pstoll

Lots of people are looking at local waste-water COVID testing data as a way to get prevalence in an area.

Here’s my local area - note the scale!

https://www.mwra.com/biobot/biobotdata.htm

And the US CDC program for this national effort: https://www.cdc.gov/healthywater/surveillance/wastewater-sur...

The “only 1% will die” argument, sure I remember that one from early days.

But hospitals are currently overwhelmed with mostly voluntarily unvaccinated people. My advice is don’t break your arm or get in a car accident or have a heart attack right now - it could be lethal because health care resources are at a breaking point (in the US).

Figure it’s relevant to this thread - Eric Topol (very credentialed medical researcher) does a pretty good overview & summary on how the vaccines are doing.

https://erictopol.substack.com/p/were-very-lucky

He didn’t go the hospital. That doesn’t change the overwhelmingly positive benefits of “hospitalization rates”.

> Roads are initially built with the promise of future returns in mind.

Roads are like startups! We should get VCs in on the action.

Link is dead.

We have a local school being built. This was the argument for no operable windows.

Except … a pandemic! We are optimizing for the wrong thing IMO at the expense of simple practical, human driven comfort. Same reason we don’t have windows on the local trains anymore.

Practically for people in classrooms at random times, being able to open a window should be an option. Only central control of when the ventilation / heat is on just fails in so many situations.

“Instead of investing in making an existing tool better, we built our own thing!”

Oye.

Same path to getting here. I hope he gets the help he needs - his work has been impressive and useful to the software industry.

Hmm… BLAS started life as Fortran codes (said intentionally to sound like a native Fortean-er).

Saying BLAS is a C library skips over much of the depth and history of its FORTRAN roots. The work that went into achieving numeric stability of this stuff is astounding. Ditto with LAPACK.

https://en.m.wikipedia.org/wiki/Basic_Linear_Algebra_Subprog...

He’s still working on it.

Disclaimer - friend of the author and have done the programming for some of his research. (Now I wish I’d cleaned it up more but oh well...)

I’d describe him as a hobby mathematician, not formally trained. But I wouldn’t put it in the numerology realm.

Short answer: yes.

tldr; answer: having engineers support their own systems encourages them to pay down operational tech debt faster.

A minor nit but the function to dedup the image ids and the corresponding comments seem to be off from a data structure/algorithmic POV.

"Looking at our deduplication function, it’s functional and performant, but not very descriptive. Essentially it checks a running set called seen, which is checked for originality as the deduplication script progresses. As python Dicts are essentially a hash map, the in check compares element hashes instead of each individual component of the dictionary object. This massively speeding up the deduplication process."

Uhm..I don't think so.

It is just extra work to check if the objects are already in the set vs just stuffing them all in and let the set handle uniqueness.

Am I missing something?

https://gist.github.com/pstoll/ae73582763540051d321a4eb15304...

Again, a minor thing. But seeing something like that makes my 'what else do I need to review' detectors go up.

Beautiful madness.

About iOS fingerprint- while a judge can not compel you to type in a password, I have heard that they can compel you to swipe your fingerprint. Something to consider when deciding whether to enable fingerprint access to your smart phone login or other sensitive credentials (e.g. Password manager keychain credentials).

http://jolt.law.harvard.edu/digest/telecommunications/court-...

(Fwiw - I use LP, no master password saved, no iOS finger print access)

$ dig origin.github.com +short

github.map.fastly.net.

199.27.76.133