Slashdot Mirror


Windows NT Turns 20

An anonymous reader writes with a link to the observation from ZDNet's Mary Jo Foley of Windows NT's 20th birthday (it came out on July 27th, 1993): ""In 1993, Microsoft launched Windows NT 3.1. It was followed up by NT 3.5, 3.51 and 4.0. Microsoft's Windows releases still rely on NT-inspired numbering conventions. Windows 7's build numbers commenced with 6.1; Windows 8's with 6.2; and Windows 8.1 with 6.3." The article also reminds us that "NT's not ancient history, in spite of its age. The NT 'core' is what's inside Windows 8, Windows Server 2012, Windows Phone 8, Windows Azure and the Xbox One.""

213 comments

  1. Lesson One by Anonymous Coward · · Score: 5, Insightful

    The article also reminds us that "NT's not ancient history, in spite of its age. The NT 'core' is what's inside Windows 8, Windows Server 2012, Windows Phone 8, Windows Azure and the Xbox One.

    Indeed. No matter how structurally sound your operating system may be, UI developers (receiving messages from on high) can still make it look like trash.

    1. Re:Lesson One by Anonymous Coward · · Score: 2, Insightful

      How's the kernel you wrote doing these days? Easy to criticise others i guess.

    2. Re:Lesson One by Jeremiah+Cornelius · · Score: 3, Insightful

      The kernel is not structurally flawed.

      It's just as sound as it was, the day Dave Cutler's team built an experimental port of VMS to CMU Mach. It's just as sound a kernel, as the day Microsoft ripped-off VMS from DEC.

      It is the perversion of microkernel VMS by a flawed loadable driver model, and the .DLL nightmare that really sucks, and introduces "unpredictable" behaviors.

      "Hey! PDP-11? Ask me how!"

      --
      "Flyin' in just a sweet place,
      Never been known to fail..."
    3. Re:Lesson One by Anonymous Coward · · Score: 4, Funny

      How's that post you wrote criticizing the post above you these days? I bet you don't think it's so easy to criticize oth-- oh, wait.

    4. Re:Lesson One by goombah99 · · Score: 1, Insightful

      BSD is the core of OSX and it's even older.

      --
      Some drink at the fountain of knowledge. Others just gargle.
    5. Re:Lesson One by EvanED · · Score: 5, Insightful

      You don't need to have done something better to be able to determine whether something is good or bad.

      Now that said, the NT kernel itself is pretty solid.

    6. Re:Lesson One by msobkow · · Score: 2, Informative

      VMS pre-dated BSD substantially, and NT is basically a rewrite of the VMS kernel.

      --
      I do not fail; I succeed at finding out what does not work.
    7. Re:Lesson One by stox · · Score: 1, Troll

      If you call 6 months substantially:

      October 25, 1977 - V1.0 VAX-11/780, Initial commercial release
      March 9, 1978 - 1BSD
      May 1979 - 2BSD
      December 1979 - 3BSD with VAX support. ie. Virtual memory, etc.
      November 1980 - 4BSD

      --
      "To those who are overly cautious, everything is impossible. "
    8. Re:Lesson One by Anonymous Coward · · Score: 1

      The core of OSX is a Mach microkernel, BSD sits on top of Mach. https://developer.apple.com/library/mac/documentation/Darwin/Conceptual/KernelProgramming/Architecture/Architecture.html#//apple_ref/doc/uid/TP30000905-CH1g-CACDCAGC

      OSX has an Unix personality but it isn't a proper one.

    9. Re:Lesson One by epyT-R · · Score: 5, Insightful

      Actually the NT kernel is probably the most well engineered component of modern windows. hell, it is what gave windows things like preemptive multithreading, proper memory protection, and hardware abstraction. The win32 base runtime sits on top of this, and pretty much everything else microsoft has released over the years acted as a wrapper for it. Windows 95 was the attempt to squeeze win32 into 4MB of ram for consumer machines while keeping hardware ports accessible by dos applications. These two goals were fundamentally in conflict with stable and reliable software. The reason we don't have to reboot windows every few hours anymore is due to the windows NT kernel.. As bad as you may think windows to be, it's A LOT better than the days of 3.x/9x.

    10. Re:Lesson One by Anonymous Coward · · Score: 0

      i'm still running nt4 in an isolated vm for a specific application ... never had a problem after all these yrs .. solid.

    11. Re:Lesson One by multi+io · · Score: 1

      BSD is the core of OSX and it's even older.

      VMS pre-dated BSD substantially, and NT is basically a rewrite of the VMS kernel.

      ...as opposed to OSX, which actually shares and reuses code from the BSD kernel.

    12. Re: Lesson One by jeffasselin · · Score: 3, Insightful

      DLL hell wasn't as bad in 2k and XP and is almost entirely gone now since Vista thanks to SxS.

      http://en.wikipedia.org/wiki/Side-by-side_assembly

      --
      If he explores all forms and substances Straight homeward to their symbol-essences; He shall not die.
    13. Re:Lesson One by Darinbob · · Score: 1

      The kernel is not bad, though really it isn't that much like VMS, it wasn't a copy but an OS from scratch. However where Windows got very weird was its wrapping the kernel with the win32 layer. Almost all OS services had to go through the DLLs. While there was a POSIX layer that also existed above the Kernel, you couldn't do very much useful with it because important stuff existed only under Win32 (ie, networking, display, full file support, etc). In essence, the DLLs were the OS.

    14. Re: Lesson One by Anonymous Coward · · Score: 0

      no, you don't need to have done something better, though in your case it would have helped you recognize nt for the intentional brokedick structure it was designed to be.
      OS/2 was and still is superior. but it was more expensive, and people love to buy stuff on sale even if it is garbage.

    15. Re:Lesson One by epyT-R · · Score: 1

      um what? Sure, there are modules (in windows\drivers), most of which are hardware drivers, but the registry is a database, not an api for devices. I suppose the kernel does use it to track device configurations and such, but that's it. Typical applications do not have kernel modules. Many do have services, now, mainly for DRM, which sucks, but services run in userland, not kernel space.

    16. Re:Lesson One by b4dc0d3r · · Score: 1

      The kernel itself is relatively solid, but there are problems.

      Interactive services can expose system-level access to users - a design flaw shich should not be allowed. I remember vaguely a hack from the logon of Windows NT which let you use the context menu, then somehow involving 'print' and/or 'help', you could get explorer.exe open without logging in. That was a flaw with the Win32 implementation, but it had to somehow allow user-level access to MSGINA and the kernel system for authentication and security - a good design would have allowed for this. This is Windows NT or 2000, ignoring XP and beyond.

      I also remember when Vista moved graphics processing into user mode, so that the usual BSOD from graphics drivers famous in XP would simply be an abnormal termination. Reading suggests this was reversed in 7 because of the slowness this added. If graphics - fundamentally the way the OS communicates with the user, since the command-line is supposed to be a second resort - has to be so close to the metal it can't be in user space without slowing it down, this is not good.

      At this point I should say that maybe a command-line interface, under the hood, may be secure. But if the intent is to provide a windowing environment, and the method of doing so is not secure, maybe the kernel has exceeded its usefulness. I should also say that a lot of uninformed people parrot the idea that the kernel is well designed, siply because it flies in the face of all the Microsoft hate. The original nerd hipster, who likes something - or believes that something can be good - even if the masses hate it. Or just because the masses hate it.

      I have personally used a Shatter attack to expose passwords masked by asterisks. There is a single byte in the window definition that says "replace every character with this one because this is a password box". If it is not filled in, the text box is normal. If it is filled in, it's a password box. Most apps that display a password set the password style (by default filling that byte with an asterisk) and put the password up. Easy to recover.

      If different processes, which should be kept separate from other processes, are vulnerable in this fashion, then either the kernel is wrong or the user layer on top of the kernel is not able to maintain the segragation - in effect the user layer is a vulnerability to the kernel.

      Just looking at Vista, and its failure due to UAC and similar security fixes, the fundamental kernel was vulnerable to a number of serious issues - not necessarily with the kernel, but because of its implementation. Having to force an abomination like UAC on programmers so that they would respect security guidelines of least privilage, shows that they did not properly restrict its operation from the beginning - not a flaw in the kernel, but in the overall design. Configured properly, Windows would not run programs, in some cases at all. Configured improperly, it would.

      The overhead of creating a fork() in Windows is ridiculous. As a Windows hacker, I found fork() abhorrent - until someone pointed out the difference in overhead.

      There are many things wrong with Windows. It is fundamentally sound, but there are a good number of other flaws I won't bore you with, that would probably be more convincing but less headline-tastic. The original conception was a good idea, and it has just gotten worse from there. Perhaps it got worse as knowledgeable people left, and as of 2003 it has stopped getting worse.

    17. Re: Lesson One by Anonymous Coward · · Score: 0

      No, it's been fucking entrenched.

      Leaky CRTs, importing dependencies, manifests, ABI changes over compiler versions, no support for fat binaries.

    18. Re:Lesson One by Guy+Harris · · Score: 1

      If you call 6 months substantially:

      October 25, 1977 - V1.0 VAX-11/780, Initial commercial release March 9, 1978 - 1BSD May 1979 - 2BSD December 1979 - 3BSD with VAX support. ie. Virtual memory, etc. November 1980 - 4BSD

      "1BSD" was an add-on to V6 UNIX (which was PDP-11 only), and 2BSD was also based on PDP-11 UNIX, so the "BSD" that contributed to OS X was more like 4.4-Lite, which dates back more to 4BSD and 3BSD than the PDP-11 BSDs..

    19. Re:Lesson One by Guy+Harris · · Score: 1, Informative

      The core of OSX is a Mach microkernel,

      Nothing "micro" about it, sorry.

      BSD sits on top of Mach.

      And rather a lot of the programming interface for kernel modules, and the system call interface to the kernel, comes from the BSD part, not the Mach part.

      OSX has an Unix personality but it isn't a proper one.

      And what might be a "proper" personality for OS X? If you've actually looked at the bits atop the core OS (yes, I have), it's a combination of BSD calls and Mach messaging to other processes.

    20. Re:Lesson One by cold+fjord · · Score: 1
      --
      much of left-wing thought is a kind of playing with fire by people who don't even know that fire is hot - George Orwell
    21. Re:Lesson One by spongman · · Score: 1

      DLLs are entirely user-mode.

      absolutely nothing to do with the kernel.

    22. Re:Lesson One by nukenerd · · Score: 2

      How's the kernel you wrote doing these days? Easy to criticise others i guess.

      Who TF modded this as "Insightful" ?

      Here is a quote (AFAIR) from the great Dr Johnson :- "I can criticise a carpenter for a badly made table, even though I could not make a table myself. It is not my job to make a table."

    23. Re:Lesson One by gdy · · Score: 1

      There is Subsystem for UNIX-based Applications, which is used now instead of the ancient posix layer you mentioned.

    24. Re:Lesson One by benjymouse · · Score: 3, Insightful

      Interactive services can expose system-level access to users - a design flaw shich should not be allowed.

      This was deprecated in Windows 2000/XP (and you had to explicitly allow a service to interact with the user) and it was disallowed in Windows Vista and has been ever since.

      The problem was that a service which typically runs with some form of elevated rights (at least to some resources) could expose those rights/permissions if there was a flaw in the interactive part that communicated with the user.

      Your complaint is interesting, because the risk is very much the same posed by sudo (and every other SUID root) utility: The user is allowed to directly interact with a process which has higher privileges than the user.

      Windows eliminated it. What about SUID root tools, which have historically *many* vulnerabilities and actual exploits and system compromises on its conscience?

      I also remember when Vista moved graphics processing into user mode, so that the usual BSOD from graphics drivers famous in XP would simply be an abnormal termination. Reading suggests this was reversed in 7 because of the slowness this added.

      Sorry, but this is BS. You need to cite sources for that "reading" you have done.

      WDDM is still very much a split driver model where the driver author has to create a (small) kernel mode part and a user model part. The graphics subsystem is still much more stable on Windows than Linux. And even when the driver or hardware fails (e.g. overheated) it merely resets and comes back up, *without* killing any processes, clipboard or services.

