Submitted by Jim Allchin, a senior Microsoft Corp. executive, walked into Bill Gates’s office here one day in July last year to deliver a bombshell about the next generation of Microsoft Windows. “It’s not going to work,” Mr. Allchin says he told the Microsoft chairman. The new version, code-named Longhorn, was so complex its writers would never be able to make it run properly.
Windows-Team Reorganization
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
Damn… seems like a step in the right direction. I’ve always been under the impression that corporate politics were a negative to the development of the computers, but in the last few years we saw intel give up the clock cycle myth and now Microsoft admitting that poorly written code can’t be recycled forever.
> Microsoft admitting that poorly written code can’t be recycled forever
Did you read the article?
They went back to Windows 2003 Server code base and not to 2000 or NT. Last time I checked, 2003 Server was their latest OS.
The problem was therefore with new code (Longhorn) and not with the Windows code before that.
Allchin is leaving and that is good: how did they get into this mess in first place, under his command? 4 years of work and then they have to ‘reset’??
Agreed; like the article said, Allchin, Cutler and a few others were from the old school of coding; clear distinct lines between the different components; lego working together to create a bigger and better thing; then re-organising the components to do something else.
You could say that the crap-o-la programming ethics were the result of the heady late 1990s computer boom; the demand for features whilst accepting buggy products; the push for bigger and fancier products, being released at an accelerated pace with little or no regards for the basic concepts of quality coding.
With that being said; the alternative Microsoft could adopt could be something closely resembling what Apple is doing; no necessarily dumping the core out there for end users to download and compile; adopt opensource components where it makes sense and developing in house components where the strategic importance is so great, it is best left a secret.
Is a HTML rendering engine really that secret and important? of course not, the important, end user stuff resides onto of it, the browser and form spell checking and numerous other fancy features.
Legit question…
So are you suggesting open sourcing these non-important technologies?
Yes and no; then again, can you replace them with opensource equivilants? how about using libxml/libxstl instead of maintaining their own library? how about using libpng/libjpeg instead of their own image libraries? in the case of IE, why maintain their own engine when Webkit and/or Mozilla could be an ideal drop in replacement?
Opensource components? depends, but then again, why not mothball some components by simply merging into the opensource components and adopting the opensource equivilant?
We’ve already seen how this can benefit Apple already; its freed up programmers to concentrate on the more important, ‘product differentiating’ parts of the product – Aqua/Quartz/Quartz Extreme 2D/3D etc. etc. whilst allowing the mundane features to be maintained collectively by the computing industry over all; in the case of libxml – maintained by SUN, Red Hat, Novell, Debian plus numerous independent programmers and organisations.
to prevent any competitor from getting a foothold. For example, to this day there is no well-developed market for replacement GUI shells for Windows PCs. Why is that? It seems like the most natural thing, something that could benefit everybody.
During the late ’90s Netscape was unable to keep up with all the wrinkles Microsoft threw into Windows (there were five versions released between 1995 and 2000, not even counting the NT/Win2K code base), they weren’t able to add all the necessary features to be competitive and still have a browser that was fast and robust. I think Bill G. saw it was an advantage that Windows internals be as complicated as possible, so Microsoft’s engineers – with source code and team member access – would have the biggest possible advantage of “local knowledge”. Later, Ballmer told the DOJ, with somewhat of a straight face, that he didn’t know whether it was even possible to remove IE from Windows.
A guy named Andrew Shulman wrote a couple books describing back-door APIs he reverse engineered in Windows 3.X and Win 95 which were being used by Microsoft’s applications teams, but not disclosed to anyone else.
Now it’s coming around and banging these guys in the butt. It turns out that Microsoft most resembles the proprietary minicomputer vendors like DEC and Wang that thrived in the late ’70s and early ’80s. They made big fat profits back in the day, but they’re all gone now.
Paul G
“A guy named Andrew Shulman wrote a couple books describing back-door APIs he reverse engineered in Windows 3.X and Win 95 which were being used by Microsoft’s applications teams, but not disclosed to anyone else.”
They weren’t disclosed because you WERE NOT SUPPOSED TO USE THEM. Raymond Chen has written a bit on this subject and how some software would try to use hidden APIs and the apps would break in the next version because of it.
They are not some super-secret optimized-only-for-ms-software APIs that were being deliberately kept from software vendors to give them an advantage. Cut the conspiracy theories, please.
You might want to actually read the books in Schulman’s Undocumented series before claiming it didn’t happen.
Or just Google for ‘microsoft “chinese wall”‘ for more information.
I’ve read and I didn’t say it didn’t happen. I said they aren’t hiding the “good API” from developers and using it only for their own software.
I’ve read and I didn’t say it didn’t happen. I said they aren’t hiding the “good API” from developers and using it only for their own software.
That’s a very sweeping statement to make. How do you know that? Given the fact that Microsoft writes software for Windows (Office, SQL Server et al) and they also write Windows, it would be rather unthinkable for them not to use their in-house knowledge to their advantage and use things other software vendors can’t. That’s called using secret APIs and getting an unfair advantage.
You’re talking total crap. Some of us weren’t born yesterday you know.
The thing is, Windows is developed by litterally thousands and thousands of coders.
Office, the same.
