brahms@NECAM.tdd.sj.nec.com (George Skillman) (02/08/91)
I just moved to Silicon Valley and I'm amazed at how many UNIX snobs hate OS/2 without knowing a damn thing about what it has to offer. I've been programming under UNIX since Version 6 and I think UNIX is wonderful because it brought fundamental operating system facilities out of ivory towers and into widespread use. I also think UNIX sucks for several reasons. First, it's merciless to novice users who end up having to hire a high paid staff to administrate the damn thing. Second, it's missing a lot of things that should have been added some time ago. For instance, OS/2 is B E T T E R than UNIX in the following respects: 1) OS/2 has threads which are wonderful when you're doing GPI programming. 2) OS/2 has a decent windowing system. UNIX's X Windows sucks! X Windows is device dependent/pixel based. Presentation Manager (PM) is more like PostScript: you have arbitrary coordinate sytems and you can define arbitrary paths that can be filled, stroked, etc. All you can do in X Windows is define polygons, rectangles and elipses. X has a few advantages, like being distributed, but the "feature" of being able to change so much of the behaviour of programs through the resource manager is dangerous at customer sites. PM also supports scalable fonts; X doesn't. 3) OS/2 has REAL debuggers instead of the wimpy crap that comes with UNIX. Microsoft's CodeView is decent enough but Logitech's MultiScope debugger is a very sophisticated debugger with a windowed (both character & PM) interface. 4) Although OS/2 inherited DOS's brain damaged command line tools, there's lots of stuff available from MKS along with fantastic text editors (Brief) from Solution Systems and other companies. 5) OS/2 has both run-time and load-time shared libraries. Gotta go with newer versions of UNIX or SUN/OS to get that. 6) OS/2 and LAN Manager are a heck of a lot easier to administrate than UNIX and TCP/IP/Ethernet. Perhaps all the Yellow Pages, etc, are more powerful, but in an office environment where your secretary often has to run things, simplicity is VERY IMPORTANT! Hey you hot shot engineers! Quit being snooty about all the neat features your products have and worry a little more about whether or not your customers can understand/use them! As to the fact that OS/2 is a single user system: hey, I want my own machine anyway! I've yet to work on a UNIX system that wasn't overloaded. Our network here at NEC is so slow I often think this diskless workstation stuff is for the birds. Look, I agree OS/2 has problems: 1) When I programmed under 1.1, the system kept crashing all the time. 2) OS/2 doesn't provide core files when your program crashes. (Fortunately, Logitech's MultiScope provides capturing core files for you.) 3) There's no command-line "ps" nor "kill". 4) The FAT file system is a joke. There's no way to really check it thoroughly with a UNIX-like "fsck". But the new HPFS file system, if it ever gets out into the mainstream, is pretty slick. 5) Lastly, as everyone knows, OS/2 is currently written in assembly for a machine running in 16 bit mode. Hopefully that will change. And now that I've bashed on UNIX, I want my chance to bash on the money-grubbers that end up steering the course of the computer industry more than we powerless engineers. I think one thing we all agree on is ANYTHING IS BETTER THAN DOS! Look what happened to the PC industry. Apple was doing just great with its innovative machines using a decent microprocessor. In steps IBM (king of hype and the inverse price/performance ratio) fumbling the ball for all mankind by picking the Intel 8088 with *segmentation* and getting Microsoft to provide a *program loader* for an operating system. It's one thing to botch a job when you're the first with a new kind of product but quite another when you have an excellent example like UNIX to follow. Was it really so important to follow CPM's example? Not much later, Apple manages to get the MacIntosh out based on technology Xerox couldn't get their act together on. How many years did it take for Microsoft to just catch up? Now, all three have a chance to atone for their sins. Intel has introduced a decent chip, the 80386, and IBM/Microsoft have introduced a *real* operating system, OS/2. So who's stinking up the works? Microsoft! After all, what besides OS/2 has Microsoft done right? Do they have the best word processor? No. Do they have the best spreadsheet? Only maybe. How's their compiler? More expensive than Turbo C++ and it doesn't even support C++! Are they at the forefront of anything in this industry? Are they any good at making projected release dates? Is their software particularly reliable? A *Best Buy*? No! And now they want us to cling to DOS with Windows patching up the holes and thank them for it? And of course their marketing department will continue to portray them as God's benevolent force in the computer industry. Hey Microsoft! I can see the future where our gandchildren are telepathic with their computers but still have to remember to put double quotes around their "find" commands because DOS 999.999 still doesn't have a command interpreter than can parse command line arguments! Haven't you guys made enough money off of us? For reasons that have nothing to do with technical merit, you're the ones who will probably decide what PCs will be like in the next ten years for the whole damn world! If you won't/can't do a good job at it, let someone else. After all, even IBM at least tried to do that! ================================================================================= George Skillman NEC America brahms@tdd.sj.nec.com Needless to say, my opinions aren't necessarily those of my company. =================================================================================
