Gigantism Is a Never-Ending Temptation for Engineers and Designers (ieee.org) 143
Vaclav Smil, writing at IEEE Spectrum: There is a fundamental difference between what can be designed and built and what makes sense. History provides a lesson in the shape of record-setting behemoths that have never since been equaled. The Egyptian pyramids started small, and in just a few generations, some 4,500 years ago, there came Khufu's enormous pyramid, which nobody has ever tried to surpass. Shipbuilders in ancient Greece kept on expanding the size of their oared vessels until they built, during the third century BCE, a tessarakonteres, with 4,000 oarsmen. That vessel was too heavy, too ponderous, and therefore a naval failure. And architect Filippo Brunelleschi's vast cupola for Florence's Cathedral of Santa Maria del Fiore, built without scaffolding and finished in 1436, was never replicated.
The modern era has no shortage of such obvious overshoots. The boom in oil consumption following the Second World War led to ever-larger oil tankers, with sizes rising from 50,000 to 100,000 and 250,000 deadweight tonnes (dwt). Seven tankers exceeded 500,000 dwt, but their lives were short, and nobody has built a million-dwt tanker. Technically, it would have been possible, but such a ship would not fit through the Suez or Panama canals, and its draft would limit its operation to just a few ports. The economy-class-only configuration of the Airbus A380 airliner was certified to carry up to 853 passengers, but it has not been a success. In 2021, just 16 years after it entered service, the last plane was delivered, a very truncated lifespan. Compare it with the hardly puny Boeing 747, which will see its final delivery in 2022, 53 years after the plane's first flight, an almost human longevity. Clearly, the 747 was the right-sized record-breaker. [...] Depending on where you stand you might see all of this either as an admirable quest for new horizons (a quintessential human striving) or irrational and wasteful overreach (a quintessential human hubris).
The modern era has no shortage of such obvious overshoots. The boom in oil consumption following the Second World War led to ever-larger oil tankers, with sizes rising from 50,000 to 100,000 and 250,000 deadweight tonnes (dwt). Seven tankers exceeded 500,000 dwt, but their lives were short, and nobody has built a million-dwt tanker. Technically, it would have been possible, but such a ship would not fit through the Suez or Panama canals, and its draft would limit its operation to just a few ports. The economy-class-only configuration of the Airbus A380 airliner was certified to carry up to 853 passengers, but it has not been a success. In 2021, just 16 years after it entered service, the last plane was delivered, a very truncated lifespan. Compare it with the hardly puny Boeing 747, which will see its final delivery in 2022, 53 years after the plane's first flight, an almost human longevity. Clearly, the 747 was the right-sized record-breaker. [...] Depending on where you stand you might see all of this either as an admirable quest for new horizons (a quintessential human striving) or irrational and wasteful overreach (a quintessential human hubris).
747 was also dual purpose (Score:3, Interesting)
The A380 couldn't carry cargo like the Boeing can
An A380 Cargo Variant was planned ... (Score:2)
... right from the beginning of the design phase. They just didn't get around to building it, because Airbus was busy delivering all ordered passenger versions and will now phase out A380 production.
Re: (Score:2)
I admit they came up with something better for cargo [wikimedia.org], but they won't build very many of those either, and it can't carry an external load [nasa.gov]. Boeing's record will remain unbroken
Re: (Score:3)
... right from the beginning of the design phase. They just didn't get around to building it, because Airbus was busy delivering all ordered passenger versions and will now phase out A380 production.
They only had something like 30 orders, al of which got cancelled when the freighter version got delayed as the passenger one had priority. It would have been tough to compete with 747, since it lacked a nose door for loading and the double deck would have complicated loading from a mechanical as well as load limit basis. My guess the orders were more to see if it was viable than a belief they could replace 747s.
Re:747 was also dual purpose (Score:4, Informative)
The A380 was designed for the hub and spoke model. Fly passengers to your hub, put them all on an A380 and take them to another hub, and then maybe on to a spoke again from there.
That model proved less popular than expected. On top of that, the A380 is so big that few airports can accommodate it. It needs special passenger bridges to board and extra space to park and taxi in.