These two teams can not mix.
The thing is, Windows is developed by litterally thousands and thousands of coders.
Office, the same.
These two teams can not mix.
Actually, they can. And they do. You ought to know that, as the MSloving guy, you are.
Stop your denial of the truth. MS has acknowledged this as truth years ago.
dylansmrjones
kristian AT herkild DOT dk
The thing is, Windows is developed by litterally thousands and thousands of coders.
Office, the same.
These two teams can not mix.
Errr, it’s called being part of the same company. Go look it up.
It’s a very sweeping statement to make that they ARE hiding the good APIs. The burden of proof is on YOU to prove they ARE, not the other way around.
It’s a very sweeping statement to make that they ARE hiding the good APIs. The burden of proof is on YOU to prove they ARE, not the other way around.
I’m afraid not, deluded MS fanboy. I always know someone has lost when they ask me to prove something that they started otherwise.
First of all, it was a part of the DoJ case where that actually came out and secondly, it’s just plain and simple logic – what’s stopping Microsoft from doing this? Err, nothing, that’s what. There are hidden APIs in Windows, that has been established. We’re all just supposed to trust Microsoft and you that they’re competing on a level playing field with everyone else? Yer, whatever.
There are hidden APIs in Windows, whether you think people should be using them or not. Microsoft does not want to release them, but of course, Microsoft’s product groups like Office etc. can still get access and knowledge to them because they are a part of Microsoft, and there’s no reason they can’t. You prove to me Microsoft’s products like Office, Exchange, SQL Server etc. has not been using anything like these (or have not been using Windows code itself directly) and we’ll call it quits. I can tell you the outcome though – you won’t be able to.
“I’m afraid not, deluded MS fanboy.”
I don’t think I need to read your comment any further. Sigh.
I don’t think I need to read your comment any further. Sigh.
*sarcasm* Well I really appreciate you being the bigger man here ;-). */sarcasm*
Sigh, why am I bothering responding to this after the name-calling.
“I always know someone has lost when they ask me to prove something that they started otherwise.”
I started it? Someone made claims that Microsoft keeps their good API from the public. I asked for proof, and none has been provided. I also agreed that there is undocumented API, but explained that does not automatically mean it is a BETTER api deliberately hidden from the public.
” it’s just plain and simple logic – what’s stopping Microsoft from doing this? Err, nothing, that’s what.”
So because they can, they did? Ok, simple logic, sure.
” There are hidden APIs in Windows, that has been established.”
I already agreed with that. Undocumented API exists, but not for the same reasons you think.
“We’re all just supposed to trust Microsoft and you that they’re competing on a level playing field with everyone else? Yer, whatever.”
As opposed to automatically assuming they are doing evil. No. You should only make accusations when you have proof.
“There are hidden APIs in Windows, whether you think people should be using them or not. Microsoft does not want to release them, but of course, Microsoft’s product groups like Office etc. can still get access and knowledge to them because they are a part of Microsoft, and there’s no reason they can’t.”
You’re making a claim with no proof, again.
“You prove to me Microsoft’s products like Office, Exchange, SQL Server etc. has not been using anything like these (or have not been using Windows code itself directly) and we’ll call it quits. I can tell you the outcome though – you won’t be able to.”
The burden of proof is on YOU, since you are the one making the accusations.
Sigh, why am I bothering responding to this after the name-calling.
If you think this is name calling, you need to get out more.
I started it? Someone made claims that Microsoft keeps their good API from the public. I asked for proof, and none has been provided.
You’re the one going off on one about it. There is proof, and large amounts of it surrounding the DoJ trial and with lots of submissions dotted around the internet. The fact that you don’t happen to know them is not my problem. I, and others, know they exist and they are easily accessible via Google. Go and have a search, and if you find nothing, please let us all know what search parameters you entered.
I already agreed with that. Undocumented API exists, but not for the same reasons you think.
The undocumented APIs exist, yes, and we’re just supposed to believe you and Microsoft that it is being used for what they say it is? Errr, the burden of proving this is all on your little shoulders.
So because they can, they did? Ok, simple logic, sure.
We’re talking about a convicted monopoly here, and there is a great body of evidence around as to what has gone on. Go find it and let us know if you find nothing.
You’re making a claim with no proof, again.
The proof is out there, and it was part of the DoJ trial and surrounding evidence. Go and find it and let us know what you find. Let us know whether you find anything that says Microsoft is not doing this.
The burden of proof is on YOU, since you are the one making the accusations.
I’m not spoon feeding links (and someone may have provided them) to a clueless twit on something that is so well known. If this was actually a reasonable discussion, and there was some point to your denials maybe. As it is, NO.
Actually hiding the “good API” is one thing MS has admitted to do, during several trials.
Look at the way MS Office and IE is constructed. Do some research on MS software building techniques, and you’ll see that they’re hiding as much as possible.
Go get the facts!
dylansmrjones
kristian AT herkild DOT dk
Ok, here’s how things work in the world. You make an accusation, you provide proof. If you don’t have any, you don’t make the accusation.
Ok, here’s how things work in the world. You make an accusation, you provide proof. If you don’t have any, you don’t make the accusation.
Wow, it’s like I’m making a personal accusation against you.
