On your first question the major problem was not necessarily speed, which did improve a bit for example the French Renault FT could make a bit over 4mph, but instead endurance. They were never able to solve the range and reliability problem given the technology at the time. This meant that even light tanks like the FT only had an operational range of about 40 miles. Breakdown rates were also incredibly high, and during the final offensives of the war in 1918 the British rapidly ran into a situation where they had to deemphasize the use of tanks simply because they did not have enough that were running.
Suggested readings: (this is a mix of books and articles)
* The Infantry Cannot Do with a Gun Less: The Place of the Artillery in the BEF, 1914-1918 by William Sanders Marble
* The Dynamics of Doctrine: The Changes in German Tactical Doctrine During the First World War by Timothy T. Lupfer
* The Marne and After: A Reappraisal of French Strategy in the First World War by Douglas Porch
* The Evolution of British Strategy and Tactics on the Western Front in 1918: GHQ, Manpower, and Technology by Tim Travers
* Operational Art and the German Command System in World War I by Bradley John Meyer
* Not Glamorous, But Effective: The Canadian Corps and the Set-Piece Attack, 1917-1918 by Ian M. Brown
I'm just gonna quote what I think is a very important part from the Graham piece itself
>"People who don't look any deeper than the Gini coefficient look back on the world of 1982 as the good old days, because those who got rich then didn't get as rich. But if you dig into how they got rich, the old days don't look so good. In 1982, 84% of the richest 100 people got rich by inheritance, extracting natural resources, or doing real estate deals. Is that really better than a world in which the richest people get rich by starting tech companies?"
And my answer to that question would be, arguably yes. Why? Because at least old money understands the concept of noblesse oblige. The real sinister psychological thing going on behind the Graham argument is that it's not at all about meritocracy, it's that this mentality of earned wealth completely rids the owner of any sort of responsibility.
The aristocrats and oligarchs of the olden days might have been corrupt, debauched and half-useless, but at least they knew it. This new, self-made entrepreneur class does not only think they have earned their money themselves, which as a sidenote is also kind of a fiction, but that they're intellectually superior, morally superior and virtuous in ways that anyone else just can't understand.
Old money might have ignored you and thrown a party, but Silicon Valley money wants to remake people in their images, they have a Protestant zealotry associated with their money that makes any oligarch look straight up sympathetic in comparison.
C.S Lewis:
"“Of all tyrannies, a tyranny sincerely exercised for the good of its victims may be the most oppressive. It would be better to live under robber barons than under omnipotent moral busybodies. The robber baron's cruelty may sometimes sleep, his cupidity may at some point be satiated; but those who torment us for our own good will torment us without end for they do so with the approval of their own conscience. They may be more likely to go to Heaven yet at the same time likelier to make a Hell of earth. This very kindness stings with intolerable insult. To be "cured" against one's will and cured of states which we may not regard as disease is to be put on a level of those who have not yet reached the age of reason or those who never will; to be classed with infants, imbeciles, and domestic animals.”
For anyone interested there're a few such tools that are modern alternatives to traditional Unix utilities. Interestingly all written in Rust.
grep -> ripgrep (command `rg`)
find -> fd
sed -> sd
diff -> delta
ls -> exa && lsd
cat -> bat
wc -> cw
du -> dust
ps -> procs
top -> ytop (**edit** archived; author recommends `bottom`)
They all offer their own modern take but I'll say the first two, which are the only ones I'm using, are the most interesting. Both ripgrep and fd offer easier usage. The ripgrep is also an ack replacement and fd can also replace parallel (even has compatible syntax). The amazing part is that both are an order of magnitude faster their counterparts. The ripgrep dev has extended analysis for how this is achieved on their blog[0].
An infinite number of mathematicians walk into a bar.
The first mathematician orders a beer.
The second orders half a beer.
"I don't serve half-beers" the bartender replies.
"Excuse me?" Asks mathematician #2.
"What kind of bar serves half-beers?" The bartender remarks. "That's ridiculous."
"Oh c'mon" says mathematician #1 "do you know how hard it is to collect an infinite number of us? Just play along".
"There are very strict laws on how I can serve drinks. I couldn't serve you half a beer even if I wanted to."
