Separate Code Files And Commingling? 222
ScottyB writes: "According to an article in the Washington Post, 'Microsoft Seeks to Revisit Code Ruling,' Microsoft is asking the Appeals Court to revisit the 'commingling' issue in its ruling." As the article states, "'Microsoft did not commingle software code specific to Web browsing with software code used for other purposes in the same files,' the company said. 'Rather, in organizing software code into files, Microsoft placed related functions close to one another' to benefit users.'" Wouldn't being in the same binaries or binary distribution constitute commingling?"
Re:Viral Code (Score:1)
What would be funny is if the put it under the GPL instead of the LGPL.... it be even MORE viral!
Re:Huh? (Score:1)
When the browser stuff was starting up, Microsoft's big core technology was OLE, being buffed up to COM with interfaces. Their new, flagship product showing off COM was their browser, complete with IBrowser interface.
All an app had to do was write to IBrowser, and all a browser maker had to do was provide an IBrowser API.
Buuuuuuhhhht waitaminute! If anyone can plug in a browser, and anyone can use a browser to write a virtual computer, complete with interface, then anyone could plug a brand new virtual OS/GUI into Windows. Then people could write to Windows or the virtual GUI, so they might as well write to the virtual GUI since that runs under Windows. Then Windows will die in a few years because a generic, public-domain definition of a virtual OS/GUI and the software to support it on any machine or base OS would render Windows irrelevant as an OS.
Soooooo, we can't have Microsoft die, so we can't allow Microsoft to be a platform that no one builds apps for directly, so we can't have a public, generic virtual OS/GUI software plug into our system that developers might develop for, so we can't allow browsers to plug into our system, but since we allow that, we un-do the flagship COM product and un-COM it such that no one but us can plug in because it's an integrated part of the OS now.
Re:A summary (Score:1)
Bitterman
Trying to shore up its weak evidence (Score:2)
Microsoft created this spreadsheet for the purposes of this litigation, in an attempt to show that the Web browser and the operating system were "integrated." But the spreadsheet backfired, because it showed that there were some functions in SHDOCVW that were used by IE, but not by the operating system. In yesterday's petition, Microsoft argues that if it had run the Windows shell through a wider variety of different scenarios, it would have used all of the functions in SHDOCVW. That would have provided a very powerful piece of evidence for Microsoft's position -- so why didn't Microsoft produce it? They certainly had the resources and motivation to do so. They even tried to do so -- and failed.
The Court of Appeals can reverse a finding of fact only if it is clearly erroneous; i.e., if there is no reasonable inference from the trial evidence that can support it. They have to defer to the trial judge, who sat through the trial and was in a position to weigh the evidence. So it should be clear that Judge Jackson was entitled to infer that the reason Microsoft failed to produce a spreadsheet proving that there was no browser-specific code in Windows was that, in fact, there is such code in Windows.
An example of commingling? (Score:5)
Everybody knows the standard C library function strcpy(). (Well, not everybody, but you get the point :)... it's been around since the year dot.)
For some reason, Microsoft introduced a new function called StrCpy. An MS-specific extension to the standard C library, no doubt a combination of convenience and locking coders in to Windows specific code... no surprise there.
What is surprising is that any application that uses this function can only be run on a computer with IE4.0 or later!
So this means that even applications that don't use IE in any way at all, suddenly find themselves requiring IE4 to be installed on the computer.
I discovered this when I inherited a small win32 application at work. It was just a small dialog-based application, but on computers without IE4 or later installed, it just wouldn't work.
Very strange, and smells of commingling to me.
(BTW, see here [microsoft.com], and here [microsoft.com], for details.)
Re:No, files ARE relevant! (Score:1)
False analogy. If you had it so that the font rendering libraries were only installed when the proprietary FooMonoPolyWare was installed, it would be a good analogy to what MS was doing.
Re:DLLs (Score:1)
http://www.crucial.com/store/partspecs.asp?imod
Some interesting implications of all this... (Score:2)
HOWEVER, let's look at this a bit closer. You can certainly re-use networking code, URL & URI parsers, font libraries, image handlers, XML, etc. But none of these constitute a "Web Browser", or web browsing capability. They are merely common functions.
"Web Browsing" implies handlers for HTML, DHTML, JScript (as opposed to Javascript), Visual Basic, Bookmarks, Web Caches, Web Proxies and other "high-level" functionality.
Now, I can understand someone putting image handlers for various image formats all in one library, maybe combined with font libraries and anti-aliasing code. These are all related, and grouping them will generally improve the performance of applications. (You'll generally want to display multiple types of image in a single application, so loading all the handlers as a single unit will make a difference.)
On the other hand, what would ANYONE be doing, putting, say, JScript handlers next to a JPEG parser? Sure, JScript code -may- refer to images, but that's not a requirement. It makes no sense, either from the perspective of the programmer or from the OS. (The user doesn't care what's combined where, because they can't see that level of detail.)
It would be comparable to Sun waking up one day and moving the Swing and AWT heirarchies in Java into the java.net package, and link them together in such a way as to make them inseperable. Sure, it'd make a marginal impact to network-related Swing and AWT code, but it would severely impact any graphical code that didn't use the network.
