How To Quash Microsoft

...you are an experienced Linux developer and have decided it's time for an all out war on proprietary systems that withhold source code from the end users. It's time to devise strategy. You have tremendous street cred and plenty of programming expertise to call on. No option can be ruled out on mere 'ethical' grounds. How do you go about your task?

Suggested Strategies Outline:

See Also HowToQuashLinux

I'll have a go:

The most ubiquitous use for computers is word processing. Right? So if Linux can win the word processor war, it will win out overall. Microsoft winword format is too proprietary and too far outside community control. Html format by contrast isn't, and in good unix tradition it's a pure text format. What's needed is some careful thinking out of how to make an html editor that's every bit as good as Word and then some. Then we leverage the market share of Apache by adding a new mode to it that promotes a wiki-style of document sharing, using this improved html editor client side. Heck, there's probably some code out there already (Amaya?) that does 80% of this and just needs packaging and promoting better.

You probably don't want a pure HTML editor, as HTML was never built for the sorts of documents that we make with word processors. However, one could write up (and publish and copyleft) a sufficiently featureful XML DTD for word processing and build a word processor that outputs that instead of Word format. The end result would be the same - one open format for multiple word processors.

-- RobMandeville

We're not going to win the intel war, not with intel being in bed with Microsoft, and certainly not by taking it on head on. So we should look one step beyond to next generation, custom reprogrammable logic. We need an initiative to work better with Xilinx. At this stage this means cranking out some really smart graphics demos using the new generation of no-processor FPGA prototyping boards. Linux will be in there running on a RISC engine using a small corner of the gate array. We're way ahead of Microsoft here as we've plenty of RISCy stuff in GNU. This is cool technology, there'll be plenty of techies in Linux world willing to put time in to making this a reality.

The virus and hacking war must go on. We must continue poking fun at Windows's lamentable security - and poking at the high profile servers, especially trying to trip up microsoft.com. Given we're an anarchic bunch, my indirect help in doing this, presented as 'tips on how to make your windows site more secure' as an anonymous contrib in a newsgroup will not be traced to me or reflect badly on the Linux community. :)

On the lobbying and PR side, Stallman is beginning to look a bit of a liability. We need some new blood, some real enthusiast who is media friendly, smart and visibly successful. Perhaps we can get someone with a good legal background to champion our cause and be seen to be coordinating not so much the techie development side, but the side of Linux that makes it usable by the man on the street who is fed up with paying for 'new' software all over again when they upgrade their hardware... With the right front man (or better woman), we'd make huge strides on the PR front.

I've noticed that recently the FSF lawyer has been writing statements more than RMS. Cool

Hmm - wait I'd say. I think it is just a matter of time, but that time may be 2-10 years! Here are several thoughts:

There are only so many features you can put in Office applications, and IMHO MS Office hit that several years ago - the rest is just fluff which may be useful for a small minority. MS is searching for revenue streams in the future, it knows Office and Windows have a limited lifespan.

Windows 9x is crap - 'Sneeze and it falls over'. This is a big disadvantage for home users, and a plus point for Linux.

Which may make you or I angry, but we know that computing doesn't have to be like this. A significant number of the computer user community believe that crashes are a symptom of using a computer, not a symptom of using an unstable OS. This is a marketing opportunity for Linux, but of course the Linux community doesn't get involved with marketing, so ...

... Windows 9x is crap and even Microsoft knows it. That's why they have been trying to end the 9x line and get everyone on 2000 and above. Then this argument is over (my 2000 server has been up for over 6 months).

Since Linux is free it has a huge advantage but really needs much better hardware support to compete with 9x. 

   Windows isn't only OS with flaws, Linux hole found
Time: 15:47 EST/20:47 GMT | News Source: Business Week Online | Posted By: Adrian Latinak On the surface, it was just another turn of the endless cycle of software release, hole discovery, and patching: operating system vendor Red Hat issued an advisory Tuesday warning the world about a serious security hole in a file transfer program that comes with Linux, and urged customers to download a patch. 

 Comment here! - 0 Comments for this story.
For home office and gaming (currently 9x users) - there is no incentive to upgrade to 98 or ME. This will at least continue until Xp is released for home users. Linux needs a hard sell to get market penetration before this event. Once non-technical users have upgraded OS they are unlikely to want to change. Gaming companies will port/develop for Linux more when there is a larger installed base.