The accusation has already been made as part of the DoJ investigation, and it has been proven. You would do well to familiarise yourself with it. There are hidden APIs in Windows, and obviously, as their divisons are part of the same company they have insider knowledge that competitors simply don’t.
You prove to use that they’re being good little citizens with all that.
“Wow, it’s like I’m making a personal accusation against you.”
Did I say that? Nope.
“The accusation has already been made as part of the DoJ investigation, and it has been proven.”
I’ve read the court rulings and saw nothing showing that Microsoft has hidden APIs that are there to deliberately hurt competition. If it’s so well known and obvious, then why can’t you provide the data?
That’s all I am asking from you. You obviously know so much about it. Provide a source, and where it says these things. Telling me to read a 200 page ruling and find it for myself won’t work. Show me EXACTLY where it is. If it’s true, you should be able to.
Did I say that? Nope.
Read the ‘it’s like’ part.
I’ve read the court rulings
No you haven’t read the court rulings, but you want to make it look as if you have.
and saw nothing showing that Microsoft has hidden APIs that are there to deliberately hurt competition.
You’re not looking hard enough. See above. There’s probably one or two links knocking about above – the rest is left as an exercise, which is probably better than huffing and puffing around this forum.
If it’s so well known and obvious, then why can’t you provide the data?
You’ve hit the nail on the head. It’s well known and obvious. Go and do some research.
That’s all I am asking from you. You obviously know so much about it. Provide a source, and where it says these things.
See above. The rest is left as an exercise. You tend to find that if you ask for something where there is vast body of information (internet, Google etc.) you don’t get it given to you. You’re the one who told us all confidently that Microsoft is not doing what everyone knows they are doing.
Telling me to read a 200 page ruling and find it for myself won’t work.
I thought you’d said you’d read it. LOL. Anyway, I would have thought you’d have been intelligent enough to read around that subject matter.
Show me EXACTLY where it is. If it’s true, you should be able to.
No one’s doing your work for you. On the contrary, you need to be producing something that Microsoft acted completely above board.
They are not some super-secret optimized-only-for-ms-software APIs that were being deliberately kept from software vendors to give them an advantage. Cut the conspiracy theories, please.
Wave everything away as a conspiracy theory. Nice. Microsoft writes Windows. Microsoft also writes Office, SQL Server, Exchange etc. Microsoft writes all that additional software without any of the in-house Windows expertise or detailed knowledge of Windows internals, or using any of the expertise and code of other divisions to integrate their software more tightly in a way competitors cannot. 2 + 2 = 5.
Cut the crap please.
Actually you’re very wrong on this subject.
According to internal MS documents they did it for one sole purpose only. To kill competition completely. And it haunts MS to this day.
dylansmrjones
kristian AT herkild DOT dk
…it is how MS handled their use–they did so recklessly and there is ample evidence to prove they have done so (it is not a conspiracy theory).
Since programmable computers came into existence there have been undocumented features in software that were intended for internal/system use only. Even open source software has “hidden APIs”–of course, in open source they are not REALLY hidden because the source is available to all, but in any well designed software there is a clear delineation between public/external interfaces and calls intended for internal user only.
In one sense I agree with you–using undocumented APIs is reckless. I remember my early computing days experienceing the same issues…In the days of CP/M strying beyond established BIOS or BDOS calls was widely regarded as taboo–everyone knew you did so at the risk of breaking compatibility going from CP/M 2.0 to 2.1 to 2.2, or even worse, becoming dependent on a particualr vendor’s CP/M hardware. Because of the fragmented nature of the CP/M market and the discipline of programmers in that day it seems that such incendents were extremely rare.
Other platforms that were single-vendor were less fortuante–a lot of developers of software for Atari, Commodore and Apple systems (among others)got burned when they ignored manufacurers warnings and did reckless things like bypassing BIOS calls or using useful but undefined opcodes in the 6502 processor. So, if a hardware maker went from release A to release B of their BIOS for their next line of computers, or switched from Commodore/MOSTek 6502 processors to Rockwell or WDC versions (because they were superior chips and were not sourced from a company owned by a competitor) software broke when it shouldn’t have.
MS is reckless because THEY IGNORED THEIR OWN ADVICE. Shulman and others investigated MIGROSOFT’S OWN APPLICATIONS and found evidence that MS used undocumented Windows API calls themselves, while castigating others for doing the same. When MS started being scrutinised by antitrust officials they established the “Chinese wall” to keep the two “empires” within MS (applications and operating systems) from “invading” each other–to the point that even some office windows were papered over. Obviously there were holes in the wall, or double-agents had infiltrated the two empires, because it didn’t work.
I love reading Chen’s work–it really illustrates what a house-of-cards Windows has become (which is suprising that MS didn’t “reset” much soomer than Longhorn). Do note, however, that Chen is very careful of mentioning specific companies or products when talking about incompatibility due to poor programming practices–when he does it is becasue the product or company responsible is defunct–and MS apps seem strangely absent. However, if you Google around or look at Windows internals long enough there is evidence MS BROKE ITS OWN RULES. The list of values in the “compatilility” registry keys is a good first sign–Excel and Word appear among the list of apps that broke and needed special treatment in future versions of Windows. While that is not conclusive evidence it points you in the direction to look.
