epmos · HackerTrans

I do not know the details of Apple's file system, but I wouldn't be surprised if it needs to allocate space for the log (journal) and can't do so.

That isn't reasonable in the sense of "this is what the filesystem should do in this situation" but if the log and user data are allocated from the same pool it is quite possible to exhaust both.

epmos

3 anni fadiscuss

Rutile is used in refractory bricks and the import was of the ore mineral. So not completely unbelievable.

https://en.wikipedia.org/wiki/Rutile#Application

epmos

3 anni fadiscuss

Due to that convoluted life cycle, we are not sure why the population is declining drastically. The numbers of young eels arriving in Europe in recent years is only 10% what it was in the 80s.

We don't know how long wild eels live (again, the life cycle) but this is much shorter than a single lifetime of captive fish.

epmos

4 anni fadiscuss

Another article on the same topic from a year ago: https://www.smithsonianmag.com/science-nature/how-can-ant-an...

epmos

4 anni fadiscuss

No-one else mentioned it, so I might suggest punched tape. https://en.wikipedia.org/wiki/Punched_tape

There is always a trade-off between mass, durability, and data density. I know you excluded those considerations, but for only 200 years in a time capsule you likely don't want to use large stone blocks or stainless steel.

The tape has several advantages. Mylar is very light and kept in the dark at a constant, reasonable temperature is going to last two centuries no problem. You can read the tape by hand, if you need. An 8-bit tape is an inch wide and can be rolled into a canister for your time capsule.

Of course, for small amounts of data print English words on acid-free paper. For large amounts, tape is going to blow out the mass budget you didn't know you had because of it's low data density. But there might be a space in there somewhere for a tech I loved as a kid.

epmos

4 anni fadiscuss

I don't think that would explain the difference. Using a quick look at https://covid.cdc.gov/covid-data-tracker/#trends_totaldeaths I see covid deaths as:

  2019 : 0
  2020 : 361,236
  2021 : 461,460 ( 1.28x  2020 )
  2022 : 191,352 ( 1.06x* 2020 )

But payouts look like this:

  2020 : 1.09x 2019
  2021 : 2.86x 2019, 2.64x 2020

If there were under-reporting in 2021 similar to 2020 then we should see about a 1.06x payout difference in 2021, not 2.64x. Not recording deaths as covid wouldn't affect life insurance payouts. So there would need to be ~2.64x more covid deaths in 2021 (after most folks got the vaccine in March-May) than in 2020 with no vaccines. This doesn't seem right to me. What would explain this is if while covid didn't get much worse in 2021 people are dying more anyway, or the covid deaths are more concentrated in insured people.

Just getting sick doesn't trigger a life insurance payout, but a glance at cumulative cases vs deaths on the data tracker suggests that in 2021 covid is less deadly than in 2020 on a per-case basis. That matches what I see on the news and people around me pretty well.

I suspect that the life insurance vs covid risks skew differently with age. Life insurance policies tend to be high for working people and lower once retired. Since the purpose of life insurance is to replace lost income for the insured's dependents most people drop or reduce their policies when they retire. So that 2.64x factor on payouts should be under the factor for deaths unless the deaths are concentrated in younger, healthier people. I have no idea what that correction would look like so I ignored it.

* Assuming the second half of 2022 which hasn't happened yet looks like the first half. Reality won't be exactly 2x the total so far.

epmos

4 anni fadiscuss

This is moderately common for environments where you are pushing a lot of startup work into the dynamic linker and will be launching processes frequently. Loading shared libraries for example.

You have a parent process which uses dlopen() to load all the libraries you want to avoid re-linking. When you want to spawn a child, rather than exec() you dlopen() an object with your child's main() and call it. For the case where you have enough libraries this is much faster than an exec(), saving tens of seconds on every application launch if you have a really bad case of C++.

There some small surprises which become obvious with a little thought. You are responsible for everything that normally happens in your process before main() is called. ASLR is only done once per session. People rarely think to fix-up argv[] for ps and friends in the first version.

epmos

4 anni fadiscuss

> But it seems odd that they wouldn't at least save a LUT relating the shortened and expanded variable name.

BASIC implementations based on Microsoft BASIC store tokenized program lines. A keyword takes 1 byte. Comments, variable names, spaces between tokens, etc take up memory on a character-by-character basis.

I didn't use Apple BASIC, but I spent too much time on various flavors of Commodore BASIC which shared this heritage. Changing a variable from "ACCUMULATOR" to "AC" to "A" saved memory and made the program run faster. The problem of not being able to make sense of your own program after was very real.

epmos

4 anni fadiscuss

Yes.

Nearly everything about vehicles has US federal and state regulation. Headlight regulations being stricter in the US are why US versions of European cars up until about the 1980s had different headlights than their home markets.

49 CFR § 393, FMVSS 108 and a crapload of other regulations and regulatory interpretations. In general, these make IRS instructions seem a marvel of clarity.

FMVSS-108 S5.5.11(a) talks about brightness.

epmos

4 anni fadiscuss

> (Empress probably)

Augustus' parents were not emperor/empress--he was the first person to have that title.

epmos

5 anni fadiscuss

> no free as in freedom C++ or Objective-C compiler

G++ was available in 1990. I don't recall when Objective C became available in GCC, but it was present in at least gcc-2.2.2 in 1992.

comments