When they started work on it they thought that would be the future. Reduce costs by increasing capacity on long haul. Didn't work out that way, so no matter how good the aircraft is, the market just isn't there for it.
Re: 747 was also dual purpose (Score:3)
Re: (Score:3)
This was partly due to 9/11 and the resulting TSA security theater making every stop take an extra 2-4 hours to get a reliable connection. Hub and spok lost The problem was exacerbated by different hubs belonging to different airlines, who did not cooperate well in shutting passengers from one airline's services to another's.
Re:747 was also dual purpose (Score:5, Informative)
Software too (Score:4, Insightful)
So much of today's software tries to do everything. The best software focuses on one core thing and does it really, really well. But too often, engineers want to load it down with all kinds of features that most users will never use. Eventually, these tools start with wither and die because they are too complex and expensive to maintain.
Re:Software too (Score:4, Insightful)
This is, for instance, why Microsoft Teams is crap.
It tries to do everything. But I don't want it to do everything. I want it to do multiplayer notepad. It is worse than IRC at that, and that's, what, 40 years old?
Re: (Score:3)
Re:Software too (Score:5, Informative)
That's viciously unfair. I can attest that MS Teams can silently update itself in the background, restart and forget audio settings, and crash Outlook when it does.
That's definitely something.
Re: (Score:3)
Hey, at least it can blow up Outlook on the way out! That's something useful, right?
Re: Software too (Score:2)
Yeah usually applications are as slow as teams only after they have feature bloat. Trans somehow manages to be slow when its feature set is total shit. Of all of the desktop sharing applications in widespread use, teams is easily the least featured. No ability to do markup whatsoever, no option for pixel perfect viewing, no option to share a selected portion of your screen... It's just total crap, and if you're in a Microsoft shop, you're fucking stuck with it.
Re: (Score:2)
Microsoft is why it's crap as are nearly all things Microsoft.
Re: Software too (Score:2)
Never have I seen a modern application with big financial backing be as slow and have as poor of a navigation interface as Microsoft teams.
Re:Software too (Score:5, Insightful)
This is what we're seeing with websites.
When sites were static HTML, they were fast and efficient
When they became loaded with javascript, they took time to load
When ads came along they became bloated and take forever to load
When "frameworks" came along, sites became huge monsters of gum-and-bailing-wire, and not you break any one of the 6000 libraries it depends on, or a CVE is found in one of them, and your entire website for your business comes crashing down.
Re:Software too (Score:4, Insightful)
What are you talking about?
Email websites SUCKED before MS made outlook work well.
Map sites sucked until Google showed us what ajax could really do.
In what would are full page reloads for every click faster?
Re: (Score:2)
Email websites SUCKED before MS made outlook work well.
Outlook works well? News to me. I avoid that atrocity whenever I can and I am always surprised how bad the functionality and convenience and safety/security of MS outlook is (web or native). I guess many people just have really low standards.
Re: Software too (Score:2)
Actually one of the things I loved about Gmail when I was in on the early beta was that its interface was so simple and responsive that I didn't feel any kind of urge to use Outlook. I could open emails faster, I could draft them faster, and best of all, I could search through hundreds of them basically instantly (outlook, to this day, still has a horribly slow search function that doesn't even work all that well once it finally does finish searching.)
Gmail was also ajax from the very beginning. Other web e
Re: (Score:2)
Email websites sucked until Squirrelmail let people self-host something better, then gmail and hotmail improved so they were better than squirrelmail. Yahoo, of course, is still not as good as the squirrel, and so isn't very popular.
Outlook is something that was popular for work email, because it did a really good job of letting your boss manage/read your mail.
Re: (Score:3)
AJAX was developed in the late 90s. Outlook Web Access was released in 2000, Gmail in 2004, Google Maps in 2005.
Since then, we've had 15+ years of web sites getting more and more bloated - often faster than hardware and network improvements could keep up - and rarely if ever getting more functional in the process.
Re: (Score:2)
and your entire website for your business comes crashing down.
