My ideas about operating systems...

[SlugFiller]

if it wasn't Microsoft Windows that won, it probably would have been Microsoft Unix, like it or not.

And then users like us would be able to switch to any UNIX version they want. Well, assuming Microsoft didn't make proprietary changes to the API, as they are known to do (See: Microsoft Java). The fact is, they opted for a proprietary API, and not for any practical advantages it has, but rather because they knew they could corner the software that way. The proof is in the fact they did the same with any other API they laid their hands on.

If a certain Linux kernel was pre-installed by enough hardware vendors, it would probably be the end for Microsoft.

Until 95% of the software you can buy off-the-shelf runs on it, it's just a dream.

Chicken and egg. If hardware vendors pushed Linux, software developers would follow suit. Software developers simply opt to support what people have, and currently that is Windows.

It does create a cycle, a catch-22 of sorts: Software developers wouldn't support an OS that isn't widely used, and an OS without enough software wouldn't be widely used. Nevertheless, the decision on which OS most users use is usually that of hardware vendors, not users, as most wouldn't know how to switch if they wanted to. That's why hardware vendors pretty much determine the market, but Microsoft's licensing policies give them a very strict control over hardware vendors.

Microsoft's business model does not include paying to support an old OS

I wasn't talking about Microsoft supporting it, I was talking about their official announcement of the matter, and the removal of related resources from their web-site. Imagine, if you will, if Microsoft simply stopped releasing updates for 9x, and had their own software NT-only, but didn't mention anything about end-of-support, and didn't remove the relevant resources and downloads from its website. Do you think software developers would have removed 9x support then?

Actually, you don't have to imagine. Microsoft actually dropped support over a year before their announcement. No one noticed. They even released NT-only programs, like MSVC .Net and installers of the .Net framework. After the official announcement, though, programs gradually began to come out with built-in OS version checks. This was even before any of them used any NT-only APIs.

Same reply to this:

Most of them wouldn't care. Windows 98 could still be supported for 150 more years, and they still wouldn't care/make apps for it anymore.

It is a fact that they only began to care after the official announcement. Before it was 10 year old, it was 9, 8, and 7, but developers didn't begin to care about that until after Microsoft officially announced it was done.

They may have just as well planted a "time-bomb" program in the kernel, at least as far as most developers are concerned.

They do, in every product they create that is initially supported - and it's 10 years for commercial products.

Uh, do you even know what a time-bomb program is, or are you just babbling? It looks alot like the latter. A time bomb program would simply say "Sorry, your OS is to old, please upgrade" in a start-up blue-screen.

While this is unheard of for OSs, I have seen commercial programs that do just that. They give you a "sorry, this is as far as we allow".

However, as 98 is gradually losing its main purpose as an OS - running programs - it might as well do the same. When that happens, there is only one solution left, the same solution which always works in this case: Cracking it.

And if you think all the developers would all of a sudden move from VB, VB.NET, C#, Delphi and what have you universally to GCC, ditch MFC/ATL/WinForms/WPF and such for GTK/QT and the like, and also ditch DirectX support for OpenGL and all that...

Mind you that they did the opposite when the above came out. The standards used by GCC are way older than anything Microsoft came up with.

Also, OpenGL is, as the name suggests, open. It is freely available in open-source OSs.

DirectX is proprietary, though, but I think there may be some form of support for it in non-Microsoft OSs.

Making apps for linux often would mean having to change your business model altogether too (dual license, or sell support or such).

I wasn't talking about dual-licenses. If hardware vendors pushed Linux instead of Windows (Microsoft license policies prevent them from doing both), software developers would simply create apps that are linux-only. The problem would be done then and there.

If you run a middle range business, bigger than the ones where the owner buys everything on a personal VISA and smaller than the ones with an IT department

Let me stop you right there. The 1% that use computers as part of their business, that is, companies, already use linux, probably on Sun computers.

