WildParser

With local superspreading-events you can get an outbreak without needing R0 > 1. There can be fire without the world burning down. Levitts math works extremely good - you can calculate the entire German death-curve perfectly. In fact there is no constant R - the model is completely different.

I'm also a bit sceptical about his conclusions. The effects of containment measures aren't really visible in the curve, but e.g. Sweden has a much worse exponent - increasing the chance for more outbreaks will also change the parameters for Levitts curves. I think Germanys curve got very slightly worse recently (shops opening - masks didn't compensate). But I'm not sure if masks did anything at all - there is no clear trend-break in the growth-rate of Germany. Also masks may just keep a local population infectious for a longer time - increasing the risk for visitors.

One biological hypothesis is that covid-19 spreads in local clusters - and has a hard time to move to other places. (e.g. it might need 20 minutes of close-contact talking) The growth-rate is shrinking exponentially (something that Levitt noticed, and can be seen everywhere)- each outbreak locally is trapped and doomed to fizzle, because it is competing with itself - it's just not mobile enough to move to another location before it burns out to maintain a constant R=1.

Btw. The statistics to calculate herd immunity to 60-90% were extremely flawed. They used numbers that came right out of superspreader-events. The most likely thing for COVID-19 is, that it has R0 < 1. It will spread by using a series of superspreader-events and will disappear after a while. Check out Michael Levitt for some understanding of the mathematics of COVID-19... https://www.youtube.com/watch?v=8aHrx68IT7o

In the Gompertz-Model (that worked perfectly for China) Sweden is converging to R~0,8 - with a total death-count <6000. Germany is going to R~0,7 / <9000.

Given the explosive dynamics of COVID-19 that soon converges to R<1 going for herd immunity was most likely a bad choice.

For Germany cutting the flu in half worked - while the bad math of the epidemiologists-community seems to last forever...

Yes, indeed. When you talk to an epidemiologist (e.g. a doctor of physics), ask him to look at f'/f (death per day/death total) in a log-chart they'll tell you that it is 'useless' to look.

If you show him a straight line with R²>0,99 (e.g. Germany using 7 day averages) - they will tell you that it all means nothing and that the curve in fact is bending - because the 'policies' must have worked.

If you tell them that working policies should have left visible marks in the curve and that the curve is converging: They will just block you.

Science is dead - at least in the field of epidemiology.

Michael Levitt has already shown the world a long time ago that COVID-19 is sub-exponential. Any competent mathematician can do that, just look at f'/f in a logchart (death per day / death total) look at the data of Germany - use 7 day averages to be surprised)

So going for herd immunity is most likely mistake. For something that destroys itself in the long term anyway (R0~0,7) something like "duck and cover" would have been the right strategy.

I've made the mistake to write a complex project in Free Pascal. As far as I can tell pretty much every release of Lazarus is breaking something. And surprisingly often those breaks are major (like e.g. broken multi-threading, broken-strings, ...).

For me the whole thing looks more like a playground for hobbyists and is not really useful for anything productive. There's not much continuity in the language. And for the devs something like 95% compatibility seems to be good enough.

For small projects it might be ok to use, but you better keep your snapshot of the compiler locked in a safe place.

On the other hand: If you're young and want to make history as the guy who replaced all those begin/end in pascal with smileys: This project might be your chance...