For server use - Linux/Apache is ok for 90%+ of web sites out there, although it lacks a number of enterprise features. The server licence fees for MS are not trivial they rapidly add up to a lot of money.

Windows has a big legacy. The fundamental API/programming model even in W2K is an antique from Windows 3.0 days. The need for backward compatibility is a disadvantage in some areas. (It is also an advantage since there is so much software written for this that can not be easily ported to Linux).

The developing world has a different balance between man power and ready cash. The TCO equation is totally different for them. The economics suggest that it is better for them to adopt Linux rather than MS products. But note: this will be crap for the US economy!

I find your support for cracking (not hacking as you call it) deplorable BTW. Grow up - that's the kind of talk I'd expect from a teenage boy, which oddly enough seems to be highly represented in the Linux community.

As opposed to the ethics-free assault proposed, which is the sort of thing well-developed adults plot while lounging around the map room?

For home office and gaming (currently 9x users) - there is no incentive to upgrade to 98 or ME. This will at least continue until Xp is released for home users.

Why wait for XP? Win2K has been around for a while now, and looks and feels essentially identical to ME.

Purely in terms of support for gaming. I have heard (and since I don't get to play games in the office I can't confirm) that DirectX support ain't so good for W2k. Also even Professional is relatively expensive, and complex.

I don't play games in the office either, but I use Win2K at home, and as far as I can tell, it works just fine for games. In fact, some of the popular hardware/tech Web sites (e.g. AnandTech?, TomsHardwareGuide?, HardOcp?, etc.) are even beginning to report that Win2K is faster for games than WinMe?. As for DirectX, Microsoft seems committed to supporting Win2K, though video card manufacturers show some variability in this respect - NVidia, for instance, has good support, while ATI doesn't do as good a job where Win2K is concerned.

I haven't found a game yet that didn't run on W2K at home and Whistler is better (except for those brain-dead companies like EA that specifically code the install to look for Win9X when they don't have to anymore!!)

Personally, I'd be inclined to really really really target the ease-of-use angle. Linux will never, ever make huge inroads into the home and small business sector while it is complex to install, use, and maintain. Large steps forward are being made, but it's still very awkward.

Ease-of-use isn't enough, however (otherwise we'd be talking about how to quash Apple or Commodore). Get a good, solid office-suite set of applications out there. Make them compatible with MS Office for document interchange; okay, the formats are proprietary and complex, but they can be analysed and broken down. If people can't easily take the memo (or the school report) they were drafting at work home for finishing up, they won't be inclined to use Linux. Again, huge steps forward are being made, but it needs more.

Then, once that's done, target DirectX! Make it possible to run those games people buy for Windows under Linux, natively (at least on an Intel platform). If you can emulate DirectX, and do it well, you'll be able to make penetrations into the gamers market.

It's not enough to point out flaws in Microsoft's products. In many ways, people either don't see or are willing to accept those flaws. You need to be able to offer a demonstrably superior solution.

Ask yourself what it would take to make Linux something that a normal human could really buy, install, and run. Hint: remove about 26^2 two-letter commands, 26^3 three-letter ones, and 26^4 four letter ones. Hint: plug and play. Hint: bulletproof and automatic install. IMO Linux is a very long way from something a normal human could learn to love. Meanwhile Win2K is actually straightforward and almost easy to install and very reliable.

They're getting closer, though. Last time I tried a RedHat install, it was straightforward up until it failed to detect the type of video driver I had. Once I downloaded the right drivers, the install worked seamlessly. With the various GUIs in place, Linux is reasonably usable by the average computer user. On a usability scale, it would exceed Windows 3.1, for example. Plug and Play support would be nice, but they'll get there.

Installation is not an issue. Most systems come with the OS pre-installed, and few users are technically proficient enough to reinstall Windows without having their hands held by a technician. (spoken from experience, I was that technician for a while)

The existence of Unix commands is as irrelevant in Linux as it is in OSX. If you want to use them, they're there. If you don't, you can effectively configure Linux using the linuxconf GUI, and do all your file-related activities using Nautilus. The reason most Linux users don't do this is just because we're Unix geeks who prefer the command line.

Although they are getting better, and I have used them all, they still have many problems and don't do a decent job of device detection.

The Linux driver problem is not a simple one:

  1. Manufacturers often won't write drivers for other operating systems; programmers cost money and the money they make from Linux users may not be high enough.