Whether these undocumented API calls were strategically kept secret to give MS an advantage in applications development is certainly debatable. I do not think it was a strategic move. However I think it is extremely short-sighted to dismiss it as a “conspiracy theory”. The truth is probably that is was more of a progmatic move on the part of developers on one of many “death marches” to get a good=performing app out the door quickly while higher-ups just looked the other way.
Thank you for your level-headed comment.
>> Do note, however, that Chen is very careful of mentioning specific companies or products when talking about incompatibility due to poor programming practices–when he does it is becasue the product or company responsible is defunct–and MS apps seem strangely absent.
Actually Chen has mentioned Word as one of worst offenders of needing hacks in Windows to remain compatible, like you said yourself.
Ok so, to make it short, to get Longhorn done, Microsoft went from a bad and buggy written code to a ripp off of Mac OS X!!!!!
Good job Allchin!!!!!!
>Now it’s coming around and banging these guys in the >butt. It turns out that Microsoft most resembles the >proprietary minicomputer vendors like DEC and Wang
>that thrived in the late ’70s and early ’80s. They
>made big fat profits back in the day, but they’re
>all gone now.
If you know anything about Microsoft and their history they are not like any of those companies at all. Microsoft every once in a while has to change and refocus, they have done this many different times for many different reasons and they are still around.
It really amazes me on how many people say something and it could be paragraphs long and totally come out looking clueless. Not trying to be mean here, but a lot of people on this OSnews website say lots of stuff and most of it is just incorrect or outright FUD.
“If you know anything about Microsoft and their history they are not like any of those companies at all. Microsoft every once in a while has to change and refocus, they have done this many different times for many different reasons and they are still around.”
I agree that Microsoft has changed in the past with good success. But they haven’t changed in a long time. The last time they had to redefine their strategy was back when they didn’t have the market position they have now, and when IT wasn’t such a pervasive aspect of the corporate world.
Microsoft changes when it is absolutely imperative to change. When Longhorn development was beginning to hit the fan, they took notice but didn’t do anything until it was clear that change was necessary, Then they instituted sweeping and drastic changes to their process. When Microsoft changes, they change big. When IT changes, they take things step by step. I believe this represents a fundamental problem.
When a competitor looms, Netscape for example, they strike tenaciously. When the competitor is gone, they settle in. They don’t change again until a new competitor threatens their domain (Firefox). Microsoft, in this respect, is reactionary, not revolutionary. Being at the top of the market is hard when your only move is too react. Everyone loves to hate Microsoft because it is not in their nature to serve the customer. They only serve the customer in the sense that they fight to stay ahead of their competitors.
For the first time in their history, they have faced a perfect storm of competitors (Apple’s desktop usability, OSS’s vast developer community, and Google/Yahoo’s web services). MS grabbed its playbook and responded to these threats with all of its might. Their ability to simultaneously take on a variety of initiatives tested their capabilities to the limit. They failed, reassessed, and restarted. At the end of 2006 we will see the results of a development effort that is essentially a 2.5-year response to over 5 years of competition.
The question is, now that Microsoft has changed, and changed quite dramatically, will the market change with them? Or will they prefer a migration path that offers consistent and steady improvements?
The last time Microsoft tried to be revolutionary, they got hit with an anti-trust suit. See a pattern?
*LOL* What kind of revolution was that? They got it by an anti-trust suit due to their illegal and criminal behaviour.
No more, no less.
BTW: Do you consider MS java, IE integration etc. as revolutionary?
dylansmrjones
kristian AT herkild DOT dk
Well, seing that for the last couple of years, other groups have been interating a browser with their UI, it must have been. (Apple, KDE)
You’re mixing up bundling and integrating.
AFAIK Mac OS and KDE does not integrate the browser with the system. Anyway. In Windows the integration is happening at lowlevel. This is not the case for Mac OS X, nor for KDE.
If IE integration is revolutionary in your eyes, then I’m feeling sorry for you. It must have been a boring life, you’ve had so far.
dylansmrjones
kristian AT herkild DOT dk
Apple has developed webcore, and it is part of the system. KDE has Konquerer, and it is part of the UI as well.
IE is not integrated “lowlevel”. It replaced parts of the shell (starting in IE4) and now Windows (and thousands of other apps) merely uses the rendering engine all over the place.
To get to using automated testing for Windows OS development on a regular basis. EEK! Testing everything purely by hand?
I’m hopeful that a desired side-effect of them finally implementing automated testing in the development of Vista makes it that much easier to use automated testing for applications running on it once it has been released. I’m also hoping that they will make complete and accurate developers’ documentation for the API as a result, since what you can’t document as having a certain behavior, you simply cannot test and say it passes or fails. In many cases, the existing documentation is fairly good, but there are still cases that aren’t correct, or are quite undocumented, despite being listed as being public API calls.
Huh? I worked at Microsoft for a year or two on the Windows 2000 team. We had automated tests for everything. I can only remember testing something by hand on two occasions.
Admittedly, I wasn’t testing the whole OS, but I’d be surprised if the team I was on was doing that level of automation and everybody else wasn’t automating at all.
Huh? I worked at Microsoft for a year or two on the Windows 2000 team. We had automated tests for everything. I can only remember testing something by hand on two occasions.