      If graphics - fundamentally the way the OS communicates with the user, since the command-line is supposed to be a second resort - has to be so close to the metal it can't be in user space without slowing it down, this is not good.

      Wrong on all counts. Windows allow multiple ways to "communicate with the user". Core versions of the servers, for instance, does not use GUIs. And it is your BS claim that it "is slowing it down". All benchmarks I have seen strongly suggests that Linux has a very hard time keeping up with Windows in this area - even when Windows uses a compartmentalized (and more stable) driver model.

      But if the intent is to provide a windowing environment, and the method of doing so is not secure, maybe the kernel has exceeded its usefulness.

      1998 called. This is not the 9X kernel. The Windows kernel of today runs services in a ''seperate session'' from the users session.

      I should also say that a lot of uninformed people parrot the idea that the kernel is well designed, siply because it flies in the face of all the Microsoft hate. The original nerd hipster, who likes something - or believes that something can be good - even if the masses hate it. Or just because the masses hate it.

      So far you have offered nothing but speculation and outdated myths. Talk about uninformed.

      I have personally used a Shatter attack to expose passwords masked by asterisks. There is a single byte in the window definition that says "replace every character with this one because this is a password box". If it is not filled in, the text box is normal. If it is filled in, it's a password box. Most apps that display a password set the password style (by default filling that byte with an asterisk) and put the password up.

      I don't know a single application that pre fills password boxes with the current password. Would you care to elaborate on that?

      Since Windows Vista, applications running with medium integrity level *cannot* freely send messages to other processes' windows. UAC integrity levels mandate that in order to send

      --
      Reading slashdot one-liner: (irm http://rss.slashdot.org/Slashdot/slashdot).rdf.item | fl title,desc*
    25. Re:Lesson One by Bing+Tsher+E · · Score: 1

      And what might be a "proper" personality for OS X?

      Anal retentive. Obsessed with image over substance. I could go on....

    26. Re: Lesson One by bigdavex · · Score: 1

      When there's a security flaw in the version of the dll a program depends on, how does Windows deal with it?

      --
      -Dave
    27. Re: Lesson One by Anonymous Coward · · Score: 0

      Have you seen the size of the WinSxS folder?
      It grows like grey goo.

    28. Re:Lesson One by Anonymous Coward · · Score: 0

      So "NT" isn't really "New Technology"?

    29. Re:Lesson One by Sponge+Bath · · Score: 1

      Now that said, the NT kernel itself is pretty solid.

      Having written device drivers for Linux, OS/2 and Windows NT-Win7 I find writing drivers for the NT kernel is the least frustrating by far. Stabler driver interfaces, better documentation, better tools (modern windbg for kernel debugging is very nice).

    30. Re: Lesson One by Anonymous Coward · · Score: 0

      Wasn't the OS/2 kernel some strange 16/32 bit hybrid? In any case, it was x86 only -- IBM had to create a Mach-based kernel for the ill-fated OS/2-PowerPC.

    31. Re: Lesson One by plopez · · Score: 1

      OS/2 was originally developed by MS. I think MS went out of their was to kill OS/2

      --
      putting the 'B' in LGBTQ+
    32. Re:Lesson One by colinrichardday · · Score: 1

      Wears black turtlenecks?

    33. Re:Lesson One by UltraZelda64 · · Score: 1

      Too bad it's just a kernel, and pretty much the entirety of the system aside from the kernel sucks ass and is getting worse at a rapid rate...

    34. Re:Lesson One by UltraZelda64 · · Score: 1

      Is the kernel-level DRM as structurally sound as it was, the day Microsoft ripped off VMS from DEC, back when, you know, such user-hostile garbage didn't exist in the system?

    35. Re: Lesson One by doccus · · Score: 1

      But isn't NT really OS/2, anyways? At least the MS OS/2...

    36. Re: Lesson One by mrprogrammerman · · Score: 1

      In theory. But in reality it just created a new DLL hell. http://www.drdobbs.com/windows/no-end-to-dll-hell/227300037

    37. Re:Lesson One by Jeremiah+Cornelius · · Score: 1

      See driver madness, above.

      --
      "Flyin' in just a sweet place,
      Never been known to fail..."
    38. Re:Lesson One by Jeremiah+Cornelius · · Score: 0

      I can see the misunderstanding. I didn't claim they were kernel - just that perceived problems with Windows were seldom actually kernel issues. At least directly.

      Kernel-level drivers, with bad or non-existant threading models, mechanically executed ports from earlier Windows versions and crappy error handling. That has polluted the NTExecutive kernel - an otherwise admirable piece of software.

      The modern Windows kernel is much improved from the original. Design ideas around NUMA that hearken to Cray and SGI are present, and support multi-core in remarkable ways, that differ from the old SMP problems. The support of virtualization is also an ESX rival.

      These ideas were also cleverly pilfered. The most unethical sorts of things happen at MS - and I happen to know of at least two competitive marketing-funded operations, which could be called "dirty tricks" or "black bag" teams.

      There have been brilliant technical leaders in this space beside Cutler - who has been at pasture for years. Bill Laing is salt-of-the-earth. I am fond to recollect the few occasions I had to meet him. Mark Russinovich is also a certifiable genius, and responsible for influencing the additional capabilities in NTExec in impressive fashion.

      We have 25 years, passing from those old days. So? NT Exec was more-or-less VMS. Shall we say a strongly typed derivative? Those who say they don't see the resemblance, are like those who don't know why OS X is like BSD. ;-)

      This brings up another comparison: Modern Windows is an updated, VMS legacy branch in much the way OS X is an updated, NeXTStep branch...

      --
      "Flyin' in just a sweet place,
      Never been known to fail..."
    39. Re:Lesson One by jon3k · · Score: 1

      You mean the one I had millions of dollars to fund the development of? Golly I'll have to check and get back to you.

    40. Re: Lesson One by hairyfeet · · Score: 1

      MSFT didn't kill OS/2, they didn't have to as it was made by IBM, end of story.

      For those that don't know their history after the courts ruled that building a COTS X86 box wasn't an illegal copy of an IBM PC so long as they didn't copy the copyrighted IBM BIOS the guys at IBM tried to take back the X86 market by pushing the MCA bus which was proprietary as hell and would have cost so much for the other OEMs to license that they wouldn't have been able to compete with IBM. To keep from getting slaughtered the "gang of nine" got together and created EISA and licensed it under RAND and it quickly became the standard everyone supported.

      So needless to say when IBM came around to offering OS/2, after it had already failed to gain traction under IBM thanks to IBM sticking with the 286 (which they had a license to second source and produce) instead of 486 (which they did not) the OEMs looked at anything offered by IBM akin to plague blankets and said "not no but HELL no" so OS/2 withered and died.

      But I do find it ironic that everyone tries to pin the death of OS/2 and BeOS on "the big bad MSFT" when if you look at the actual history of their company nearly every success is preceded by "and then the other guy did something REALLY dumb" such as IBM trying to screw the OEMs with MCA and sticking with 286 after it was obsolete to BeOS starting out only supporting a rare niche processor (the AT&T Hobbit chip) to PPC (too expensive and never got the economy scale than X86 has).

      Hell even the ones everybody knows about, wordperfect being replaced by MS Office and IE killing netscape can be traced back to dumb moves on behalf of the parent companies like Wordperfect having bad DOS ports of their WP Office Suite long after Windows became a hit to Netscape shitting all over itself with NS4 thus letting MSFT have the browser win by default, it ALL comes back to "And then the other guy did something REALLY dumb" which let MSFT have a free shot at the title.

      --
      ACs don't waste your time replying, your posts are never seen by me.
    41. Re: Lesson One by Anonymous Coward · · Score: 0

      My guess would be that Windows allows whatever functionality is permitted by the broken library. It's the operating system's job to allocate resources to requests, not to second-guess unintended problems by third-party software.

      When there's a security flaw in the version of a .so a program depends on, how does $otheros deal with it and is it structurally different?

    42. Re:Lesson One by red+crab · · Score: 1

      The comment you posted will be visible only after a System Reboot. Reboot Now?

    43. Re:Lesson One by hobarrera · · Score: 1

      So if I buy a crappy monitor, I need to assemble my own monitors en able to criticise? The same with a car, wine, etc?

    44. Re:Lesson One by Guy+Harris · · Score: 1

      The comment you posted will be visible only after a System Reboot. Reboot Now?

      And you're absolutely certain those messages from Windows are the result of structural flaws in the NT kernel (rather than problems well above the kernel)? If not, that's the equivalent of saying whatever annoys you about {GNOME,KDE,whatever command-line shell you're using,etc.} on Linux is the fault of structural flaws in the Linux kernel.

    45. Re:Lesson One by Guy+Harris · · Score: 1

      Wears black turtlenecks?

      Says "boom!" a lot?

    46. Re:Lesson One by drsmithy · · Score: 1

      It's just as sound as it was, the day Dave Cutler's team built an experimental port of VMS to CMU Mach. [sympatico.ca] It's just as sound a kernel, as the day Microsoft ripped-off VMS from DEC.
      And by "ripped off" you mean "hired the team responsible for building it and employed them to build an OS", right ?

    47. Re:Lesson One by drsmithy · · Score: 1

      At a programmatical level, there's no such thing as "user-hostile".

      Do you have a better way of meeting the requirement that DRM presents ?

    48. Re:Lesson One by Jeremiah+Cornelius · · Score: 1

      And duplicated work belonging to DEC, and went through losses in court because of it. Yes.

      --
      "Flyin' in just a sweet place,
      Never been known to fail..."
    49. Re:Lesson One by Anonymous Coward · · Score: 0

      It's just as sound a kernel, as the day Microsoft ripped-off VMS from DEC.

      Didn't Cutler design a replacement for VMS but in their wisdom DEC cancelled it to save money which lead to Cutler moving to Microsoft?

    50. Re:Lesson One by Jeremiah+Cornelius · · Score: 1

      That is the pro-MS spin, on the publicly produced history - written up by Helen Custer as history and accolade. Like all lies and cover stories, it has verifiable elements of veracity.

      She was paid on MS dime, and published by MS press.

      Victors write the history, YMMV, etc.

      PRISM was the "VMS.next" Cutler was working on. Three aims separated it from earlier VMS, from which it was strongly derived/forked:
      - 64 Bit
      - Portability for RISC
      - Posix

      All of these are targets for which DEC released various "OpenVMS" versions, beginning in 1990.

      --
      "Flyin' in just a sweet place,
      Never been known to fail..."
    51. Re:Lesson One by durval · · Score: 1

      The article also reminds us that "NT's not ancient history, in spite of its age. The NT 'core' is what's inside Windows 8, Windows Server 2012, Windows Phone 8, Windows Azure and the Xbox One.

      Indeed. No matter how structurally sound your operating system may be, UI developers (receiving messages from on high) can still make it look like trash.

      Structurally sound? Are you nuts? NT (and all the crap MS did since they abandomned Xenix) is unsound structure with unsound UI on top of it...

      --
      Best Regards,
      Durval Menezes.
      I have never met a computer that didn't like me.
  2. NT 3.51 was the best kernel by JoeyRox · · Score: 4, Interesting

    I've seen the source and it's a work of art. Whoever they had working on NT 4 for the PnP and other additions really massacred the code.

    1. Re:NT 3.51 was the best kernel by Anonymous Coward · · Score: 0

      I'm thinking that Dave Cutler had something to do with the 3.5.1 kernel work

    2. Re:NT 3.51 was the best kernel by Anonymous Coward · · Score: 4, Insightful

      I didn't see the source but could definitely tell. I remember NT 3.51 being very responsive even when a program was misbehaving. If I recall I had it run 16-bit programs in their own memory so they didn't affect each other. NT 4.0 did seem like PnP was just crammed in along with the Windows 95 interface. I feel old.

    3. Re:NT 3.51 was the best kernel by Hartree · · Score: 1

      That was my sense as well. Especially for the DEC Alpha version, NT 3.51 seemed quite stable.