  2. Manufacturers who do support Linux often want to do binary-only drivers because they are afraid of exposing proprietary technology. (I submit that in the long run, it's the non-proprietary stuff that survives, but that's probably anathema to the people who have to report profits at the end of the day - and would like to sell new hardware every few months.) This creates problems because Linux is often subject to changes that make said binary drivers incompatible. The manufacturers that realize this more often than not just don't release drivers rather than have to deal with the headache.

  3. Manufacturers like to cut costs by making the device as simple as possible (hence Winmodems). These need herculean development support to support on the platform they were not developed for, and often don't even fit into the alternative platform's design.

  4. Linux coders who do write the drivers often write only what they need for their own application and code it only well enough that they can get it to work.

If you want easy install MandrakeLinux is king. But better hardware support is required. Windows have millions of drivers. Linux doesn't. I don't know how DirectX could be copied - it's proprietory isn't it? What about OpenGL?

OpenGL is fine as far as graphics goes (some heavy hitters in the game industry like JohnCarmack even prefer it to DirectX), but DirectX includes more than just graphics - sound, input (e.g. joysticks, gamepads, etc.), networked multiplayer support, etc.

DirectX job in life is to provide a closer, faster access mechanism to the hardware without going through all the layers of the OS

Linux command line will never catch on. It will be KDE or Gnome. That's another problem - having 2 of them.

Actually there are more and they will never agree. That's what killed CDE in the UNIX world.

A Gnome developer quotes the following, presumably he agrees with it:

Does that put us back down to one then?

A good start to helping Linux become a better user-oriented OS would be: Take X. Stab it repeatedly until it dies. Then burn it, and jump on the ashes.

Yes

I get the point. Branched into XwindowProtocolShouldBeStabbedAndBurnt

Another point is that Microsoft seems to be shooting itself in the foot with the MicrosoftXpLicensingModel?. Talk about big brother... I'll not buy XP unless they change this.

It's not possible to kill Microsoft. For one thing, the windows operating system and various derivatives are not the issue, the issue is what other kind of stuff do you get ForFree bundled with windows? Microsoft will just keep adding on "innovations" to windows so that when you "buy" (read: when the cost of the OS is passed on to you in the retail price of the PC) the operating system, you get the other stuff that you would otherwise have to pay for. Stuff like web browser, firewall, chat software, word processor, spreadsheet etc. etc. etc. Since this keeps the size of the OS way up, and the speed of the programs way down, the people who sell bigger memory and faster CPUs are totally delighted to bundle the Microsoft OS with the hardware. Anything not bundled with hardware is never going to be able to compete with the same kind of thing that is bundled with hardware, even if the bundled stuff is inferior and the non-bundled stuff really is free.

CategoryLinux!

Ease of use. Reduction of complexity. Adoption by industry will win the day.

Release a stripped down, "Enterprise Desktop", Linux distribution just for the office desktop. Throw out all bells and whistles, only offer a single window manager, and throw out most network servers. It needs Word Processor / Spreadsheet / Presentation / Web Browser / File Manager / E-Mail and a choice of network authentication clients. Include tools to read and convert several competing proprietary file formats. Include tools to automatically push upgrade/install software via the business LAN. Start a company to sell PC's preconfigured with this Linux distro, partnering with DELL, Gateway, IBM, HP, Compaq as the actual hardware suppliers, so you offer a linux desktop office productivity solution, pre-installed, easily customizable to use the clients existing business LAN server environment, (Novell, NT, Solaris, etc). Obviously release a server product as well, capable of interoperating with a maximum of already installed clients, this part is already taking place, just not with your brand. Start producing specialty killer business apps with commercial support, (help and e-mail line, knowledge database, CD helpline subscription, etc.), for your new linux distro. Offer extra killer apps like SAP. Fix GUI user interfaces for existing killer apps like GIMP, and CDRECORD, removing dependencies on command line operations wherever possible. Make the interfaces consistent with the window manager chosen, make sure the window manager is intuitive to use, even similar to windows look and feel, etc..

I'm a gamer - I hate the way I am forced to use windows to play my games. I would much rather be using Linux, but until someone comes up with a program that will let me play any Windows game on a Linux system, I'm gonna have to stay with windows. So long live linux and I hope we get rid of this buggy program called Windows soon, :-)

..switch to MacOsx

See Also: MicrosoftSecretFiles for reasons not not to use MicroSoft products.

EditText of this page (last edited October 19, 2008) or FindPage with title or text search