That’s because this is marketing. There is no way you could build anything like Windows, with multiple developers, without automated tests. Microsoft allegedly changed a things in the run up to Windows 2000 as well, just to convince us all of the quality.
You’d be surprised by a lot of practices in the proprietary software industry. I’ve been hired by two of these companies, and my general response to my job description at both was, “you guys don’t have this already?” Both positions were geared towards improving the software development processes.
I’ve found that the practices put in place through the ad-hoc management of OSS are vastly better than in proprietary outfits. Of course, when I say better, I mean given the available resources and the reality of the situation. Developing quality software is a process that is more than a bunch of good developers, and there isn’t a formula that scales well. The best solution for a startup might be peer review, for open source possibly static analysis and unit tests, for large houses all of the above plus functional verification and integration testing.
Just : can we laugh at Microsoft for their impressive work on OS ? ( not able to do a better OS than Unix 30 years after … but still gaining rank in : lack of security, impressive instability, crap licences … )
Poor guys, just let us all cry for 2seconds because they realise they have done a crap software … bouh … normally they don’t care ?!
at least microsoft can get serious sometimes and actually changes how things work.
the unix world hasnt moved for 20 years now. take X11 for example, they never tried to change it a bit to make it y2k ready (im not talking about y2k bugs here).
instead of getting back to work, they still try to adapt it (and they fail) for today’s uses. they just recycle all the old crap and write new crap on top of it. no surprise why there’re always like 5 layers on top of X11 on most people machine. kinda pathetic… 5 layers away from X11 and there’re still problems with keyboard, mouse, fonts, ect. things that i consider basic.
anyway, thats why people will buy vista, because it will be just better. think about it for a minute. people always bitch about windows but they always end up buying it. why? they could get a unix derivative for free. so why they still buy it? theorically its just a proof that microsoft can deliver a better solution. i dont know many people that would waste money on something inferior. think about it for a minute.
ironically, i think apple learned from microsoft success and from their own failure and unix failure.
even if you have already a good product (ie: xp/2003), you cant just keep fixing it (ie: fixing security issues), you have to make it more attractive in general. i think the unix world never understood that simple idea. maybe they think all they have done is already perfect? sorry, but nothing is perfect, never. i have an old linux box here (1997?) that i never updated (not worth it…old hardware) and when i compare it to my current linux box (2005), i dont see a difference. that directly mean that nothing really worth mentionning happened in 8years. oh and please dont tell me there are a million more gtk/qt apps out there that just doesnt count as an integral part of the os. im talking about the os here of course, not about what others have to offer.
We really live in a crapy world !
Despite the fact that my comment was ripped for no apparent reason, I feel really sad to read such comment proning Microsoft leadership in software !
On the one hand people suffer the stockholm syndrome: I have a windows alternative ! I can do everything I want !! Sorry I have plenty : FreeBSD / Linux ! They work in 64bit !
On the other hand I agree that plenty of crap remains and a little “standardisation” could help instead of making stupid conf files in diffrent places and in diffrent style.
stupid conf files in diffrent places and in diffrent style.
Well, the same goes for windows, after MS admitted that the registry was a complete screw-up.
Does ‘Documents And Settings’ ring a bell? There you can find tons of configuration files in a many different styles, and add to that the whole nonsens in the registry, which is still used quite heavily, in many different ways.
Everytime you can find a weak spot in the GUI of GNU/Linux you can find a similar week spot in Windows, OS/2 and most other systems.
You would know that, if you knew a bit more. (And yes, I’ve been using windows for many years, as well as OS/2, DOS, GNU/Linux’es, Mac OS Classic and Mac OS X, and Commodore 64 and Amstrad 664 and a couple of other old systems I’ve forgotten the name on.)
dylansmrjones
kristian AT herkild DOT dk
“the unix world hasnt moved for 20 years now. take X11 for example, they never tried to change it a bit to make it y2k ready (im not talking about y2k bugs here).
instead of getting back to work, they still try to adapt it (and they fail) for today’s uses. they just recycle all the old crap and write new crap on top of it. no surprise why there’re always like 5 layers on top of X11 on most people machine. kinda pathetic… 5 layers away from X11 and there’re still problems with keyboard, mouse, fonts, ect. things that i consider basic.”
X11 progress stagnated for about a year under XFree86, but now Xorg is one of the most active areas of development in the free software stack. X11R7 will be released before the end of the year, which is a complete restructuring of the codebase into a modular tree. Which brings me to the topic of layers. Layers are a basic divide-and-conquer strategy that has always been indigenous to software engineering. Not only do asbtraction layers simplify development, but they also allow for modularity and flexibility. If there was a single codebase that went from display driver all the way up to GUI toolkit, I wouldn’t use it or develop on it.
“ironically, i think apple learned from microsoft success and from their own failure and unix failure.
even if you have already a good product (ie: xp/2003), you cant just keep fixing it (ie: fixing security issues), you have to make it more attractive in general. i think the unix world never understood that simple idea.”
