Path: utzoo!attcan!uunet!husc6!ukma!rutgers!aramis.rutgers.edu!athos.rutgers.edu!hedrick From: hedrick@athos.rutgers.edu (Charles Hedrick) Newsgroups: comp.windows.news Subject: Re: Is NeWS UseABLE? Keywords: framebuffer /dev/fb /dev/win* Message-ID: Date: 3 Aug 88 06:36:19 GMT References: <2135@ssc-vax.UUCP> <5870@orstcs.CS.ORST.EDU> Distribution: na Organization: Rutgers Univ., New Brunswick, N.J. Lines: 60 In response to the question as to whether NeWS is UsEAbLE, I think the answer is a qualified yes. It got a bad reputation for crashing under 1.0. There are no doubt still ways to make 1.1 crash, but I haven't ever seen it crash, and I don't think our users find that a problem anymore. The tendency around here is to do any new applications for it, as it seems to be easier to program than other windows systems. The main problem is that there aren't many normal system applications for it. X has versions of most of the standard Sunview tools. NeWS does not. However Sunview applications run under NeWS, though in a slightly clunky way, so that isn't as bad a problem as it might be. The biggest problem, as everyone agrees, is that there's no decent terminal emulator. If the one that is going to be posted Real Soon Now solves that problem, I think NeWS will be ready for serious use by real users. I'm still not convinced that there's any generic performance problem with NeWS itself. Nterm is certainly unacceptably slow, but I don't find psterm to be. I'm sure that like many other things it will depend upon the application, and also on how it is structured. However having said that, I'm still beginning to wonder whether NeWS isn't sort of an orphan. I kept hoping that Sun would be coming up with a basic set of NeWS applications. But the presentation at Usenix said that they were doing Sunview 2 in X, and the final NeWS development system wouldn't be out until sometime in 89. NeWS 1.1 doesn't have any more applications than 1.0. It's clear that like computer systems, languages, etc., a window system is going to win based on the software that is available for it, not its theoretical merit. If Sun themselves aren't willing to start producing applications for it, I think it's bound to end up as sort of a backwater (a very nice backwater, understand). I'm not worried that NeWS will vanish. Merged X/NeWS will at least make sure that the final Sun window system has the capability to run NeWS. So I don't feel badly about doing applications based on it. But it's beginning to look like the "software gap" is becoming impossible. The X community already has a long head start, and if Sun is depending upon System V release 4 to get NeWS out, that's going to hold things up long enough that I just can't see it catching up. I suspect we'll still use NeWS for things that we don't mind seeing run only on the Sun. But I'm beginning to think that where we have to choose a window system, it's likely to be X. There are several contexts in which we really do have to choose one window system, and can't depend upon merged X/NeWS. I'd like to use micros as window servers, and I also like the looks of the new Visual Technology X terminal. I think it may be asking a bit much for these low-end products to run merged X/NeWS. Oddly enough, my feeling that we should be using X will take effect only when we get the merged X/NeWS. For the moment, I'm inclined to recommend that people use NeWS. I think our users would react very badly to the rather hacker-oriented style of the X tools. So I think our large-scale X use is likely to start with Sunview 2/Open Look. For the moment, we have both, but I'm suggesting that people continue using the Sunview tools under NeWS and do their own development work in NeWS. When vendors ask us which window system we'd like them to support, I just don't know what to tell them. If anybody at Sun wants to change this, I advise them to get a usable terminal emulator out immediately, and within the next few months come up with NeWS versions of the major Sunview tools. But I think we'd know already if such an effort were underway.