Please create an account to participate in the Slashdot moderation system

 



Forgot your password?
typodupeerror
×
Windows Operating Systems Software

Microsoft's New Core OS Team Learning from Linux 732

sokk writes "Seems like Microsoft is paying attention to the Linux way of doing things. According to itworld.com, a new central engineering division will work on the core of Windows: "The Windows Core Operating System Division (COSD), within the company's Platforms Group, will be responsible for the core OS platform, including development, program management and testing, Microsoft said in a statement sent via e-mail.". A little further down the page analyst Rob Enderle: "They have been studying Linux extensively. Part of their study has been on how Linux has been able to maintain a high level of consistency in the kernel while groups around it maintain maximum flexibility,". "
This discussion has been archived. No new comments can be posted.

Microsoft's New Core OS Team Learning from Linux

Comments Filter:
  • Comment removed (Score:4, Informative)

    by account_deleted ( 4530225 ) on Wednesday December 17, 2003 @02:27PM (#7746830)
    Comment removed based on user account deletion
  • Release Date (Score:4, Informative)

    by thebatlab ( 468898 ) on Wednesday December 17, 2003 @02:28PM (#7746839)
    "By closely controlling the OS core, Microsoft will be able to better ensure that Longhorn will arrive on time and meet its quality and security objectives, Enderle said. He expects Longhorn to come out in the fourth quarter of 2005, provided that a beta becomes available as planned in 2004."

    Looks like they're still looking at a 2006 release. (Come on, fourth quarter 2005 always means 2006).

    The question is, will assembling this team help them meet that goal or will the initial organization of it take away some time and delay the project more?

    I can't say from experience on what the effects of forming a new style of management to a project, no matter how capable that style is, will do as a project is underway but I'd assume there would be some hassles to start things off with and get the ball rolling.
  • by gekkotron ( 641131 ) on Wednesday December 17, 2003 @02:36PM (#7746927) Journal
    Obligatory posts saying that "COSD is dying."
  • by Animats ( 122034 ) on Wednesday December 17, 2003 @02:55PM (#7747114) Homepage
    Windows NT worked that way until the end of NT 3.x. Then Microsoft demoted Dave Cutler, turned the code kiddies from the Windows 95 team loose on NT, and messed it up. It's prettier now, but less stable.

    Microsoft has had a terrible time transitioning people from the DOS-Win3.1-Win95-Win98-WinME family to the NT-based systems. More than half of Windows-based desktops worldwide are still running DOS-family OSs. Even though they've all been discontinued. Even though they have zero security and crash constantly. They're still out there.

  • by ChaseTec ( 447725 ) <chase@osdev.org> on Wednesday December 17, 2003 @03:04PM (#7747193) Homepage
    The Core OS team is at least 4 months old. I run a site about operating system development and I just finished a run of a Microsoft banner advertising OS developer positions. The banner linked to here [microsoft.com] if you're interested.

  • by fitten ( 521191 ) on Wednesday December 17, 2003 @03:09PM (#7747257)
    The reason that Linux's codebase has remained so cohesive, focused, and flexible is that Linux has so many really skilled developers

    That's partly (if not more than partly) because of the (traditionally) high cost for your ticket to enter Linux/Un*x Land (both monitarily and intellectually). Keep the entry barriers high (steep learning curve, tools that are difficult to use, etc.) and the only ones who can stick it out will typically be the above average folks.

    On the other hand, make it so that any Joe Shmoe can cobble together *something* that works (at least works a fair amount of the time) and you get more people developing on it (ever hear of VB?).

    Things are changing with Linux dropping the cost point of entry and the intellectual point of entry dropping (but still having a good deal of a ways to go, IMO) with the development of more tools on Linux but Linux isn't there yet until Joe Shmoe can throw together a toy app quickly and easily that can keep track of his beer, cigarette, and pork rind expenditures.
  • by revividus ( 643168 ) <phil...crissman@@@gmail...com> on Wednesday December 17, 2003 @03:12PM (#7747303) Homepage
    Funny, I can switch between tabs with ctrl-tab. It's very convenient.
  • by brain159 ( 113897 ) on Wednesday December 17, 2003 @03:15PM (#7747334) Journal
    No No NO. Did you not read the Groklaw article yesterday or thereabouts pointing out that this scenario is not true?