In light of them dropping Java -- in essence, killing it off on the desktop and relegating it to non-Windows servers -- this could spell doom for Microsoft. IBM tried similar stunts, and that damn near killed them. This is a dangerous gamble.
Given that Microsoft ==IS== the computing world, to many people, this could also cause a complete re-think on how widely-used computers are. Think about it. IT Department invests millions in migrating a company or a University to the web, using Java, IIS and Internet Explorer. Head of IT Department then goes to the boss and says "sorry, you now have to give us twice as much again to re-write everything in .NET, or we become unsupported."
Because everything is so tightly merged, don't even think a 3rd-party can write a Java add-on. There's nowhere to add it!!!
This time, next year, either .NET or Java will be dead, and the future of Microsoft will have been decided, one way or the other. If .NET wins, Microsoft will have total power. The UN and NATO might as will surrender to it, if that happens.
If Java wins, Microsoft will either have to completely re-implement Java (something they just DON'T do, and may not have time to do), OR they will have to eat humble pie (which BG3 has declared inedible) and include AS-IS the standard JAVA JVM and development tools. Since nobody likes a loser, that'll kill any hopes they have on the Anti-Trust appeals. Which, in turn, may lead to an even-more radical breakup, or even being disolved, in view of their attempt to turn the entire Internet into a Microsoft VPN.
Re:Some interesting implications of all this... (Score:2)
For example, a TCP/IP stack is useless, in and of itself. It can make/break connections, but it has no capacity to handle name resolution, to talk to a server, or create suitable packets. It's simply a means of translating one stream into another. It doesn't PRODUCE anything.
High-level libraries have the distinction that they are either sinks or sources. They don't translate, per-se. An HTML renderer, for example, is a sink. Information goes in, but no stream comes out. That data is displayed on the screen and then forgotten about.
A diagramatic way to say the same thing is:
Do we want to win this one? (Score:1)
--
$you = new YOU;
Re:Some interesting implications of all this... (Score:2)
If the HTML rendering engine is componentized, your web browser is really just a front end that displays that and has a text box to enter a URL into.
This is way IE is written.
Re:No, files ARE relevant! (Score:3)
Yeah like that damn HTML rendering engine that is so easy to be leveraged!
Those bastards, we need to go back to the good old days when programming was hard work. To hell with component based development!
Don't Forget The Java Plugin and Java Web Start (Score:2)
Actually, Sun has provided a couple of different ways to integrate their JVM into IE and Netscape 4. There's the Java Plug-in [sun.com], which makes it possible to write applets that will be run inside an add-on Sun JVM.
Even better, there is now the Java Web Start [sun.com] application manager utility, which provides great support for deploying Java based applets and applications to people's desktops.
We use these methods to deploy Ganymede [utexas.edu] to folks here at the laboratory, and everything works great, be it in IE, Netscape 4, or whatever.
- jon
Re:...specific to... (Score:1)
Re:Some interesting implications of all this... (Score:1)
Nope. Next year, Java will still be doing exactly what it has done for the last couple of years---provide developers with a stable, well documented, easy to use platform. It's not perfect (especially GUI stuff) but it's quite good. The people who will use .NET are the people who use VB now. Corporations don't survive if they make substantial investments and then throw them away. You said so yourself.
Also, your comment about MS threatening the U.N. or NATO is a joke. The U.N. is much more then just the General Assembly, but it's far from any sort of enemy with Microsoft. Why would Microsoft consider taking on something like the UN? Enlighten me.
NATO is a military organization. They have: guns, soldiers, ships, tanks, airplanes, satelites as well as conventional and nuclear missiles and bombs. Additionally, they have the comunications and computational resources to manage all of that. They fight wars. Microsoft has a bunch of desktops and servers.
Re:No, files ARE relevant! (Score:1)
Who said that?! A component system that requires you to install an application is broken. Why not simply allow someone to install the components required without making other judgements?
The wheel is turning but the hamster is dead.
Re:An example of commingling? (Score:1)
You mean "the shell guys" reimplemented strncpy for no good reason as well as strcpy?
--
Re:Don't use strncpy! (Score:1)
strlcpy. http://www.usenix.org/events/usenix99/full_papers
--
Re:An example of commingling? (Score:2)
Which simply serves as further evidence that Microsoft botched it. Nowhere is it written that libc needs to be implemented as a single entity. As an example, you could have a "veneer" link library (conveniently named libc) which offers up all the libc functions, but implements none of them. Instead, it re-binds to sub-DLLs which implement libc operations separated according to general functionality (one DLL for string operations, another for file operations, another for printf() and friends, etc.). Thus, your program will only consume memory for the functions it's actually using.
This is not rocket science. Hell, it's not even computer science. It's just good, solid, clear thinking.
Schwab
Re:An example of commingling? (Score:3)
Ha ha, very ha. Microsoft "invented" shared code very late in the game, well after Amiga and Mac, and probably after Sun as well. The .so extension on UNIX systems canonically means "Shared Object".
