100% tariff on the MM/DD/YY date format.

@wendinoakland @renwillis @Natasha_Jay @stux

NASA lost the Mars Climate Orbiter in 1999 due to a fatal mix-up between English (feet, pounds) and metric (meters, Newtons) units, where contractor software used English measurements for thruster force while NASA expected metric, causing the orbiter to enter Mars' atmosphere too low and burn up, a classic example of unit conversion error.

https://science.nasa.gov/mission/mars-climate-orbiter/

$327.6 million

@Natasha_Jay@tech.lgbt I still have my proposal for separator to indicate format as a compromise: DD.MM.YYYY, YYYY-MM-DD and MM/DD/YYYY. This disambiguates the ones with YYYY at the end, as long as we can get americans to stop using . sometimes randomly.

@amiserabilist @renwillis @Natasha_Jay @stux Wasn’t that a critical error with the Hubble Telescope lens? (I’m suggesting you do the research…)

@amiserabilist @renwillis @Natasha_Jay @stux Wasn’t that a critical error with the Hubble Telescope lens? (I’m suggesting you do the research…)

@wendinoakland The Hubble Telescope uses a large, 2.4-meter (8-foot) primary mirror, not a lens

@amiserabilist @renwillis @Natasha_Jay

@stux @amiserabilist @renwillis @Natasha_Jay Maybe it was the mirror…? There was some metric/imperial measurement error on a critical Hubble element

@wendinoakland @amiserabilist @renwillis @Natasha_Jay I'm gonan do some checks Those are veeeery expensive errors woahh

@wendinoakland @amiserabilist @renwillis @Natasha_Jay I'm gonan do some checks Those are veeeery expensive errors woahh

@amiserabilist @renwillis @Natasha_Jay @stux Wasn’t that a critical error with the Hubble Telescope lens? (I’m suggesting you do the research…)

@Natasha_Jay @renchap YYYY/MM/DD is the only acceptable format.

@UkeleleEric But dd/mm/yyyy (regardless of separator) is not meaningfully easily sortable. yyyy-mm-dd is.

@WTL @Natasha_Jay @renchap

@WTL @Natasha_Jay @renchap there are arguments for either that or DD/mm/yyyy. Both are equally logical - just increasing or decreasing.

@WTL @Natasha_Jay @renchap I’ve never understood the U.S. obsession with getting it wrong. It’s not even simple to sort by date.

@mkj @WTL @Natasha_Jay @renchap why not? You can sort it numerically in the same way. 30122019 is obviously after 29112019 - you can, quite easily, separate the date into three separate variables (DD), (MM), (YYYY).

@wendinoakland @glasspusher

>the primary cause of the failure with the Hubble Space Telescope's mirror was spherical aberration caused by a miscalibrated, faulty testing device during manufacturing.

Here are the key details regarding the failure as discussed by users:

The Root Cause (The "Why"): The primary mirror was ground and polished by Perkin-Elmer to the wrong shape because a device called a "null corrector"—used to measure the mirror's surface during polishing—was assembled incorrectly.

The Specific Error: A spacing error of 1.3 millimeters in the null corrector (caused by a missing cap on a rod) led to the mirror being polished too flat at the edges by roughly 2.2 microns (1/50th the width of a human hair).

Ignoring Data: Users noted that a second, independent testing device actually indicated the mirror was flawed, but this data was ignored because the, albeit faulty, "new" null corrector was trusted more.

The Fix: Because the mirror was ground so accurately (just wrongly), scientists knew exactly how to fix it. The 1993 servicing mission (STS-61) installed COSTAR (Corrective Optics Space Telescope Axial Replacement), which acted as "glasses" to correct the aberration.

>One report explained it this way;

If you took a finely polished mirror and enlarged it to the size of the Gulf of Mexico, you'd have a surface with +/- 80 foot swells.

If you enlarged the Hubble mirror to the same scale, you'd have a surface with +/- 3 inch swells.

https://www.reddit.com/r/todayilearned/comments/cjw5az/til_that_after_almost_20_years_of_battling_for_a/

>Imagine you are trying to make a perfectly shaped, giant, curved mirror for a space telescope (like Hubble). Because it’s so large, you need to check if you are grinding the glass into the exact right shape.

The Problem: The mirror is a weird shape (aspheric). If you shine a light on it to check the shape, the light bounces back looking crazy and blurry. You can’t tell if the mirror is wrong, or if it just looks that way because of the weird curve.

The Solution (The Null Corrector): You put a special lens or mirror between your eye (or camera) and the giant mirror. This device takes the messy, distorted light and "fixes" it.

"Null" Means Zero Error: If the giant mirror is ground perfectly, the null corrector makes the reflected light look completely flat (or "null" of errors).

https://en.wikipedia.org/wiki/Null_corrector

>Why an Aspheric Shape is Needed

Spherical mirrors (like a perfect bowl) focus light from distant objects to different points depending on where the light hits the mirror, causing blur.

Aspheric mirrors, with a more complex curve, can focus all light rays to a single, sharp focal point, producing clearer images. Hubble's design used this superior shape.

http://astro.vaporia.com/start/asphericmirror.html

@renwillis @Natasha_Jay @stux

@amiserabilist @wendinoakland @renwillis @Natasha_Jay @stux

+5 for Gulf of Mexico

I dunno what "finely polished mirror" they're talking about. Maybe they're talking a bathroom mirror? That mirror would be unsuitable for astronomical purposes

I've got a 47 cm diameter mirror with an error of about 20 nanometers.
My error translated to 1500 km diameter would be about 6 cm, so yeah. If you can't piss with the tall dogs, stay out of the high grass.

the Hubble mirror had 2 null testers, one all lenses, one lenses and mirrors. The latter one showed an error, the former one did not. They got lazy and said "the second null tester is screwed up". They were wrong.

@mitsunee @Natasha_Jay

And what do you with "DD/MM/YYYY"?

100% tariff on the MM/DD/YY date format.

@zed @bsdphk @Natasha_Jay Wristwatches solve a different use case. When you look at them, chances are you know if it's a quarter to midnight or to noon.

When you represent date/time in a globalised world, it's an entirely different matter. An event at 11:45 is totally ambiguous, so you need 4 digits + the AM/PM bit to be unambiguous. With 24h time, you only need the four digits.