"But that's not a problem" mathematician #3 chimes in "at the end of the joke you serve us a whole number of beers. You see, when you take the sum of a continuously halving function-"
"I know how limits work" interjects the bartender.
"Oh, alright then. I didn't want to assume a bartender would be familiar with such advanced mathematics".
"Are you kidding me?" The bartender replies, "you learn limits in like, 9th grade! What kind of mathematician thinks limits are advanced mathematics?"
"HE'S ON TO US" mathematician #1 screeches
Simultaneously, every mathematician opens their mouth and out pours a cloud of multicolored mosquitoes. Each mathematician is bellowing insects of a different shade.
The mosquitoes form into a singular, polychromatic swarm. "FOOLS" it booms in unison, "I WILL INFECT EVERY BEING ON THIS PATHETIC PLANET WITH MALARIA"
The bartender stands fearless against the technicolor hoard. "But wait" he inturrupts, thinking fast, "if you do that, politicians will use the catastrophe as an excuse to implement free healthcare. Think of how much that will hurt the taxpayers!"
The mosquitoes fall silent for a brief moment. "My God, you're right. We didn't think about the economy! Very well, we will not attack this dimension. FOR THE TAXPAYERS!" and with that, they vanish.
A nearby barfly stumbles over to the bartender. "How did you know that that would work?"
"It's simple really" the bartender says. "I saw that the vectors formed a gradient, and therefore must be conservative."
If you're interested in a relatively inexpensive oscilloscope to play around with, check out the Rigol 1054Z (or 1074Z Plus if you want MSO capabilities). It's four channels and has more than enough features for playing around with, especially for the price. Using the Riglol website, you can unlock all software options on it, including increasing the bandwidth to 100mHz and the memory to 24MP.
It was the first 'real' scope I bought and I still use it a fair amount despite having upgraded to a Rohde and Schwarz MSO model. I've been amazed how much I use an oscilloscope after getting one, from measuring ripple on power supplies to diagnosing serial communication issues.
I found out about it a few weeks ago and am ridiculously happy I did. It immediately jumped in my top 3 favourite channels. Lots of deep dives into how last-century technologies worked, that neither talk down nor overwhelm. And the guy's really funny as well. Look at his videos and choose one you like, or start with one of his playlists such as the story of Laserdisc: https://www.youtube.com/watch?v=Eg8tK1LpLS8&list=PLv0jwu7G_D...
This guy was a badass. He Mcgyvered a 1 kilometer data link using LIGHT because he was annoyed his employer, PARC, opened up a new office up the road and he couldn't test his printer.
Excerpt from Dealers of Lightning:
------
Starkweather and Rider worked together on coordinating the SLOT and character generator until early 1972, when they were stymied not by a technical obstacle but one entirely man-made. This was the relocation of more than twenty of PARC’s seventy scientists up the hill to a building newly rented from the Singer Company and known as Building 34 (because its address was 3406 Hillview). The Computer Science Lab, including Rider, got bundled off to the new quarters while everyone else, including Starkweather, temporarily stayed behind on Porter. The move separated the two by a kilometer of real estate—too far to string an overhead line and, with the four-lane Foothill Highway in the way, impossible to link via a ground cable.
“The administrators said, ‘Don’t worry. You’ll be back together in another year,’” Starkweather recalled. “I said, ‘Great, what are we supposed to do in the meantime?’”
But one Sunday afternoon shortly after the move Starkweather got a brainstorm while sitting at home. He immediately jumped in his car, drove to Porter Drive, and mounted a stairwell to the roof.
Just as he had thought, he could take line-of-sight aim from where he stood to the rooftop of Building 34. He might not be able to span the distance by cable or wire—but he could do it by laser beam.
The next day he ordered four telescopes from Edmund’s for about $300 apiece. He and Rider replaced the eyepieces of two with low-power lasers and the others with sensitive photodetectors. They bolted one laser scope and one detector on each roof, aiming each at its complement across the way, to create a visible light data link.
The circuit worked flawlessly in almost any weather, even fog, although minor adjustments were often necessary after a rainstorm, when the weight of accumulated water made the roofs sag slightly.