    4. Re:NT 3.51 was the best kernel by 0123456 · · Score: 0

      That was my sense as well. Especially for the DEC Alpha version, NT 3.51 seemed quite stable.

      Not for me. At the time, I was working for a mixed Linux/NT company and the NT machines would regularly blue-screen just reading email, whereas my Linux machine was only rebooted for hardware upgrades.

    5. Re: NT 3.51 was the best kernel by jeffasselin · · Score: 2

      Oneof the major issues wit 4.0 was how they switch the video drivers to ring 0 for better performance, but most video drivers in those days were total crap. Things have gotten better.

      --
      If he explores all forms and substances Straight homeward to their symbol-essences; He shall not die.
    6. Re:NT 3.51 was the best kernel by Hartree · · Score: 2

      We were using it for DEC Alpha servers doing credit card transactions rather than user machines. at least for that it did pretty well. We only took it down once in 3 years to upgrade the drives.

      Some form of Unix could have done the same work with half the machine, but others wrote the system before I came on the scene.

      I had Linux as one of the OS's on my workstation and liked it a lot. For the Unix servers processing check transactions and doing database work (Oracle) we used SCO instead as we were using X25 and multi-serial port cards for some of the communications and terminal serving. Linux just didn't have the drivers for them in those days.

      The place was an absolute dog's breakfast of different OS's from SCO to OS2 to every version of windows and dos you can imagine.

      It was one of those "character building experiences".

  3. It's no OS/2 by Anonymous Coward · · Score: 0

    Where's my HPFS?!

    1. Re:It's no OS/2 by Anonymous Coward · · Score: 0

      I'd take HPFS/386. A filesystem that can be set so secure, one can deny all access to a file... and no user on the box could delete that file, ever. It even encrypted stuff, so a hex editor couldn't recover files that one didn't have permissions for.

    2. Re:It's no OS/2 by yuhong · · Score: 1

      Early versions of WinNT did support it.

  4. Not the worst thing ever... by Anonymous Coward · · Score: 0

    At least I could run an Oracle data base on it and it had NTFS... The Virtual Memory/Swap file behavior had me pulling my hair out for a while... sigh
    Damn I hate it when I wax nostalgic

  5. OS/2 was better by jfdavis668 · · Score: 1

    I used both NT 3.5 and OS/2, and OS/2 was much better. Microsoft put more effort into support for the long term.

    1. Re:OS/2 was better by symbolset · · Score: 1

      IBM contracted Microsoft to do the update to OS/2 and that was the end of that.

      --
      Help stamp out iliturcy.
    2. Re:OS/2 was better by jfdavis668 · · Score: 0

      Microsoft made OS/2 1.0-1.3. I am referring to 2 on. Those were the versions competing with Windows NT.

    3. Re:OS/2 was better by yuhong · · Score: 1

      Ah the MS OS/2 2.0 fiasco. FYI, even MS never intended OS/2 NT to replace OS/2 2.x initially.

    4. Re:OS/2 was better by dfghjk · · Score: 1, Informative

      No they didn't. OS/2 1.x was primarily done by IBM using IBM's tools. Many portions were so poorly understood by MS that they wouldn't change any of it or release its source to OEMs.

      OS/2, from the very beginning, was by IBM for IBM. It was an OS designed for the 286, a processor designed by IBM for IBM, and it was defeated by Intel who took ownership of the 32-bit follow-on processor and by MS, who took ownership of the 32-bit follow-on OS. IBM was solely responsible for the trash that was OS/2 and the 286. MS was on the good side of that fight and we are better off for it.

      As an engineer that had to fly to Redmond to work on pre-1.0 OS/2 ports, I am quite familiar with who wrote it and what MS thought of it. It's quite clear that you are not.

    5. Re:OS/2 was better by Anonymous Coward · · Score: 3, Interesting

      OS/2 1.x was primarily done by IBM using IBM's tools

      After v1.0. Microsoft was entirely responsible for the initial releases of OS/2.

      the 286, a processor designed by IBM for IBM

      Wrong. Intel designed the 80286. IBM (along with other chip makers) made their own version later on.

    6. Re:OS/2 was better by yuhong · · Score: 1

      by MS, who took ownership of the 32-bit follow-on OS.

      Of course, this was before they turned this OS/2 2.0 project into an entire fiasco that I have very bad words against MS about.

    7. Re:OS/2 was better by TapeCutter · · Score: 1

      IIRC it was called a "partnership", IBM wanted a piece of the Microsoft's O/S pie, MS told them it was down a dark alley. I recall seeing an early version at a trade show running four video files simultaneously, which believe it or not was an impressive stunt for a desktop back then. From what I've heard OS2 was reincarnated for use in ATM networks.

      --
      And did you exchange a walk on part in the war for a lead role in a cage? - Pink Floyd.
    8. Re:OS/2 was better by Anonymous Coward · · Score: 0

      You're a clown. The 80286 processor was designed by Intel.

  6. It was originally a pretty good design by msobkow · · Score: 5, Interesting

    Originally it was a pretty good design, based on the concepts implemented by DEC's VMS system. It only got butchered later by people who didn't know their stuff as well as the original engineers.

    Warts and all, Windows owes it's lineage to VMS and the once mighty DEC.

    I've heard there are still places running VMS-based hardware.

    --
    I do not fail; I succeed at finding out what does not work.
    1. Re:It was originally a pretty good design by Anonymous Coward · · Score: 0

      I've heard there are still places running VMS-based hardware.

      A certain national medical lab -still- uses VMS systems to issue lab reports to thousands of medical facilities. One reason is that they use Digital Standard MUMPS for scripting the collection and issuing of the data. It's worked well for a quarter century, why replace it with the unknown and potentially buggy?

    2. Re:It was originally a pretty good design by Trepidity · · Score: 4, Informative

      I've heard there are still places running VMS-based hardware.

      Hell, VMS-on-VAX was Digital's replacement for the PDP line of minicomputers (phased in in 1977), and even their predecessors are still running in a few places.

    3. Re:It was originally a pretty good design by SlashDread · · Score: 1

      I'm sure OpenVMS is in use, running some COBOL for a bank orso. In 1995 VAX/VMS was HUGE in network cluster tech.

      I was right in the middle of that, since we were pitching a VAXcluster box against Appleshare and WinNT for serving Computer Based Training for Fokker Aircraft (now sadly defunct).

      Our requirement was to serve a classroom full of Apple Mac's (Apple OS 6.x, oh good times), Appleshare won.
      VAX/VMS choked on 2Mb/s on a 10Mb ethernet. Still I loved the architecture, and it sparked by furure *nix interest.

      The rumour in the day was that the WinNT team was basically the former VMS team, and that WNT was really VMS-rot-1 :)

    4. Re:It was originally a pretty good design by TAZ6416 · · Score: 1

      I'm the sysadmin for a cluster of 3 VAXserver 4000's, they haven't been switched off since 1994. They once supported over 1000 users doing various stuff but now sadly all they do is some legacy COBOL stuff and print jobs.

    5. Re:It was originally a pretty good design by cold+fjord · · Score: 1

      I've heard there are still places running VMS-based hardware.

      There is still plenty of VMS running out there. There are still places that have made serious investments in new VMS hardware within the last couple of years. From what I have seen VMS sites/users tend to be as loyal as Unix users. There are still some things about VMS that Unixland lags on.

      That is going to start drying up now that HP has announced EOL for VMS.

      It would be interesting to see HP sell it to someone. They never really seemed to care for it.

      --
      much of left-wing thought is a kind of playing with fire by people who don't even know that fire is hot - George Orwell
    6. Re:It was originally a pretty good design by Anonymous Coward · · Score: 0

      I detest even logging onto it.

      But one of our customers has it in production. It;s running Oracle 9...

      They cant turn it off. They are locked in for a few years. This server/database combo (both heavily out of support now...) are running BioTech medical experiments. The Experiments last for a duration of 10 years and any interruption to service jeopardizes the whole experiment.They would spend weeks ... months maybe ... trying to justify that an experiment should not be abandoned with "unverifiable data" for an outage.

      The cost of keeping this rubbish up and running is huge. The cost of replacing it before end of lifecycle tho... thats almost incalculable.

  7. Here's to Kernels by lobiusmoop · · Score: 4, Funny

    The Linux kernel would have bought it a beer, but it hasn't turned 21 yet.

    --
    "I bless every day that I continue to live, for every day is pure profit."
    1. Re:Here's to Kernels by SimonInOz · · Score: 2, Interesting

      Legal drinking age is 18 in Finland (and much of the civilised world, actually. USA is kinda weird. Mind you, allowing driving and drinking at the same time, does that sound like a good idea? Maybe they are right. No, surely not).

      No, wait, Linux first release was 1991, that makes it, um, 22. What the heck is wrong with your arithmetic? What do they teach at school these days? Bah. Get off my lawn. (And yes, I did program PDP 11s back in the day. Why do you ask?)

      --
      "Cats like plain crisps"
    2. Re:Here's to Kernels by Anonymous Coward · · Score: 0

      And yes, I did program PDP 11s back in the day. Why do you ask?

      Which in Finland was likely the late 90s. Country was pretty primitive until Nokia's money permeated out and it joined the first world.

    3. Re:Here's to Kernels by colinrichardday · · Score: 3, Informative

      Perhaps the GP meant that the NT kernel hasn't turned 21 yet.

    4. Re:Here's to Kernels by thegarbz · · Score: 1

      USA is kinda weird.

      No kidding. You're allowed to vote, own a gun and are considered an adult. But don't you dare to drink a beer.

  8. Windows NT's name by norite · · Score: 5, Interesting
    In case anyone was wondering what NT stood for

    V +1 = W
    M + 1 = N
    S + 1 = T

    --
    -- Fuck Beta
    1. Re:Windows NT's name by EvanED · · Score: 1

      Haha, that's pretty funny. I hadn't seen that before.

      Doubt it was their motivation, but it's funny regardless.

    2. Re:Windows NT's name by Anonymous Coward · · Score: 0

      That's like IBM:

      H+1=I
      A+1=B
      L+1=M

    3. Re:Windows NT's name by Tynin · · Score: 1

      In case anyone was wondering what NT stood for

      V +1 = W M + 1 = N S + 1 = T

      I forget where I read it, but I recall NT standing for New Technology.

    4. Re:Windows NT's name by norite · · Score: 1

      That's what a lot thought....but when you boot up NT 4.0 you get the 'Based on NT Technology' splash screen.

      --
      -- Fuck Beta
    5. Re:Windows NT's name by 0123456 · · Score: 1

      I forget where I read it, but I recall NT standing for New Technology.

      My, possibly time-addled, memories from the OS/2 days were that the future of the PC was going to be OS/2 NT (OS/2 New Technology), but then Microsoft announced that they were releasing a new OS called Windows NT, which was nothing to do with OS/2 New Technology at all, no sir.

    6. Re:Windows NT's name by Anonymous Coward · · Score: 0

      You're confusing NT 4.0 with NT 5.0 (Windows 2000). The NT 4.0 splash screen never had any such verbiage, as a quick google search will confirm.

    7. Re:Windows NT's name by CannonballHead · · Score: 1

      Well, you know, that New Technology Technology is what powers Automated Teller Machine Machines.

    8. Re:Windows NT's name by Anonymous Coward · · Score: 0

      > Doubt it was their motivation,

      none the less, this explaination has been around since the
      end of the collaboration with OS/2. look it up!

    9. Re:Windows NT's name by dgsoftnz · · Score: 1

      In case anyone was wondering what NT stood for

      V +1 = W M + 1 = N S + 1 = T

      Not quite. Windows NT was initially developed for the Intel i860 CPU Architecture and ported to MIPS/i386/Alpha/PowerPC later. The codename for the i860 XR processor was "N10" or N-Ten.

    10. Re:Windows NT's name by Guy+Harris · · Score: 1

      Well, you know, that New Technology Technology is what powers Automated Teller Machine Machines.