    The penalty for abusing GPLd code is not the compulsory re-licensing of everything. If you're in breach of the terms of the GPL license, then you're breaking copyright law. Nothing makes their proprietary code suddenly open, unless they decide to comply with the GPL rather than fight/settle/re-code.
  • by Sevn ( 12012 ) on Wednesday December 17, 2003 @03:46PM (#7747575) Homepage Journal
    That would be the major reason to ignore this story completely. The *laugh* analyst *laugh* in this case is a known douchebag. The "Enderle Group" is made up of exactly one person. Wild guess who. If anyone takes offense at my use of the word douchebag, you come up with a better word to describe someone that creates a "group" that contains only themselves and puts their last name in the title of said group. Perhaps he has imaginary friends or multiple personalities or pets he counts in his membership totals.
  • by Anonymous Coward on Wednesday December 17, 2003 @03:55PM (#7747650)
    I think, at the kernel level, you`ll find that Microsoft programmers are top notch. If you read something like Showstopper which chronicles the development of NT you will get some idea of the calibre of people they use.
  • by red floyd ( 220712 ) on Wednesday December 17, 2003 @04:04PM (#7747739)
    AFAIK, They had a POSIX layer in NT4, deprecated it in 2K, and removed it completely in XP.
  • Re:Not new (Score:3, Informative)

    by IvyKing ( 732111 ) on Wednesday December 17, 2003 @04:26PM (#7747926)
    It was at one point clear (DOS/Win3.1), but then the GUI started to "own" many features (net support, and even CD-ROM access!) from 95 on - and they finally did away with the separated "core system" from ME on.

    The earliest versions of Win NT kept the GUI code out of the kernel, but that was changed in later versions of NT to improve performance. Bear in mind that NT was much more of a resource hog than OS/2.

    More importantly is the corporate culture - M$ has played fast and loose with programming in the past, e.g. the original IBM PC and their first software for the Mac.

    IBM PC - Intel's databooks for the 8086 specifically stated that interrupts below 20H were reserved for future versions of the 8086 family - IBM and M$ then proceed to use those interrupts for the BIOS routines.

    Mac - Apple had many programming guidelines that were intended to allow migration to future versions of the 68000 family (e.g. don't use the upper 8 bits of addressing for flags). When Apple came out with a 68020 version of the Mac, a lot of the M$ software was broken by the upgrade.

  • by Theatetus ( 521747 ) * on Wednesday December 17, 2003 @04:29PM (#7747946) Journal

    They already do shared source for their "partners". I got to see some Windows code when I was trying (and eventually failing) to write a driver.

    And don't forget that you can see the source code of Windows CE if you want to.

  • by gatkinso ( 15975 ) on Wednesday December 17, 2003 @04:41PM (#7748056)
    1) Having ported lots of code between VC++ and gcc I can tell you that there would be some modification, especially down that low.

    2) POSIX subsystem compilance was never very advanced, so obviously not much being pulled from there

    3) Certainly the file systems are very different

    4) The header files aren't even close in the headers that are common between the two

    Recognize this directory tree?

    http://www.sysinternals.com/ntw2k/info/xpsrctree .s html

    Didn't think so. (Means very little however.)

    Socket implementation is different, threading model is different, task scheduling is different...

    So while there could be plagarism, it would not be without so much effort that you may was well just write the stuff yourself, after being inspired and pointed in the right direction by the GPL'ed code.
  • by Net_Wakker ( 576655 ) <puddingdepot@yahoSTRAWo.com minus berry> on Wednesday December 17, 2003 @04:48PM (#7748120)
    Funny, I can switch between tabs with ctrl-tab. It's very convenient.
    In KDE ctrl-tab will switch desktops, so mozilla won't switch tabs. Konqueror will, if you use ctrl-alt-tab.
    Just adding noise to the discussion.
  • by snatchitup ( 466222 ) on Wednesday December 17, 2003 @06:48PM (#7749099) Homepage Journal
    What M$oft doesn't wanna admit is that they are scared s'less of what IBM is doing with th Eclipse project. It's a bit of an Enigma.