“When SLOT was running I’d send a pulse of light up the hill to signal the character generator to send a line of data down to the detector on my roof, which would send it down to this laser and then to the printer,”
Starkweather recalled. “After all, we were only encoding ones and zeros. It was like sending binary data on a long wire made out of light, instead of copper.”
William Langewiesche is a pilot himself, and has written many other articles over the last 2 decades detailing accident investigations (among other things). As much as the technical details, he often explains the organizational and political circumstances that are just as interesting. Some that stick with me: the late-90s ValuJet and EgyptAir crashes, and the Space Shuttle Columbia.
What do everyone use for feed reader !? My old phone used to have a built in feed reader. And my old browser also used to have a feed reader. Looked on Google play and there where very little to choose from. Thinking of creating my own feed reader ... Or have "blogging" moved over to "Youtubing" and the occasional podcast !?
Where do you find them? I keep trying to get into podcasts (tech, computer science, military, history...) but I really struggle. A sea of ancient abandoned shows when I search on my iPhone, seems nothing still running? Where should I be looking?
Slate columnist Sara Dickerman wrote a critique of pepper as a table spice a few years ago, while acknowledging its historical medicinal qualities, as the OP does:
> Why should this brawny spice be kept on the countertop at all? Why not stash it in the rack with the fennel seed, the mustard seed, and the cinnamon—all the wonderful spices that add life to our food but are by no means all-purpose? I think we’d appreciate pepper's qualities all the more if we used it just for specific dishes, not universally.
How does one go about building a custom mechanical keyboard? Are there any good resources to get started? I've long fantasized about building a standard-layout keyboard with a couple of extra unobtrusive shift/modifier keys.
Thank you for bringing up Copernicus EMS and HDDS, I've added it to my observation notes. That data would be definitely be useful, but no one in the effort knew about these resources at the time.
Water rescuers were given a general idea of an area that could be assumed flooded throughout, but boaters had to exercise caution at all times because of water hazards (both stationary and mobile) and flood waters could subside without warning, leaving some boats stranded on shallow mud flats.
While it was somewhat life-threatening with water rescues, the lack of data could be worked around because the water rise to the point it was life-threatening was reckoned in 15-30 minute intervals, and people could coordinate rescues of rescuers around that timeframe. However, in a fire response version of such Net-based coordination, accurate, up-to-the-second assessment would be life-critical. Not sure how to address that without industrial-grade drones or even UAV's, and hella-expensive grade FLIR, and the FAA (and possibly DHS) would then be deeply involved.
Not sure why UAV-based realtime GIS mapping wasn't deployed. The hardware tech exists, but perhaps the pipeline for data ingestion, cleanup, processing and publishing isn't up to realtime demands, even for a limited scope of "show me what is different from this satellite GIS map of the same area". That would be an interesting problem to solve, correcting for daylight shadows, transient objects like cars and planes, etc.; I'm sure the military intelligence folks in the IMINT sector have long since worked out the initial challenges of this problem space, but open algorithms for applying to this problem space (not to speak of code) don't seem readily available to me after searching around, so it might be a long slog up the learning curve.
These disasters might be a good application of temporary air mobile vehicles supplying some form of wireless signal. If you're going to loft up a vehicle into the air to collect realtime GIS, you might as well blanket the area with an emergency wireless signal. Would have to solve problems around rate-limiting, dynamic prioritization based upon confirmed emergencies, etc.
The flooding points out to me the lack of research into flood waters diversion [1]; there is still a lot we don't know about what happens when we try to mitigate excess flood waters. We know at least some flooding is necessary for managing topsoil erosion. But even the $26B estimate to bring Houston up to a more hardened flooding posture [2] seems like cheap insurance compared to the $100B estimates being thrown around now. I've wondered if it is possible, and what it would take to build a national water grid that intakes excess flood waters in one region of the nation, processes it so it is ecologically "clean" of invasive species, and shunts it to a drier region. The capacity of such an infrastructure within the channels themselves might lessen the need for designated floodwater storage plains, and simultaneously vastly increase land utilization. However, megascale infrastructure projects still seem out of our grasp in this era.