      ...into which you type your Personal Identification Number Number.

    11. Re:Windows NT's name by Frankie70 · · Score: 1

      You are missing the Point. The architect of VMS became the architect of WNT - Dave Cutler.

    12. Re: Windows NT's name by Anonymous Coward · · Score: 0

      That was the point??

    13. Re:Windows NT's name by Anonymous Coward · · Score: 0

      Which is verified on-line using an Network Interface Card Card

    14. Re:Windows NT's name by jon3k · · Score: 1

      I can't imagine they would actually have so little foresight as to call it "New Technology". I mean, what comes next? "Newer Technology" ? "New-New Technology" ?

    15. Re:Windows NT's name by kriston · · Score: 1

      Windows NT was called that because it was being targetted to a CPU family architecture nicknamed "N-Ten" after the Intel i910 processor. Intel produced a more suiltable follow-on processor in this family called the i860 which was the chip the Microsoft team eventually settled on, but the "N-Ten" name stuck.

      The newer marketing name "New Technology" was a so-called "backronym."

      --

      Kriston

  9. Seriously? by BobNET · · Score: 4, Informative

    An article for WinNT turning 20, but nothing for Slackware when it did the same 10 days ago? What is wrong with you, Slashdot?

    Wait, don't answer that...

    1. Re:Seriously? by Anonymous Coward · · Score: 0

      And which one has turned out to be more relevant to the industry? Yup, Windows. And on the Linux front (for all it's Unity based sins) Ubuntu.

      And yes, I was using Slackware in 1994. Before I started using NT 4 in 1995/6.

      Slackware, great at the time. Pretty much irrelevant now.

    2. Re:Seriously? by peragrin · · Score: 1, Insightful

      It takes a lot of skill to keep windows running for 20 years.

      a 24 year old could have run slackware for the last 20 years.

      --
      i thought once I was found, but it was only a dream.
    3. Re:Seriously? by SJHillman · · Score: 0

      If you're going to say Ubuntu is more relevant, you'd have to include the qualifier of "currently". Given how distros rise and fall, for all we know, a Slackware distro could be what eventually brings the year of linux on the desktop. And I'd say Android has made a *much* bigger splash than Ubuntu.

    4. Re:Seriously? by Anonymous Coward · · Score: 0
      Which one is more relevant to Slasdot's original purpose?

      The drift towards Slashdot's endorsement of proprietary in general and Microsoft in particular is no coincidence.

    5. Re:Seriously? by jones_supa · · Score: 2

      An article for WinNT turning 20, but nothing for Slackware when it did the same 10 days ago? What is wrong with you, Slashdot?

      Wait, don't answer that...

      You could have submitted it to Slashdot. That's how this place runs...

    6. Re:Seriously? by Anonymous Coward · · Score: 0

      Slackware is alive?!

    7. Re:Seriously? by Anonymous Coward · · Score: 0

      a 24 year old could have run slackware for the last 20 years.

      Oh no ... my secret is out. I installed SW 20 years ago when I was 24. And I'm still 24 today.

      Slackware: the fountain of youth.

  10. DLL nightmare by ArchieBunker · · Score: 1, Insightful

    I have yet to experience this DLL nightmare you speak of. I've had way more dependency hell on Linux than anything. Say you find a great program that does exactly what you need. Well the author based it off some obscure library that needs a dozen other dependencies. One of those said dozen fails to compile. I'm not a CS major so the story pretty much ends there.

    The only DLL issue I've had was getting some of the cygwin tools. It needed some DLL and their site was useless for supplying it. I just need the one file, not their installer giving me the entire dev environment. In the end I searched for the filename and "index of" and found a copy that way.

    I still don't understand how VMS can be compared to NT. They don't even seem remotely similar.

    --
    Only the State obtains its revenue by coercion. - Murray Rothbard
    1. Re:DLL nightmare by cartel1982 · · Score: 1

      The parent post was talking about engineering issues, not end user issues. Computers look different if you're working on their insides.

    2. Re:DLL nightmare by Jeremiah+Cornelius · · Score: 2

      Debug an NT device driver.

      Hey! I recognize this!

      --
      "Flyin' in just a sweet place,
      Never been known to fail..."
    3. Re:DLL nightmare by 0123456 · · Score: 3, Interesting

      I have yet to experience this DLL nightmare you speak of. I've had way more dependency hell on Linux than anything. Say you find a great program that does exactly what you need. Well the author based it off some obscure library that needs a dozen other dependencies

      Fortunately, 'apt-get install great-program' always works for me.

      I still don't understand how VMS can be compared to NT. They don't even seem remotely similar.

      VMS
      WNT

      Just as HAL was one step ahead of IBM, Windows NT is one step behind VMS.

    4. Re:DLL nightmare by Anonymous Coward · · Score: 0

      Some of the people that make comments here have obviously not coded in the platform for at least 10 years.

    5. Re:DLL nightmare by Anonymous Coward · · Score: 1

      I think you need to brush up on your English. The letters HAL precede IBM, which means it's one step behind.

    6. Re:DLL nightmare by Anonymous Coward · · Score: 0

      Heh thanks, I never noticed that, puts HAL's whole personality issues in perspective

    7. Re:DLL nightmare by ArchieBunker · · Score: 1, Informative

      Oh why didn't I think of that? Because a package was not available! Hence my need of compiling from source.

      VMS=Virtual Memory System
      WNT=Windows New Technology

      Your point?

      --
      Only the State obtains its revenue by coercion. - Murray Rothbard
    8. Re: DLL nightmare by Anonymous Coward · · Score: 1

      Just ask Tron.

    9. Re:DLL nightmare by 0123456 · · Score: 1

      I think you need to brush up on your English. The letters HAL precede IBM, which means it's one step behind.

      I think you need to read more Clarke.

    10. Re:DLL nightmare by noh8rz9 · · Score: 0

      VMS=Virtual Memory System WNT=Windows New Technology

      Your point?

      the GP is being obtuse, but he's making a snarky comment that the letters in the acronym WNT are offset by one from the letters in the acronym VMS (V->W, etc). this is similar in vaguely tenuously stretch of a way to the novel (and movie) 2001 featuring the computer HAL. Wags at the time noted that HAL was a similar letter shift from IBM. That's it, nothing to see here, move along.

      --
      let's have a conversation! let me know what you think.
    11. Re:DLL nightmare by Anonymous Coward · · Score: 1

      That might prove difficult if he uses the English alphabet from Z to A.

    12. Re:DLL nightmare by benjymouse · · Score: 1, Insightful

      Fortunately, 'apt-get install great-program' always works for me.

      Talk about a flawed model. Instead of dealing with the problem (reference management), Linux repositories sweeps the problem under the rug.

      Instead of coming up with a solution which can actually *both* allow shared libraries to be used *and* support a heterogeneous software environment, Linux repositories forces an artificially homogeneous environment.

      Instead of a software package supporting d distros and v versions by virtue of an OS supported library broker, you are forced to create d*v versions of the software package. This is the reasons so many drivers and applications break between versions: Nobody can test every distro and every (supported) version of every distro. The software authors typically will only test a few of the most popular distros and often only the most recent version. Compatibility errors crop up, but the exercise of finding them is left to the users.

      The basic idea of a software repository with signed packages is sound. (Mis)using software repositories by cloning them to address the problem with dependency hell is stupid. The model does not scale. And it *still* does not solve the problem. Step outside of the distro and version specific (!) software repository and you are back in dependency hell, knee-deep. Software repositories are the Matrix. It is a dream world that has been created to disguise the ugly problem which still exists in Linux: Dependency hell.

      Contrary to the ignoramuses still beating the DLL hell drums, the problem has been solved the right way in Windows: Side by side assemblies (SxS). From "Windows Internals":

      Fusion (SxS) Redirection. Fusion (also referred to as side-by-side, or SxS) is an extension to the Windows application model that allows components to express more detailed binary dependency information (usually versioning information) by embedding binary resources known as manifests. The Fusion mechanism was first used so that applications could load the correct version of the Windows common controls package (comctl32.dll) after that binary was split into different versions that could be installed alongside one another; other binaries have since been versioned in the same fashion. As of Visual Studio 2005, applications built with the Microsoft linker will use Fusion to locate the appropriate version of the C runtime libraries.

      Basically, as an application developer you have two choices which both avoid dependency hell: Ship the specific version of the DLL with your application (but do *not* install in Windows/System32) or request the installer installs the DLL as a SxS assembly and then reference it from your application manifest. The SxS cache can hold multiple versions of the same DLL and the manifest can describe the policy for brokering the version (whether the app always wants the latest or a specific version, for instance). This allow patching to be performed on the central SxS cache instead of each application with separate libraries.

      There is a reason GP has not experienced "DLL nightmare". It doesn't exist any more. These days it is but a myth that detractors desperately want to keep alive.

      --
      Reading slashdot one-liner: (irm http://rss.slashdot.org/Slashdot/slashdot).rdf.item | fl title,desc*
    13. Re:DLL nightmare by serviscope_minor · · Score: 2

      Talk about a flawed model. Instead of dealing with the problem (reference management), Linux repositories sweeps the problem under the rug.

      I strongly suspect that you do not understand the problem fully.

      I then read your post and became convinced.

      Linux has no .SO or dependency hell as you seem to think.

      On a fine grained level, libraries support symbol versioning (e.g. libstd++) so a new libstdc++ can link against an old program, provided the versions do not drift too far. On a coarser level, you have shared object versioning, so you can have multiple different versions of a .so in a system all within the same path.

      On the coarsest level, you can override linker paths manually, so if you want to override a SO include or the other mechanisms can't cope for some reason then it still works.

      Once you understand this it is easy to package a binary that runs on basically every Linux distro out there. People who understand include Mozilla, libreOffice, OpenOffice, Mathworks and a whole slew of others. I clearly understand because I can ship binary code with cross distro compatiblity.

      Basically it works and it works well.

      The thing is that if you're not using a system library, then you need to ship the libraries you're using. This is a minor hassle (compared to the normally trivial ./configure&&make&&sudo make install) and worse it means every tiny utility loads its own copy and takes up that little bit more memory. The thing about having everything on the same .so version if possible is it is really much more efficient for memory use. To package everything separately would involve having vast numbers of duplicates of libraries and would have the lovely upgrade hell that you get on OSX and Windows where every program nags you for updates.

      --
      SJW n. One who posts facts.
    14. Re:DLL nightmare by benjymouse · · Score: 1

      All popular package managers, including APT, RPM and the FreeBSD Ports Collection, suffer from the problem of destructive upgrades. When you perform an upgrade -- whether for a single application or your entire operating system -- the package manager will overwrite the files that are currently on your system with newer versions. As long as packages are always perfectly backward-compatible, this is not a problem, but in the real world, packages are anything but perfectly backward-compatible.

      (from http://archive09.linux.com/feature/155922)

      We have all been bitten by this. It is not that prevalent on servers (what I use Linux for), but desktop distros definately suffers from this.

      also:

      http://www.linuxquestions.org/questions/linux-newbie-8/dependency-hell-233892/

      http://fixunix.com/suse/265243-dependency-hell.html

      http://forums.fedoraforum.org/showthread.php?t=289520&goto=nextnewest

      http://www.brighthub.com/computing/linux/articles/52282.aspx

      Dependency hell is real. Package managers help mitigate the problem locally, but software repositories are distro and version specific. Once you step outside of the official repository you risk stepping into dependency hell, as many of the above experienced. I fully concede that it *shoudn't* be a problem if every piece of software author did what they were supposed to do. It is the OSs job to provide a service and guidance so that authors do the right thing. Linux is not quite there yet, unfortunately. That's why you see the typical advice to install from the "official" distro and version specific repository. But that model is flawed.