Really? Can you point to this? I can't recall any major website crashing down simply because a CVE was published in a library. What I do recall is a CVE causes a bit of effort for system administrators to update said library when it is discovered, naturally this is all completely transparent to the user.
But you also forgot something:
When sites were static HTML, they were small niche and weren't able to serve the economy like the internet of today. You may as well read a book.
When sites became loaded with ja
Re: (Score:3)
When sites were static HTML, they were fast and efficient
I built websites in those very early days. The reason you remember them as being fast and efficient was that people like me spent days working on them to get them load in under 10 seconds in dial-up connections or whatever the particular requirements were.
JS didn't change that much. Because we hand-crafted the early JS as much as we did the HTML.
What changed things was broadband. Suddenly you didn't need to optimize your images anymore, and compared to them, the actual text/code was small anyways, so why bo
Re: (Score:2)
yeah, I'm sleepy. That should be dot-com not /. era and of course there was CGI before PHP, but the people who wrote CGI in those early days still understood that their output was going across the wire. The issue with PHP was that you SEEMINGLY had an HTML page in your hand, but with just a bit of loops and expansion, your 5 KB page turns into 100 KB being sent to the visitor. But your mental model is that there's a 5 KB file being served.
Re: (Score:2)
I don't agree that websites / javascript are an example of bloatware.
An operating system is intended to be able to run literally ANY software that can be built. Today, the web browser has become an operating system all its own. It's literally built to run anything, up to and including other operating systems like Linux. https://geekflare.com/run-linu... [geekflare.com] The web browser is the one operating system you can count on being available to all users, whether they are on Linux, iOS, Windows, or a Chromebook. To me a
Re: (Score:2)
An operating system is intended to be able to run literally ANY software that can be built.
No it isn't, there's all kinds of limitations from architecture to API that limit what software an operating system can run.
Today, the web browser has become an operating system all its own. It's literally built to run anything,
The modern web browser can't even open an ftp connection and if isn't running exactly the right browser, web pages are broken all over the internet. It is frustrating how many web pages no longer work due to people making assumptions that there is only one operating system with one browser, something that started in the '90's and as your comment shows, is still ongoing.
Re: (Score:2)
The current web frameworks are the road to hell. Complexity is way outside of what can be handled. External dependencies are through the roof. Maintainability is crappy short-term and long therm probably just not there anymore. Security is a joke and only somewhat there because attackers have trouble finding the vulnerabilities as well in all that mess.
This whole approach has to be scrapped and replaced by something that actually respects KISS, like all good engineering does. Will take a while to happen bec
Re: (Score:2)
That depends entirely on the type of software, and what it's doing. For someone working singly, with a very simple workflow, the UNIX religion works fine, but just like other "ideal solutions", it breaks down in the face of reality. Using hyperbole such as "most users will never use" falls flat, because you're only extrapolating from your own perspective.
From MY personal perspective, some examples I can give where the more complex integrated suites have made life simpler as a USER are for example Content Cr
Re: Software too (Score:2)
If the unix model breaks down in the face of reality you'll have to explain to us why it's still being used for most web servers, corporation back end systems and embedded systems (in the form of linux) almost 50 years after AT&T created it. Oh , and let's not forget MacOS - a version of BSD - which rules in the design arena.
Re: (Score:2, Insightful)
Either you missed the context, or you deliberately ignored it(which in your case is a 50/50 probability), but the point was about the "Do One Thing And Do It Well", which is part of the UNIX religion, and it works well for tools like ls, cat, ps etc. But people quickly found that they needed the extra power of perl etc over tools like grep, because it was far more convenient. If you look at webservers, they use fairly complex daemons now, and call upon databases, rather than flat files directly served by th
I hear you, bro (Score:2)
Pay no heed to the ankle biters.
Re: (Score:2)
"Do one thing and do it well" was for CLI apps, especially systems programming.
It was never a general-purpose rule for everything *nix. Take a look at the 90s *nix graphics workstations, for example.
Your screed is ignorant.