What we are talking about is that 90% of the computer-user population commonly known as "end-users". They use whatever comes with their computer, and never even realize alternatives exist. They are controlled by hardware vendors which are, in turn, controlled by Microsoft. That's how Windows becomes the most commonly-used OS on the planet. That's how Solitaire becomes the most commonly used application.

People which will use XP for sure next decade have several reasons for it:

(Snip)

2. Most apps they use work with it

I give it 5 years before that reason disappears. Wanna guess what happens after that?

Finally:

@crahak:

Regarding your last post. It seems with every sentence you are digging yourself into a deeper hole: Compiz, the GL-based GUI for linux, already contains all of the features Vista has, and then some. Yet linux in its entirety takes less resources than XP. With each new version, the resource usage decreases. Mind you that there are linux desktops that make Compiz look like Win3.1 in comparison, and they also use very little resources.

That you think resource usage should increase just goes to show how far you've been brainwashed by Microsoft: It is nothing more than a combination of poor programming and an aim to push for better hardware (to promote the symbiotic relationship between Microsoft and hardware manufacturers).

It's not just the desktop either. Drivers, stability, security, networking, linux has Vista beat hands-down on all of these, even if you look at a 10-year-old linux.

And yet, linux versions do not require progressive hardware updates. They do not require paradigm shifts. They maintain full backwards compatibility, with both software and usage. If you want, you can configure your latest linux to look exactly like one that is 10 or 20 years old, down to the boot options.

That Vista could never be 98, that it can't be configured to do what 98 could do no matter what, that it wouldn't even run on the same hardware, that just goes to show how fundamentally flawed it is.

Vista will prevail, though. Despite being far worse than the alternatives, and possibly worse than even its predecessors, it does have a few undeniable advantages that would give it market domination. In increasing responsibility for the aforementioned, they are:

1. The strongest alternative, linux, requires manual editing of poorly documented text configuration files to work properly.

2. Software developers continue to use the latest version of Win32API as determined by Microsoft. As Microsoft continues to add changes, however impractical and useless they may be, to that proprietary standard, latest software would simply require the latest OS.

3. Hardware vendors will continue to push the latest Microsoft OS with any pre-built box. Since custom built boxes are a niche market, this alone would be sufficient to ensure widespread use of Vista.

[CoffeeFiend]

None of Windows' source code came from CP/M. Design and such, yes, like everything else. Doesn't make it the same product.

The point was, they ALL inherit stuff from older/other OS'es, be it code or otherwise.

Design plans are something different. A new plan might be based on an old one, but it's still a new plan.

How's that any different? You take the old and modify it a bit. But just because you decided so, editing plans makes it a new plan, but editing code doesn't make it new code? Do you keep a list of everything where making small evolutionary changes makes it new or not? Say, I open house plans in autocad, and move a window to the left by one foot, this is a new house somehow, but if i open the source code for some GUI, and move a button 50 pixels over, then now it isn't new? It's a double standard if anything. Makes no sense to me. It's the EXACT same thing. Both are updates on a prior design, and selling products based on that still makes it a new product regardless of what the product may be.

It's not efficient at all, considering how little more it does for the average user.

It DOES more for most users (ease of use, new features, etc), but maybe you just don't notice it... Just like cluberti mentionned.

There's a mistaken belief that a new major version of a product should use up much more RAM, and that therefore it's great. This is nonsense.

I've never seen anyone, or heard of anyone with such a belief. But ~99% of new apps do use more RAM. What makes it great is the new features. Extra resource usage is just something we more or less expect to happen with every new version (new features take resources). FF3 is an exception here, their main savings comes from not keeping all the images uncompressed in memory, which was due to a poor design decision in the first place (the kind most systems haven't done)

[CoffeeFiend]

It is a fact that they only began to care after the official announcement. Before it was 10 year old, it was 9, 8, and 7, but developers didn't begin to care about that until after Microsoft officially announced it was done.

Care to back up that "fact" with solid evidence? Developers make software for platforms with a large user base, because there's where the money is. If 90% of users were still using windows 3.11, we'd code for it, support or not. Win9x having < 1% of a install base is why, not support.