      --
      Reading slashdot one-liner: (irm http://rss.slashdot.org/Slashdot/slashdot).rdf.item | fl title,desc*
    15. Re:DLL nightmare by benjymouse · · Score: 1

      http://forums.linuxmint.com/viewtopic.php?f=47&t=99890

      http://fixunix.com/redhat/489285-rhel-5-1-rhel-5-2-dependency-hell.html

      http://www.freelists.org/post/oracle-l/Linux-X86-26-Oracle-9204-and-RPM-Dependency-Hell

      http://unix.stackexchange.com/questions/53077/firefox-circular-dependency-hell-on-linux-mint-13

      http://linuxgazette.net/issue71/tag/3.html

      http://lwn.net/Articles/198455/

      http://forums.opensuse.org/english/get-technical-help-here/applications/406017-does-dependency-hell-still-exist-2.html
      * "The best (worst) way to get dependency hell is to add too many repositories. Eventually one will find the right combination of conflicting repositories to create a situation of dependency hell."
      * "The problem is the more repositories that are added, the less and less likely that applications are built against the same set. "
      * "Dependency hell ALWAYS exists. ALWAYS (Yes, Debian users, I'm talking to YOU). The question is whether or not if somebody tries to NOT go to dependency hell, is the system going to take them there anyway."

      http://unix.stackexchange.com/questions/32594/fedora-16-dependency-hell

      http://www.eonlinegratis.com/2013/dependency-hell-trying-to-install-gcc-on-centos4/

      Oh, dependency hell is real alright!

      --
      Reading slashdot one-liner: (irm http://rss.slashdot.org/Slashdot/slashdot).rdf.item | fl title,desc*
    16. Re:DLL nightmare by Anonymous Coward · · Score: 0

      You got modded down. I have no idea why.

      Moderators of Slashdot: grow the fuck up, you are the cancer that is killing this board.

    17. Re:DLL nightmare by Anonymous Coward · · Score: 0

      I have yet to experience this DLL nightmare you speak of. I've had way more dependency hell on Linux than anything. Say you find a great program that does exactly what you need. Well the author based it off some obscure library that needs a dozen other dependencies

      Fortunately, 'apt-get install great-program' always works for me.

      I still don't understand how VMS can be compared to NT. They don't even seem remotely similar.

      VMS
      WNT

      Just as HAL was one step ahead of IBM, Windows NT is one step behind VMS.

      Um... You might want to review the alphabet.

    18. Re:DLL nightmare by klui · · Score: 1

      > I have yet to experience this DLL nightmare you speak of

      Haha. https://en.wikipedia.org/wiki/DLL_hell

    19. Re:DLL nightmare by Anonymous Coward · · Score: 0

      I use Linux professionally and the few times that I had a package that was not available, it was compiled and installed in less than 2 minutes. In a source tree, there's usually a file called "INSTALL" which clearly lists all dependencies. Dependencies can usually be installed with an 'apt-get'.

      The fact that on a linux system every piece of reusable software has to be installed only one leads to a dramatic amount of saved disk space. Linux is much cleaner in this way. This smart way of doing things is exclusive to open source software, however. Every compiled library that is shared among programs or other libraries has to be of the same version.

      On Windows, every application installer (.exe) comes with its own version of the library it was compiled for, and puts it in "Program Files/Application Name".

      But disk space is sooo cheap? A fully functioning linux system fits on a CD. Ever seen a windows live cd?
      And disk space is NOT cheap if your using SSDs.

      There you go, Mister complainy pants.

  11. Windows OS version numbers from NT4 onwards by Lieutenant_Dan · · Score: 2
    --
    Wearing pants should always be optional.
  12. Missing releases by dolmen.fr · · Score: 1

    A few releases are missing in the partial Windows NT history mentioned in the article:
    - Windows 2000
    - Windows XP
    - Windows Server 2003
    - Windows Server 2003 R2
    - Windows Vista
    - Windows Server 2008
    And the Xbox.

    1. Re:Missing releases by SJHillman · · Score: 1

      You forgot 2008 R2. But I think their point was the Windows that are most current. Besides, MS has client and server go hand-in-hand so it's a little redundant to mention both.
      - 2000/2000 (obviously)
      - XP/2003/2003R2
      - Vista/2008
      - 7/2008 R2
      - 8/2012

    2. Re:Missing releases by jones_supa · · Score: 1

      The current NT 6 core of Windows is quite nice.

    3. Re:Missing releases by Anonymous Coward · · Score: 0

      You forgot XP x64 aka 2003 x64 Workstation.

  13. I remember the good old days by WaffleMonster · · Score: 3, Insightful

    When every new release of NT brought with it new and useful features at least I was always excited to upgrade from 3.5 on till about 2K8.

    Now nobody seems to care about technology anymore... It is all politics, marketeering and guarding the table to ensure no excess value is ever left upon it. Innovation is now measured by games with shells, errecting walled gardens and fresh paint of questionable quality. Sad to see so much potential go to waste.

    1. Re:I remember the good old days by SJHillman · · Score: 3, Interesting

      2008 R2 has some nice improvements over 2008, and 2012 has a lot of nice new stuff, but I *hate* that they decided to use the Metro UI for Server 2012. I liked having the Windows Classic theme for 2003/2008. The Metro UI makes it more difficult to get to some basic settings. It's a server, make it simple and straightforward and aesthetics be damned.

    2. Re:I remember the good old days by Anonymous Coward · · Score: 0

      Is the server really that different from the client? I always found previous versions of the server just as easy to get around as the client. I haven't used 2012 yet, but Windows 8 is the easiest Windows yet in terms of getting to basic settings and useful tools. Not everything has migrated to the new interface yet, but it is still easier that previous versions. Based on past experience, I would expect the server to be the same way.

    3. Re:I remember the good old days by je+ne+sais+quoi · · Score: 1

      Do you realize that you just spent your whole post describing the detail of what you hate about metro, but never actually mentioned what it is you like about 2012 and the "nice new stuff"? As an OS X/linux user who has not yet even seen Win8, I was all ready to hear about what is nice & new in it, but I never found out. This isn't a criticism, just an observation: an uninformed reader might take away that your emphasis on what is wrong with Windows 8 drowns out any benefit.

      --
      Gentlemen! You can't fight in here, this is the war room!
    4. Re:I remember the good old days by thegarbz · · Score: 1

      An uninformed reader may be right. I have a friend who raves about Windows 8 and all the features under the hood and the better management and control over group policies and better network stack, and and and all I could do was sit there saying Metro, Metro, Metro.

      Internal improvements can be dammed if the user facing components suck.

    5. Re:I remember the good old days by Anonymous Coward · · Score: 0

      I *hate* that they decided to use the Metro UI for Server 2012.

      As someone who hasn't worked on Windows servers past 2008 and barely tolerates the Metro UI on Win8, all I can say is W. T. F.

      I kinda-sorta understand why they want Metro UI on a desktop, but a tablet interface on a FRIGGIN' SERVER?! What the hell is wrong with them?

  14. Doesn't it go further back? by Sir+Holo · · Score: 0

    FTA: "NT's not ancient history, in spite of its age. The NT 'core' is what's inside Windows..."

    And isn't DOS's 'core' what's inside NT?

    and QDOS's 'core' what's inside DOS?

    1. Re:Doesn't it go further back? by devman · · Score: 4, Informative

      DOS stopped being in the core with WinME. WinNT was based on VMS and never had DOS lineage.

    2. Re:Doesn't it go further back? by SJHillman · · Score: 2

      The DOS line ended with Windows Me. For a while, MS had the home line based on DOS and the business line based on NT. At the turn of the century, they dropped the DOS lineage and just made different versions of the NT lineage (Home and Pro).

    3. Re:Doesn't it go further back? by Rockoon · · Score: 1

      You need to check your history.

      There were two different lines of Windows kernels until Windows 2000. It was with 2000 that the WinNT kernel finally had all the API wrappers needed to run all regular 32-bit windows programs.

      The people on slashdot are also wrong about the origins of NT, which was in fact rooted in OS/2 NT aka OS/2 3.0 .. when IBM contracted Microsoft to develop OS/2 there was to be a consumer line (2.0) and a professional/server line (3.0) developed simultaneously. During the contract dispute negotiations, IBM walked away with the 2.0 code base while Microsoft walked away with the 3.0 code base.

      --
      "His name was James Damore."
    4. Re:Doesn't it go further back? by Anonymous Coward · · Score: 0

      WinNT was "based on" VMS in even less sense than Linux was "based on" Unix. The NT kernel shares some architectural features with VMS, but no source code or even APIs. Compare that with Linux, which may not share source code with Unix, but share so many architectural features and APIs that they're almost indistinguishable.

      dom

    5. Re:Doesn't it go further back? by msobkow · · Score: 1

      The NT kernel was based on VMS in it's core design, which is what matters, not the APIs. VMS APIs were based on FORTRAN interfaces, while NTs were not, so of course they're not "compatible."

      The key feature that I remember from my operating system class is that instead of semaphores, NT/VMS were based on the concept of "critical code sections" -- a completely different approach for implementing the locking needed for parallel code.

      --
      I do not fail; I succeed at finding out what does not work.
    6. Re:Doesn't it go further back? by zlogic · · Score: 1

      Have you actually used WinME? It's the same Win98, but with DOS hidden and locked down and a icons/sounds backported from Windows 2000. Windows 2000 is a much more modern NT-based OS and having used it alongside WinME I can tell the difference is clearly visible, with Windows 2000 winning in almost every test except memory consumption and compatibility with Win9x apps.
      Windows 2000 was originally planned to replace the DOS-based 98, but application/driver compatibility was not perfect, so Microsoft instead produced the WinME abomination before finally moving everyone to XP.

    7. Re:Doesn't it go further back? by benjymouse · · Score: 1

      The NT kernel was based on VMS in it's core design, which is what matters, not the APIs. VMS APIs were based on FORTRAN interfaces, while NTs were not, so of course they're not "compatible."

      BS. The NT kernel was created based on experience with VMS and a lot of new ideas. For instance, permeating the entire NT kernel is the concept of handles. A handle identifies a kernel object with a jump table. What we today call an object (in the OO sense), as it identifies and encapsulates an piece of memory with data as well as method references.

      One of the advantages of this design is that the desired access level (read/write/delete/...) can be expressed when the handle is originally opened. If access is allowed, the kernel object identified by the handle will have the corresponding method pointer reference the actual kernel method that performs the operation. If a given access was not requested the method pointer will simply point to an generic "access denied" method. The upshot is that permissions are checked once when the handle is created. After that the granted access is implicit the the function indirection. If you didn'r request write access, the "write" method is mapped to "access denied". Not having to check permission of the user on each use of the handle yields much better performance than having syscalls check permissions upon each access to a file.

      VMS was a much more traditional OS. It didn't have this object-oriented view of its internal structures.

      --
      Reading slashdot one-liner: (irm http://rss.slashdot.org/Slashdot/slashdot).rdf.item | fl title,desc*
    8. Re:Doesn't it go further back? by nukenerd · · Score: 1
      Zlogic wrote :-

      Have you actually used WinME? It's the same Win98, but with DOS hidden and locked downWindows 2000 is a much more modern NT-based OS

      I read Devman as that is what he meant - that WinME was the last OS to be based on DOS.

      Windows 2000 was originally planned to replace the DOS-based 98, but application/driver compatibility was not perfect, so Microsoft instead produced the WinME abomination before finally moving everyone to XP.

      I understand that it was not because "app/driver compatibility was not perfect" (it still isn't) but because of MS internal politics. The DOS/W9x team and the NT team were run in rivalry and the former lasted for years after they should have been taken out and shot. By about 1998, even entry level PCs could easily have run a lite version of NT (I ran NT4 then on a far-from-cutting-edge PC). The excuse was that games needed the direct hardware access that W9x allowed, and that the games writers could not adjust to NT. Eventually, with XP, the games writers had to adjust - or get lost.

    9. Re:Doesn't it go further back? by Smurf · · Score: 1

      I agree with everything you said, but it really seems that you did not understand devman's comment to which you replied.

    10. Re:Doesn't it go further back? by thegarbz · · Score: 1

      Have you actually used WinME? It's the same Win98, but with DOS hidden and locked down and a icons/sounds backported from Windows 2000.

