Date: Fri, 1 Sep 2000 02:40:57 +0200 From: Shaun Jurrens <shaun@shamz.net> To: arch@FreeBSD.ORG Subject: sysinstall specification (was: build tools as...) Message-ID: <20000901024057.B3659@dakota.priv.shamz.net>
next in thread | raw e-mail | index | archive | help
Although it is hardly my place to enter into the discussions on -arch, I would like to offer an organizational suggestion, that being a greater strength on my part than writing code. A generally recognized method of solving complex problems is to subdivide them into manageable parts. Modularity in programming, design, and problem solving have enough aspects in common, that this little treatise might help. My suggestion is simply that those "fathers" of the current sysinstaller make the design of a new sysinstaller simpler to those not possessing the years of experience with the pitfalls of the gradual "monster" now known as sysinstall. As disgusting as the word rings in most coders ears, it would beneficial if the small group of sysinstall coders/release engineers would put on their "manager" (sorry if that hurt) hats and write a specification of that which would be accepted as a new sysinstaller. That means: 1. coding requirements (language, and general, i.e. "modular" approach) 2. user-interface requirements 3. size requirements 4. list of required functions 5. list of optional functions 6. a time-frame for development (subject to abuse as time allows) Point 4 being the most complex list, it should be divided into sublists of 1) currently available functions including how they are currently (inadequately) resolved, perhaps with references to code, and 2) new and desired functions and how these need to work in conjunction with the other portions of the code. This might seem as difficult as just writing the code, but I see that most of the required authors are sufficiently talented to compose long works in English. It has the advantage of serving as a public checklist for acceptance of the finished "modules" of the code as well, and may interest a greater number of people to take on smaller amounts of the work. The final specification will need to be announced and a moderated mailing list for contributors, tester, etc,. As much as I would like to avoid stepping on anyone's toes/ego, I think this would offer a better chance of getting a new sysinstaller than the current "waiting for divine intervention" method. It would also demonstrate a certain professionality in organizing a much needed (depending on your views, perhaps) development project. I personally think there are enough people willing to work on the project, but the constant warnings about its complexity as well as the "single point of failure" pressure makes people shy away from volunteering. Grant me a bit of charity in your criticisms, the small points have been omitted in favor of keeping this from getting any longer than it already has. -- Yours truly, Shaun D. Jurrens shaun@shamz.net shamz@freenix.no To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-arch" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20000901024057.B3659>