mcdonald@fornax.UUCP (Ken Mcdonald) (02/20/89)
After finally having had a chance to play with a Next box for a couple of hours, I was, well, underwhelmed. So, I thought I'd shoot off about what I see as its problems (and pluses), and find out if maybe some of these will disappear in the future. --Number one: it uses UNIX. OK, I don't expect this problem to disappear, but it is worth noting, because I suspect it causes many of the other problems that make me less than completely enthusiastic about the machine. UNIX is big. That means thay you pay (in terms of hardware necessary) for the ability to use UNIX, even if you hardly use any of its features, and the vast majority of UNIX users don't. I think that to many people, the incredible power of UNIX is actually a detriment--if you can do anything, but can't figure out how to do it, then you CAN'T do anything. Most people I know who use UNIX, including most of the computer science professors at my university, would probably be far happier with a much simpler system that did basic computation and networking. The exceptions to this rule seem to be UNIX support people--hackers, gurus,etc. Which raises an interesting question about the ultimate utility of UNIX, but hey, I'll leave that for another time. --Number two: functionality has been sacrificed for somebody's conception of appearance. As in, icons in the icon dock don't have any names. As in, hierarchical menus come up with their tops aligned with the tops of their parent menus--a real pain if you have a short hierarchical menu which pops up in response to the last item on a long parent menu. The little dot you drag on to move the split in the window of one of the standar applications (quotations? I forget) is a pain--why not drag anywhere in the split? That weird monitor stand might irritate some people (though, to be honest, I don't mind it.) --Number three: this machine can be really slow! Slow to bring up apps, which is a real pain, because its not even giving you feedback that it's at least working on starting up the app. Slow in execution--I quite handily got several words ahead of the computer in Write Now, and I'm not a very fast typist--45wpm, max. I can't even get a single character ahead, on my mac. (Of course, this doesn't apply to everything--more comments in the 'pluses' section.) On the pluses... --One: display PostScript is good. Peformance is quite adequate, and will improve when (I expect) they bring in special hardware to do colour DPS. Its really nice to KNOW that what you see on the screen is what is going to print out, and from a programmer's point of view, it must be heaven not to have to write code to print to both the screen and the printer, seperately. --Two: The sound capabilities, plus the DSP hardware, make a nice package, and about time someone did this! --Three: all the other hardware, of course It seems to me that what NeXT did right was in the low-level stuff--hardware, display software, etc. (If there is anything else to the low-level). What they did wrong was at the higher levels--OS, to some degree the interface, maybe a few other things that I'm not aware of yet. This means they can still rectify their errors. Comments, anyone? Ken McDonald
fargo@pawl.rpi.edu (Ethan M. Young) (02/21/89)
(Although, to my knowledge, there are no NeXTs here at RPI to play with...)
UNIX is the number one problem of NeXT?
Mild flame: Are you an MS-DOS freak or something?
Mild chill
Although UNIX is big, and although not everyone will make use of many of the
features of UNIX, not everyone will be purchasing a NeXT. The NeXT, if you
recall, was designed as a high-end technical/research workstation. Thus, the
main body of NeXT users will be college students, professors, graduate stu-
dents, research oriented businesses, etc. And not just general research, more
in the area of programming and electrical research. And what better OS for
a programmer to work in than UNIX?
Plus, even though the little tricks of UNIX may be known primarily by the
guru's and wizards, there happen to be alot of 'em around. (comp.unix.wizards)
Plus, there are TONS of programs and utilities available for UNIX. If you
can't find something you need in the standard library of functions, you can
almost always find it on another machine. Not only can you find all these
functions, but most of them come with the source code. Something which is in-
valuable to any programmer. (Especially those of us who do alot of porting)
Overall, I think that UNIX was a very good choice for the NeXT.
I just wish IBM thought the same way. (Dreaming......)