Re: (Score:2)
Historical examples of such software might include the operating system Multics, which was, from what I've read, very ambitious. Unix was named as a play on 'Unics' the opposite of Multics by Ken Thompson who had worked on the Multics project. Multics was successful though in sparking and promoting new concepts in operating system design. If you read the wikipedia article on Multics, it points out that by todays standards Multics is pretty small. But a lot of growth in the size of operating systems sinc
Re: Software too (Score:2)
Dont forget Plan 9, another overcomplicated beast designed by people who'd spent too much time in ivory towers reading sci fi instead of looking at what real world requirements were.
Re: Software too (Score:2)
Re: Software too (Score:2)
See ITunes for more reference
Re: (Score:2)
So much of today's software tries to do everything. The best software focuses on one core thing and does it really, really well. But too often, engineers want to load it down with all kinds of features that most users will never use. Eventually, these tools start with wither and die because they are too complex and expensive to maintain.
I guess this explains 'systemd'.
Re: (Score:2)
File Systems are another example (Score:2)
Today you can buy a 20TB hard drive and if you put a bunch of them in a RAID you can g
Re: (Score:2)
But too often, engineers want to load it down with all kinds of features that most users will never use.
I don't think it's the engineers driving featuritis, its the users (who always want "just one more thing" to perfect their particular workflow/use-case) combined with the marketing and sales department (who needs to be able to list reasons why people should pay for the new version every year, otherwise revenue dries up and everybody gets laid off; "now more streamlined and with fewer bugs" isn't considered sufficient reason).
Hindsight is 20/20 (Score:5, Insightful)
The premise is dumb. You don't know if it's too big until you actually do it.
In fact, every project has its share of naysayers, and if you listen to them you'll never do anything.
Re: (Score:2)
Re: (Score:2)
The premise is dumb. You don't know if it's too big until you actually do it.
That's what they said about your mom.
Re: Hindsight is 20/20 (Score:2)
Fred Brooks, author of The Mythical Man Month said to beware of âoesecond systemâ syndrome. Once a leader succeeds, he gets a big ego, and then needs to complete his next masterpiece with every darn feature he had to cut from the original while no longer accepting any moderating feedback from the mere mortals who have been working with him all along.
Example: Star Wars episodes 1, 2, 6.
Brooksâ(TM) example were cpu architectures and Early iterations of IBM operating systems.
A more interesting
Re: (Score:2)
Indeed. Brooks had most things about software projects (with some things he found applying generally to projects) figured out accurately way back. The MMM should be required reading for anybody leading a software project or deciding about budget for one. Instead it is a somewhat obscure book only known to those that really did dig into the matter on their own initiative. Management education (not only) in this space is a complete failure that ignores things that are critical and _known_.
Well, at least there
Re: (Score:3)
The premise is dumb. You don't know if it's too big until you actually do it.
Furthermore, none of the examples listed in TFA were actual failures.
The pyramids are not only still standing, they are the only thing left from that era. Everything else is lost in the sands of time.
The Cathedral in Florence is still standing as well.
Even the Tessarakonteres [wikipedia.org] was not a failure. It was built as a ceremonial prestige vessel and was never intended to be a warship. So saying it was a naval failure is silly because that was never its purpose.
Re: (Score:2)
The pyramids are not only still standing, they are the only thing left from that era. Everything else is lost in the sands of time.
Stonehenge
Hypogeum of Ħal Saflieni
Ebla tablets
Skara Brae
Peruvian pyramids
Uffington White Horse... wait, no, that's older
Newgrange... wait, no, that's even older
I mean, I can go on with an almost endless list of things that are as old or older than the great pyramid.
And, hell, the great pyramid was originally white and smooth sided, not the bare stone jagged mess that it is today. The entire outer limestone surface is gone.
In fact, the oldest artifacts we find tend to be tools and small figurines
Re: (Score:2)
We've found stone tools that are a several million years old.
I was with you up until this. Humans haven't existed that long.
Re: (Score:2)
The pyramids weren't made to last for lasting's sake. They were built to preserve and protect the treasure and mummy of the ruler. By that yardstick they proved to be abject failures very quickly, so quickly that rulers stopped building them.
