|
From: | David Ayers |
Subject: | Re: Documentation |
Date: | Tue, 24 Sep 2002 00:27:14 +0200 |
User-agent: | Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.1) Gecko/20020826 |
Alexander Malmberg wrote:
I very much agree! I also believe the OPENSTEP spec is very "underdefined" and that the Apple docs and release notes aren't all that bad, yet they don't suffice. For example the text system, which is supposed to support subclassing of NSCell's and NSText classes extensively isn't defined well enough to actually create clean classes that make use of setUpFieldEditorAttributes:, as the state of the field editor comes into play for it standard attribute setting methods, which trigger notifications and the like.What I'm looking for is detailed specifications of exactly how different classes are supposed to work, including how they're supposed to work with other classes, what all the different options really mean everywhere, in what order things are supposed to be done, and things like that (or statements that something is implementation-/un-defined). Since GNUstep seems to be neither exactly OpenStep nor Cocoa, I guess you could call this the GNUstep specification.
I would volunteer to supply some input. I would have an OPENSTEP Enterprise 4.2 to verify and test certain aspects and offer documentation. (I would appreciate some samples to get an idea of detail and tone though.)So I guess what I want is a complete and sufficiently detailed GNUstep specification (from a developer pov; implementation notes wouldn't belong there). I realize that it'd take time to completely write such a thing, and it wouldn't be fully static, but I think it'd be very useful for many things (developing for GNUstep; making sure GNUstep does what it's supposed to; making sure the design fits together), maybe even critical. If persons who know how things are 'supposed' to work (eg. to match unspecified OpenStep behavior) could help verify things, I could help write it.
I'm totally aware of the fact that anything Apple didn't document, may change anytime, but experience also shows, anything they did document can change anytime, so I don't think this is an argument.
Cheers, Dave
[Prev in Thread] | Current Thread | [Next in Thread] |