Besides, there's no legitimate reason the Windows shell DLLs couldn't themselves have dynamically linked against a libc DLL -- which itself could then be shared amongst all apps -- rather than exporting incompatible substitutes.
And no, libc is not too large/monolithic to make this work in a 4M system. The only difficult part is getting the functions and data that rely on global state (printf(), errno, et al) to work. But strcpy() is stateless, and as such would drop into a DLL trivially. That Microsoft screwed up something so agonizingly basic I feel speaks volumes about their, "modern software engineering practices."
Schwab
If I had mod points you would get them (Score:1)
+1 Insightful
I got a good chuckle out of that one.
Re:An example of commingling? (Score:2)
Re:Huh? (Score:2)
A friend and I just had a discussion with some girl over ICQ that was of the belief that a person should not be 'tied to the meanings of words' and that we should 'open our minds'. I believe her mind was so open that her brains flew out the door.
There is something inherently wrong with randomly giving a meaning to a word for ones own use. It rather defeats the purpose of words, language, thought, etc, since society must usually universally agree on the meaning.
Microsoft: We do not commingle code, we put related functions close to each other for ease of use
Translation: I have a fuzzy puppy dog, and he likes to eat socks.
see how that doesnt work? might as well throw away the language if one cannot assign a specific set of meanings to words and have society universally agree on the meaning.
Re:An example of commingling? (Score:2)
This post and others like it have absolutely convinced me that:
Just remember, every post you make just gives one of us a soapbox to stand on. And, you think you're doing any good trying to sway opinion here? What a waste of effort. It's much more likely it's *you* who will be persuaded. I'm not 100% sure, but perhaps there is an interesting new trend: former Microsoft astroturfers suddenly going off the radar and later reemerging as opensourcers. Hey, it feels good when that icy little lump in there melts.--
Re:One word - comctl32.dll (Score:1)
I'm not saying that that the specially restrictive terms of IE redistribution lasted long -- 3rd party vendors yelped and MS backed off after 6 months or so. But there was a window when the only way to get those DLLs was to get into the IE distribution business.
--
Re:One word - comctl32.dll (Score:2)
Initially, those distribution contracts had requirements such as using IE-specific features on the company's public website and other IE cross-marketing.
(Don't forget the anti-trust case is all about actions in the past - the IE3/IE4 era.)
--
More on it from the Register... (Score:2)
My head is kind of turning in circles on what exactly Microsoft is trying to say...let's see if we can decode this...
So first Microsoft tried to say that they could not take IE out since it was an integral part of the OS, which was mostly shot down in trial by government lawyers with the "Look, you can uninstall IE, and the OS still works fine, so you guys were bundling the browser with the OS."
Then the Appeals court happened, and the court decided concerning commingling that MS had illegally thrown source code of IE into other parts of the OS (i.e., in a somewhat random manner as to "tie" the browser to the OS).
So now Microsoft is what, trying to say, "No, actually we did not tie IE into the OS, but rather IE's sources/functions were put into separate files"?
I guess this means that the last quote would be conveniently missing the "... which means that you can easily separate IE from the OS," which would go against their initial argument and show them to be trying to deceive the courts and the public.
Go figure.
OT: Captive car parts brands (Score:1)
Motorcraft is tied to Ford.
Mopar is tied to Chrysler.
Who's tied to the other auto makers?
one f'cking button on their mouse (Score:1)
Huh? (Score:4)
Microsoft's argument that mixing IE and other code in libraries wasn't comingling sounds an awful lot like arguing that you didn't comingle salt and water when you poured salt in the glass of water and stirred it: it conveniently ignores the very definition of "comingle".
Re:Whaaa? (Score:2)
Obligatory Monty Python quote:
Inspector: Then we have Number Four, Number Four: "Crunchy Frog".
Mr. Hilton: Ah, yes?
I: Am I right in thinking there's a *real* frog in 'ere?
H: Yes, a little one.
I: What sort of frog??
H: A... dead frog.
I: Is it cooked??
H: We use only the finest baby frogs, dew-picked and flown from Iraq, cleansed in the finest quality spring water, lightly killed, and sealed in a succulent Swiss quintuple-smooth treble-milk chocolate envelope, and lovingly frosted with glucose.
I: That's as may be, but it's still a frog!
H: What else?
I: Well, don't you even take the bones out??
H: If we took the *bones* out, it wouldn't be *crunchy*, would it?
From the Crunchy Frog sketch [city-net.com]
Files irrelevant; User Interface is all important (Score:5)
What matters here is the user interface. If Microsoft had included Internet Explorer, but required users to download one key piece (that added the icon to the desktop and start menu; nothing else), then Netscape would have been on a level playing field. The issue is what's visible to the user, not the nature of obscure files that Joe User will never care about.
Of course, Microsoft's problem is that they bought into the government's arguement that the browsers and operating systems are legitimate markets to distinguish. Microsoft should have argued that they are really in the user interface market, and as such, the operating system and browser are naturally related.
Re:Whaaa? Like saying a car won't work w/o a radio (Score:2)
have it both ways the whole time.
