Mailing-List: contact cygwin-apps-help AT cygwin DOT com; run by ezmlm Sender: cygwin-apps-owner AT cygwin DOT com List-Subscribe: List-Archive: List-Post: List-Help: , Mail-Followup-To: cygwin-apps AT cygwin DOT com Delivered-To: mailing list cygwin-apps AT cygwin DOT com Subject: RE: prev/curr/test Date: Tue, 26 Mar 2002 11:31:22 +1100 MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Message-ID: X-MS-Has-Attach: X-MimeOLE: Produced By Microsoft Exchange V6.0.5762.3 content-class: urn:content-classes:message X-MS-TNEF-Correlator: From: "Robert Collins" To: Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by delorie.com id g2Q0Vac23373 > -----Original Message----- > From: Christopher Faylor [mailto:cgf AT redhat DOT com] > Sent: Tuesday, March 26, 2002 11:21 AM > >Hmm. I think that unclicking bin should uninstall - leaving it there > >would be counter-intuititive. > > If you have the word install next to a box, I don't think it > follows that it will be uninstalled. Of course, my turn for stupidity :} > >Tick Bin to install foo. Untick Bin to remove foo. Tick Source to > >trigger a download of the source (download only mode) or extraction > >(install from x mode). > > And when you just don't want a package? What do you click to > get the equivalent of skip? Don't click either? In this example perhaps the bin column should be labelled "install". > Given your above comments, I think we still need another > clickable "thing" next to the Bin/Source, unless you have > some way of getting the equivalent of a "skip/keep". Ahh, so if you don't want to update you can stay put. Hmm. What about [X] - install [H] - hold [ ] - uninstall I know, it's heading back to the circular clicking thing :[. > >However, I think some folk will want the current interface, > so perhaps > >we offer a 'basic' and 'advanced' download of setup, or a [Advanced] > >button somewhere on the chooser. (To start with I'd suggest two > >downloads because the chooser doesn't use a factory yet, so I can't > >parameterize the display at runtime. That is a goal though.) > > I don't see any gain in keeping the old interface if we make > the above changes. There is equivalent functionality and, > imo, better clarity. You always get into trouble when you > start trying to maintain disparate views. Actually there is less functionality - no one-step reinstall, no selection of arbitrary versions. Currently I can install any version found on my hard disk, these proposed interfaces remove that ability. > I haven't looked into the code recently, but I think this > gets rid of some of that state machine stuff that (used to?) > was never quite right. I think that nuking that logic would > be a big enough goal for going to a plan like this. It would remove the need for the state machine code, but that is very stable now anyway (see package_meta if you are interested). Rob