Dude, you don’t make any sense, and it’s not because of the writing proficiency. The conclusion doesn’t have anything to do with the premise. Was MS successful because WindowsME was more attractive than 98? Isn’t improved security item number one on Vista’s agenda? Did Apple fail (whatever you’re referring to) because they didn’t make MacOS attractive enough to compete with Windows95? Or was it the vast quantity of cheap IBM PC clones that did them in? Did UNIX fail because it was less attractive than MS-DOS? Or was it because of the standards wars between AT&T and Sun?
“i have an old linux box here (1997?) that i never updated (not worth it…old hardware) and when i compare it to my current linux box (2005), i dont see a difference. that directly mean that nothing really worth mentionning happened in 8years. oh and please dont tell me there are a million more gtk/qt apps out there that just doesnt count as an integral part of the os. im talking about the os here of course, not about what others have to offer.”
You’re honestly trying to argue that Linux 2.0 is essentially the same as Linux 2.6? Unless I misunderstand your intended scope of comparison… Between 1997 and 2005, Linux gained all of the following capabilities:
USB, PCMCIA, SMP (without crippling spinlock), audio support, RAID, three different file caching schemes, POSIX shared memory, LVM, firewire, APM, ACPI, Plug-n-Play, and many others, including countless devices drivers, filesystems, and performance improvements.
That’s just in the kernel. Just for a little bit of timeframe comparison, KDE 1.0 was released in mid-1998, GNOME 1.0 in early 1999. Red Hat was in the 4.x series of releases in 1997, and transitioned from SysV libc to glibc (version 2.0) towards the end of the year.
So you, my friend, have got to be joking…
Obviously Microsoft dredged up some marketing smarts. What do you do if you just wrote off several years of development as a bad job? Spin a story to the WSJ about how it’s a good thing, really, because it’s teaching you to be less sucky in future. Nice job of damage limitation…
How true how true.
Actually what kills a company is its bad vision by its superiors. small guys do not make horrible mistakes. Windows should be completely rewritten since xp was out to fix the gigantic old wrinkled code dating back to the ninties. Even if your next software is not secure it will take time till hackers and virii writers catch up with the new OS destruction process. Bill Gates is a blind corrupt guy himself, he demand improvements on his OS without allowing any of his advisors to speak about rewritting the code. MS will die for sure and what they did to vista is not sufficient; Can you imagine that when booting Vista it will give you greating that says your OS might be at risk, you are not running any antivirus software?! What a load of crab after all these years of hesitancy.
My friends this time is apple’s time not MS, maybe next decay it might be Microsoft or who knows it might be a new emerging UNIX company, Linux will forever be the basic OS. Linux will dominate in poor countries or countries and companies that looks for value more than quality.
My ten cents that by summer 2006 the very latest Vista will be generally perceived as a flop. Of course Bill Gates will do everything in his power not to admit to it.
Well, there are times when rewrites are good (Mac OS X), and there are times when rewrites are bad (the apparently ever-changing API of GCC).
I guess Microsoft is hoping that this rewrite will be good, to the point of betting the company on it. (Not that it’s a complete rewrite if they’re starting over from Server 2003).
And look, according to this it’s more like UNIX now- specific compartmentalized functions… Well, they’re fixing the problem that everyone thought was there. It remains to be seen how well they do at it.
The biggest reason why the problem was there to begin with was because of speed. The system ran faster with all this stuff in the kernel. Now, they can put things out of the kernel and get the same performance, and so this is what they are doing.
It was really just a trade-off of stability for speed. Did them well in the short-term, but in the long-run, it bit them in the butt.
Maybe I’m going to write an article about “integrated complexity” for OSnews. About how the *nix philosophy of little building blocks (LEGO) works better in the long run than integrating everything, because when your OS becomes both integrated and complex, there is little hope to stay organized.
Oh so you can’t see a difference between a linux installation from 1997 and 2005! Come on how much bullshit is that. It’s about as stupid as saying there’s no difference between windows 95 and windows xp.
Some day I wish some people get a clue on this forum…
The reality is that Microsoft just underestimated the giant task ahead for Longhorn/Vista and are now trying to spin their failure to deliver the goods in a positive way. What do you expect? They’ve always been a marketing company before being a technology company. If only they replaced Balmer the monkey with a tech guy maybe they’d stop spinning so fast it makes me dizzy!
Gates is a programmer, and not a bad one. Maybe that explains why MS has much market share: they really listen to their tech guys.
Agreed. As a programmer, he’s able to understand what they’re doing and, of course, influence the whole thing.
For example, I cannot understand how people can claim that tens of integrated services inside an OS is not ages better than tens of isolated softwares. That’s just crazy.
The problem with Windows is its original design (even NT-based one) has not been drawn with Internet-power in mind because it originates in a period of time when Internet was marginal. So they got hurted by Internet success because it was not easy to convert the whole company strategy (and tens of services) to a different shape in a couple of years while still mantaining compatibility.
The importance of Vista to MS goes much further than its UI or stuff like that. If you go under the hood, you will see that there will be major changes going on which users will never know of but will surely benefit.
For example, you won’t need to run your system as an Admin to perform your daily tasks which is not possible now on WindowsXP. Now try to imagine converting a system which requires you to be an Administrator to a system where you don’t need that anymore still mantaining compatibility. This is not for casual developers…
I think that MS has hell of techs and company management can actually understand what the hell they’re doing (though they might be wrong, sometimes, as most people are…). Plus, having a roadmap for a platform is a lot harder than having a roadmap for a single product.
