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: setup goals Date: Tue, 7 May 2002 01:14:35 +1000 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" X-MS-Has-Attach: X-MS-TNEF-Correlator: X-MimeOLE: Produced By Microsoft Exchange V6.0.5762.3 content-class: urn:content-classes:message From: "Robert Collins" To: "Cygwin-Apps" Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by delorie.com id g46FEgC26363 I thought I'd make the setup design and coding and make goals clear in one place - I don't think it's been done recently. I don't anticipate this document being controversial, but if you do disagree with what I say, please speak up! Setup's primary goal is to be: A mingw application to bootstrap a cygwin install from ftp and http servers. This means that size matters. As does 'intuitiveness'. It should 'just work'. Setup's main secondary goal is: To allow incremental installs and updates of the cygwin net distribution for cygwin users, in an easy, intuitive fashion. As long as those two goals are consistently met, setup can enlarge to accomplish other things, such as: * rebasing * rebinding * command line options * optional console based version. * rsync support. * mirroring. * scripting. * 'native' compilation. * WINE compilation/execution. * code base reuse for other setup.ini/cygwin package tasks. And I'm sure there are more. Patches to HEAD that break the first two goals, will only be accepted on a 'temporary break until xyz' basis. Patches to HEAD (or checkings from non-reviewed committers such as Chris) that break the first two will be accidental (I hope!). The point of this is that the only thing I *always* test is mingw targeted compilation, and install from net. Other things may break, and if they do, they will get fixed, but not at the same priority as if a break occurred to the first two things. Second to last, patches to 'release' branches, such as setup-200205, will only ever be for critical bug fixes, not for UI enhancements and the like. Lastly, on development branches, anything goes, I don't care if a development branch even builds. (Currently we don't have any development branches). Rob