mcdonald@fornax.UUCP (Ken Mcdonald) (02/16/89)
I have some questions about the NeXT which I hope some kind soul will
find time to answer...
1) When and how (in terms of hardware) will NeXT support colour?
2) When will the Renderman standard be incorporated? How will it be
integrated with PostScript? (i.e. will renderman translate three-d
images into PostScript to display them on screen?) Is renderman
powerful enough to be a permanent base for all three-d graphics, or
will it only be a stopgap solution? What the heck is it, anyway?
3) Can NeXT support multiple screens a la the MacII?
4) What is performance like with only the optical drive? A hard drive
would be nice, and I agree their hard drives are a heck of a bargain,
but unaffordable is still unaffordable. At $6500, a NeXT may be just within
my reach. At $8500, I doubt it.
5) What will be the policy towards support of students who buy NeXTs, and then
take the machines with them when they leave campus? If NeXT wants any students
to buy a machine (and I hope they do), this question needs to be answered.
6) How easy is it to program a NeXT? This, of course, is a highly subjective
question, but hey, I'll take highly subjective answers.
7) How well is UNIX hidden? How well will it be hidden in the final release?
Will I be able to say, program, debug, compile, etc, without really having
to think about being in a UNIX environment (aside from I/O, I mean). For
people who have used LightSpeed Pascal 2.0 on the Macintosh, that or something
better is the kind of thing I'm looking for--not just the power, but the ease
of use. I did some programming on a VAX running UNIX a while back, and while
all the tools were there (sybolic debuggers, smart editors, etc) I never really
used them, because wading through the documentation was just not worth it. Thatis the kind of thing I want to avoid. So I guess this question could be, will
the NeXT be like the Mac in that to use the major part of most programs, you
don't even need to read the docs?
8) Any chances any of the compnent prices will come down in the near future?
(Say the next six months?) I realize this is kind of looking a gift horse
in the mouth, but like I say, an unaffordable bargain is unaffordable first,
and a bargain second.
9) Are there any plans to introduce a LCD shutter high-res printer for the
NeXT. Given the inherent simplicity of these devices as compared to the
laser printers, it seems to be the way to go, for that quality of output.
Also, since everyone else is commenting on the source code issue, I suppose
I might as well say something too...
In my opinion, providing source for the NeXT on any except a very restrictive
basis will be death to this machine, at least in terms of the market it is
intended for. Most of the people who are supporting easily available source
seem to be very familiar with UNIX, at least to the point that they start
talking UNIXese, and I can't follow past the first two sentences--and if I
can't follow, do you think an anthropology professor will be able to? If
NeXT releases their source code to all and sundry, then the NeXT box will
become "just another workstation"--a wonderful toy for those willing to leanr
and hack UNIX, and useless, to a large degree, to anyone else. Take a look
at what the lack of standards did to the IBM micro graphics world. Take a
look a what standards, with a strong incentive to follow them, did for Macintoshsoftware. In order for decent commercial programs to be developed, NeXT
MUST NOT be a "moving target". It must be stationary in the same sense the Mac
is, i.e. if you follow the guidelines laid down by NeXT, your program will work
on all NeXTs, and I just hope NeXT has a lot fewer guidelines than Apple!
The need to modify the way the machine works is important, of course, but I
don't think access to the source is necessary to accomplish this. The Mac
OS has lots of "modifications" available to its OS, which all work by
intercepting OS calls, and doing something extra in addition to what the OS
call normally does. This isn't even a formal scheme, it is kind of an ad-hoc
setup to let you do things that couldn't otherwise be done, and it generally
works very well. And if there is a problem, you can just remove the programs
that do the intercepts, and be back to a vanilla, but guaranteed working, system. If NeXT was to adopt a similar idea, but in a more formal manner--i.e. do
compile-time type checking to ensure OS patches take the proper number and
types of args, provide rules to handle what happens when there is more than
one patch to apply to an OS call, and in some manner ensure that the a modified
OS call will always carry out at least the standard functionality built into it,
then I think that would provide a solution for most cases where source would
otherwise have been needed, without compromising the NeXT core.
Well, 'nuff said. Thanks for any comments on the questions I asked.