Mind you that they did the opposite when the above came out.

In some cases, they did. Doesn't mean they'd do the inverse again.

I wasn't talking about dual-licenses. If hardware vendors pushed Linux instead of Windows (snip), software developers would simply create apps that are linux-only. The problem would be done then and there.

Wishful thinking. And dual licenses and selling support is a huge part of how companies live off of making software for linux regardless. Hardware vendors already offer solutions with Linux, but no one is buying. People want their windows apps, so they buy a Windows PC.

Let me stop you right there. The 1% that use computers as part of their business, that is, companies, already use linux, probably on Sun computers.

Business/servers represent FAR more than 1% of computers, and Sun servers aren't quite that common no. We do have a handful of Linux boxes here and there, but they're outnumbered about 100:1 by Win 2003 boxes. It's like that for most companies I've seen (if you put aside the cheap LAMP hosting companies, where it's mostly linux)

It seems with every sentence you are digging yourself into a deeper hole: Compiz, the GL-based GUI for linux, already contains all of the features Vista has, and then some. Yet linux in its entirety takes less resources than XP.

Eh? Where exactly did I mention the Vista GUI was great? Or that Aero Glass (which I don't use currently) uses reasonable video resources? Right, I haven't!

Compiz Fusion runs FAR better on sh*t onboard Intel GMA 950 video than Aero Glass does on a 100$ vid card. You're trying to put words in my mouth there, and saying I'm the one digging myself in a hole, funny!

In fact, that's my main deception with Vista (2nd one being the lack of a SSH-like console, for which I now use CopSSH and PowerShell remoting), I had expected Aero Glass to be somewhat closer to Compiz Fusion. But it basically lacks almost everything CF has, and there still aren't any viewports or such.

Drivers, stability, security, networking, linux has Vista beat hands-down on all of these, even if you look at a 10-year-old linux.

Drivers? Certainly not. Stability? Both are plenty stable... Networking? Both work fine (altough Linux's is admittedly superior, I do use linux for NAT/DHCP/DNS, 6to4 tunnelling + radvd, squid + ad blocker, etc). Linux only has DIFFERENT issues. It's not better than windows, just better suited at other tasks, and the apps available for it are vastly different.

They maintain full backwards compatibility

Not. A lot of linux apps much younger than 10yo just won't run on linux for various reasons (glibc changes, etc). And there is no stable ABI either.

that it wouldn't even run on the same hardware, that just goes to show how fundamentally flawed it is.

Right. And you try to run a modern distro with KDE or such on your old 2000-vintage Win98 box. It'll be slow as molasses too. And not running on 10 year old hardware means fundamentally flawed exactly how? It doesn't HAVE to.

Edited June 18, 2008 by crahak

[BenoitRen]

Vista security (heck, even XP security with SP2/SP3) plus additional features are what people want, not like people who post on forums like this.

People *want* media center. People *want* photo and video to be easy. People *want* to be able to use almost any app or hardware device they can find on shelves, and have it work either by just plugging it in, or putting in the CD and then plugging it in. People *want* a machine that they don't have to think about security, although UAC is a bit obtrusive in this aspect (good design idea, bad implementation).

Which Windows XP already does fine. Even Windows 98 SE did to a certain extent.

Applications supporting an operating system isn't a feature, anyway.

It's the EXACT same thing. Both are updates on a prior design, and selling products based on that still makes it a new product regardless of what the product may be.

I'm starting to think we're not getting further than arguing semantics, so I'll leave it at that.

It DOES more for most users (ease of use, new features, etc), but maybe you just don't notice it... Just like cluberti mentionned.

And how, pray tell, does rearranging the user interface for ease of use require more RAM? Why should new features I don't use use RAM?

But ~99% of new apps do use more RAM. What makes it great is the new features.

More features does NOT necessarily translate into more memory usage. It should especially not use more memory if I'm not using the **** feature in the first place! That's just sloppy programming.