I think with current BETAs, Microsoft actually showed its competitors that they might be late, but they will deliver so that HUGE task has been almost completed. If I were MS competitors, now that I know that they can deliver what they promised, I’d start worrying about what to do to keep up with a Windows system which will be more secure, faster, snappier and with lots of innovation inside, even for hardware which hasn’t been released yet.
Perhaps that’s the dilemma bill is having in mind when he resisted the idea of making windows more like lego. If you do the os that way, there is no theoretical problem for a 3th party to replace or remove some of its features, without taking it down. Then it will be difficult to argue that IE (for instance) is an internal part of the OS and cannot be removed.
This brings the whole problem of antitrust cases against MS in new light. Imagine some “smart” guy deciding to ban MS from shipping the build in firewall because of anticompetative behavior.
On the other side the os is so complex now, that it is becoming practically impossible to build it the old way.
It’s a shame that this artical wasn’t written sooner, as lot of sites and forums have been asking “Where
s the eye candy?” This explains that they are working on the core and then on the presentation.
Why did you remove my post?
I was right in my comment. So i tell it again!!! To tell in short what Microsoft did to get Vista done is as follows: they went from a a buggy and unstable code to a ripp off of Mac OS X. Its true, really!!!!
For this, i guess that we all need to thank Mr Allchin for his great vision of how to develop an OS: copy what someone has already made.
Great jog of innovation!!!
They have already done that for windows ! when copying on Mac first ever window/icon desktop !
It’s impressive !
They do tha for a lot of stuff, when they don’t know how to do, they buy a company tha do the trick or stole technology.
Then they claim it’s cool to have licence and patents !!!
( don’t stress they have remove mine too !! )
No they don’t and even if Gates was/is a good programmer he’s long stopped doing anything technical on Windows. You’re talking maybe back in the days of DOS… He’s largely been on the marketing side overhyping everything Microsoft.
He is still a tech guy, and that is what truely intrests him.
Balmer, on the other hand, is a mathematical genious, and a good business leader.
They could have just reengineered everything in Eiffel. That process alone would have given them an excellent set of reusable, relatively bug-free components that could have been pieced together Lego-style.
Yeah I know that’s not realistic, it’s just a pleasant thought.
We had all this exact same stuff coming out in the run up to Windows 2000, about how they were focusing on quality and how they’d reorganised their programming teams.
This is just pure BS marketing designed to convince of the quality of Vista, aka Lonhorn.
Yet everyone on this site claims that Win2000 is the best release of Windows, ever….
Yet everyone on this site claims that Win2000 is the best release of Windows, ever….
Go figure.
Thats because Windows 2000 is the best OS so far that Microsoft has come up with… notice i said so far
This idea is quite old indeed. Sad to see something already known in the 80s to be considered in the 21st century..
I guess that means that there is really is a reason to upgrade. Before now, the conventional wisdom was: “pfft, might as well stick with XP.”
But, now that we all know that XP is cr@p, and new Vista was so carefully designed, and completely rebuilt. There is now a compelling reason to upgrade.
Amazing that msft would confess about all previous version of windows being cr@p – just before msft releases the first non-cr@ppy version.
Even before Windows 3.1, it was always the same story with msft: “nirvana is just an upgrade away.” All of the old problems will be fixed, and of course, there will not be any new problems.
Hmmm.. Lots of small programs with well defined functions designed to take input and produce output that can in turn be used by other programs. Doesn’t that sound like Unix? And how long has Unix been doing that? Want to start a pool on how long it takes before Gates tries to
claim he invented this technique and tries to patent it?
You know it’s going to happen.
Go googling, dammit!
Or read the halloween documents, or the findfact.
It was documented during the NS-MS battle that MS was hiding good api for themselves (look @ findfact)
There, you have your documentation. Go read it.
dylansmrjones
kristian AT herkild DOT dk
I HAVE read that stuff and I do NOT see what you are talking about. Why can’t you link stuff and be more specific?
I’ll tell you why. Because when you say “go look for yourself”, it implies the truth is there and I’m just not looking for it. It puts the burden of proof on me even though the burden of proof is on you.
It’s a common debate tactic, and it’s disgusting.
There is another debate technique, which is to deny all facts and claim you’ve never heard about it before, so people can keep proving the same things over and over, thereby getting nowhere.
It’s also a common debate tactic, and this time it’s being used by you.
The DoJ-trial concluded that MS was keeping central parts of it’s API hidden in order to prevent competition. In other words: MS was using this API in it’s own applications, while securing that other developer parties had no access.
It’s being repeated constantly in findfact, so if you haven’t seen the statements, you haven’t read them. The same goes for the first 2 or 3 halloween documents.
I HAVE read that stuff and I do NOT see what you are talking about.
No you haven’t.
I’ll tell you why. Because when you say “go look for yourself”, it implies the truth is there and I’m just not looking for it. It puts the burden of proof on me even though the burden of proof is on you.
It’s a common debate tactic, and it’s disgusting.
Oh, poor you.