      That can't be all the "improvements." At some point they must also have added the code which causes the OS to self destruct. Never before (and I hope never again) have I heard the line "Don't reboot it'll only make it worse."

  15. Last revolutionary M$ product by michaelmalak · · Score: 3, Insightful

    NT was the last revolutionary product put out by Microsoft. VB3 came out the same summer, and was also revolutionary. Excel 4.0 and Word 2.0 were the only other two revolutionary Microsoft products, and those came out the year previous.

    All of these products are essentially unchanged over the past 20 years, with even the same codebase, with the exception of VB 3.0, concepts of which continue in the 2nd generation Visual Studio product (based on the late-90's Visual Interdev platform, chucking the highly responsive 1st generation that ended with Visual Studio 6.0).

    1. Re:Last revolutionary M$ product by mlts · · Score: 4, Informative

      Being my usual devil's advocate, there have been some innovations which have been useful that MS has made. They are not revolutionary as NT (which was nice at the time because it was completely pre-emptive, while Windows and System 7/8 were cooperative multi-tasking operating systems.)

      One of the bigger ones was the jump from NT 4.0 with all its service packs to Windows 2000. The old domain structure got tossed for a new directory server model, which has proven to stand the test of time in companies. Nothing is perfect, but AD has been decently reliable and secure. I don't often hear about complete compromise of AD unless someone managed to get complete rights on an AD server.

      GPOs are another item. This is something that has zero value to all but enterprises, but are extremely useful when they come to play. The enterprise-tier management tools in Windows are not perfect, but they are extremely useful. If I want to lock access to USB flash drives to certain users, I can easily do that with security groups and OUs. This isn't flashy, but it makes life easier to turn the legal department writings into stuff I can say I can implement.

      Then, there are some cool features. Windows Server 2012 has disk deduplication. This will come in handy on VM servers. It isn't perfect deduplication, as it is a two stage thing (writes are done normally, and a background task removes the duplicated blocks with links), but it is something useful.

      There are also things that get the "A for best effort" award. .NET comes to mind because it does help with some basic security issues, and allows one to use their language of choice (I even remember visual ADA.)

      To me, MS is a mixed bag. They do some cool things in the enterprise. However, on the user front, they need some help/polish. They need to focus on developer morale so a new platform would get a critical mass of apps/games on it when it comes out.

    2. Re:Last revolutionary M$ product by mendax · · Score: 2

      I would opine that Windows 95 was actually the last innovative product that Microsoft brought out. Think about it: 16- and 32-bit program support, plug and play (or "plug and pray" as we called it early on), protected memory support, good virtual memory support, pre-emptive multitasking, networking support. Yes, as a multitasking operating system is was mediocre and its stability left little to be desired when compared to Windows NT, but compared to available alternatives (Windows 3.11 and MacOS) it was miraculous. AND it ran on just about every configuration of PC hardware imaginable.

      --
      It's really quite a simple choice: Life, Death, or Los Angeles.
    3. Re:Last revolutionary M$ product by cartman · · Score: 1

      C# is an excellent language, and is superior to Java, in my opinion.

      C# hasn't really gone anywhere, because MS isn't really pushing it anymore, but it was well-designed.

      There are some other MS products which were pretty good. SQL Server was fine (I realize it was based on Sybase 4). Visual Studio is pretty good.

      MS's worst products, in my opinion, were Exchange and Outlook. MS should have fired everyone that was working on either of those. I was astonished that Outlook still sucked so badly after a decade of development. Basic features still didn't work well, in my opinion, around 2008. It's not that hard to write a mail client.

    4. Re:Last revolutionary M$ product by Alioth · · Score: 1

      The trouble with GPOs is that many of them are extremely porous and can be bypassed with minimal effort. They are enforced mainly by the programs that they apply to. For example, the GPO to stop people using the command prompt can be easily bypassed simply by having a cmd.exe binary that ignores the setting. Similarly with regedit.

    5. Re:Last revolutionary M$ product by unixisc · · Score: 1

      I was really excited about NT when it first came out - I wanted there to be a common OS, living on different microprocessors, not just x86, and NT was that for me. The Unixes were fragmented, each w/ a different one for each platform, but NT had the potential of being common - b/w Intel, MIPS, Alpha and later, maybe others.

      I was however disappointed that NT never progressed on the RISC side at all. NT-RISC on both Alpha, as well as MIPS R4x00 was a golden opportunity for Microsoft to develop a win64 subsystem long before AMD came up w/ AMD64. Also, had Microsoft ported all their mainstream apps to that, those would have had a good start. As it is, some of the engineering workstation apps, like Pro/Engineer, were ported there, and the platform could have done with other CAD and engineering applications going to it. It would also have kept real competition between the CPUs alive, although Intel would still have had its fab advantages.

      Also, doing that work then would have made MIPS today a major player in the tablet market. The Alpha would have had to go thru some major design changes to get to the ARM like power consumption levels, which they may have achieved, having established the performance levels. Microsoft would have had win64 apps ready long before AMD came up with the 64-bit version of the CPU.

    6. Re:Last revolutionary M$ product by benjymouse · · Score: 1

      The trouble with GPOs is that many of them are extremely porous and can be bypassed with minimal effort. They are enforced mainly by the programs that they apply to. For example, the GPO to stop people using the command prompt can be easily bypassed simply by having a cmd.exe binary that ignores the setting. Similarly with regedit.

      Uhm. Not all GPOs are security policies as you seem to believe. If you do not want users to access resources on the local computer then use permissions to restrict users instead of trying to stop them from going to the command prompt. This is not your stupid sudo. There are other and better ways to secure resources than trying to prevent users from running applications that can access them.

      You may hide the command prompt if you don't want your users to strafe out where they don't know how to come back from. But if the intentionally launch an application (or a copy of the shell) they go there themselves.

      It should definitively NOT be your security boundary.

      You also fail to understand how GPOs are applied. They are *NOT* enforced by the programs they apply to. Policies are a declarative way to configure local state, primarily the registry. Instead of a script trying to run remotely (or on every boot), policies are enforced at boot time (machine policies) *or* at logon time (user policies).

      Individual apps do not check up against GPOs. They read config state from the registry (or old style config files). This provides a clean interface where GPOs are not *necessary* (you can use other means to alter the registry) but *available* if you need to enforce a set of configurations consistently across all or subsets of machines and/or users.

      You can think of GPOs as declarative scripts. The group policy service reads/interprets policies and performs actions such as setting registry values, changing text/config files or (seldom) running scripts. But because they are declarative they can inherently be applied and re-applied and rolled back.

      On Linux/Unix you will need something like puppet or chef to match the declarative power Windows has had with GPOs since the introduction of AD.

      --
      Reading slashdot one-liner: (irm http://rss.slashdot.org/Slashdot/slashdot).rdf.item | fl title,desc*
    7. Re:Last revolutionary M$ product by nurb432 · · Score: 1

      I tend to agree about NT. Once you add the GUI component Win2k was the pinnacle of their OS line. It was uphill to that point, and has been downhill ever since.

      I disagree on excel being revolutionary however, it was evolutionary, from competitors products and ideas.

      --
      ---- Booth was a patriot ----
    8. Re:Last revolutionary M$ product by nukenerd · · Score: 1

      I would opine that Windows 95 was actually the last innovative product that Microsoft brought out.

      WTF ?! - a kludgy OS that was outdated at its launch - it had been preceded by better OS's in Windows NT and OS/2. MS should have been producing a Lite version of NT in 1995, but someone in MS was still in love with DOS and wanted to keep building OSs on its ricketty foundation - as they did with Win95/98/ME for 5 wasted years.

      AND it ran on just about every configuration of PC hardware imaginable.

      That is not thanks to MS, it is thanks to the hardware makers who knew that without W95 drivers for their kit they would never sell it. MS was in the fortunate position that they could lean back and let the hardware makers do the grunt work of writing drivers for everything except the generic stuff like keyboard and mouse.

    9. Re:Last revolutionary M$ product by mlts · · Score: 1

      GPOs are one tool in the toolbox. If I wanted to block users from running cmd.exe, I'd not bother with GPOs, but use AppLocker policies and block from there. Of course, that means that another shell could be used, but if I'm limiting what users can run, I would be going with a whitelist policy, as opposed to a Swiss-cheese blacklist policy.

    10. Re:Last revolutionary M$ product by michaelmalak · · Score: 1

      Excel 4.0 was revolutionary because it introduced autofill.

    11. Re:Last revolutionary M$ product by Anonymous Coward · · Score: 0

      AD is just LDAP.

    12. Re:Last revolutionary M$ product by drsmithy · · Score: 1

      WTF ?! - a kludgy OS that was outdated at its launch - it had been preceded by better OS's in Windows NT and OS/2. MS should have been producing a Lite version of NT in 1995, but someone in MS was still in love with DOS and wanted to keep building OSs on its ricketty foundation - as they did with Win95/98/ME for 5 wasted years.
      It wasn't Microsoft that kept DOS alive (with the possible exception of ME), it was customers.
      The kind of compatibility customers wanted, couldn't be provided without a hybrid like Windows 9x.

    13. Re:Last revolutionary M$ product by Anonymous Coward · · Score: 0

      > MS's worst products, in my opinion, were Exchange and Outlook.

      You forgot SharePoint. We spent over $250k last year maintaining it for a 30-person company. Quite a few IT people love SharePoint because it creates jobs.

  16. Proper one... by Anonymous Coward · · Score: 0

    http://blog.opengroup.org/2012/07/25/apple-registers-mac-os-x-10-8-mountain-lion-to-the-unix-03-standard/

  17. Saddled with rubbish by Anonymous Coward · · Score: 0

    Its too bad NT 3.1 was saddled with trying to maintain a semblance of backward compatibility with the existing Window user mode API. WIN32 should have been a fresh start rather than being burdened with the legacy rubbish now known as WIN16.

  18. Re:Not ancient history by Anonymous Coward · · Score: 0

    They've done everything they can to promote Android adoption, not Linux.

  19. NT was originally a 'stop-gap' OS by Anonymous Coward · · Score: 0

    It is now long forgotten, but Microsoft had plans for a brand new OS based on the 'semantic' theories that were dominant in the 80s, and a new approach to the kernel/driver system. NT was to be the easily hacked together stop-gap product until the new OS was ready for the big-time.

    The final part of this failed OS mega-project was the recently cancelled 'semantic' file system that Microsoft wasted billions on across the years, and promised to release with each new version of Windows.

    It should be known that almost everything significant Microsoft has 'achieved' across the last 20 years was actually code Microsoft originally bought from other companies. Microsoft's billions have produced very little internally. When Microsoft has attempted to originate software concepts, the results have usually been disastrous. Metro/RT/New-UI is but the latest example.

    Windows serves as a universal 'good enough' base platform OS, but it's long past the time when a free OS should be doing the job better. Part of the horrific collapse of the PC market is down to the Microsoft tax that bangs an extra 100-150 dollars onto the cost of every PC. The PC market is no longer strong enough to tolerate this unnecessary expense. It is downright frustrating that Google is holding back on 'Android for Desktops', partially to promote the lousy ChromeOS platform.

    1. Re:NT was originally a 'stop-gap' OS by Locutus · · Score: 0

      not too unlike how MS-DOS was a hack of the cheap and dirty DOS they purchased to sell IBM so they could get MS-BASIC on IBM computers. I heard lots of stories of how the delivered MS-DOS was crap and IBM had to rewrite lots of it to get a functional system and PC-DOS was born. Even DOS/Windows inside of OS/2 ran better than directly on the hardware.

      Microsoft is well known for stop-gap software. IMO

      LoB

      --
      "Anyone who stands out in the middle of a road looks like roadkill to me." --Linus
  20. Re:Not ancient history by Anonymous Coward · · Score: 0

    No.

  21. Oops, forgot Visual C++ 4.0 by michaelmalak · · Score: 1

    I forgot to mention Visual C++ 4.0 from 1995, which introduced the modern IDE. I just now updated Wikipedia.