I wrote an article on this a while back [osnews.com]... it seemed to me Microsoft's arguments were a lot like having a car mechanic say he coldn't remove
your car stereo because that would entail removing the battery too and then the car wouldn't work.
An application? Or integral part of the OS? Which way is it, Redmond?
--
Re:Viral Code (Score:2)
Windows links to IE base code libraries only
IE links to Windows base code libraries only
They both make calls to each others basic code libraries
Figuring out which libraries to LPGL for IE could be, um, problematic at best.
I'd preffer that they be forced to LPGL both, of course. That would allow some one to write an OS that could compete with Windows in it's own space (Would you like some Wine [winehq.org] with your cheese [microsoft.com], Mr. Gates?) ^_^
Re:Whaaa? (Score:4)
I think they've painted themselves in a corner here. Remember (paraphrased) "the browser and the OS are tightly integrated and cannot be separated "? Well, if the files aren't commingled (eck, hate that word) but are indeed just "related functions close to one another", then they damned well could have taken the IE functionality out, couldn't they have? Judges will pick up on that at least, I hope.
Re:An example of commingling? (Score:1)
Microsoft also renamed many of the standard C library functions with a leading underscore, such as _strdup(). Presumably, this is because some of the common C library functions are not "officially" defined by ANSI C, but conveniently the Microsoft function names make porting Win32 software more difficult.
btw, you shouldn't be using strcpy() anyways. Use strncpy() instead to avoid buffer overflows.
Re:Do you believe that? (Score:2)
Microsoft isn't the only organization that believes that. GNU is an operating system. GIMP is a part of GNU. Thus GIMP is a part of an operating system. Wow!
Re:Do you believe that? (Score:2)
Re:Files irrelevant; User Interface is all importa (Score:3)
Slightly more relevant, though still a bad analogy, would be package management. RPM is an integral part of Redhat. The DEB format is an alternate format that has a few advantages over RPM. Yet you cannot totally remove RPM from you Redhat system and replace it with DEB! You would then be unable to install any Redhat packages. Even if you used alien or other similar tool, you still need RPM around.
Or maybe even more germane: GNOME. Many of the Redhat utilities use the GNOME libraries. You must have parts of GNOME installed even if you use the KDE desktop, or you have to forego the use of those Redhat utilities. If I use Netscape on Windows I still have to have Exploder installed. If I use KDE on Redhat I still have to have GNOME installed.
Re:Can't wait til they break up.. (Score:1)
Courtrooms (Score:2)
First is the definition of evolution, all the way up to the definitions of "is".
I say get those fools a dictionary and make them stick to it.
-Ben
Re:Explorer == Shell; Konqueror != Shell (Score:1)
And why would you want that? The integrated shell into windows made the user interface good. The 95/NT shell is such a horrible, unusable piece of junk (IMHO).
The one thing that really annoys me about this though is that it's really difficult to get the good shell in NT4/95 now. You used to be able to download IE w/ ActiveDesktop and install that (which is what gives you the new shell), but that's no more. Nowadays you gotta hope you have an old copy of IE4SP2 laying around....
But to the point, I personally think the integration of IE into the OS was a good thing. If it's not obvious on it's own merits the fact that the KDE crowd has done the same should be a sign of this. I mean come on, Netscape wanted the browser to become the platform, right? The fact of the matter is that Microsoft did a vastly superior job of making the browser the platform.
Re:An example of commingling? (Score:2)
2. StrCpy is included with StrCpyN in the shell api because - well - the shell guys were nuts. You don't have to use it. strcpy still works. The only advantage you get with using it is if you use StrCpyN, you can prevent buffer overflows.
Besides... for maximum portability between platforms, and for unicode/MBCS/ANSI portability, it should have used _tcscpy()
Oh... and just because the app you have to maintain was written by an idiot who didn't think before using an API call, doesn't mean that it's Microsoft's fault.
Simon
A summary (Score:5)
Judge 1: You're guilty. And you suck.
MS: You're wrong. I want to ask Judge 2.
Judge 2: You're guilty. Judge 1, that "you suck" was out of line.
MS: What a victory! Now, Judge 2, we still think you're wrong. Can you reconsider?
My prediction
Judge 2: You're guilty. And you know what? You do suck.
They are still doing this. (Score:2)
Of course your app won't work on anybodies machine unless they have ie5.5 too. Face it MS owns us. They can make us install whatever they want whenever they want. It will only get worse with XP.
Whaaa? (Score:4)
How will the users know where the code is? The in-house developers, maybe, if they are the ones developing the code, but not the users. Developers like myself don't particularly care where the code is either, as long as we know which library to link against. Of course, the judges don't know that.
Re:You didn't expect them to roll over and die ? (Score:2)
Mormonism is, along with Scientology and Islam, one of three religions officially recognized by the editors this web site
Eh? Wtf? Scientology is half pyramid scheme, half mind control cult and half horse crap. And how would you possible reconcile Mormanism and Islam? Is Mohammed coming to take the faithful to their home planet, just like in Scientology and Mormanism? News to me.
