From owner-freebsd-chat Wed Jan 22 11:29:40 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id LAA21797 for chat-outgoing; Wed, 22 Jan 1997 11:29:40 -0800 (PST) Received: from fallout.campusview.indiana.edu (fallout.campusview.indiana.edu [149.159.1.1]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id LAA21680 for ; Wed, 22 Jan 1997 11:27:43 -0800 (PST) Received: from localhost (jfieber@localhost) by fallout.campusview.indiana.edu (8.8.4/8.8.4) with SMTP id OAA06798; Wed, 22 Jan 1997 14:26:51 -0500 (EST) Date: Wed, 22 Jan 1997 14:26:51 -0500 (EST) From: John Fieber Reply-To: John Fieber To: Larry Lee cc: chat@freefall.freebsd.org Subject: Re: Commerical applications In-Reply-To: <199701221749.KAA15464@usr07.primenet.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-chat@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Wed, 22 Jan 1997, Larry Lee wrote: > I do think that most people use their computers as communication > and organization devices, and what is needed in the basic offering > are applications such as word processors, internet access, fax machines, > as well as installation procedure to make this stuff work. > > I don't think FreeBSD needs to be a windows clone to be accepted. > Specifically, I don't agree that icons, point/click, and drag/drop > are _requirements_ to doing useful work. Although I will go see what > I can find about CDE. Well, sure. A keyboard isn't a _requirement_ doing useful work. There are many people, probably some right here on this list, that have done lots of useful work using punched cards. :) You seem to be saying that FreeBSD needs the functionality and ease of use of windows to attract people who might otherwise use Windows. Then you turn around and say that you don't think that the GUI model Windows provides is necessary to accomplish that. To be blunt, you are dead wrong and have completely missed the fundamental change that interface design took when the GUI was developed at Xerox and later marketed by Apple. For better or worse, computer interfaces are not intuitive; they are idiomatic. People invest time in learning the idiom and what the GUI offers is consistent use of idioms with visual affordances from application to application. The most widely accepted interface model today is the windows/icon/mouse interface and if you want people to use your system, you adhere to that model, or you demonstrate that your model is superior. The average Joe has long ago rejected the traditional Unix command model, so what do you propose? The other thing to note is that the Mac and Windows applications are successful because they (usually) fit into a well defined application framework that permits smooth interoperation. Much of this framework is directly visible only to the developers, but the users reap the benefits. What a lot of Unix developers seem to miss is that the advantages of using the same idiom consistently across applications almost always outweighs the advantages of providing a new idiom that may even be slightly more efficient for a given task. The latter only becomes worthwhile if the advantage is very large and the learning time can be recouped over extensive use. Another issues is that just like some ties just don't go with some shirts, not all interface idioms are compatible; when designing a new idiom, it must blend well with the existing framework. X, with its motley collection of home-brew toolkits just doesn't cut it in this respect. This application framework is a necessary prerequisite for the emergence of quality applications that work well enough together to provide a critical mass to attract more applications and users. Unfortunately, X by itself provides **no** such framework. Xt adds a thin layer (thick, actually, but that is a technical issues), but not nearly enough. Motif adds a fair amount on top of that, including standardizing things like window management, and drag and drop, and a more usable clipboard model. CDE is the only thing to date that offers the sort of GUI application framework that Windows and Mac offer. Tk appears to be the runner up but operates at about the same level as Motif. Java has some interesting potential but still needs work. Since Motif and CDE are not free, there has been a well unified opposition to it among many developers. Unfortunately, that unification evaporates instantly when presented with the task of providing an alternative. :( If you are at all familiar with the Amiga, it had a similar framework problem. Prior to version 2.0 of the operating system, basic mechanisms for writing GUI software were in place, but no application framework. GUI applications, outside of games, basically sucked. When 2.0 came out, it came with a style guide and a toolkit which made following that style guide the path of least resistance. Unfortunately, it was too late as would-be developers of productivity applications had long ago dismissed the platform. As a final note, I don't think anybody would argue that Microsoft has been successful by releasing technically elegant, or even stable products. Their interface design isn't that great either. What Microsoft offers is an assurance of continual improvement. Windows95 isn't great, but it is a hell of a lot better than Windows 1.0. Microsoft is persistent and although it may take them a number of tries to get something right, they usually will get it right eventually. In the Unix/X context, if we want the sort of applications you are talking about, we need a robust, modern application framework. The most comprehensive one is CDE. It is not free, there are many rough edges, it is incredibly bloated, some think it is ugly, but with a unified effort, these liabilities can all be minimized (even price), and some eliminated. The commercial Unix world is just now recognizing this as being essential in their defense against the persistence of Microsoft and NT. -john