Hardware vendors already offer solutions with Linux, but no one is buying.

Excuse me? The Linux computers at Walmart have been selling like hot cakes, and inexpensive laptops with Linux are all the rage recently.

[CoffeeFiend]

And how, pray tell, does rearranging the user interface for ease of use require more RAM? Why should new features I don't use use RAM?

And because YOU don't use the feature, and that such a feature uses RAM, it shouldn't be there for everyone else?

That's just sloppy programming.

Yeah right. That explains why 99% of new versions of 99% of apps use more RAM -- sloppy programming! There sure must be an awful lot of sloppy developers out there.

Excuse me? The Linux computers at Walmart have been selling like hot cakes, and inexpensive laptops with Linux are all the rage recently.

Who's surprised a 200$ PC would sell? (even if it's a P3 class CPU with 512MB ram) , from a place where most shoppers don't really know what they're buying no less. What I really wonder is, what percentage of those run Windows now? 95%? And "selling as hot cakes" probably means FAR less units sold there in a year, than Dell shipped in a single day (I've seen ppl mention 10000 total, a VERY small amount when you look at the big picture). Walmat also doesn't sell them in-store anymore either because of poor sales. Oh ya, and the only guy I know with a Eee PC, runs Vista on it

[GrofLuigi]

Reading the topic title... I see nowhere "my idea about Microsoft's profit". No offense, but too many people worry about that. I never could understand the urge some ppl feel to defend Goliath from David. Microsoft invests millions of dollars in that anyway.

GL

Edited June 19, 2008 by GrofLuigi

[georg]

Quote (Slugfiller) "The 1% that use computers as part of their business, that is, companies, already use linux"

http://www.informationweek.com/news/intern...cleID=201807072

By Alexander Wolfe, InformationWeek, 19 Sep 2007

According to Gartner's figures, 67.1% of servers shipping during the second quarter of 2007 were fitted with a Microsoft OS; 22.8% had Linux. Interestingly, that was down slightly from the 23.1% share Linux had in the year-earlier period. According to the W3Counter Web stat site, Linux recently achieved a 1.37% share to inch past Windows 98 (Windows 98!). Perhaps the best reality check I found comes via an N.C. State University survey of users of ResNet, the residential network for students, which found that "Windows usage has consistently been over 90%. Linux usage was at its highest in 1998 at 2.47%. It dropped to less than 1% in 2003 and plateaued [in 2006] around 1.5%."

Quote (BenoitRen) "The Linux computers at Walmart have been selling like hot cakes"

http://www.desktoplinux.com/

By Steven J. Vaughan-Nichols, 13 Mar 2008

On March 10, Wal-Mart spokesperson Melissa O'Brien told the Associated Press that Wal-Mart had decided not to restock its in-store gOS Linux-powered Everex Green gPC TC2502. "This really wasn't what our customers were looking for," O'Brien said. I decided to do a little old-fashioned reporting to see if I could get to the bottom of why Wal-Mart will no longer be carrying cheap Linux systems in its stores...so I visited several of my local Wal-Marts. Here's what I found out. Customers did want to buy the computer, but they, and all too often the workers, were thoroughly confused because the PCs came with Linux and not Windows. The word "clueless" comes quickly to mind. The one technically adept customer service representative I met told me, "These are the same people who really can't tell the difference between the computer and its software. At best, they know they need Windows to run Quicken, Office and games. That's it." ...all too often, "They thought it was a normal -- read, Windows -- PC and they exchanged it."

