35,698,655 total cases
1,045,953 dead
26,867,153 recovered
In the US:
7,679,644 total cases
215,032 dead
4,895,078 recovered
Yesterday’s numbers for comparison:
Good news.
Just to clarify, good news that safety is not being compromised in response to political pressure and that even in the face of strong attempts at compromising safety the FDA is standing firm for safety.
I sense some firings on the horizon…
36,044,736 total cases
1,054,604 dead
27,149,067 recovered
In the US:
7,722,746 total cases
215,822 dead
4,935,545 recovered
Yesterday’s numbers for comparison:
Wisconsin:
Desperation is the world’s worst cologne.
Well, his adoring acolytes, apparently immune to science and medical advice, adore him even more now.
36,394,149 total cases
1,060,462 dead
27,412,267 recovered
In the US:
7,776,224 total cases
216,784 dead
4,983,380 recovered
Yesterday’s numbers for comparison:
36,753,910 total cases
1,066,856 dead
27,667,554 recovered
In the US:
7,833,763 total cases
217,738 dead
5,025,193 recovered
Yesterday’s numbers for comparison:
Wasn’t sure where to put this, so apologies if it’s not considered breaking news.
Considering how much people are relying on accurate data about the rate of infection and the death rates to make informed policy decisions, the UK has realized that using Excel to aggregate this information left them with problems with the limits of old Excel templates. (the newer xlsx files don’t have the same problem).
I think about this often in regards to Excel. It’s used so often for data mining and data collection when there are better database tools available.
While I agree that Excel is not great at many things and that it is horribly overused, there’s really no excuse for this particular issue. The .xlsx format has been around for 13 years now, so there’s no good reason to hit the 64K row issue. It also generates an error when you try to exceed it. A lot of balls had to be dropped for this to not get caught right away.
More importantly, why in the fuck isn’t Public Health England using an actual database for this?
I’m guessing we’re looking at an office here that never had to deal with more than a few hundred rows of data before, and they just adapted on the fly. I’m not saying it’s right, but you can see how it happens.
But certainly they’re not the first people to screw up with Excel…
Let’s face it, NO ONE likes to install Microsoft updates. They often wreak some kind of secret havoc that you don’t even discover til waaaay later. Best just to ignore those updates, amirite? 
That article gives some great examples. Especially the aftermath. Economic policy was based on the Reinhart-Rogoff Economic study with the wrong data. In the Barclay case, a $1.75B bankruptcy deal was based on incorrect data. The TransAlta’s spreadsheet mistake wiped out 10% of the company’s profit that year.
smh None of that surprises me at all. That’s only the stuff that got caught. I suspect there are way bigger errors out there based on flawed Excel spreadsheets. Yet people rarely crosscheck with a database or another cross check of the numbers. I wouldn’t be surprised if a lot of Covid-19 data that is captured on Excel spreadsheets is inaccurate as well.
…this is probably the best explainer.
Long story short: Public Health England do use an actual database. And existing labs within the public health sector (Pillar 1) feed their results directly into the database. The mostly private sector labs though (Pillar 2) can’t talk to the database so send in their results via CSV files that got fed into an excel template that then got fed into the database. So while Public Health England deserve criticism for not keeping their excel versions (etc) up to date: I think the UK government and the private sector cronies deserve equal amounts of the blame IMHO for a variety of reasons related to the billions of dollars invested in the private sector that could have been put to better use.
37,110,995 total cases
1,072,712 dead
27,897,034 recovered
In the US:
7,894,478 total cases
218,648 dead
5,064,300 recovered
Yesterday’s numbers for comparison:
Today the world had over 350,000 new cases (a new one day record!).
If that continues, we will be adding over 1,000,000 new cases every three days. That works out to about 28,000,000 new cases by the end of the year (for a total of around 65,000,000 cases).
That’s if the current rate holds. Unfortunately the trend has, thus far, been ever upward.
Speaking of upward trends, the US is back up to 60,000 new cases per day.
I don’t know any more than you, but dollars to donuts they exported/saved the results from something (probably Excel) as a spreadsheet in the well known common format, and all they got was the generic warning “features will be lost”
And you send in the old format because that’s still the generic/common format.
Even if they are using an actual database for this, they would still export as CSV or use an export tool to export as Excel 97. Because CSV and 97 are still the common interchange formats.
CSV is perfectly fine to export (and import). Including Excel in the process is asking for problems, but my guess is that they are using it for some sort of T in the ETL process. For an example of the database related issues with Excel that goes far beyond versions, just load a CSV with zip codes in it and say goodbye to New England. On the other hand, database designers (at least those with any experience at all) always make zip fields have character types, so when importing a CSV, it retains the leading zero just fine. Standard rule of database design is that if you’re not going to sum it, average it, or do any other sort of math or aggregations in it, don’t store it as a number, with the possible exception of a unique identifier field. Excel, on the other hand, sees digits, assumes numeric.
At this point, this seems like the wrong place for this, so I’ll stop the hijack.
Heh. As late as a couple years ago I’ve run into web sites for ordering stuff that failed at that. Incredibly annoying to go through the shopping process all the way to entering your address and have it refuse to accept your Massachusetts zip code with a snarky “must enter 5 digits.” Grrr.