tom@mims-iris.waterloo.edu (Tom Haapanen) (02/08/91)
[ Note: I've tried to present a reasoned response to George Skillman's ] [ article. Please direct any flames to alt.flame, or to me by email; ] [ they're not needed in comp.os.os2.misc ] George Skillman <brahms@NECAM.tdd.sj.nec.com> writes: > [Unix vs OS/2 comparison deleted] > And now that I've bashed on UNIX, I want my chance to bash on the > money-grubbers that end up steering the course of the computer > industry more than we powerless engineers. Oh boy... > I think one thing we all agree on is ANYTHING IS BETTER THAN DOS! Look > what happened to the PC industry. Apple was doing just great with its > innovative machines using a decent microprocessor. In steps IBM [...] You call the 6502-equipped Apple II an "innovative machine using a decent microporcessor"? No? Then you must be getting confused by the 1981 launch of the PC, and the 1984 launch of the Mac. > In steps IBM (king of hype and the inverse price/performance ratio) > fumbling the ball for all mankind by picking the Intel 8088 with > *segmentation* and getting Microsoft to provide a *program loader* for > an operating system. It's one thing to botch a job when you're the first > with a new kind of product but quite another when you have an excellent > example like UNIX to follow. Unix wasn't really a reasonable example to consider at the time. Back then, it was still a minicomputer operating system which certainly couldn't be made to run on a 16K IBM PC with floppies (remember, the original PC did not support a hard disk). [I can hardly believe that I'm actually justifying the existence of DOS... :)] > Now, all three have a chance to atone for their sins. Intel has introduced > a decent chip, the 80386, and IBM/Microsoft have introduced a *real* > operating system, OS/2. So who's stinking up the works? Microsoft! Here is where you go completely off track. Microsoft has been offering OS/2 for, what, about three years now, with steadily improving versions (1.0, 1.1 and 1.2). However, there have been few major applications written, and the market has responded by not buying OS/2. On the other hand, when Windows 3.0 was introduced in May (with much hoopla, yes, but no more than at the OS/2 intro back in '87), hordes of developers and users rushed to Windows. How can you say that it's Microsoft that's making us use Windows instead of OS/2? We certainly have a choice as consumers. Microsoft is, like most succesful companies, a market driven company. > [Anti-Microsoft flames deleted] > For reasons that have nothing to do with technical merit, you're > the ones who will probably decide what PCs will be like in the next ten > years for the whole damn world! If you won't/can't do a good job at it, > let someone else. After all, even IBM at least tried to do that! Microsoft originally wanted to create OS/2 using the Windows API, but IBM refused, as they basically wanted to go proprietary. Think about that --- what if all those Windows applications out there could just be recompiled to run on OS/2? Might not OS/2 have been more succesful? Microsoft has not abandoned OS/2, and they are working on OS/2 3.0. I can see it becoming a success, as it will run on a variety of platforms, and provide Microsoft's original vision of a Windows API (alongside the PM API), thus allowing thousands of Windows applications to run on OS/2 natively. In conclusion, don't flame Microsoft for doing Windows 3.0; flame all the millions of users out there who are buying Windows instead of OS/2. And maybe even give some thought to *WHY* they are still using DOS and Windows instead of switching to OS/2. [ \tom haapanen --- university of waterloo --- tom@mims-iris.waterloo.edu ] [ comp.binaries.os2 moderator --- comp.windows.ms.* faq list maintainer ] [ "i don't even know what street canada is on" -- al capone ]
resnicks@netcom.COM (Steve Resnick) (02/09/91)
