Monthly Archives: October 2020


The news is now telling us there’s a likelihood of another covid “lockdown”. France has just gone full-on prison camp, and other countries are tightening measures.

Last time we gave them the benefit of the doubt: lockdown might just serve a purpose. But that’s conditional on there being an exit path: without it, lockdown is real pain for illusory gain. Six or seven months ago it seemed somewhat implausible, but evidently the government’s advisors thought otherwise and who was I to argue? At least, beyond reservations expressed in that post, notably:

… if the lockdown proves worse than useless in the longer term – perhaps because return to normality proves impossible without the Herd Immunity of most of the population catching it.  But if that happens we’re in good company, with much of the world likely to be in similar trouble.

Now we have experience: “lockdown” failed for lack of an exit strategy. Sweden got it closest to right (at least among European countries) when they introduced more relaxed guidelines that had the effect of lockdown in reducing case numbers but with less damaging side-effects.

Macron appears to have gone mad, repeating an already-failed experiment. Will our own powers-that-be behave more sensibly in this matter? If they lockdown now it’s far worse than in the spring, when we were at least free to go round the supermarket without wearing a germ-incubator. And when indeed it became for some weeks a particularly pleasant experience, with staff and shoppers sharing a “blitz spirit” of cheerful bonhomie.

A question that I’ve wondered about a few times, and raised elsewhere but not on this blog. I’m no historian, but I’ve heard said that Native American populations were devastated by the Common Cold when the arrival of European settlers introduced it. Is this not an interesting historical parallel (bearing in mind that the Common Cold is a generic description for a wide range of lurgies, some of them coronaviruses), and why is noone discussing it? A possible inference is that what makes covid worse than a regular cold is precisely its novelty to our populations.

Concluding ApacheCon

Three days of ApacheCon@home – not far short of a full regular ApacheCon. A comparable number of presentations, far more attendees, but missing some fun elements like lightning talks. I felt too knackered to blog on Wednesday or Thursday evenings, but a few more thoughts bear recording.

In fact it wasn’t just the evenings I felt knackered: I felt diminishing returns on the contents, both presentations and social (not helped by more glitches with the technology). If a change is as good as a rest, the changerest associated with a conference venue and atmosphere is perhaps essential to the familiar experience. Which leads me to the thought: this would have worked better as three separate one-day events. Shorter events could perhaps be themed further for different timezones.

And of course, the primary reasons for making it a multi-day event – to get a decent return on the overhead of travel and accommodation, and to make the most of the in-person event – is gone. It seems to me that more and shorter events make a lot of sense in the online space!

Regarding the substantial contents, I attended several talks in the Geospatial track. Of particular note was one by Lucian Plesea, whose interests seem to have a huge amount in common with my own. His work on accessing and visualising huge datasets look a lot like what I was aiming for with the HyperDAAC, a couple of generations on in terms of both computing and Big Data. Dr Plesea is working at ESRI, so hopefully his work shouldn’t languish unnoticed as HyperDAAC did!

Furthermore, at the core of his implementation (and talk) is a set of Apache modules, practicing what I preach in terms of making it his application server. I was gratified by his reference to my book in the BoF at the end of the day (after his talk). His modules are open source at github, and look interesting, and perhaps deserving of packaging for a wider audience.