Ken McDonald
{...ubc-cs!mcdonald@fornax.uucp}wetter@cit-vax.Caltech.Edu (Pierce T. Wetter) (02/20/89)
> > 6) How easy is it to program a NeXT? This, of course, is a highly subjective > question, but hey, I'll take highly subjective answers. > I get asked that question all the time about macintoshes. Here is my answer: It is easier to write good programs and harder to write bad programs. Easier: I dare you to write a windowing system for all your programs. Harder: Whatta mean there's no printf? (Actually, all of the major programming environments come with shells for lamos who can't use TransSkel.) The NeXT machine simply adds UNIX onto all of the mac stuff, (if you ignore the DSP stuff for now) and organizes it somewhat by using Objective C. Did that help? Pierce -- ____________________________________________________________________________ You can flame or laud me at: wetter@tybalt.caltech.edu or wetter@csvax.caltech.edu or pwetter@caltech.bitnet (There would be a witty saying here, but my signature has to be < 4lines)
phd_ivo@gsbacd.uchicago.edu (03/08/89)
> Harder: Whatta mean there's no printf? (Actually, all of the major > programming environments come with shells for lamos who can't use TransSkel.) I am just about to go out and buy myself a NeXt computer---withou having seen one yet. I want to use it as a vanilla UNIX for a while, and slowly get familiar with Next peculiarities. I would therefore appreciate information about: Software: (1) Can I write simple UNIX C programs (i.e. using stdio.h), and run them from some sort of command line? (2) Is there a good symbolic debugger on Mach? (I am currently more familiar with VMS than the various UNIX brands.) (3) Has anyone tried using S on it? Is it difficult to install? Costly? (4) Has anyone used SAS on it? -"- -"- (5) Is there a good Fortran Compiler? -"- -"- (6) Is there a TeX/LaTeX version for it? -"- -"- Does it work well with the NeXt printer? Hardware: (1) How reliably is data stored on the OD? I realize that the head shouldn't crash, but do sectors go bad (often) ? (2) Is the OD usable for normal applications? How does its feel compare, say, to a Macintosh 65ms harddisk? (3) Has anyone made a 9600baud modem out of the DSP? Can it be used to communicate with a USR Courier HST? Any information would be appreciated. ivo welch phd_ivo@gsbacd.uchicago.edu
khera@juliet.cs.duke.edu (Vick Khera) (03/09/89)
In article <2174@tank.uchicago.edu> phd_ivo@gsbacd.uchicago.edu writes: > >I am just about to go out and buy myself a NeXt computer---withou having seen >one yet. I want to use it as a vanilla UNIX for a while, and slowly get familiar >with Next peculiarities. I would therefore appreciate information about: this is absolutely silly. you are about to go spend several thousand dollars on a machine you have never seen nor used. would you go out and buy a car based solely on the manufacturer's description--from a brand new company? i strongly suggest that you spend some significant amount of time with the machine before purchasing one. i was very impressed with what i read about the machine and the demonstrations i saw. when i actually sat down and spent a couple of hours working on the machine, i had some serious doubts about owning one at its current state (i tested it running version 0.8 of the system). i really like the hardware and the user interface, but it needs some serious work in the software speed department. i am sure that it will take some getting used to-- i have used suntools, X11, and MGR on sun's and i like the MGR model the best. the NeXT interface is a bit too Macintosh like for my tasts. just be sure you know what you are getting when you buy the machine. don't trust the sales hype. but you should know that, being in the UofChicago Graduate School of Business ;-) >Software: > (1) Can I write simple UNIX C programs (i.e. using stdio.h), and run them > from some sort of command line? yes. there is a shell window you can run in. >Any information would be appreciated. > >ivo welch phd_ivo@gsbacd.uchicago.edu =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= ARPA: khera@cs.duke.edu Department of Computer Science CSNET: khera@duke Duke University UUCP: {mcnc,decvax}!duke!khera Durham, NC 27706
fad@think.COM (Franklin A Davis) (03/13/89)