In article <451@nec-gw.nec.com> brahms@NECAM.tdd.sj.nec.com (George Skillman) writes: >Look, I agree OS/2 has problems: > >1) When I programmed under 1.1, the system kept crashing all the time. > I program under 1.2 with my BBS running in the background all the time. It works well. >2) OS/2 doesn't provide core files when your program crashes. > (Fortunately, Logitech's MultiScope provides capturing > core files for you.) > I agree, MultiScope is a cool debugger. I use it frequently. >3) There's no command-line "ps" nor "kill". > There is a program called PSTAT.EXE for OS/2 1.2 + I have a command line kill (I wrote this for my cron deamon to kill it when needed =]) >4) The FAT file system is a joke. There's no way to really check it > thoroughly with a UNIX-like "fsck". But the new HPFS file system, > if it ever gets out into the mainstream, is pretty slick. > HPFS is nice, and OS/2 will magicly check it if it wasn't shut down. There are a lot of DOS utilities which will deal with the FAT system and to some extent, it's a must the DOS be used to do this. For example, if drive C: is a FAT system, you cannot run chkdsk becuase the system has the drive locked. I think a maintenance mode like with UNIX would be handy. >5) Lastly, as everyone knows, OS/2 is currently written in assembly for > a machine running in 16 bit mode. Hopefully that will change. > > OS/2 2.0 is a hybrid of 16bit and 32bit code. >After all, what besides OS/2 has Microsoft done right? Do they >have the best word processor? No. Do they have the best >spreadsheet? Only maybe. How's their compiler? More expensive >than Turbo C++ and it doesn't even support C++! This is something that really annoys me. People who bitch that a C compiler isn't C++, or people who recomment going to C++ instead of C. *I* don't like C++ and could care less that MSC doesn't support it. I do, however, keep hearing claims that Microsloth C is faster, and generates faster and smaller code the Turbo C. This is true on trivial little hacks, but I have taken large masses of code (Like MicroEMACS). If I used the makefile from Dan Lawrence for Microsoft C, it blows up entirely (internal error in code.c - compiler bug??? ) If I turn off optimizations I get a bigger, slower executable for DOS in comparison to the one I get from Turbo C. Too bad there isn't Turbo C for OS/2 ... -- ------------------------------------------------------------------------------- resnicks@netcom.com, apple!camphq!105!steve.resnick, IFNA: 1:143/105.0, USNail: 530 Lawrence Expressway, Suite 374 Sunnyvale, Ca 94086 - In real life: Steve Resnick. Flames, grammar and spelling errors >/dev/null 0x2b |~ 0x2b, THAT is the question. -------------------------------------------------------------------------------
cur022%cluster@ukc.ac.uk (Bob Eager) (02/10/91)
In article <23245@netcom.COM>, resnicks@netcom.COM (Steve Resnick) writes: > In article <451@nec-gw.nec.com> brahms@NECAM.tdd.sj.nec.com (George Skillman) writes: >>2) OS/2 doesn't provide core files when your program crashes. Core files are history. Interactive debuggers are here to stay! > HPFS is nice, and OS/2 will magicly check it if it wasn't shut down. > There are a lot of DOS utilities which will deal with the FAT system > and to some extent, it's a must the DOS be used to do this. For > example, if drive C: is a FAT system, you cannot run chkdsk becuase > the system has the drive locked. I think a maintenance mode like > with UNIX would be handy. By maintenance mode I assume you mean single user mode (in UNIX terms). The OS/2 installation disk makes a good maintenance disk. Use ESC at the IBM logo to escape back to an OS/2 prompt with drive C unlocked. You can make your own maintenance disk using this as a basis, tailoring it to your system. I changed the PMSHELL= in CONFIG.SYS to be CMD.EXE (instead of the installation program), and added the HPFS stuff. I also deleted unwanted drivers (the installation disk needs to boot anywhere and has all of them) and added things like CHKDSK. Magic! >>5) Lastly, as everyone knows, OS/2 is currently written in assembly Is it? Do you have proof? I am sure it is mostly C. Anyway, UNIX has *some* assembler too....I speak as a UNIX user of 16 years. > Turbo C for OS/2 ... I have Zortech C++ for OS/2..... -------------------------+------------------------------------------------- Bob Eager | University of Kent at Canterbury | +44 227 764000 ext 7589 -------------------------+-------------------------------------------------
jwohl@csserv2.ic.sunysb.edu (Jeremy Wohl) (02/10/91)
In article <1991Feb8.142945.14944@watserv1.waterloo.edu> tom@mims-iris.waterloo.edu (Tom Haapanen) writes: >> Now, all three have a chance to atone for their sins. Intel has introduced >> a decent chip, the 80386, and IBM/Microsoft have introduced a *real* >> operating system, OS/2. So who's stinking up the works? Microsoft! > >Here is where you go completely off track. Microsoft has been offering OS/2 >for, what, about three years now, with steadily improving versions (1.0, 1.1 >and 1.2). However, there have been few major applications written, and the >market has responded by not buying OS/2. On the other hand, when Windows 3.0 >was introduced in May (with much hoopla, yes, but no more than at the OS/2 >intro back in '87), hordes of developers and users rushed to Windows. How >can you say that it's Microsoft that's making us use Windows instead of OS/2? >We certainly have a choice as consumers. Microsoft is, like most succesful >companies, a market driven company. Here where I think you are wrong. Microsoft decided their current strategy some time ago. They did *not* provide nearly the "hoopla" that was provided with Windows. In addition, there was an enormous behind-the- scenes developer push with Windows that just wasn't there with OS/2. They left it to IBM, which already has a stereotype to large to shake off and