Qoute (Fredledingue) "install only what is relevant to your...needs...update only the parts which you had installed"

http://arstechnica.com/articles/culture/mo...s-will-suck.ars

By Peter Bright, Ars Technica, 5 Apr 2008

"modular Windows will suck for Microsoft and suck for you" (long article with many responding in forum)

Comment by Uatu in forum:

"If you're a big fan of modularization, you can already roll your own with nLite, which can strip an XP iso to under 100MB (around 250MB install on HDD, 35MB Ram utilization with all optional services stopped). Or W2K to 70MB, less than 200MB on HDD, 30MB Ram, which is in the ball park of **** Small Linux. You can strip IE and Media player entirely, including the Html libraries and the DirectShow framework etc, and Firefox and VLC still work. Of course, most of your other software and hardware won't work, because they were written with the assumptions that tons of components are present. You can then add component by component to find out the dependencies of the software. And after you've added them all to ensure a broad range of compatibility, you're more or less where you started with a standard XP install. And if MS were to do this modularization thing, that task will be shifted to software writers instead of geeks playing with nLite for fun. So I basically agree with the article conclusion. Modularization has minimal benefits and lots of drawbacks for consumers and software vendors, and probably even MS itself in the long term. It might be beneficial for very specialized applications, but MS already offers XP Embedded (basically an MS sanctioned nLite) for that."

[CoffeeFiend]

Modularization has minimal benefits and lots of drawbacks for consumers and software vendors, and probably even MS itself in the long term. It might be beneficial for very specialized applications, but MS already offers XP Embedded (basically an MS sanctioned nLite) for that.

He TOTALLY hit the nail on the head!

Windows is already plenty "modular" for most people's tastes. You can already stop services you don't want, and there's add/remove features too. You can regsvr32 /u some unnecessary components. And then there's nlite and vlite for extremists. And if you need a very specialized stripped down to the bone OS for special purposes, there's XP embedded. Just what else do some people expect?

The only thing I can think of that they don't let you easily remove is IE (nlite will do it...), and their awful media player -- but then again, just look at the total non-popularity of the "N" versions of windows, and you'll already see people DON'T want such features-removed crippled OS'es, it just doesn't sell.

On the other hand, developers, from small ISVs to big corps would have to test for the presence of various components (e.g. IE, for it's ActiveX controls), making installers more complex, take longer, have more conditions, and then force people to install them anyway. Otherwise every program would come up with their alternative to the feature that possibly wouldn't be there, spending time and money reinventing "the wheel". And with each vendor shipping their own version of "the wheel" (for the microsoft one is optional and might not be installed, and they sure won't want to try to support 2 different ones either), you will be stuck with 50 other different wheel-like contraptions from various vendors, that each has their own bugs, quirks, vulnerabilities, and need their own patches. Each "wheel like contraption" would use their share of disk space (more binaries to install) and memory (more private bytes, less shareable ones), using more of both in the end, which kind of defeats the whole point of it. It just goes against basic ideas like code reuse. You can be sure all developers would be *so* happy about that (I can already see that being the OS taking the most time to have apps updated to work with, if ever). It sounds like extra complexity, and extra configuration overhead too -- something businesses and admins would sure love! And power users would sure love to be nickel and dimed for every feature too. And even end users are going to love either have apps not work because of missing features (I can already hear them say "windows sucks"), or have lots of popups telling them to install the missing features -- kind of like this: you click on a shortcut, and it pops up "would you like to install tcp/ip stack?", then 5 minutes later "would you like to install http protocol handler?" then a couple later "would you like to install the html rendering module?", etc (sounds like fun!) It would be the worst thing ever.

I can't really think why they'd go that way at all. Server core made sense for servers (no GUI at all, headless servers, so a bit less overhead), even though server core is kind of crippled. But it's not like you're gonna want a GUI-less headless desktop. And again, with the ridiculously low sales of the Windows N series, the world showed us they DON'T want crippled featureless OS'es:

"XP N sales represent 0.005 percent (1/20,000th of one percent) of overall XP sales in Europe.

No PC manufacturers have ordered or preinstalled Windows XP N on PCs."

[cluberti]

This has devolved into a Win9x vs Linux vs Vista vs kitchen sink flame fest, and convincing one side or the other that their opinion is incorrect is crazy (and it never works). Since there is some good stuff in here, I'm not trashing it, but I am closing the thread.