- - - - -
Re:A summary (Score:5)
Judge Sporkin's comments [jas.com] on the consent decree:
"It is clear to this Court that if it signs the decree presented to it, the message will be that Microsoft is so powerful that neither the market nor the Government is capable of dealing with all of its monopolistic practices. The attitude of Microsoft confirms these observations. While it has denied publicly that it engages in anticompetitive practices, it refuses to give the Court in any respect the same assurance. It has refused to take even a small step to meet any of the reasonable concerns that have been raised by the Court."
Judge Sporkin: Microsoft's Unwitting Ally [coudert.com], an interesting article from The Computer Lawyer (March 1995).
"On February 14, 1995, Federal Judge Stanley Sporkin refused to approve the consent decree negotiated between the U.S. Government and Microsoft to settle the antitrust complaint filed against Microsoft by the Government. Many think that this denial is a harsh blow against Microsoft. In fact, it may be very helpful to Microsoft for several reasons:
*It started an appellate process during which Microsoft will not be bound by any decree and after which any decree might be obsolete.
*It has caused an important adversary of Microsoft -- the government -- to become an advocate for Microsoft.
*It may well lessen the likelihood that the Government will investigate or take enforcement actions against other Microsoft anticompetitive practices.
*It may focus on a practice -- vaporware -- which is not illegal and which is fairly commonplace."
Civil Action No.: 94-1564 [antitrust.org], UNITED STATES OF AMERICA, Plaintiff, vs. MICROSOFT CORPORATION, Defendant. Stanley Sporkin, Judge.
The Judge Who Rejected Microsoft / Stanley Sporkin is known as aggressive, unpredictable [sfgate.com], S.F. Gate, 2/16/95
"Assigned to review the settlement, he could have rubber-stamped it, which was what both sides wanted. Instead, he peeled the respectability off the agreement like layers of skin off an onion, exposing it as an unenforceable deal that let the government save face while letting Microsoft off the hook.
In his lengthy, fiercely worded ruling, Sporkin characterized Microsoft as having monopolistic practices that pose ``a potential threat to the nation's well-being,'' and he called the consent decree ``too little, too late.''"
- - - - -
Viral Code (Score:4)
Microsoft should consider licensing its IE code under LGPL so that the community can help to better enforce a clear distinction between the browser libraries and the applications which use them.
It's About Add/Remove Software (Score:2)
The issue, it seems to me, is not about whether IE was commingled into the distribution, but as MS says if parts of IE were in the same binary file as some non-IE functions. The effect being that to remove all IE binaries would not be possible if they were commingled in this way. This appears to be an Add/Remove Software issue.
On a related note, as I understancd it, what we call IE is a relatively small amount of code to give a UI to the relatively large amount of code which comprises the internet functionality of Windows. When people talk about "removing IE", they seem to be talking about removing the small amount of code which is the IE UI. I don't think anyone really wants internet functionality removed, and I think adding it to Windows was a big gain for the user. Steve.
Re:More on it from the Register... (Score:2)
"Fraud"
Microsoft is attempting to change its sworn testimony by the sounds of it
Re:Commingling ain't so bad (Score:2)
When you buy a car from Ford do you have to use a Ford stereo? Is there a component in the stereo which is used by your anti-lock brakes, so that if you don't own a Ford(tm) stereo your ABS will not work? Sure, you could install a second stereo with better features, but why bother if you must have the Ford one as well.
What would be great.... (Score:2)
--
blueprints (Score:2)
Having the courts or the government rule on methodology of software engineering is a very bad precedent.
Source code is akin to blueprints. Follow the instructions to make the actual product.
If the blueprints for your landscaping are mixed with the blueprints for your house, or the wiring within the house, how does this damage anything that the user does?
If the blueprints for ParseHtmlTable() are mixed with the blueprints for an FTP file server, again, how does this put the user at risk? Ultimate file size of the binaries is the only downside. If they expose the API to the developers, then every developer on the platform can use it, regardless of what bloated binary holds the bytes.
I think that the court shouldn't even tell the manufacturer what features can and cannot belong in a product: that's what the marketplace is for. Telling them HOW to build code, where it holds no relevance to the safety or choice of the user and infringement of copyright, is outside the expertise and purpose of the court.
You can change Microsoft and Adobe and Doubleclick's ways. Change them with your dollars, with your own quality competition, and with your voice.
Re:blueprints (Score:2)
our antitrust laws are leaning toward the more evolved EU laws; that being, a company's practices need not be bad for consumers in order to be bad; the health of the competitive market must also be taken into account, and if a company's practices are destroying competitors, they company is breaking the law
Hm, tell me how an API in the "wrong" DLL would be harmful to competition?
There are three cases:
If the API were actively breaking competitor's products, that's got nothing to do with what DLL embeds the API's code, and is actionable separately.
good sw engineering practices (Score:2)
It sounds to me like they're trying to confuse the court with technospeak (for BOfH readers... kinda stuff), obfuscating an issue that has NOTHING to do with how the software is presented to the user (i.e., in an integrated manner).