convert customers and most importantly: the press. >Microsoft originally wanted to create OS/2 using the Windows API, but IBM >refused, as they basically wanted to go proprietary. Think about that --- >what if all those Windows applications out there could just be recompiled >to run on OS/2? Might not OS/2 have been more succesful? > >Microsoft has not abandoned OS/2, and they are working on OS/2 3.0. I can >see it becoming a success, as it will run on a variety of platforms, and >provide Microsoft's original vision of a Windows API (alongside the PM API), >thus allowing thousands of Windows applications to run on OS/2 natively. Ick. The Windows API at that time was disgusting (I think it is now,too). And extentions (as will now take place) would not have done much except kludge the system. PM, although not perfect, provides sophisticated and consistent graphics models. Stick the Windows API on top of a real OS? -- Jeremy Wohl / wohl@max.physics.sunysb.edu / jwohl@csserv1.ic.sunysb.edu
resnicks@netcom.COM (Steve Resnick) (02/12/91)
In article <21787.27b47dd7@cluster@ukc.ac.uk> cur022%cluster@ukc.ac.uk (Bob Eager) writes: >In article <23245@netcom.COM>, resnicks@netcom.COM (Steve Resnick) writes: >> In article <451@nec-gw.nec.com> brahms@NECAM.tdd.sj.nec.com (George Skillman) writes: >>>2) OS/2 doesn't provide core files when your program crashes. > >Core files are history. Interactive debuggers are here to stay! > >> HPFS is nice, and OS/2 will magicly check it if it wasn't shut down. >> There are a lot of DOS utilities which will deal with the FAT system >> and to some extent, it's a must the DOS be used to do this. For >> example, if drive C: is a FAT system, you cannot run chkdsk becuase >> the system has the drive locked. I think a maintenance mode like >> with UNIX would be handy. > >By maintenance mode I assume you mean single user mode (in UNIX terms). The >OS/2 installation disk makes a good maintenance disk. Use ESC at the IBM >logo to escape back to an OS/2 prompt with drive C unlocked. You can make >your own maintenance disk using this as a basis, tailoring it to your system. >I changed the PMSHELL= in CONFIG.SYS to be CMD.EXE (instead of the installation >program), and added the HPFS stuff. I also deleted unwanted drivers (the >installation disk needs to boot anywhere and has all of them) and added >things like CHKDSK. Magic! > Although that's a good idea, building the "maintenance disk" is convulted at best, since MS doesn't provide a standard meathod of making a bootable OS/2 diskette. I realize this can be done, my complaint is that MS didn't provide the means of doing it. (I know, samantics. =]) >>>5) Lastly, as everyone knows, OS/2 is currently written in assembly > >Is it? Do you have proof? I am sure it is mostly C. Anyway, UNIX has *some* >assembler too....I speak as a UNIX user of 16 years. > Looks more like C to me. Imagine doing all that structure manipulation in assembler! Ugh! >> Turbo C for OS/2 ... > >I have Zortech C++ for OS/2..... [Flame: ON] Why, when I talk about *C* compilers on the various newsgroups here, do I get told, "Well, so-and-so's C++ compiler is great!" I don't like C++, I will not program in C++, and I will not spend the additional money on a C++ compiler just because it can support C. (Personally, I don't like Zortech, but that's my opinion.) [Flame: Off] C and C++ are different langauges. One supports the other, with a cost. If I were going to spend the money on an OOPL, I would spend it on a *real* OOPL, Smalltalk. :) Cheers! Steve -- ------------------------------------------------------------------------------- resnicks@netcom.com, apple!camphq!105!steve.resnick, IFNA: 1:143/105.0, USNail: 530 Lawrence Expressway, Suite 374 Sunnyvale, Ca 94086 - In real life: Steve Resnick. Flames, grammar and spelling errors >/dev/null 0x2b |~ 0x2b, THAT is the question. -------------------------------------------------------------------------------
fly@ccwf.cc.utexas.edu (neil aggarwal) (02/13/91)
I think that someone out there needs to produce OS/2 32-bit (or a reasonable
facsimile) FAST. With Microsoft jacking around with Windows, they will be able to get
in the market with the product that everyone is looking for. Waiting too long may
be a problem because many people who are using os/2 are forced to find other
operating systems.......
jwohl@eeserv1.ic.sunysb.edu (Jeremy Wohl) (02/14/91)
In article <44067@ut-emx.uucp> fly@ccwf.cc.utexas.edu (neil aggarwal) writes: > > I think that someone out there needs to produce OS/2 32-bit (or a reasonable >facsimile) FAST. With Microsoft jacking around with Windows, they will be able to get >in the market with the product that everyone is looking for. Waiting too long may >be a problem because many people who are using os/2 are forced to find other >operating systems....... I agree that the faster IBM releases 2.0, the better. However, I'm not sure the damage hasn't already been done (or at least when that press conference hits the mags), or that the press hasn't already been sufficiently influenced towards MS' solution. Does anyone know IBM's latest shipping prediction? -- Jeremy Wohl / wohl@max.physics.sunysb.edu / jwohl@csserv1.ic.sunysb.edu