For what its worth, I think the bigger record companies are right to be cautious about all-you-can-eat. As Sony and Warner's men at MIDEM rightly pointed out, it is actually a minority interest product, possibly of interest to two groups of music fans - ironically the record industry's favourite group - the aforementioned big-spend music fan - and their least favourite - more prolific file-sharers.

It's the latter group the service has really been designed to appeal to - it being a proposed solution to that question that always crops up at music business conventions these days: "how do we compete with free?" But it ignores the fact that, despite IFPI's scary stats on this subject, the prolific file-sharing community is still relatively small compared to the wider music market. It is also arguably a little naïve in its assumption that if you offer every song ever made as MP3 for fifteen quid a month then suddenly every file-sharer everywhere will go legit.

Some use file-sharing networks because large parts of the record industry's collected catalogue still isn't available via legit digital music platforms, certainly in higher quality formats. All-you-can-eat in itself doesn't address this issue.

Some use file-sharing networks because when they first hit the net two to ten years ago, looking for music, there were simply no decent legit services on there. Certainly none that weren't hindered by digital rights management or artificial blocks to on-demand play. Such services are now available, certainly in the UK, but given file-sharing had a ten year head start, more needs to be done to promote them. Such promotion doesn't necessitate adding all-you-can-eat to the legit menu. 

Some file-sharers are teenagers and students whose predecessors in the 1980s and 1990s would have shared music with their school and college mates by distributing illegally-made home-taped copies. Many such teens of the 80s and 90s grew into fully-fledged paying customers of the record industry once they reached their mid-twenties, and there's every chance the Napster generation will follow suit - even without all-you-can-eat being an option.

It is also worth noting, some file-sharers are just selfish, and would rather artists, songwriters, producers and label people slog their guts out for free in order to provide them with some light entertainment at no cost. Fifteen pounds a month will still be off putting for these people.

And finally, some file-sharers wrongly believe their file-sharing kicks it to the man. Given it's the man providing all-you-can-eat, they're unlikely to sign up.

Of course, some file-sharers might be attracted to all-you-can-eat - especially if three-strikes becomes a reality and actually works (though that's a sizable 'if') - but it's wrong to assume such a proposition would be a panacea.

The record industry should, of course, consider a multitude of different digital music business models because it is unlikely any one will be enough to keep the business afloat. But it seems to me all-you-can-eat is more risky than most, with the potential to jeopardise the other models being developed.

Out of ten years of panic, false starts and bad strategy, a working digital music market is slowly starting to emerge, albeit in the main pioneered by venture-capital-backed net-pioneers rather than the major music companies (or the ISPs, who most actively support all-you-can-eat). This emerging marekt has three strands.

First, the ad-funded on-demand streaming services. So, Spotify, We7, MySpace Music, MUZU, Vevo. For some consumers this is the be-all-and-end-all service, but for many it provides the suck-it-and-see component that many say is the role P2P file-sharing often plays. 

Second, the subscription services. So, Spotify Premium, Sky Songs, Napster. Ad-free streaming, and possibly a set number of MP3s bundled in each month for keeps. A more full-on service for more full-on music fans, and a business model with endless potential for niche-genre reinvention.

And third, the now old fashioned a-la-carte download store. iTunes, Amazon MP3, 7Digital. Generally for more causal music buyers (ie the majority), and probably used by many in unison with the ad-funded streaming services.

This is a nice set up that might just work in the long term. And I worry that throwing all-you-can-eat into the mix will seriously destabilise it all. Which seems like quite a risk just to satisfy a small group of music fans, some of whom will become legit customers over time anyway, and some of whom will never spend money with the record industry, even if you start putting them and their families in prison. 

For my mind, the record industry would be better off working out a licensing model that makes the three above mentioned strands of the emerging digital music market work and prosper long term (because I think we all know the licensing deals currently offered to Spotify et al by the record companies and collecting societies are ultimately unviable).

And then the wider music industry needs to work out whether the combined sound recording and music publishing sectors - based on this three-strand digital market - will be profitable enough to continue to saddle the majority of the cash investment that needs to be made into new musical talent. If not, we need to jump start one variation or another of the 360 degree model - probably by having all interested parties in any one artist (so label, publisher, promoter, merchandiser, sponsors) putting money into a central business venture which then controls an artist's brand, marketing and fan engagement, while dishing out rights, products, performances and partnership opportunities that can be monetised by said enterprise's various business partners. 

And here ends my MIDEM-inspired polemic for 2010. Though there's much from Cannes to come below.