    Microsoft wanted it easy for developers to get its tools. They were never free, however. Eclipse is free. But IBM's version (Websphere AppDev) for the enterprise is basically Eclipse, with additional plugins.

    I can run Eclipse on Linuz, etc. Same engine. IBM is or has overaken WebLogic in the AppServer market.

    ----
    The idea that MS wants a better OS, so it's looking at Linux is an understatement. Basically, now its... "Okay, we bodged up our OS to justify violating Monopoly laws. Now that we've won.... We just need to undo the code intermingling of End User Applications with the kernel."

    Eclipse is the next "killer" app, by going back to the fundamentals of how to build a huge software business.

  • by dschl ( 57168 ) on Wednesday December 17, 2003 @06:56PM (#7749146) Homepage

    Short story in semi-point form. No names, lest people I know lose jobs. Car accident, patient in intensive care unit. Young woman, has small children, I believe. ICU doctors want to pull the plug (think it was a neurologist, ie, brain surgeon). No brain scan completed, doctor makes recommendation to family to turn off ventilator, family approves.

    Nurse mentions to doctor that patient appears to be responding, and the doctor should get a scan to measure brain activity. She is overridden by the doctor, who maintains that the patient is brain dead, a vegetable. Doctor orders that patient is to be removed from life support. Nurse decides to adminster medication to assist breathing based on a standing order (blanket prescription for the ICU),and then turns off life support as ordered bvy doctor. Only because of medication administered, patient continues to breath. Nurse nearly loses job over this.

    A few days later, patient is awake. A few months later, patient is nearly fully recovered, possible loss of recent memories, but is up walking about, part of her family, leading a useful and productive life. As far as I'm concerned, the above doctor should not be practicing, and should be sued for everything he owns, but doctors protect their own.

    I don't trust doctors anymore. Period. You should get to know some ICU nurses, and you might want to revise some of your opinions. Personally, I would cause severe physical harm to any doctor with a god complex, before I would let them touch anyone I care about. I would also ask an experienced third party nurse to do an assessment, review a brain scan, and provide a second opinion (insamuch as the law permits nurses to have an opinion), before I would pull the plug on anyone based on a doctors advice.

  • by utlemming ( 654269 ) on Wednesday December 17, 2003 @07:23PM (#7749318) Homepage
    Personally we ought to do away with the "Use or Loss 'Em" policy -- why? Because sometimes people just don't have the time to use them, or they don't want to use them on the dicussions that are avialbe for the those three days. Point in case -- I lost a couple mod points because I did not want to mod a weekend discussion, since there was nothing going on (it was a weekend when the articles were not personally interesting to me) At least, they ought to give a longer time to mod so that you can take your time and puruse when you have the time. Rushing people gets the first posts points, and those with something intellegent, don't have the opportunity.

    At the very least, give the mod points out and then have it so people can not get more mod points until they have used the ones they have. I think then people will get moderated in a way that will reflect the quality of a comment.

  • by Pr0xY ( 526811 ) on Wednesday December 17, 2003 @08:34PM (#7749784)
    I thought it was worth mentioning that the Microsoft Windows source code can be liscenced from them to a company for a certain price (i forget how much but it's a lot). It's the restrictions on the liscensing that prevents it from getting out really. If you find any bugs using such source code, you are legally obligated to immidiately notify Microsoft or they can sue you into obvilion, also if you ever were to release it to the public, consider yourself and your company out of business as they have the legal right to take every penny you have ;)

    Finally, you need to take a course on how to properly compile the source since it is seriously MASSIVE (spans several CDs, rememeber not just a kernel, but a web browser, cd-burner, all bundled applications, etc...it's BIG).

    just some food for thought, the source is available, but if you do anything not approved by MS while in possesion of it, you are pretty much screwed.

    proxy

I have hardly ever known a mathematician who was capable of reasoning. -- Plato

Working...