There were four stand-out conclusions for me. Firstly, the explicit statement that IMS is a long-term strategic investment. This ought to be self evident, but of course we all know that IMS has gone through a rapid curve of hype and deflation. The optimism about IMS providing a much needed fillip to the industry that emerged a few years ago rapidly grew out of all proportion. As Marc LeClerc from Ericsson noted, internet time may be a useful term in some contexts, but it isn't for IMS, and to apply the same time expectations that people applied during the dotcom bubble are completely unrealistic. Of course, that is precisely what the industry did and is now paying for it. But long-term in this case means for the next 10 or 15 years.

It's clearly not enough to expect IMS to fly on the basis of either the technology vision or the initial service rollout alone - a point I heard eloquently made in 2005 by a veteran of the IN world, who argued that much the same mistakes had been made at the outset of that era. Plus ça change...which brings me to the second point. In Paris, this issue was explicitly addressed and tied to the strategic dimension. That is, several operators argued that there is a defined number of services that would be required in order to build the business case for IMS. Further, the business case would work provided that the operator stuck to plans and executed this strategy to build a level of critical mass in their market.

In fact, the number of services that were identified was relatively modest, with a range of 3 - 5 providing some kind of consensus (granted, from a limited sample of subjects). That doesn't sound much given the potential of IMS to be a platform that can enable a huge range of services. At least it should be achievable. If not, why bother at all?

The third point to note was that the arguments that have sometimes been apparent regarding the respective roles of IMS and SDP seemed to have reached a quiet resolution. At previous conferences, questions have often been framed to position IMS and SDP as competing technologies and, to some extent, people have been happy to promote this conflict. This was definitely absent in 2008. In numerous presentations, I saw diagrams that showcased operator's SDPs. But, the crucial point here is that the service enablers that feed into the SDP to enable it to support the (hopefully) wonderful array of services (whether traditional telco or web 2.0-orientated) were provided by the IMS. Thus, IMS is seen in what is, in our opinion, its true light - as a standardised means of exposing service enablers to applications. It doesn't matter if these applications are hosted within an operator's network, or outside, it's a vital part of the overall SDP framework. And, given the complexity of the network information that needs to be available to applications, it needs to be a pretty demanding beast. And perhaps not a malicious squid.

So, a more realistic view of IMS seems to have emerged. It's not a panacea, but it's a solution that will probably be at the heart of operator strategies for many years to come. It will defintely continue to evolve and may change in some aspects, but the basic architecture offers much. It's just that it isn't a quick fix - it needs investment and development to achieve critical mass. And there's probably a good deal of work to do - as Telus noted, there's some way to go before vendor interoperability is really achieved. But, the IMS vision is one that is completely complementary to the overall SDP framework an operator adopts; indeed, it's an essential element of the SDP strategy. You can't build services that really leverage the network unless you can access core capabilities of that network.

Which brings me to our final point. There were several presentations that highlighted the APIs that certain operators intend to expose to third parties - Telefonica's presentation really stood out here. These APIs expose capabilities from the operator network that are unavailable from other sources. The idea, or so it seemed to me in the case of Telefonica, is not to compete with the social networking sites or over-the-top providers, but rather to expose capabilities that further enrich applications that are emerging, presumably so that Telefonica can extract some value in tandem with such application vendors. Is IMS necessary to support these APIs? Probably not - but it would be extremely hard to envisage how every operator, particularly those with multiple networks, could create their own infrastructure for capturing and exposing those capabilities. They probably could, but it might just end up looking uncannily like IMS.