From: Alec Cawley Newsgroups: comp.editors,comp.sys.transputer Subject: Re: Folding editors Date: Fri, 30 Oct 1998 17:44:35 +0000 Organization: Lost in the Wilderness Distribution: inet Message-Id: References: <36374B57.168B9492@pact.srf.ac.uk> <71aj0c$oc$1@fangorn.moria.de> <6r1znq7264.fsf@belhaven.spider.com> <3639BEFD.2CC25E3C@esgem.com> <6r4ssm59ay.fsf@belhaven.spider.com> Mime-Version: 1.0 Xref: ukc comp.editors:32001 comp.sys.transputer:8891 In article <6r4ssm59ay.fsf@belhaven.spider.com>, Iain A F Fleming writes > >The entity calling itself Stephen Maudsley wrote: >> >> Iain A F Fleming wrote: >> >> > folding is evil >> >> Certainly I've seen code where scoping of variables has been so convoluted >> that it was unmaintainable. > >Mainly that, and the allied fact that it allows large blocks of (often >conditional) code to be folded away rather than being split off into >subroutines/functions. > >Programs composed of small numbers of large functions with complex control >flow tend to be less comprehensible that a large number of small functions. > >Folding encourages the hack-it-and-see incremental mentality. My experience is the opposite. I have had modules which had essentially got out of control, and were made comprehensible by folding. I go the other way around, if anything. Programs composed of large numbers of functions can become incomprehensibly fragmented. Folding up groups of related functions can reduce fragmentation. I particularly value folding at the top level: my whole module fits on one screen and reveals itself to consist of maybe a dozen classes. Each of those classes can then be expanded if necessary. Without folding, I can see no way that I can overview a whole module. I see folding, thus employed, as a merging rather than a splitting tool. Folding should never conceal control flow - that is simply abuse of a tool, like using a chisel as a screwdriver. I find that I can organise my code such that I can nearly always see the whole of flow of a system *at a particular level* on a 60 line screen. On such a display, all closed folds would be true "black boxes". Whether such black box is inline code or a procedure call is, IMHO, irrelevant. As a related thought, it seems to me that there is an opportunity for using folding as a metric of code quality. It seems to me that variables should not be affected at many levels of a program: a high level variable should not be tweaked many levels down. If a module is "fully" folded, the number of folds which have to be opened to reveal all write accesses to a variable would, IMO, be related to the "dangerousness" of that variable. Returning to the original question, I use a folding editor at work (FED) written for by a colleague for X-Windows. I will ask him if he is willing to release it. It improves somewhat on F by handling multiple files, macros, colouring folds etc. -- Alec Cawley