In article <2174@tank.uchicago.edu> phd_ivo@gsbacd.uchicago.edu writes: >I am just about to go out and buy myself a NeXt computer---withou having seen >one yet. !!!!!???! >Software: > (1) Can I write simple UNIX C programs (i.e. using stdio.h), and run them > from some sort of command line? Yes. They have both a shell window and a terminal emulator window. They're not as nice as X or even SunTools, but they'll do. From there, it's standard unix. > (2) Is there a good symbolic debugger on Mach? (I am currently > more familiar with VMS than the various UNIX brands.) Yes, they use GDB, which is the Gnu debugger. They also use Objective C, which is object-oriented C based on gcc, the Gnu c compiler. GDB has been enhanced to understand all of the Objective C syntax, and it's an excellent debugger. > ... > (5) Is there a good Fortran Compiler? -"- -"- No -- there's not a good ANYTHING on it yet -- it's a "baby" machine. Well, the C compiler is good, because it's based on Gnu, and GnuEmacs is pretty good (but you can't set the mark with Ctrl-SPACE, so even that software is brain-damaged for the time being). Also, it runs Sun3 binaries (if they don't do graphics) so that gives you some leverage. But there ain't much software -- yet. > (6) Is there a TeX/LaTeX version for it? -"- -"- > Does it work well with the NeXt printer? This might be running at some university already, but I haven't heard -- anyone?? >Hardware: > (1) How reliably is data stored on the OD? I realize that the head > shouldn't crash, but do sectors go bad (often) ? I believe it's very relieable. Time will tell. > (2) Is the OD usable for normal applications? How does its feel > compare, say, to a Macintosh 65ms harddisk? It's a *lot* slower, for intensive disk access. Steve Jobs says that as a vanilla user he doesn't mind it, and I believe him. But for a programmer, the hard disk is a requirement. > (3) Has anyone made a 9600baud modem out of the DSP? Can it be > used to communicate with a USR Courier HST? Wow, what do you want in a couple of months?? Do you have any idea how hard this is??!? Sure, someone will do it... but don't hold your breath! >Any information would be appreciated. I second an earlier comment -- don't buy before you try. I like the machine, but at version 0.8 I would NOT want to rely on it as my workhorse machine (I'm sitting at my Sun). --Franklin franklin a davis Thinking Machines Corp. Cambridge, MA 02142 617-876-1111 <fad@think.com> {topaz, harvard, uunet}!think!fad If you plant ice, you're gonna harvest wind
jhc@vax5.CIT.CORNELL.EDU (03/16/89)
In article <37418@think.UUCP> fad@think.com (Franklin A Davis) writes: >... GnuEmacs >is pretty good (but you can't set the mark with Ctrl-SPACE, so even >that software is brain-damaged for the time being).... Well, seeing as C-SPACE is just a euphimism for C-@, or \000, why don't you try to use C-@ for set-mark? You'll find that it works quite well in Terminal. -JimC -- James H. Cloos, Jr. "Entropy isn't what it used to be." jhc@Crnlvax5.BITNET --c/o Fortune @ batcomputer.UUCP jhc@Vax5.CIT.Cornell.EDU #include <std_disclaimers.h> cornell!vax1!vax5.cit.cornell.edu!jhc@rochester.UUCP B-7 Upson Hall, Cornell Univ., Ithaca, NY 14853 +1 607 272 4519
fad@think.COM (Franklin A Davis) (03/17/89)
In article <18168@vax5.CIT.CORNELL.EDU> jhc@vax5.cit.cornell.edu (James H. Cloos, Jr.) writes: >Well, seeing as C-SPACE is just a euphimism for C-@, or \000, why don't you >try to use C-@ for set-mark? You'll find that it works quite well in >Terminal. Yes, but C-@ is shifted, so it uses three fingers. Miserable hack. --Franklin franklin a davis Thinking Machines Corp. Cambridge, MA 02142 617-876-1111 <fad@think.com> {topaz, harvard, uunet}!think!fad If you plant ice, you're gonna harvest wind
jhc@vax5.CIT.CORNELL.EDU (03/22/89)
In article <37584@think.UUCP> fad@think.com (Franklin A Davis) writes: >In article <18168@vax5.CIT.CORNELL.EDU> jhc@vax5.cit.cornell.edu (James H. Cloos, Jr.) writes: >>Well, seeing as C-SPACE is just a euphimism for C-@, or \000, why don't you >>try to use C-@ for set-mark? You'll find that it works quite well in >>Terminal. > >Yes, but C-@ is shifted, so it uses three fingers. Miserable hack. Except that you don't have to hit the shift key. Just CTRL & 2/@. You reach up rather than down, and it is harder to use your right hand for the second key, but you don't need to use 3 keys. Sorry that I didn't point that out precisely. -JimC -- James H. Cloos,
Gerben.Wierda@samba.acs.unc.edu (Gerben Wierda) (04/03/91)