In contrast the relatively simpler rock cut tombs in the Valley of the Kings lasted far longer, and there are several tombs which must have been made which have not yet been found yet.
Build Better, not Bigger (Score:2)
Sometimes bigger is better and so there are the examples given in the summary. Other times smaller is better - look at how computers shrank in size over the last 50+ years. Sometimes size is largely irrelevant - once mobile phones shrank to hand-sized they have remained at roughly th
Re: (Score:2)
Computers shrunk, but the biggest got bigger.
The current top one draws a whopping 30MW.
Roadrunner drew 4 (2008, picked because it's old and I knew people working on it)
ASCI red (1997) drew 0.87.
The cray 2 (1985) drew a puny 0.2
But yes I agree that most of the building bigger is because we want a thing and we get progressively better at figuring out how to build them which also allows us to scale them up. While there is a prestige element to supercomputers, they are actually built for a purpose of solving pa
Re: (Score:2)
I find that totally insane and and incredible waste of resources. These large things are basically not useful for a lot of things and they usually get partitioned into much smaller parts in actual use anyways. This is pure "mine is bigger" on the side of the politicians, nothing else.
Re: (Score:2)
Bigger things usually means the guys running big things do need less staff.
E.g. if a big tanker can be run with 12 men, an even bigger tanker usually can be run with 12 men as well, or one more.
Re: (Score:2)
Often it's about poor use of space, rather than just size. There are a lot of large cars that are no more roomy inside than much smaller ones. The most extreme examples are Japanese Kei cars, which despite being tiny even by European standards, somehow feel spacious and comfortable and practical.
Re: (Score:2)
Yes. Something like the Kei cars is really all that is needed to transport people and smaller goods. Everything larger is just a personal fetish. Of course, fetishism accepted by society often finds a lot of takers.
A fetish about not being smashed to bits (Score:2)
There is a "knee of the curve" effect of cars below a certain size becoming death traps.
A Camry has a pretty good real-world safety rating, and going smaller does not appear to save much gas.
Re: (Score:2)
I think it's not so much safety issues (those can be addressed with good engineering) as cost issues.
For example, the 451 Smart Car was about 50% the size of your canonical Camry, but sold for about 90% of the price. For most people (i.e. those who consider more interior room to be better than less interior room) that wasn't a good value proposition, and therefore the Smart Car never sold well in the US.
I think there are a lot of fixed costs for a car that don't scale down as you scale the car down. It's
Re: (Score:2)
The premise is dumb. You don't know if it's too big until you actually do it.
And in actual reality that statement is dumb. You can pretty much know before if it is one or at most two enlargement steps, but you need to want to know. The A380 is basically a "my dick is bigger" story (see George Carlin for his "bigger dick theory"), nothing else. Leadership wanted this thing and was not interested in whether it would work well. Sure, them being Airbus, the A380 is actually safe and reliable (unlike the crap Boeing "designs" these days) but pragmatism and economic viability had left the
Re: (Score:2)
Here's the thing; the A380 has been, at the time of its release at least, the most efficient aircraft available, despite being overbuilt (Airbus originally planned an even larger version).
Only two years later the engine manufacturers started offering a new engine generation and then the whole A380 project was fucked. And yet, despite being overbuilt and having inferior engines, it actually sold much better than Boeing's answer to it.
Very large aircraft only makes sense at congested hub airports lacking slot
Re: (Score:2)
Very large aircraft only makes sense at congested hub airports lacking slots. Airbus assumed such airports were the future, but it turns out the future was direct flights from secondary airports. VLA is simply too much aircraft for these.
Yes. But I am pretty much sure a major part for that assumption was that they wanted to have the biggest one.
Re: (Score:2)
Here's the thing; the A380 has been, at the time of its release at least, the most efficient aircraft available, despite being overbuilt (Airbus originally planned an even larger version).