    1. Re:Oops, forgot Visual C++ 4.0 by Locutus · · Score: 1

      Because Borland C++ wasn't a modern GUI IDE and predate MS Visual C--( remember it was not C++ compliant for about 2 decades ).

      I do remember when Microsoft got rid of the multi-window IDE and everyone else followed. Hated it and still do.

      LoB

      --
      "Anyone who stands out in the middle of a road looks like roadkill to me." --Linus
    2. Re:Oops, forgot Visual C++ 4.0 by spongman · · Score: 1

      huh? everything in that wikipedia (except the name 'developer studio' which was decided on for 4.0 by Julie Larson in a rather contentious meeting) was part of Visual C++ 2.0.

  22. really? NT still lives inside each new Windows? by Locutus · · Score: 0

    And to think I believed their marketing rhetoric all the times they said Windows was a complete rewrite. Except all those times viruses were found to effect versions of Windows from the current all the way back to DOS/Windows versions.

    NT should have never had a 5th birthday let alone a 20th but we all know how far a monopoly gets you when you use it with a club and hatchet.

    LoB

    --
    "Anyone who stands out in the middle of a road looks like roadkill to me." --Linus
    1. Re:really? NT still lives inside each new Windows? by Anonymous Coward · · Score: 0

      From a functionality POV, WNT 4.0 would be great even today. Win7 does not bring lots of new stuff. But MSFT needs to generate revenue, so they create a new GUI periodically, to make people fork over money. Mr Ballmer will always go into the "totally excited" salesman mode, when a new GUI for WNT is available. It's just one of his lame salesman tricks.

  23. proof... by pbjones · · Score: 1, Interesting

    ...that 640k was not enough!

    --
    There was an unknown error in the submission.
    1. Re:proof... by Barlo_Mung_42 · · Score: 1

      Who said that?

    2. Re:proof... by Anonymous Coward · · Score: 0

      Steve Jobs said 64K is enough.

  24. OS/2 is better, um... by xdor · · Score: 1

    Maybe for services, but it still baffles me how even with the latest OS/2 release the mouse still didn't track smoothly and the GUI was easily process locked. You couldn't work in that environment without pulling your hair out trying to get anything done.

    Compared to NT, for user experience, OS/2 never had a chance.

    1. Re:OS/2 is better, um... by DrXym · · Score: 1
      Can't really blame the kernel for the brain damaged PM that sat on top of it. In their infinite wisdom IBM made it such that there was a single message queue across all windows processes and if one of them got stuck while handling a message the entire UI locked up. You had to have a console based process monitor like Watchcat installed to call up in these situations to kill the offending process.

      Anyway I didn't consider the OS/2 kernel to be anything special. The NT kernel actually had a Posix and OS/2 personality and was portable across hardware architectures so it was pretty sophisticated for its time.

    2. Re:OS/2 is better, um... by plopez · · Score: 1

      OS/2 was originally developed by MS

      --
      putting the 'B' in LGBTQ+
  25. Minor correction... apk by Anonymous Coward · · Score: 0

    SafeDLLSearchMode = Win7 default - You just don't have to SET it in the registry manually here anymore (like you had to in older Windows NT-based OS models pre VISTA/7):

    HKLM\System\CurrentControlSet\Control\Session Manager

    SafeDllSearchMode = 1

    * Another one I always felt was a "decent protection" (and saved a bit of memory too if you don't load that std-classic lib/dll ever again from other apps etc./et al)?

    The "AlwaysUnloadDLL" registry option -> HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\CurrentVersion\explorer\AlwaysUnloadDll

    Not SURE if it is still functional in Windows 7 though, as it was in older bulids of Windows (not by default either, it was another you had to set yourself, manually, iirc), but, that's another option to ASSURE that a required version of a DLL & its function/method required.

    (As that's all that loads, not the entire thing "all or nothing" of the ENTIRE dll body for loads for usage into the calling app's memory space - this IS touchy though: IF other apps call that lib concurrently also, creating reference counts to its functions - then, yes the entire DLL is loaded but again, ONLY the parts the calling apps need are used by said apps (for DOS coders, this much like how overlays worked))

    I don't KNOW if it will still work though on more modern Windows like VISTA/7 onwards. Probably does, but there's "downsides" too, noted below.

    In fact - This I wager STRONGLY is NOT ON BY DEFAULT (like SafeDLLSearch is in Win7) in more modern versions of Windows - simply because of what it does and "downsides" of it in load/reload latency performance:

    I.E.-> You have to flush reference counts to the model currently in RAM/calling app memory space functions/methods used by the calling app, then reload from disk again from it (or another model if necessary and has the same name, which used to cause "DLL HELL" with std./classic dlls (vs. OLEServer types) on Win9x & earlier builds of NT-based OS before the protections I noted in my last post).

    APK

    P.S.=> So bad choice of words on my part - the REGISTRY ENTRY is 'deprecated', but its on by default (it's just not there in the registry BY DEFAULT, like for example, many TCP/IP parameters are for tuning entries on IP stack tweaking are for 'system tweaking types') However - Yes, you CAN enter them anyhow, but if they're active by default? You're really just "bloating" your registry (with no gains & because of said 'bloat', you slow up some (miniscule)) - so correcting that for posterities' sake (and IF they are deprecated? They're bloat AND useless)... apk

  26. Windows NT 3.51 by ta0 · · Score: 2

    Windows NT 3.51 was the most stable operating system I have ever used.
    I miss it sometimes.

    1. Re:Windows NT 3.51 by shikaisi · · Score: 0

      Windows NT 3.51 was the most stable operating system I have ever used.

      I guess you never used Novell NetWare v3.12 then. Now that was stable !

      --
      No left turn unstoned.
    2. Re:Windows NT 3.51 by Anonymous Coward · · Score: 0

      So is my abacus.

    3. Re:Windows NT 3.51 by msobkow · · Score: 1

      Start up Visual Studio. Enable a breakpoint in C/C++.

      Insert a CD and start playing it.

      Run your program and wait to hit the breakpoint.

      Now wait until the CD player skips to the next track.

      Guaranteed blue screen.

      --
      I do not fail; I succeed at finding out what does not work.
  27. Did some checking, (Re)learned a "new thing" by Anonymous Coward · · Score: 0

    SafeDLLSearchMode was turned on by DEFAULT in Windows Server 2003 too (makes sense since it was the codebase used to build Windows 7 initially), so my statement of "Pre VISTA/7" in my last post? Not entirely 100% accurate (we strive for perfection, perfection being a road NOT a desitination, & technical accuracy).

    * I don't KNOW if I just forgot that since it's been so many years since I last ran Server 2003 (here @ home in default workstation type install it does anyhow where you add server features after), or that I am just getting "old & senile" but there you are (again, for posterities' sake).

    Another fellow cited "side-by-side assemblies" here too in this exchange - there's a point I missed... good catch by jeffasselin, here -> http://tech.slashdot.org/comments.pl?sid=4020769&cid=44396911

    APK

    P.S.=> I haven't looked at this lunacy in awhile, good review (because of all the changes in this during the 1992-present day versions of Windows NT-based OS). Personally, IF I use DLL's (rare for me on smaller projects or ones I don't need to send out periodic updates to smaller parts of them or data? I go "single 'stand-alone' self-contained" exe design - less moving parts, & less hassles)? If I used another version build, that say, returned an integer 1/0 return vs. another build say, returning a boolean T/F return (common enough difference along with stub/deprecated API stuff)? I just stuff it into the appdir with my app - makes for tremendously LESS hassles vs. "DLL-HELL" potential issues that way & is simpler (simple, is good - this is one of the "nuttier" things to get down cold and RETAIN imo as far as possible caveats building with Windows & "oldschool" classic style DLLs, which OLE Server type DLL marshalling by GUID cures imo)...

    ... apk

  28. Re:Not ancient history by Anonymous Coward · · Score: 0

    I think your tin foil hat needs a muzzle. Stop the crazy getting out as well as in.

  29. Show Stopper! by andrewa · · Score: 1

    I might just crack open my old copy of the book "Show Stopper!" by G. Pascal Zacharay. A great read on the design and development process involved in releasing Windows NT.

    --
    :(){ :|:& };:
    1. Re:Show Stopper! by Barlo_Mung_42 · · Score: 1

      Just read this over the past week. Never knew Cutler was such a douche. Reminds me of the argument that came up here a few weeks ago about Linus's management style. Cutler makes Linus seem like a real hoopy frood in comparison.

    2. Re:Show Stopper! by Anonymous Coward · · Score: 0

      Theo de Raadt, the creator of OpenBSD, is also a douche -- and that's a good thing. To lead a large software project with lots of people who aren't as skilled as the leaders is not an easy task. Engineering management isn't really compatible with political correctness. Spineless leaders produce spineless creations.

  30. Windows Reboot by Anonymous Coward · · Score: 0

    And you thought you had to reboot Windows 7 too much, try NT :)

    NT has great kernal and sub system concepts and more uncommon features now it seems MSFT is removing a lot of the things that made NT great, in new runtimes.

    Pity, this was the last great Server OS from Microsoft.

  31. Re:Not ancient history by Anonymous Coward · · Score: 0

    Android is no less Linux than anything else using the Linux kernel.

  32. @ArchieBunker - Re:DLL nightmare by nukenerd · · Score: 1
    ArchieBunker wrote :-

    I still don't understand how VMS can be compared to NT. They don't even seem remotely similar.

    Then follow this link.

    From the link : " Most of NT's core designers had worked on and with VMS at Digital; some had worked directly with Cutler. ... Many users believe that NT's developers carried concepts from VMS to NT, but most don't know just how similar NT and VMS are at the kernel level (despite the Usenet joke that if you increment each letter in VMS you end up with WNTWindows NT). "

    And : " [the] similarities could fill a book. In fact, you can read sections of VAX/VMS Internals and Data Structures (Digital Press) as an accurate description of NT internals simply by translating VMS terms to NT terms. "

  33. Guy Harris by Anonymous Coward · · Score: 0

    Whoever modded this down to -1 doesn't know who Guy Harris.

    1. Re:Guy Harris by Anonymous Coward · · Score: 0

      On the other hand, some people seem to have modded the original AC up. There's no accounting for idiots, I guess.

  34. NT turns 20 by Anonymous Coward · · Score: 0

    But that's 140 in software years.

  35. Comment removed by account_deleted · · Score: 1

    Comment removed based on user account deletion

  36. NT 3.51 = Graphics subsystem in usermode by Anonymous Coward · · Score: 0

    Ring 3/RPL 3/usermode (vs. Ring 0/rpl 0/kernelmode). Cutler almost quit MS when he got wind they were going to move it to kernelmode (introducing potential instabilities) - but between DirectX (gaming/multimedia) & the stable templates provided by the MS DDK, it kept it stable enough (oem vidcard maker drivers & what they add beyond that could be problematic only).

    Fork? That's *NIX stuff man... I know of NO Win16/32/64 API call of that name. I use ShellExecute or CreateProcess though in code. Why do forks though, when you can do lighter-weight threads?

    APK

    P.S.=> Only thing I've found "wrong" in Windows 7's that the hosts file after 12/09/2008 was patched badly on that account, no longer allowing 0 as a valid "blocking IP address" (whereas any NT-based OS pre-VISTA still can though, oddly) - MS' own VP of the "Windows Client Performance Division" who posts here (Richard Russell/Foredecker) even conceded that to me here http://slashdot.org/comments.pl?sid=1467692&cid=30384918 since it makes for a smaller file, which loads faster off disk AND parses faster line-by-line internally during the File Open/Read-Write/Close I-O cycle with text files like hosts. Reported to MS directly too, never was fixed. Makes a 30% difference in size for me with a custom hosts file I've built since 1997 (due to fastflux + dynDNS botnets that recycle/reuse their malicious hosts like mad "reincarnating" them as needed since they paid for them no doubt) @ 2,075,107++ entries & growing currently, via this 32/64-bit app I built to import/sort/normalize & deduplicate them for better online speed, security, reliability, + even added anonymity to a degree -> http://start64.com/index.php?option=com_content&view=article&id=5851:apk-hosts-file-engine-64bit-version&catid=26:64bit-security-software&Itemid=74

  37. Re:(The Angels) Tourism by Anonymous Coward · · Score: 0

    Have you seen the (The Tar) Tar Pits?

