mhandley

42 ms·

421.
by mhandley 6y ago
I first saw this on Twitter four days ago, but only one example was given there. Turns out that priming from a single example four days ago did hold for me - I got the second half of all the sentences today without any further priming, and
422.
by mhandley 6y ago
Looking at cumulative graphs like that is misleading. It's also impossible to conclude anything about peaking in the UK from the total figures, as the baseline for testing has been changing as the number of tests has increased rapidly
423.
by mhandley 6y ago
From the paper: "Interpretation: We were unable to confirm a benefit of hydroxychloroquine or chloroquine, when used alone or with a macrolide, on in-hospital outcomes for COVID-19. Each of these drug regimens was associated with decr
424.
by mhandley 6y ago
Agreed - really wish more people would use ISO date format: https://en.wikipedia.org/wiki/ISO_8601
425.
by mhandley 7y ago
The incubation period runs out to around 27 days, though above 2 weeks is uncommon. It seems possible one of them already had it before quarantine, and just happened to be one of those who had a long incubation period. https://w
426.
by mhandley 7y ago
I don't think you can assume that non-covid-19 causes should be lower than previous years. For example, the main cause of deaths in normal years is cardiac problems. If Covid-19 is discouraging people from seeking help, and especiall
427.
by mhandley 7y ago
The Royal Society's policy briefing report on Green Ammonia goes into all this in detail, and makes the case pretty well: https://royalsociety.org/topics-policy/projects/low-carbon-e...
428.
by mhandley 7y ago
Is the 1M/day figure for individual tests? When the expected number of positives is low, can't you run group tests? Mix the first samples from 16 patients together and test them as a batch. If the result is negative, all 16 are
429.
by mhandley 7y ago
Sorry about that! It did rather morph from our nice simple initial design into a monster once it met the real world.
430.
by mhandley 7y ago
I spent a decade working on multicast, helped write some of the second generation multicast conferencing tools, and co-authored a number of the standards. Generally, the issue was that multicast requires per group or per-source state in ro
431.
by mhandley 7y ago
Is this any more useful as a comparison? http://nrg.cs.ucl.ac.uk/mjh/covid19/
432.
by mhandley 7y ago
The problem is that for this virus, "just enough" immune people is about 60% of the population. If 1% of that 60% die, as will happen with COVID19, that's a lot of deaths. The other alternative is social distancing, so you c
433.
by mhandley 7y ago
Agreed, but in the end what will make most difference is getting off of exponential growth soonest. If you don't do that, no country can care for the surge of patients. South Korea is the only country to shut down exponential growth
434.
by mhandley 7y ago
At the moment it's looking like France is doing better than Germany. Many Eastern European countries acted early too, and it looks to be working. Lots of graphs here: http://nrg.cs.ucl.ac.uk/mjh/covid19/
435.
by mhandley 7y ago
I just posted today's update: http://nrg.cs.ucl.ac.uk/mjh/covid19/ Denmark and Norway are less than a week behind Italy in terms of confirmed cases per million inhabitants, as is Switzerland. Spain moved clo
436.
by mhandley 7y ago
Denmark's situation is very unusual. This graph gives an idea of where Denmark lies relative to Italy: http://nrg.cs.ucl.ac.uk/mjh/tmp/covid-eu-norm2.png This is actual case data, plotted with the time offse
437.
by mhandley 7y ago
When is too late to act? Clearly Italy acted too late. I've been plotting the actual case data, in an attempt to get some idea of where countries lie with respect to Italy: https://twitter.com/MarkJHandley/status
438.
by mhandley 7y ago
Tradition has it that it was Kings that stole his head. After all, Kings was founded as a religious alternative to "that godless institution on Gower Street". UCL: proudly secular since 1826.
439.
by mhandley 7y ago
My inaugural lecture as a professor back in 2003 was held as part of a celebration of 30 years of the Internet at UCL, where Vint Cerf and Bob Kahn received awards. Pretty daunting audience for a new networking professor. Afterwards there
440.
WHO Director-General's opening remarks at Covid-19 media briefing – 5 March(who.int)
1 points by mhandley 7y ago|0 comments
441.
by mhandley 7y ago
A year or so back my son and his friends built a muon detector as a high school project, and then flew it from London to Germany and back to measure muon rates at altitude. Basically two metal boxes glued together, with some hand cut holes
442.
by mhandley 7y ago
Back in 1993 or so, I was doing my PhD and also ended up doing a lot of videoconferencing demos at various conferences. This really required a Sun Sparcstation 20, maxed out with dual CPUs, as much RAM as it would take, and a video capture
443.
by mhandley 7y ago
Yes, that's correct, but the interesting part is not the footprint, but the latency and variability of the dynamically routed paths you can obtain. That's what I simulated. I also made some modifications to how you'd normal
444.
by mhandley 7y ago
Elon Musk has stated that until they have inter-satellite laser links deployed, they can get wide area low latency communications by relaying via ground stations at a high slant angle. I built a simulator and ran an analysis of this, and a
445.
Astronomers detect biggest explosion in the history of the Universe(phys.org)
1 points by mhandley 7y ago|0 comments
446.
by mhandley 7y ago
Ammonia, produced using renewable energy, is one carbon-neutral fuel that might work for aviation. The Royal Society just produced a full report on green ammonia and its possible role in a carbon-neutral economy: https://royalso
447.
by mhandley 7y ago
We ought to ship all our kids off to some boarding school somewhere, and deliberately infect the lot of them. Once they've recovered and are immune, they can come home again! Unsurprisingly, I've not actually suggested this to my
448.
by mhandley 7y ago
For some reason, we all know the right order of magnitude for a car in horsepower. But most don't know that 1hp is roughly 750w. It's odd because we all know roughly the power rating of a microwave or a lightbulb or a kettle. A
449.
by mhandley 7y ago
802.11 AMPDUs already tie up the link for ~4ms in normal operation. Without this, the medium acquisition overheads kill throughput. But you're correct that a single 64KB packet sent at MCS-0 would take a lot longer than that. 802.11
450.
by mhandley 7y ago
For 802.11, the biggest overhead is not packet headers but the randomized medium aquisition time so as to minimize collisions. 1500 bytes is way too small here with modern 802.11, so if you only send one packet for each medium aquisition,

More ›