ident=mail) by crack.them.org with asmtp (Exim 3.12 #1 (Debian)) id 18mIe6-0003mh-00; Fri, 21 Feb 2003 13:15:42 -0600 Received: from drow by nevyn.them.org with local (Exim 3.36 #1 (Debian)) id 18mGkt-0003wN-00; Fri, 21 Feb 2003 12:14:35 -0500 Date: Fri, 21 Feb 2003 17:14:00 -0000 From: Daniel Jacobowitz To: Andrew Cagney Cc: Jim Blandy , Elena Zannoni , gdb-patches@sources.redhat.com Subject: Re: [drow@mvista.com: Re: RFA: LOC_COMPUTED support] Message-ID: <20030221171435.GB14877@nevyn.them.org> Mail-Followup-To: Andrew Cagney , Jim Blandy , Elena Zannoni , gdb-patches@sources.redhat.com References: <20030213211157.GA13537@nevyn.them.org> <20030221152420.GA32260@nevyn.them.org> <3E565E09.9020706@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <3E565E09.9020706@redhat.com> User-Agent: Mutt/1.5.1i X-SW-Source: 2003-02/txt/msg00523.txt.bz2 On Fri, Feb 21, 2003 at 12:12:41PM -0500, Andrew Cagney wrote: > Just a generic heads up on the structure of this code and these, er, > batons (from the point of view of the architecture). > > This is implementing something like (I've not hacked C++ in > 10 years): > > // some base classes > class A; > class B; > > // a derived class (wonder if I got the order right). > class B::class B1; > > // A `uses' B but is parameterized with the specific instance > class A->method (class B B); > > in C. > > I honestly think that using baton's distract from what is a simple O-O > construct and standard O-O terminology. > > The frame and architecture code both reflect this structure - pass in > the object and then use methods supplied as part of the object. > > Once all this is settled, I think I'll look to re-factor (hmm, buzword) > the code so that its structure better refects what is going on. I don't think I agree with you on this, but if you want to change it by all means post a patch for some concrete discussion :) This is another point when I am willing to sacrifice a certain amount of clarity to not carry around an extra set of method pointers; there are a _large_ number of these batons. I'd rather discuss a transition to C++ than any particular instance of this problem. Do you think it's feasible? -- Daniel Jacobowitz MontaVista Software Debian GNU/Linux Developer