  38. NT 3.51 not too bad, but 4.0 started a decline by Anonymous Coward · · Score: 0

    Anecdotally, NT 3.51 was reliable for small tasks. It didn't have to be rebooted for stability reasons (for driver / install reasons, yes)

    But NT 4.0 was a step backward. It started a decline that Windows never really recovered from in serious companies. I recall architects taking a very long look at NT in the 1990s, but after NT 4.0, the die was cast. Whatever was going to come next was not going to be Windows. Then Linux got traction and the rest is history.

  39. OS/2 - The T-Shirt by middenview · · Score: 1

    I still have the T kindly handed out by IBM as we filed into the launch presentation at COMDEX - "OS/2 - up and running not up-and-coming". Tempted to frame it and send it back.

  40. OS/2-PPC by unixisc · · Score: 1

    Wasn't the OS/2 kernel some strange 16/32 bit hybrid? In any case, it was x86 only -- IBM had to create a Mach-based kernel for the ill-fated OS/2-PowerPC.

    Yeah, it was x86 only, which is why OS/2-PPC was a totally different project - Presentation Manager personality sitting on top of Mach. There is a more modern liberated version of it, which is Presentation Manager sitting on top of the L4 microkernel. Too bad the PPC is almost dead by now, otherwise that would have been a great candidate for POWERstations, PPC Amiga boxes and even older PPC based consoles.

  41. Good coding practices stop it... apk by Anonymous Coward · · Score: 0

    Quoting your article to reinforce what I myself do that works:

    "Include all DLLs privately in the applications bin-directory and use private assemblies as far as possible if assemblies can not be avoided. The best way to defeat all DLL conflicts is don't use any DLLs at all, but use static libraries instead"

    And, what did I write that I myself do?

    "Personally, IF I use DLL's (rare for me on smaller projects or ones I don't need to send out periodic updates to smaller parts of them or data? I go "single 'stand-alone' self-contained" exe design - less moving parts, & less hassles)? If I used another version build, that say, returned an integer 1/0 return vs. another build say, returning a boolean T/F return (common enough difference along with stub/deprecated API stuff)? I just stuff it into the appdir with my app - makes for tremendously LESS hassles vs. "DLL-HELL" potential issues that way & is simpler" - http://tech.slashdot.org/comments.pl?sid=4020769&cid=44397369

    IF I can figure that much out, even though I know what else exists to help stop it that I noted here http://tech.slashdot.org/comments.pl?sid=4020769&cid=44396873 & I am FAR from "the world's greatest programmer" (there is no such animal - only guys that focus on and look @ a problem and process + its data long enough to architect, or if you are allotted time & budget, re-architect your solution for safer + more efficient operations)? Others can & WILL, too... you don't have to depend on the OS covering you - you cover, yourself, way beforehand. It comes with experience (punches in the head in other words).

    APK

    P.S.=> It works, it's simple, and less "moving parts complexity" WITHOUT having to depends on side-by-side assemblies, OLEServer GUID marshalling, Windows File Protection etc./et al...

    ... apk

  42. Jeremiah Cornelius - quoting a hero of mine by Anonymous Coward · · Score: 0

    "We all stand on the shoulders of giants" - Mr. Anders Hejlsberg Borland/Microsoft (architect of .NET & my fav. coding tool of ALL-TIME, Delphi/Object Pascal).

    "Pilfered ideas", as you call them, are HOW progress occurs - you described it yourself, with improvements to the core OS (NTOS). This goes for FAR beyond computing, right into life itself and anything around it man (even personal growth by following the example of others). I mean, by YOUR reasoning? There should only be 1 auto maker: FORD!

    I love the Mustang (or rather the fuel injected, roller overhead cam/roller rocker arms, dual exhaust tech in its proven 4.6L-5.0L V8 motor - it's racing tech put to the street in essence & it wins more "street races" than any other make out there mind you), lol, so that wouldn't bug me but... their being the "sole provider" would block out the progression I speak of above - ideas, good ideas, come from many spots, many people, and often in quite unexpected areas from folks you'd LEAST think would come up with them (e.g. a patent clerk and relativity theory being a prime widely recognized one). No one set of engineers would come up with them all. In fact, that very premise IS what makes the internet great - sharing ideas.

    DEC wouldn't have HAD VMS (which is the 1st OS I ever used circa 1982 - 1987 thru highschool into collegiate academia on my 1st of 2 degrees), without Mr. Cutler and his architecting it based on prior experience in VMS!

    Yes - it was HIS intellectual ideas - yes, yes - the "lawyers" probably *might* say otherwise, but it's truth... fuck them, lol! Besides, I'd be willing to bet my bottom-dollar that DEC probably tried to STOP Cutler from using those ideas (unless you can show me otherwise, & I'd believe it for once outta you, lol) - only problem, from a "legal" standponit too no less?

    NDA's notwithstanding - IF a company tries to make it so you can't use what you know to make a living, in a "right to work state" (most are iirc), they would be sending you into poverty potentially... that, is not allowed, & WHY NDA's fail 90% of the time. They're bullshit... I've seen it happen to others & that's how it's gotten around (I am sure you've seen or heard of it as well).

    You bitch about err-handling in the kernel... want to know WHY it was so? C language itself... one of the BIG areas where C++ knocks the snot out of it, isn't just in objects, but err-handling! C natively has no such facilities, you have to "rig" them "paper clips/rubberbands/superglue" style.

    APK

    P.S.=> You have to "cut them some slack" on that account - they did the best they could, with the tools they used (C's used over Assembler for portability and speed of development, and can optimize out JUST ABOUT as fast performing code too) - in fact, I am AMAZED TO THIS DAY that the Microsoft software (or any others really) run as well as they do with such a mass of complexity in software alone, let alone the range of hardware they do... the entire thing's a testiment to the ingenuity of humanity imo!

    ... apk

    1. Re:Jeremiah Cornelius - quoting a hero of mine by Jeremiah+Cornelius · · Score: 1

      Right you really are on most of this.

      You cite imitation, derivation, improvement and tribute as the way that progress is made, in the arts and sciences. I actually advocate that.

      But in the instance of Microsoft as a company, there are endless back-room rip-offs, double crosses and secret handshakes. They extend from (at least) QDOS through VMS and VINES to SGI (Rick Belluzzo, look closely) - right to the present day: Nokia and Win7 stink to heaven. This was a weird deal in the back room, if ever their were one.

      Nokia is essentially a subsidiary of MS, without the legal tripwires. They ditched a successful if future challenged business, to sell devices by the folks behind Zune and Kin.

      I know of at least TWO Sr. Execs at MS, who were dismissed because of flagrant sexual harassment, who each went to "penitence" jobs at minor MS partners, only to surface 6-9 months later in Sr Exec jobs at Nokia - where their function relates to Microsoft, corollary partner to their former roles.

      Gates stole BASIC, in the beginning. Everything thereafter followed this pattern.

      --
      "Flyin' in just a sweet place,
      Never been known to fail..."
  43. The "big 'FIX'" is in, everywhere... apk by Anonymous Coward · · Score: 0

    Gates "stole" DOS too (Tim Patterson iirc) - heck, they ALL do it, & nobody does anything about it... why? We ALL gain by it is why, for the reasons noted already by us both. That is or should be the END GOAL/GAME though. Overall boosts for everyone.

    On what you saw in MS: Sure, there I believe you - Unskilled douchebags with too much time on their hands and not enough REAL work on their product breeds it, along with their sociopathic tendencies (which get them to the 'top' via "the ends justify the means - fuck the slaves" philosophies) + methods. HOWEVER: That's not the guys creating the wares/OS, right? They're busy actually doing good. NO "idle hands are the devil's workshop".

    Yes - I've also seen "shenanigans" like that too!

    1 directed my way I was terminated for spotting security holes that I suggested simple ways to fix no less - didn't matter: I was back @ work right down the road oddly shortly enough after, & MADE SURE the bastard(s) CIO & his henchman got nuked out of THEIR jobs, very shortly afterwards with their own 'criminality' (unlike you, I don't hold back on pulling the trigger when it was pulled on me first, & worst for doing what I knew in my heart was the RIGHT thing to do!).

    ( & they're STILL looking for work: Why? They don't have any more relatives who are major stockholders/owners, lol, or that are part of the masonic temple like they are placing them but above ALL else, & lastly they don't have any REAL hard skills in the computer sciences)

    After all, & Yes, I know this & was one in my time (extremely successful one in loss prevention) - anyone, can be a "mgt. babysitter", anyone (they ought to be the LEAST paid in fact, imo & experience).

    However, I had to leave it. Bugged my conscience in the things the job made me HAVE to do @ the behest of my 'superiors' (who were anything but on many levels) - worst being lying to subordinates. Also found out that of 218 other competitors in that company, I was outperforming them all 1.5 yrs. straight month in & month out, but I was NOT the highest paid... well, fuck that.

    That made me hate myself (the lies I had to tell, lives I ruined possibly, & worst getting PLAYED on pay). I had to go & grow past and away from it... Comp. Sci. was the answer. It's also a potential route to bettering the world around you imo.

    A lot of my pals still in mgt. roles were like "don't be a fool man, stay up here @ the top, IT is dead, & the stockmarket makes SURE it's dead" (payroll being the single easiest thing to control overheads-wise, real "business 101" microeconomics class stuff)... no thanks. Ask me to sell my soul too, ok?

    There's NOT ENOUGH promote from within the ranks, & too much "promote from the billionaire boys club frat" out there. Results are showing what the end result of THAT is, macro-economically in the USA, cascading to the rest of the planet (mostly).

    Man - it's what you GET, with people leading people who have NEVER DONE THE JOB OF THEIR SUBORDINATES... head doesn't know what the ASS is doing or rather how it's done (makes for shit decisions/bad product). 1 look @ Mr. Ballmer, says it ALL. Heck - why look so small here? The nation's economic, & credit status show it too RIGHT TO THE TOP (fake top, it's the 1% that run the show behind the curtain).

    IF you watch StarTrek? You'll "get this":

    Imo @ experience: Too many "geeks" are of the "Mr. Spock Mindset" (for lack of a better analogous comparison here): They much prefer their scientific duties to the Captaincy. By way of comparison & contrast?? Too much mgt. is Lt. Commander Gary Mitchell ("Morals... are for men: NOT 'Gods'"). We need "mixtures" of that.

    You cited Russinovich - I agree, & used to be a 'co-worker' of his via Sunbelt, & have corrected his work before telling him WHAT (pagedefrag) but not only that but how to & why/where as well, which he thanked me for in email (and got the best of him in a debate @ Windows IT Pro too (using MS own docu

  44. Advance NT Fair! by Anonymous Coward · · Score: 0

    NT is being used in remote communications infrastructure by a specific pacific country in Afghanistan.

  45. Additionally for posterities' sake (static lib) by Anonymous Coward · · Score: 0

    Why I use Delphi's VCL (statically compiled) http://it.slashdot.org/comments.pl?sid=3958509&cid=44243717 per your Dr. Dobbs journal article - makes for decent "single moving part self-contained" executables, like this one -> http://start64.com/index.php?option=com_content&view=article&id=5851:apk-hosts-file-engine-64bit-version&catid=26:64bit-security-software&Itemid=74 from "yours truly".

    AND, again - per my last post http://tech.slashdot.org/comments.pl?sid=4020769&cid=44402835

    I.E.-> On what to do to AVOID "DLL HELL", IF you use "extra moving parts" DLL's!

    (imo, something to be avoided but, easily gotten around using static libs + doing placement as I noted, "seconding your article's motion" in the same stroke)

    APK

    P.S.=> Experience, even BAD experience (& yes I've had them) is like a punch in the head - you either learn by it, and do better, or keep getting them!

    ... apk