REPLIES TO: gerben@rug.nl (please) 1. Hey, I don't want to bother this group too much, so I got the NeXTAnswers and the Techdocs from sonata.cc.purdue.edu. But my release 1.0 writenow cn't handle even the INSTALL.wn file, it does not understand most of the pictures. That is fun, because most of the info is essentially 1.0 with a line (valid for 2.0 or not checked for 2.0) added. Anybody any suggestions how to solve this? 2. Sometimes I suddenly cannot dim or lighten the screen anymore, not even from Preferences. Rebooting helps. How come? 3. I did the change "chown uucp Kermit; chmod u+s Kermit" instead of making /dev/cua writable readable for everybody. But now any file I get to the system is owned by uucp instead of me. Isn't chmod on /dev/cua a better solution after all? Or is there another fix? 4. Has anybody a working version of ARC? (I know: use zoo, zip etc, but ARC for unix has a big advantage, it understands the CRNL and the NL textfile representation so I can easy move between various systems) 5. I assume that the Stuart I picked up cannot write the ~/.Stuartrc file because it isn't valid for that (registrate and pay $400 or something like that for just a terminal program) when I haven't registered? Btw. is there a way to keep Stuart from activating when the mouse moves across the window? Gerben -- ============================================================================= Extended Bulletin Board Service, Research & Development Office of Information Technology, University of North Carolina at Chapel Hill internet: bbs.acs.unc.edu or 128.109.157.30
scott@texnext.gac.edu (Scott Hess) (04/05/91)
In article <3210@beguine.UUCP> Gerben.Wierda@samba.acs.unc.edu (Gerben Wierda) writes:
5. I assume that the Stuart I picked up cannot write the ~/.Stuartrc
file because it isn't valid for that (registrate and pay $400 or
something like that for just a terminal program) when I haven't
registered? Btw. is there a way to keep Stuart from activating when the
mouse moves across the window?
Please talk directly to me about Stuart problems. Just to defend myself
in public: There are no copy-protection style limits imposed on Stuart.
When/If you register, I do not send a magic number, or a new version, or
anything of the sort. I do not hold with that type of distribution.
I do not write the .Stuartrc file because there are a couple security
considerations having to do with the Stuartrc default and running Stuart
as setuid root. If you had read the manual pages, you would see this
(to be fair, I've gotten entirely too many questions of this sort :-).
The mouse-autofocussing is done by the MouseFocus default. Clicking
that off with cause future windows not to have that attribute. At
this time, you can't "fix" already running windows - but that will
come in the future (all together now "in Stuart3.0" Due out sometime
last week, but apparently late due to various other work I've been
doing).
Later,
--
scott hess scott@gac.edu
Independent NeXT Developer GAC Undergrad
<I still speak for nobody>
"Simply press Control-right-Shift while click-dragging the mouse . . ."
"I smoke the nose Lucifer . . . Banana, banana."preston@ll.mit.edu (Steven Preston) (04/06/91)
>>>>> In article <3210@beguine.UUCP>, Gerben.Wierda@samba.acs.unc.edu (Gerben Wierda) writes: > 5. I assume that the Stuart I picked up cannot write the ~/.Stuartrc > file because it isn't valid for that (registrate and pay $400 or > something like that for just a terminal program) when I haven't > registered? It is $40 (forty dollars) for an individual license to use Stuart, NOT 400. That may just be a typo, but I thought I would make the point so that people reading this are not scared away from either using or registering Stuart. -- Steve Preston
geoff@ITcorp.com (Geoff Kuenning) (04/08/91)
In article <3210@beguine.UUCP> Gerben.Wierda@samba.acs.unc.edu (Gerben Wierda) writes: > 3. I did the change "chown uucp Kermit; chmod u+s Kermit" instead of > making /dev/cua writable readable for everybody. But now any file I get > to the system is owned by uucp instead of me. Isn't chmod on /dev/cua a > better solution after all? Or is there another fix? I prefer to do "chgrp uucp Kermit; chmod u-s,g+s Kermit" and make things writable by the uucp group. This still means that files will be members of the uucp group, so you probably should remember to do chgrps on them, but at least they're owned by you. -- Geoff Kuenning geoff@ITcorp.com uunet!desint!geoff