Re:No, files ARE relevant! (Score:2)
MS could have simply distributed the new comctl32.dll in the latest service packs, but chose to include them only in IE, apparently so that those applications would require IE.
One word - comctl32.dll (Score:4)
DLLs (Score:2)
DLLs are an interesting idea--and it's kind of true that a desktop window looks a lot like a browser window--sometimes you can even type a URL into the desktop window if you like... it is, effectivly, a browser window.
But microsoft kinds screwed themselves when they said that they couldn't remove the browser without harming windows. That's what it really comes down to. Most people would have been happy if MS would have just removed the damn E from their desktop!
On the other hand, enforcing this ruling might lead toward a plug-in windowing system. Remove your IE style desktop windows and install netscape style windows. The exact opposite of what MS was trying to accomplish in the first place. Cool.
Re:Ha Ha! (Score:3)
Re:In other words... (Score:2)
My rommmate at the time figured out that if you used the IE3 uninstaller, you weren't left with the fatal aftereffects that the IE4 uninstaller had.
In other words, it looks like Microsofft wilfully broke many peoples systems just so that they could make this argument.
--
Delaying tactics (Score:5)
Re:Whaaa? (Score:2)
If MS wanted to give users the choice, this would be exactly what they should have done.
Re:Some interesting implications of all this... (Score:2)
Well, since they probably all use PCs with MS Windows pre-installed, they've already surrendered. ;^) After all, isn't that what Linux and Mac users are preaching all day long?
Re:Whaaa? (Score:5)
For instance the httpRequest module. It used to be that the only software that needed to make httpRequests was the web browser and so just making those functions a part a stand alone application was all good. MS however noticed that you could use http as a perfectly valid protocol for many other network requests (often that would have proprietary modules to make) and so made a module that could make those requests and because it was useful to all applications (including IE) they put that module in the OS instead of a separate distribution. Again the same thing happens with the html parser/renderer, it works great for help files and documentation for a vast host of applications, so useful in fact that they write it into its own module and include that with the OS so all other applications can use that functionality for free. Same happens for many other modules that can be used for multiple applications but started as functions of IE.
Now we go back to IE and look at what it needs, something to make network requests, something to render the content it recieves, other things that are not available as modules in the OS. So MS writes a wrapper (which turns out to be like 500k or something silly and small like that) that hosts all those controls and makes calls between them and whoa its a web browser.
If an OS were to have made all those functions (httpRequests, html rendering, ftp requests, menue options) part of the API before ever having written a web browser noone would argue that they were just trying to give benifit to as many applications as possible. Noone would say they are trying to corner the browser market would they? They would be touting the ease of application development on that platform because of all the neat high level modules that are included.
Granted MS stole ALL the code for these modules from the IE project. I will give you that they DID try to become the dominant browser by including that wrapper in thier OS so that people could have a web browser with out downloading a single thing. I think they big mistake they made is by saying IE couldn't be removed, they should have maintained from the VERY BEGINNING that sure it can be taken out and shown people how to delete the icons and wrapper DLL and then everyone would be happy at this point.
Now back to the commingling charges, if the modules in suspect (which just have to be the httpRequest and html parser/renderer) were ONLY useful to IE and were not use in ANY other programs by ANY other manufacturers then there would be a great argument that MS only moved them to the OS for the benifit of IE. I maintain however that there are MANY programs that use the modules and some of them are not even web browsing applications and so it should be apparant that these modules are of value to an OS that wants easy development to be possible.
Its sad that what is arguably the largest monopoly case in history could well have been never started (they would have come with another reason to go after MS I'm sure) if MS would have let people delete less than 1 meg of files from there system.
It would have been great if MS would have just added the modules to the OS and then advertised worlds smallest web browser, only 500 or 600k total installed :)
Re:An example of commingling? (Score:2)
Re:Whaaa? (Score:2)
I think that the argument is that by putting closely related functions together in a DLL they can minimize DLL memory consumption. If you separate code according to which MS program needs it, then programs that need code for both Windows and IE will need to load both the windows and IE DLLs. If, OTOH, you break the code down into different DLLs by functionality, you can load only the DLL containing the functions you need. This presumes that MS can predict which functions are likely to be used in the same program, though. Of course there's no reason that they couldn't do it both ways and break up the Windows specific and IE specific code into DLLs by functionality and Win/IE, so that you could load only some really small DLLs if you needed only a few functions.
Re:An example of commingling? (Score:2)
Re:Function Proximity Benefits Users? (Score:2)
Microsoft says "to benefit the user" about as often as a politician says "to protect the children" and with about as much sincerity.
Nuke'em from orbit (Score:2)
I am really tired of them trying to finnagle their way out of legal judgements and traps in a way that is based on only legal loopholes.
This is just something that convicts them more and each day in my eyes.
It is like happened with Clinton, who got nailed for parsing every last syllable in his statements.
Bill Clinton's old arkansas nickname was "Slick Willy". Looks like Bill Gates is a "Slick Willy" too.
With this much bad karma accumulating this quickly, the MS management and legal staff could bypass re-incarnating as Bugs, and go straight back to being mud and pond scum.