My SO helped with the crowdsourced civilian effort to dispatch water rescues of stranded people. The effort is wound down now; it will likely be hailed as a coup of "social media", but that's overhyping the social media aspect. Social media got the word out of where to go online to enter the coordination, but had little to do with the actual coordination itself. The group my SO worked with coordinated mostly through the Zello mobile app, Glympse mobile app, houstonharveyrescue.com (going dark soon, due to PII concerns), and a Google Sheets to track water rescue requests.
TFA is mainly discussing use of drones in the recovery phase, and only tangentially touched upon drones for rescues. Drones were not used much for long-range water rescues, because they were a hazard for the many volunteer helicopters that responded.
This pointed out the need for a solution (preferably as automated as possible) that allocates helicopter and drone flight paths in a disaster area.
The whole experience was very eye-opening for me. There isn't a good solution for coordinating disaster response by civilians, but even just the ad hoc quick-and-dirty collection of apps used by the various civilian groups that responded showed how much leverage Internet-enabled coordination delivered. The latency of civilian response is much lower than government response, but once the government landed resources, the government response had much greater volume. Mix both groups at the right times, and you'd have an admirable disaster response, pretty much what happened in Houston.
Observations from listening in on my SO during meal times (the only times I could break from work):
* Misinformation is rife. This is a difficult problem to address. Example: rumor starts that a rescuer was slashed with a machete. Story morphs into shot and slashed, then slashed-got-sepsis. Turns out a guy stepped on glass and got a nasty gash.
* No good solution to map rapidly-changing road conditions. Piles of rescuers with valuable boats in the first critical hours of response were diverted to drive around to find a way into the right areas of Houston to deploy. Need a way to effectively intake reports from people with just trucks (lots of citizens responding with no boats wanted to help in some way), snapping pictures at a specific location, giving location and time, and reporting road closures due to specified height of water, electrical line, etc. Bonus for AR-enabled measurement of water depth, based upon baseline measurement of vehicle. Extra bonus for measuring water speed by tossing a recognized object into the water and tracking it. Then people who pull up the heat map of closures will flood-fill (pardon the pun) out possible routes, avoiding lots of redundant checks of possible routes. A lot of valuable time was wasted on this, the first few hours were filled with civilians an hour from arriving at the area (as instructed over social media) calling in and asking how they can reach where they can drop their boats, because the main routes were all closed.
* No good solution to map flooded areas, how deep, and forecasted levels. People pieced it together by hand and passing along the grapevine. Depth matters: below a certain level, outboards were getting stuck. Below a different level, and all boats had to watch for fences they could get snagged on (had a few that capsized on such obstacles). Ideal: remote-reporting gauges scattered in a grid pattern throughout the area, or gauges that can be dropped down during the initial rescue efforts, and reclaimed later.
* We reached out to Uber and Lyft. IMHO, this was a PR coup sitting around for the taking. You have a system that optimizes for efficiently tracking and queuing requests, matching requests to vehicle capacity, directing the closest vehicle to the request, and showing requesters the live status. This was precisely what the water rescue coordination needed. Uber gave a canned "we're standing down for the safety of our drivers, for those who are outside of the areas of Houston that kicked us out that we can still operate in". Lyft said great idea, but the conversation black holed after that.
* Any app-based solution will have to be very sensitive to energy usage. Rescue requesters ran out of power on their phones distressingly often. Zello was established early on as a bad way to communicate with requesters; it drained batteries very quickly. Instead, requesters reached out to relatives/friends they knew who were safe, instructed them how to get Zello and get on the rescue channels, then put in a request, and then those relatives/friends would periodically query for a status update on the request. Use strongest WiFi if available, fall back to cell data (lowest-tech with strongest signal available), then SMS, then voice.
* These status update requests (see previous entry) took up a lot of bandwidth at the height of rescues, and added to the stress on the rescuers. A queuing system that operated over WiFi if available, then over cell data if not, telling requesters they are number N in line for the nearest rescuer, would have made the coordinating a lot easier.
* A unique ID was eventually established for assigning each request. A voice recognition system could easily listen in on a group and automatically assemble in timeline form all conversations that mention a particular ID, so anyone looking at a particular rescue request could see all historical discussion about that request.