Thank you and happy hunting! Internet: fargo@pawl.rpi.edu
____ [> SB <] fargo@{paraguay|uruguay}.acm.rpi.edu
/__ -=>??<=- Bitnet (??): usergac0@rpitsmts.bitnet
/ ARGO : 3000 years of regression from the year 4990gore@eecs.nwu.edu (Jacob Gore) (02/21/89)
/ comp.sys.next / mcdonald@fornax.UUCP (Ken Mcdonald) / Feb 19, 1989 / >... you pay (in terms of hardware necessary) for the ability >to use UNIX, even if you hardly use any of its features, and the vast majority >of UNIX users don't. Sure they do. Thay may not know it, though. For example, when they recompile their program after adding 1 to an array index, or adding a variable, or changing a variable from single-precision to double precision, and their program still compiles OK (hmmm... what lousy excuse for an OS could he be talking about... :-) >Most people I know who use UNIX ... would probably be far >happier with a much simpler system that did basic computation and networking. >The exceptions to this rule seem to be UNIX support people--hackers, >gurus,etc. >Which raises an interesting question about the ultimate utility of UNIX, but >hey, I'll leave that for another time. Stop. The question you raise is CRUCIAL, and you need to understand the answer, or you will continue to be confused on the "Why UNIX?" issue. Because it's us "support people" that people who think they need a "simpler" system (read: "A system that does exactly what I need, exactly the way I need it done, and with no overhead for stuff I don't use") run to whenever some "simple" thing doesn't work on their "simple" system. And it usually turns out that getting the "simple" thing to work on their "simple" system requires a major rewrite of their system. To people who can be self-sufficient with whatever system they use, I NEVER give advice to switch systems (though I may advise them to investigate alternatives). But people who expect things to just do what they expect them to do, and expect others to make it so... well, they really SHOULD listen to the choices that those "others" would make when choosing their environment (hardware, system software AND application software). So even if I accept your argument that UNIX is useless to end-users (and I do not accept it), the fact that it eases the job of people who have to keep a large number of stations working and a large number of users satisfied is an enourmous plus in its favor. [Very self-constrained flame:] If you could only imagine how tempting the proximity of Lake Michigan is every time I have to deal with a "simple" problem that has a very hairy, time-consuming solution (if it has one at all) -- and I know damned well that the problem wouldn't exist at all if this wasn't a PC in front of me, running MS-DOS! (Oops, I gave away the name of the lousy excuse for an operating system :-) If we could afford alternatives, we would probably have very computer-literate fish in the lake by now. Jacob Gore Gore@EECS.NWU.Edu Northwestern Univ., EECS Dept. {oddjob,gargoyle,att}!nucsrl!gore
breck@aimt.UU.NET (Robert Breckinridge Beatie) (02/23/89)
In article <888@fornax.UUCP>, mcdonald@fornax.UUCP (Ken Mcdonald) writes: > ... > It seems to me that what NeXT did right was in the low-level stuff--hardware, > display software, etc. (If there is anything else to the low-level). What > they did wrong was at the higher levels--OS, to some degree the interface, > maybe a few other things that I'm not aware of yet. This means they can > still rectify their errors. Well, I'll start off by saying that I've never so much as laid hands on a NeXT box. But I am opinionated :-). I'll avoid arguing about whether putting a UNIX-like operating system on the NeXT box was a good idea or not. (But I will say that I think it was a good idea). What I would like to comment on is something more to do with what I guess people call "Software Engineering". Isn't the whole point of implementing systems as layers to allow the designer/implementor to change the LOW LEVEL layers without affecting the layers that the user interacts with? Once people become used to the user interface or the OS, I would assume that you would be reluctant to change the layers that people use. Go ahead and change the low-level layers that people never actually see, but leave the high-level layers alone. Of course higher level layers must be subject to improvement. But if in a new release of a system I am no longer allowed to run more than one process at a time, or pipes no longer work, or awk no longer works, well I'm going to be one unhappy camper. -- Breck Beatie (408)748-8649 {uunet,ames!coherent}!aimt!breck OR breck@aimt.uu.net "Sloppy as hell Little Father. You've embarassed me no end."