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.

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

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.

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.

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

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

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.

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.

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.

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.

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

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.

  FS: ^\ GS: ^] RS: ^^ US: ^_

Other posters have noted that they have ASCII separators recently, I have as recently as 2018 internally at a FAANG. So they are not completely unknown in the wild.

You are correct, of course.

I don't think however this is as crazy as it seems at a glance. The author(s) of the law appeared to be trying to prohibit devices that try to call many telephone customers as opposed to contacting a specific number. Calling/texting everyone in town to sell overpriced pre-paid auto repair, illegal. Calling/texting John Doe about his upcoming scheduled visit to his MD, legal. In the later case, the phone number is stored (in a database record of John's contact information) and the system "dials" that number to send him a reminder text message. Paragraph (b) in the syllabus gives two reasons why--dialing emergency numbers and tying up all lines assigned to some entity.

Remember, from Facebook's point-of-view what happened here was something like that second case. They had an existing customer account with a given contact number and used that number to communicate about that account. Deguid did not in fact have a Facebook account. Perhaps as Facebook suggests a previous user of that number did.

This is not an unknown problem. I have a phone number that gets messages about a specific apartment complex frequently. Or did, I muted that sender so if they stopped sending half a year ago I would not know. I suspect some previous user of the number lived those apartments. If Facebook had lost their case those messages would stop. But so would my reminder about office visits to my doctor and I think I would be worse off overall.

The options in gnu tar are -L and -M for specifying tape length (in KB) and creating multi-volume archives.

The -L version is handy as it can prompt you to change the "tape". It has built-in tape handling as well, if you are using a real tape that can be controlled via mt.

Apparently (just checked info) it supports an arbitrary command to execute at the end of each archive. That would be handy for burning CDs, though I don't remember that feature from when I used floppies or QIC-40 to move files from one machine to another.

Around 1991-ish I would use these options to put archives onto stacks of floppy disks. The floppies didn't have a filesystem, just blocks of tar archive. In the days of 9600 bps modems, this was often the fastest way to move stuff to another machine. I had an often-reused stack of 3.5in HD floppies labeled with numbers for this purpose.

About that long, though I had commented before that.

Most of the sites do not have a known location. Or at least, not known to the people producing the map. Click a few markers and you will see lots of "Unknown Location" or "Off $location" which usually means the dumping ship left from that location.

It is true that many munitions were dumped close to shore, but there is a certain amount of making the map as scary as possible going on.

As an aside, GNU libc keeps ( or at least used to keep, I haven't checked in years ) the pointers used by malloc()/free() next to the blocks themselves, which gives really bad behavior when freeing a large number of blocks that have been pushed out to swap--you wind up bringing in pages in order to free them because the memory manager's working set is the size of all allocated memory. Years ago I wrote a replacement that avoided this just to speed up Netscape's horrible performance when it re-sized the bdb1.85 databases it used to track browser history. The browser would just "go away" thrashing the disk for hours and killing it just returned you to a state where it would decide to resize again an hour or so after a restart. Using LD_PRELOAD to use a malloc that kept it's bookkeeping away from the allocated blocks changed hours to seconds.

I have never been quite as excited over a machine as when I got my Amiga 500. I had owned a Commodore Plus/4 and a 128 since middle school and high school, and bought the 500 to start college. It was quite an amazing little machine for it's day.

Then again, I haven't ever been 17 again either, so that might explain it.

I like your implementation. To everyone asking about context, I believe this tool is intended to be used when you have already created such context, where coworkers want to know group salary information while still protecting their personal information from each other.

I had imagined the use would be where people were already in a group and wanted to share information anonymously. I had a brief moment before I fell asleep wondering how to prevent someone from just reloading the page as others submitted their salary information, thus being able to know what salary mapped to what user by the order they show up. I see that you set a minimum limit on the number of users in a pool, have you addressed this once the pool grows beyond that size?