* The Zello conversations quickly got unwieldy when there were too many people vying for "the microphone". Fortunately, people figured out how to manage this somewhat, splitting into Port Arthur and Houston-specific channels, for example. An app to auto-split by role (dispatcher, rescuer, requests, etc.) and density-based geography (bounded by neighborhood boundaries, perhaps) would have helped some of the confusion.
* A voice recognition system to simply assign people to the right channel based upon their initial request would be helpful. There was an opportunity here for someone like Twilio to set up a single phone number that did this. In the first few hours, people did this through their personal lines: "Call me at xxx-xxx-xxxx when you are an hour out on I-10 from Conroe to get the current rally point." Then you hear later: "I'm an hour out, called xxx-xxx-xxxx number as instructed, and it's been busy for the last 15 minutes, what else can I do to find the current rally point?"
* Most useful feature of Zello: historical recording of every single transmission while you were listening. This let people go through them and follow up on water rescue requests, then mark them safe if they were rescued. This was a big problem at first: rescuers were pulling up a map on the web app, rushing to a request, then getting disappointed when they find out the rescue request was long since taken care of by another rescuer. Zello could improve on this: the historical recording was only for the duration you were listening; a feature (even paid) that pulled last N minutes/hours from their servers would be even better. Even better is a solution that tracks a rescuer to a rescue request, then presents a simple confirmation screen (# of adults, elderly, children, disabled, babies, pets rescued, any variation from pre-arranged drop-off point, any voice notes required), and auto-marks a request.
* Need a solution that maps water conditions at a specific location, ideally with tagged input of submitter, time, audio/picture/video, and NMEA data. At one point, the flat-bottomed boats were having a lot of trouble navigating choppy waters as Harvey came back in and churned up the flooded areas. Fold in with weather data, and predictively age out the conditions if possible, displaying that the computer model thinks conditions might be so-and-so but be careful because it could still be the reported condition, until another submission confirms calmer conditions.
* People REALLY love to help. But if their efforts go unappreciated, or go to waste (about the same as unappreciated), they will get dejected very quickly. This is why precise, comprehensive coordination is so critical to manage.
* There was initial concern about fake rescuers. This concern should not be dismissed, but as far as we could tell, this didn't happen.
* Need a solution that maps shelter facilities / government resources as they come online, capacity, and current utilization, so rescuers can efficiently forward rescuees to the best available facilities, most of whom are somewhat in a state of shock. Many shelter facilities early on were just school gyms, churches/temples/mosques, warehouses, and retail stores.
* A large number of private helicopters volunteered early. Knowing the most urgently medical-critical requests to prioritize was all manually performed.
* I suspect that the ad hoc, thrown-together approach of apps to coordinate the rescues is close to their scalability limit. About 10K rescues were logged, in round numbers. I don't think the same approach will work beyond 3-50K rescues, because bottlenecks were becoming apparent to me even with what we had.
* Best part of Internet-enabled rescue coordination: anyone now has a choice to actively participate in the rescue no matter where in the world they are. That's incredibly powerful and a game-changer.
That's all I have off the top of my head.
All in all, I'm quite impressed how well this went, despite the difficulties and setbacks I saw, and it shows some of the best humanity has to offer. There are some really interesting, deep CS and software engineering problems to solve in disaster response management and coordination.
Special shame to Joel Osteen: after his weak response compared to the area churches and mosques with far less funds who threw their doors open within the first hours after Harvey hit, he should be ostracized, as if "prosperity ministry" wasn't bad enough on its own. Didn't know who this bloke was before Harvey, other than "some guy who runs a megachurch", but after reading the news stories, I can't believe he convinces so many parishioners into following him.
Suggested readings: (this is a mix of books and articles)
* The Infantry Cannot Do with a Gun Less: The Place of the Artillery in the BEF, 1914-1918 by William Sanders Marble
* The Dynamics of Doctrine: The Changes in German Tactical Doctrine During the First World War by Timothy T. Lupfer
* The Marne and After: A Reappraisal of French Strategy in the First World War by Douglas Porch
* The Evolution of British Strategy and Tactics on the Western Front in 1918: GHQ, Manpower, and Technology by Tim Travers
* Operational Art and the German Command System in World War I by Bradley John Meyer
* Not Glamorous, But Effective: The Canadian Corps and the Set-Piece Attack, 1917-1918 by Ian M. Brown