It would be an act of mercy to nuke them from orbit. It would stop their every more rapid decline.
Check out the Vinny the Vampire [eplugz.com] comic strip
Hey, we need an acronym for this! (Score:2)
In the spirit of the term FUD, we need an acronym for this MS behavior.
SCLAT: Selective Creation of Legal Terminology
or perhaps more clear at a glance,
LIE: Leveraging Ignorance Everywhere
In more seriousness though, that's what they're doing; hoping that they can establish a legal definition for what "comingling" means, so that they can then prove themselves innocent of the crime they just (mis)defined.
Does anybody know how a technology expert can suggest an alternate definition of how dynamically include()d/require()d code works to the Appeals (or soon-to-be assigned Circuit) court? Not that I'd be that expert, but someone who is ought to draft a proposal.
Re:Do you believe that? (Score:2)
Re:DLLs (Score:2)
Re:Some interesting implications of all this... (Score:2)
MS should know - GPL covers this (Score:5)
2. You may modify your copy or copies of the Program or any portion of it, thus forming a work based on the Program, and copy and distribute such modifications or work under the terms of Section 1 above, provided that you also meet all of these conditions:
[...]
b) You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License.
[...]
These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it.
Thus, it is not the intent of this section to claim rights or contest your rights to work written entirely by you; rather, the intent is to exercise the right to control the distribution of derivative or collective works based on the Program.
In addition, mere aggregation of another work not based on the Program with the Program (or with a work based on the Program) on a volume of a storage or distribution medium does not bring the other work under the scope of this License.
I think this is quite clear in delineating what is "commingled" vs. what is not - and it's on their website!
I'm helping M$ commingle (not by choice) (Score:2)
not_anne
Re:No, files ARE relevant! (Score:2)
An HTML engine can be distributed without requiring that the entire browser be installed and running on the system at all times. Take a look at Mozilla if you need an example.
Those bastards, we need to go back to the good old days when programming was hard work. To hell with component based development!
The problem is not the component nature of MSIE, the problem is that Microsoft required people to install the entire browser to legally use the HTML renderer with their application.
Wait a minute! (Score:4)
Microsoft argued in court that MSIE could not be removed from the operating system, or the operating system would no longer work correctly.
Now they are telling the appeals court that code for the operating system is not commingled with code for MSIE.
Oh what a tangled web we weave...
No, files ARE relevant! (Score:5)
The way Microsoft forced MSIE onto everyone before they could integrate it with Windows was to include the MSIE install on all Microsoft software. Worse yet, practically every Microsoft product in that time period required MSIE to be installed, because commingled with those MSIE DLL functions were general-use functions that the other software took advantage of, and Microsoft would not allow people to install just the DLL's required. This is the same reason why programs like Quicken required MSIE to be installed, and didn't just ship with the "free" Microsoft DLL's that contained the code they needed.
THIS is how Microsoft used it's monopoly to force everyone to install MSIE, regardless of whether you ever upgraded Windows to a version that included MSIE.
And once you've got MSIE installed and constantly loaded into memory, it becomes real easy to eventually give up using Netscape et. al.
Re:One word - comctl32.dll (Score:2)
That's funny, because you can grab COMCTL32.DLL from the Microsoft PlatformSDK without nary a hint of Internet Explorer. It's located in--
[install root]\PlatformSDK\Redist\ComCtl32
Once you install the SDK of course (you can choose to just install the redistributables, none of the documentation or source files).
Plus, you can also download it seperately directly from Microsoft [microsoft.com] as a 400-500k download.
The only thing IE includes with it that I'm aware of is the MSHTML COM libraries, as well as the WebBrowser control (and it goes without saying that you should need IE for this).
Anger...subsiding...blood pressure...normal (Score:2)
You still can't stand the things he does [ridiculopathy.com]- you just can't talk to him about it anymore.
Re:Files irrelevant; User Interface is all importa (Score:2)
---
Re:Huh? (Score:2)
Re:Some interesting implications of all this... (Score:2)
This is quite true. Take MacOS X for example which supports more "Web buzzwords" than I even care to understand....but. Several applications can render HTML right out of the box, the Help systems is heavily based of HTML (and related technologies) and Flash. I get the impression that Apple supplies the ability render Web Documents, without having their own Browser. Something MS claimed was impossible to do (If you remove IE all sorts of things would break..Help, software update...)
Okay, that's strange (Score:5)
Now they are claiming that the location of the relevant function calls right next to each other in the source code is something they did to help the *users*? Errmm, excuse me, but even a Windows *developer* doesn't care about the location in code of the function calls, only whether they are well documented in how you call them (which is a whole different rant). What the hell would a *user* care?
Basicly, it looks to me like they're trying to avoid admitting they scattered IE functions throughout un-related code in order to bolster their original claims, now that those claims have been found to be bogus and that was upheld on the appeal.
--Dave Rickey
Re:Wasting your time... (Score:2)
The better products do not win simply by being better. That is what this is all about. Microsoft's abuse of monopoly power to maintain their monopoly.
Tell me...what happened to word perfect? Corel? Citrix? Borland?
