Thursday: was frantic the night before, but had an excellent (and hugely productive) journey down on the West Coast line (zip-pow-blam-oh, we’re there already?). The Royal Institution was nigh-on deserted, it being the lovely Olympia’s wedding, so we snagged an office and set to work on my revised Christmas Lecture scripts.

Last year’s set are being toured to Japan and Korea, and muggins here is rewriting them. I was a bit nervous – I’ve found it terribly hard to get back into it, and as a result the couple I’ve attacked so far have felt like rush jobs. Several things were heartening, however. Firstly, watching a couple of the shows for the first time, on the way down (thank you UKNova), I discovered that they really weren’t half bad. For the record, David Coleman is a genius – and if you saw Question Time last night, that shamelessly bravura steadicam sweep behind the panel near the end would have been his doing.

Then, the scripts, for all the frantic surgery right up to transmission, were actually pretty coherent. And finally, leafing through them, I spotted all manner of more pleasing ways of saying things.

It was gratifying, then, to have Sir John pick through what I offered up, and profess wild enthusiasm. He must have been heartily sick of the damned things too, but we’ve managed to enthuse each other all over again, and found new ways of nuancing some aspects, new thoughts to draw out, and in one case I think even spotted – somewhat belatedly – what the lecture is about. Heck, John even approves of my recasting of the last lecture, which is going to become such a stirring clarion call for global scientific cooperation I swear we’ll have a string section in there by the time we’re done…

Friday: the How2 website dispatched for another week (what is it with organisations having rubbish upstream connections?), we dove into the months-old SciCast budget and schedule drafts to see what we could salvage. The good folks searching for industrial funding need to know how many people we might reach, which of course means brazenly making something up careful estimation and planning.

It was, frankly, a depressingly difficult meeting. Not because it revealed any underlying weakness in the project (there’s some tricky balancing to be done between mildly tangential objectives, but that’s a rolling ‘taking a step back and thinking really hard’ problem); rather, it underlined just how far we still have to go. It’s been four months since we first raised the idea, and it occurred to me that I could actually have taken that Birmingham TV job and would have finished in time to rejoin SciCast around the earliest point that’s now plausible for a formal ‘Go’. Eek.

So, with a heavy heart, I realise that I’m going to have to plot an escape route. It could easily take another two months to secure the industrial support we need, and I’m not sure I can hang around that long. Which would be tragic. On the other hand it might all fall into place with three phone calls. Still… bugger.

One rousing day, one melancholic one. On balance, life could be worse.