re Lemon Law
Refusing to agree to software licenses that prevents us from holding these companies accountable is the quickest way to motivate change.
Since purchase and use involve agreement, you are advocating a boycott. Some organizations already boycott MSFT products on grounds that range through and include price, security, lack of source for customization and/or improvement/extension/bug-squashing, philosophy, and incompatibility with standards (e.g., web standards, accepted Java "standards", etc.). MSFT still finds enough willing victims, though, due to a range of factors that include MSFT-created barriers to exit (file format lock, undocumented APIs, etc.), lack of economic incentive for free solutions to advertise themselves and consequent customer ignorance, and the self-interest of MSFT-trained and MSFT-dependent "experts" who pitch MSFT to capture ongoing revenue fixing MSFT's self-destroying product environments.
Movement away from MSFT is building momentum, but will not be a quick trip. The thing most in favor of abandonment of MSFT solutions is MSFT's own ineptitude: if MSFT actually delivered security or standards-compliance or reliable operation, the wind might be taken out of some of its critics' sails.
We need a Lemon law for software.
This actually has a side effect you might not anticipate. Liability under the "Lemon Law" could create a very serious barrier to hosting or contributing to no-fee open source software projects: you get no money from people to whom you acquire serious liability ... you re-think release/participation/etc.
An exception for software to which the user has the code and the right to promulgate altered versions of the code would protect (a) open source projects from liability and (b) the public from software that they can't fix if the vendors skip town. It would then be up to implementors to offer contractual terms guaranteeing the performance of open-source solutions, not the legislature. Do-it-yourselfers would get what they paid for, and people wanting a guarantee could buy it from vendors.
The utter lack of interest at MSFT to have a working product, combined with its scorn for its victims, became evident to me in 1995 when I attempted to use a MSFT scripting tool in a MSFT product on a MSFT operating system to solve a large-scale data-handling problem in a laboratory which was performing research at a state-run medical school. Calling for support essentially led you to a laugh track. No fix was available for any of the critical bugs that plagued the various MSFT wares; victims were expected to pay upgrade fees to get screwed by yet another no-promises-made-or-honored sales/service experience. You would think that MSFT would be interested in a customer the size of a large state government, or even a high-profile medical school, but in fact it was clearly not the case.
MSFT has made its bed and after a while, if things still don't change, the momentum among people who don't want to pay to be screwed in that bed will reach a point that MSFT's practices will have made it irrelevant in the markets in which it hopes to maintain a monopoly. Sure, Linux isn't "ready for primetime" on the desktop, but compared to where it was 10 years ago, it's clear between MSFT and freeware which one is has the positive relative velocity.
Consumer protection laws requiring release of file format specifications would be a substantial advantage to consumers, allowing third parties to facilitate migration and conversion to/from closed formats in which businesses can have data locked. You don't have any idea what a PITA it is on my Mac to want into a file created in Lotus WordPro on OS/2 Warp v.4. People with old versions of Word, WordPerfect, and the like suffer this all the time in business. File format lock, obsolete software unable to run on modern hardware, and the like really do create a pretty nasty cost of business. Addressing this would be of real utility in opening migration prospects and re-invigorating the competition for desktop productivity applications. Until MSFT cannot continue to keep its file formats a moving target, no-one will be able to offer real data migration, and the barrier to conversion will lock people into old solutions or a path of MSFT expense.
Take care,
--Tex.