Please get yourself some original comments and arguments and stop repeating stuff I and others have said, will you? It’s not moving you any further along. Here’s a link to click on, if your pea sized brain is up to it:
http://news.com.com/Microsoft+highlights+antitrust+progress/2100-10…
Microsoft also noted that on Aug. 6, it released communications protocols used to allow the Windows desktop OS to work with Windows server products. But unlike the APIs, which are available royalty-free, the communications protocols must be licensed from Microsoft. The company does not publicly disclose what the fees are. In all, Microsoft released 113 communications protocols and 290 APIs. Earlier in the month, Microsoft said it would disclose 272 previously undocumented APIs.
And this is just the stuff they admitted to. Did you read that bit? Communications protocols used for communication between server and desktop (stuff like Exchange – which is closed to Exchange and Outlook), and 272 undocumented APIs that cannot be just ‘undocumented stuff you’re not supposed to use’.
Keep in mind that this is just a news.com article as well with no real detail. I will leave that as an exercise for you.
Have you read the file called findfact.html?
Exactly THAT document with the court ruling concludes MS was hiding API. It’s wasting pages and more pages all about how MS was hiding central API in an effective and illegal attempt to prevent competition.
I have it somewhere in my archive. If I find it within the next week, you can have a copy incl. info about where to find the evidence.
Until then I will recommend you read findfact.html (it’s also exists as a pdf-file). You obviously haven’t.
dylansmrjones
kristian AT herkild DOT dk
I’ll read that.
Though I do find it funny that you say “I have it somewhere… I’ll send it if I find it…”. I don’t see you sending anything, because in a week this topic will be forgotten and no one will remember.
Right now I’m looking at “findfact.html”, but it’s a pretty big document. A little assistance as to where I can find what you are talking about would be nice of you.
More insults being thrown my way. Can’t you debate civilly?
And stop saying “It’s out there, you’re just too stupid to find it”. PROVIDE THE DATA or shut up.
I thought you’d said you’d read it. LOL.
I did, even if you want to claim I didn’t. I read it and I honestly do not remember anything saying Microsoft purposefully withheld APIs to hurt competition. I’m NOT saying it wasn’t there. I’m saying I do not remember it and asking you to point out exactly WHERE it says that. If you can’t do that and wanted to keep telling me “it’s out there”, fine, but you look like an ass by doing so.
No one’s doing your work for you.
It’s not my work, it’s yours. Make accusation, provide proof. What part of this is hard to understand? I am researching it as I type this, but I shouldn’t have to, since you obviously know it’s there, and could do it a LOT faster than me.
Here’s what I’ve fund after looking around:
http://msdn.microsoft.com/library/default.asp?url=/library/en-us/dn…
Now we can look at these APIs ourself and make our own judgements as to whether or not these gave Microsoft an advantage of competitors.
>This idea is quite old indeed. Sad to see something
>already known in the 80s to be considered in the
>21st century..
yeah, that is exactly how I feel about linux trying to be a unix clone instead of being their own OS. Sad to see something raped from the 70’s and rehashing old technology trying to make something new.
At least BEOS was something new.
>BTW: Do you consider MS java, IE integration etc. as revolutionary?
yeah, because we know Linux which is based off of Unix which is a early 70’s OS by AT&T is so revolutionary for the 21st century.
You trolls really need to start trying harder.
What is really funny is the distros trying to be like Windows and trying to run windows apps, trying to look like windows and trying to install as easily as windows does and outside of the linux community, nobody cares.
That is sad, troll all you want, but dude, the truth will set you free.
yeah, because we know Linux which is based off of Unix which is a early 70’s OS by AT&T is so revolutionary for the 21st century.
Poor idiot. Linux is Unix like and not based on Unix.Back to your homeworks dumbass troll.
I was very interested in this article but I had to register to be able to read it
.
From the little bit on top it seems to me that Microsoft hasn’t come up with sufficient checks and balances to make sure they can deliver.
Of course, the software they make is not a trivial thing, but I’m not convinced they have used ‘best practice’ methodology as they climbed to the top.
Anonymous wrote:
> Microsoft admitting that poorly written code can’t be recycled forever
“Did you read the article?
They went back to Windows 2003 Server code base and not to 2000 or NT. Last time I checked, 2003 Server was their latest OS.”
Don’t you know the lineage of Windows? Server 2003 and 2000 Server are all based on NT 4 code.
And NT4 was based on NT 3.51 (actually they were pretty much the same, with the shell as the only major difference).
NT 3.51 was based on 3.5 based on NT 3.1 which was the first release of NT (marketing trick to match the versioning of 16bit Windows running on DOS).
NT was written from scratch, but with Windows API in mind. NT’s low level design was inspired by older systems, which was based on or inspired by even older systems.
The Windows codebase was rewritten a lot in regard to Windows 2000, XP isn’t much different that Win2K (when talking low level) and the same goes Windows 2003 Server.
When MS went back to Win2K3 Server it means that they skipped the source they had written (Longhorn) and started from scratch with Win2K3 Server (which is quite different than NT4 Server).
dylansmrjones
kristian AT herkild DOT dk
and NTx was a pile of crap (last MS garbage I ever had to use) = server must be crap. So if a+b=c, then Longhorn (based upon crap) will be crap. End of thread, let’s move on folks.
Sir, your logic is impeccable.
http://windows.czweb.org/show_article.php?id_article=99