I certainly don't think a donation model would work. I *do* think a tiered pricing system would! The commercial vs. home user paradigm should be much more widely adopted. Yes, there are home users who want to use Photoshop but will not be making money off it! I would much rather pay $100-200 or something for Photoshop (normally like $700) and be entitled to updates, support, etc. than use a pirated version. It should not just be confined to academic versions. They could even have tiered support, giving priority to commercial customers. I'd have no problem with that.
"Light" versions of applications don't really do it for me either, not by a long shot. Most of the time if I want a big, expensive app, I want all of it.
But it's very true that my use of for example Cinema4D is going to be far less intensive and financially lucrative for me than someone using it professionally. Again I would pay 1-$200 for it and be glad of it.
I think what they need to do is look at what the real value of both their applications *and* services are. If they can parcel it out a bit more and make some things optional, maybe it'd be more "ok" to lower the price. I honestly do not *want* Adobe Bridge, at all. So if that's the big new feature they spent a year making for CS2, they can leave it out of my version and just sell me whatever updates there are to the core Photoshop app. I'd never use ImageReady either (I would just use the web bits of Photoshop - my needs are not that demanding for web use), so they can leave that out. Let's forget light versions, just strip out all the extra crap and sell me the core app for less, as a non-commercial user. By agreeing to the license I am as much legally bound to not use the output commercially as a commercial user who only buys one license is to not install it on multiple systems. If you are worried about the non-commercial user making money, then worry about the commercial user also installing more licenses than they should. In reality they shouldn't worry much about either one. Those who will be trustworthy will abide by the rules, those who wouldn't anyway (and would be more prone to piracy) are basically a lost cause.
The library analogy is a really interesting one. A lot of places do now provide online downloadable demos, and more and more they are making them fully featured I think, instead of cutting them down. I got Adobe Audition and it seems to be fully functional. So that's kind of like the "library" approach - you "check it out" (download it), then you have to "return it" (it expires) in x number of days. You can check it out again (uninstall/reinstall, or at most redownloaded - unless the app is really a bugger and you have to hunt down its install key and kill it or similar). If they made this more condoned it might be a good thing. The constant check-out might entice people to buy. That seems like the idea of a demo anyway, it's just supposedly not for longer-term use.
Personally I think the most fair approach would be some kind of micropayments system. It could be done per company/application, or everyone could go through some standard micropayment gateway, or several options of them (let's say - god forbid - Paypal created one for example). The payment gateway would be a separate piece of software you installed on your system and applications that supported it would detect it and hook in. Then for every say hour you used Photoshop it would charge you $1 or 10 cents or whatever. At the least it would make the actual price a whole lot more bearable by spreading it out over a long time. It would also net *new* income from those who don't actually need to own Photoshop but definitely need it for one or two things here and there, or just for this one thing *right now*. Of course we all know the issues with such systems and why they haven't been done before, but it's still an interesting theoretical approach anyway. It seems quite appealing to me.
The point about "leading by example" in schools, etc. is a very interesting one too. I agree that this can be "setting a bad example" in that it makes people desire expensive software. There's nothing really wrong with using commercial, expensive software provided A: it gets the job done noticeably better/faster/easier than a free/OS alternative and B: it fits within the institution's budget. However at the least the free/OS alternatives should be made aware to students. I do work in an educational institution as the Technical Coordinator and although I don't interact with students much, I am definitely trying to make the move to free/OS software. I have spent years finding the best free/OS programs (this place has helped a lot of late, too), so I have a good stockpile. Now that I'm working for a single company I have to make sure the license allows for corporate use - being an educational institution *and* a non-profit definitely makes that easier. In many cases I've needed to contact authors to get a precise answer on licensing terms. In most cases I am given the ok to use it, which is great. We now use Open Office and Thunderbird on a number of machines and I'm working toward moving to Firefox as well. We use Tugzip or Zipgenius, Foxit PDF Reader, XnView, etc. And notably all of them are vastly superior tools to what was being used before! The point anyway is that using free/OS software doesn't have to be a sacrifice. It's a genuine alternative and it can actually improve productivity and capability.
Btw, having recently tried Sound Forge to work on an audio problem I have, I actually would far rather use Audacity! Audicaty is way faster, believe it or not.
I intend to post a thread about that soon (speed of commercial apps vs. free/OS, specifically in media editing). The point though is that the commercial product is *not* always superior! There is Kristal and Reaper as well for multi-track editing, both seem to do what they do very well, possibly as well as Adobe Audition/Sound Forge/etc. I think in certain market segments there are free/OS alternatives that are genuine competitors. This is not true in all areas, and it probably has something to do with how unusual the UI is, how readily copyable an existing product's UI/approach is, etc. I think good, original UI design is not the strength of free/OS software, in most cases.
Cross-platform needs also play heavily into open source issues. The Gimp is as clunky as it is partly because Windows is not its native platform.
There are a number of freeware paint apps that are much nicer though, like Photofiltre, Paint.NET, etc.