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, 12 Nov 2001 18:40:52 -0500 From: Christopher Faylor To: cygwin-apps AT cygwin DOT com Subject: Re: Setup snpshot bug no skip/keep Message-ID: <20011112234052.GA25682@redhat.com> Reply-To: cygwin-apps AT cygwin DOT com Mail-Followup-To: cygwin-apps AT cygwin DOT com References: <200111091357 DOT IAA29622 AT cs DOT umb DOT edu> <20011109111617 DOT F24506 AT redhat DOT com> <042f01c16b06$67ae3350$0200a8c0 AT lifelesswks> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.3.23.1i On Mon, Nov 12, 2001 at 01:01:34PM -0500, Brian Keener wrote: >Robert Collins wrote: >> The problem with 4 buckets is that either 1 is a pointer to another or >> you have to remember to change 2 whenever you change one. Both of which >> will be annoying. When this becomes a bit more oop, this can be made >> transparent, and then it will be easier. > >I'm confused - remove the current Info for installed and add a new bucket for >installed in the Info structure that keeps track of Prev, Curr and Test. You >still are only updating one installed bucket. >> >> The current UNKNOWN bucket, IMO should be used to store the installed >> package details IFF it's not listed in setup.ini. > >That works for me - I guess I just don't see why the installed Info Structure >is separate from the info Info Structure as we discussed above. But I think in >some earlier versions before the heavy rewrites we did use Unknown for an >installed version not in setup.ini - at least that was the trust we used >although we may not have used that part of the structure. I believe that the unknown part is my invention. It's there to allow indexing by trust levels. It's currently unused but you could shove install in there. I am not sure why I didn't do that to begin with. In earlier revisions all of this was stored in two unwieldy arrays that shared a common index and the same for loop was copied to ten different places throughout the code. It's only slightly better now but at least with just one structure to worry about it is easier to think about things like this. >> Also local tarballs should be included, and the bucket count should be >> dynamic. > >Absolutely. I'm not sure what you mean by bucket count but, IMO, there should only be two fixed categories "curr" and (optionally) "test". All of the "prev"ious versions should be retrievable, not just one of them. cgf