Re:An example of commingling? (Score:2)
If compiler vendors choose to implement their version of strcpy() as a call to StrCpy() then you should blame them, not MS.
What? You use VC++? Well of course it is going to implement strcpy() using StrCpy(), that is almost a given. Last time I checked, VC++ was not part of a standard installation of any version of Windows, and it can quite easily be added/removed at the user's convenience (well, at least as easily as any other Windows program).
You are quite free to choose a different compiler that implements library functions the way you would like them implemented. There is no lack of competetion here, and many great Win32 compilers are even free (beer, not speech)!
If you can't find a compiler that implements strcpy(), or most of the other library functions for that matter, in the manner you prefer, then write your own version of those functions. It doesn't take a highly skilled programmer to move a block of null-terminated memory from one location to another.
Additionally, VC++'s header files contain furious amounts of #ifdef's dealing with versioning. Chances are that the dependency on IE you found in the particular application you mentioned was due to its vendor incorrectly compiling (or misjudging their audience).
Alan
Re:Delaying tactics (Score:3)
You didn't expect them to roll over and die ? (Score:2)
But this is kkkorporate Amerikkka we are dealing with, so it should come as no surprise.
Check out the Internet's most controversial website: adeqacy.org [adequacy.org]. Like a breath of fresh air, and trolling is not tolerated!!
Wow... (Score:2)
--- My Karma is bigger than your...
------ This sentence no verb
------ Ths sntnc n vwl
Re:I may be just a bit too sensible here but... (Score:2)
First things first, you *can* replace the TCP/IP on windows, it's quite easy and painless.
All you need to do is write your own WinSock2 library (the spesification is free, as well as any implementations).
Second, yes, it *is* easy to replace IE with Mozilla.
If Mozilla implement *all* of IE's interfaces.
IE is implemented as a seria of COM objects, with well-defined and *un-changable* interfaces.
I understand that Mozilla implement much of IInternet interface, so it's a good start.
Next, you register Mozilla with IE's GUID, and instantly you get Mozilla where-ever you used to get IE.
--
Two witches watched two watches.
Re:An example of commingling? (Score:2)
MS has been trying to get developers to start using unicode for quite some time, and nt/w2k is for example using unicode the full way. All relevant (string-using) function calls in libraries have two versions, one unicode (with a W suffix) and one non-unicode (a 'A' suffix) that converts the passed string to unicode, and then calls the unicode version.
Anyway, to avoid having to change anything in a program to get a unicode version, MS has introduced macros for all relevant data types and function calls. By changing a #define before you include the windows headers, you select which functions you want to link in and how the data types should be declared. If you want to go the unicode way, a function such as StrCpy() would be #defined as StrCpyW(), while if you didn't specify the unicode define, it would be StrCpyA(). This way, if you have used these functions calls and the relevant data types consistently, getting a unicode version is just a #define and a compile away.
So no plot nor hidden agenda:
Re:One word - comctl32.dll (Score:3)
I was working on an application that included the updated comctl32.dll, and this played merry hell with our distribution. We really didn't want to force IE on our customers, but Microsoft and the vendor of the control we were using said we had no other choice.
Microsoft = Innovation... (Score:5)
Hrmm... (Score:4)
commingle (k-mnggl)
v. commingled, commingling, commingles
v. intr.
To become blended.
Yes, I suppose it would.
Screw 3...
Re:Files irrelevant; User Interface is all importa (Score:2)
I think this was really Microsoft's argument to begin with. The starting assumption for anyone trying to market something is market segmentation:
Q: Do you sell soap and water?
A: Oh no, sir! We sell shampoo, and dish washing liquids, and body gels, and bubble solutions (you know, for kids!), and industrial grime removing agents, and inedable emlusifiers, and...
But then the internet came along, and someone in marketing realized that they could implicitly take credit for all that content, if they made it seem like it was "part" of their core product. So the company that makes edit, notepad, wordpad, word, wordview, etc. sudenly decided that there was no internet segment after all; the internet, they decided, was just part of Windows.
But the forgot to put on their turn signals when they reversed course, and now they get to tap dance.
-- MarkusQ
Comingle This: (Score:3)
With an unlimited "Appeal" operation, users may simply sit back and watch their control-z-initiated laywers "revisit" all those annoying unwanted automatic modifications to their Word documents and the like, and keep on "appealing" until they end up with document they had in mind in the first place.
Hmm. What an "appealing" idea...
Konqueror/Netscape (Score:2)
Commingling ain't so bad (Score:4)
But what I'd rather see in my computer is more commingling, because frankly, that seems to be the best thing for getting things to work. When you buy a car, do you buy a car with a steering wheel from a Mercedes, an engine from a Ford, brakes from a Toyota? No, your car was an integrated object designed by one company. However, it seems like computers are just the opposite. Everyone has a piece of the action and often times these things don't interlock too well. I would be all for commingling if it actually fostered a more integrated product, so I dont have things conflicting with each other.
It's ashame that this enforces Microsoft's monopoly, but I think it's truly better for the consumer.
An answer for everything (Score:3)
Whatever you say Bill!!
~Windfinder