It seems when designing a new airplane, the world can change during its multi-year development time. A380 is great for hub and spoke but people want to go direct flights. SST was viewed as what will be common passenger transport (747 can be used for cargo), so much for that prediction. Then we have the car, "a six lane superhighway! traffic jams will be a thing of the past, and it will be beautiful!" Yeah, so much for that prediction. There is the object itself (the car, the plane, the ship, the widget, the
You don't know you've gone too far ... (Score:2)
until you get there.
Paid to do it. (Score:2)
How many of these oversized projects were commissioned, approved, and funded by the engineers? Not counting those who were engineers, got an MBA and moved into executive roles.
Re: (Score:3)
Itâ(TM)s hard to know. (Score:2)
Generally, they just build bigger until they reach the practical limit, because itâ(TM)s hard to know the limit before youâ(TM)ve reached it. There are a lot of instances where bigger looks better on paper, but turns out not to be worth the effort in practice. Nothing wrong with that, planners are just following the science.
Iâ(TM)m curious about this effect when it comes to SpaceXâ(TM)s starship. For expendable and partially reusable vehicles, the most practical size is in the 10 to 20 t
The 747 was the perfect plane for its time (Score:2)
The DC3 and 747 are probably the greatest commercial airplanes ever made. The advancements that came with the 747 can not be overstated.
Re: (Score:2)
Same goes for the Airbus A320.
Re: (Score:2)
Also Airbus can still develop actually new planes, while Boeing apparently now fails catastrophically even at only modifying existing designs.
Often driven by Wall Street obsession with growth (Score:4, Insightful)
Re: (Score:3, Insightful)
Re: (Score:3)
You can't just have a nice, profitable business that pays salaries and dividends. Wall Street insists that you keep growing.
That's not true. There are lots of businesses that pay salaries and either pay dividends or just are privately owned and pay their owners directly without it officially being "dividends" in the strict stock market sense. Wall Street generally doesn't pay any attention at all to these businesses, far from insisting that they do anything.
Of course if you happen to be running a business that you think might catapult you to the next level of wealth then you might be trying to get Wall Street's attention, in whi
Hindsight (Score:2)
Yeah (Score:2)
So what? Anybody heard of the pyramids of Lilliput? No? Cause nobody gives a shit about a tiny pyramid. We need giant spaceships if we are going colonize Mars, it would be shit expensive and polluting to do it any other way.
Re: (Score:2)
What we need to colonize Mars are giant radiation shields. Without those, you cannot get there from here.
Computers too. (Score:2)
As an expert in the field (as most of us) I find it obvious that computing hardware has been optimising for the wrong things lately. Regular laptops are in multi-tera-flop territory now and do 60+fps in resolutions that go way beyond print. It's getting quite silly.
At the same time, if one single component in a consumer computer goes bust there is no feasible way to repair, replace our even diagnose the problem.
Optimising for low power, sturdyness and easy maintenance computing is overdue. In power efficien
Re: (Score:2)
Apple fanboys man...
You complain about repairability then laud Apple. The hell?
Also, remember netbooks? Cheap, low power, consumer-rugged, repairable... That wasn't apple. And that had nothing to do with x86. None of the things you mention do.
If you want very rugged and repairable, get a toughbook. Also not apple. Also x86.
Re: (Score:2)
If you want very rugged and repairable, get a toughbook. Also not apple. Also x86.
Does not run Mac OS X or macOs - or if you indeed can hackintosh it, it takes more time than it is worth.
In my experience there are no Apple fanBois. Only Apple haters.
I use Macs because they run Mac OS X / macOS - which is unix. And my Macs are all repairable - no idea where this misinformation is coming from.
Re: (Score:2)
You moved the goalposts so far they're on the moon.
You're giving Apple credit for being over a decade late to the party because they're the first people to do it with OSX.
Since you were replying up my post about toughbooks: sure they don't run OSX, butt neither does a broken mac.
The fanboyism is strong here.
Re: (Score:2)
Re: (Score:2)
Re: Computers too. (Score:2)
I remember watching Apple announce the MacBook Air as the worlds thinnest laptop on my Toshiba R100, which was thinner and lighter.
Re: (Score:2)
Same here. And I am also hoping for RISC-V, because while x86 / AMD64 has had a good run, there is not much to be gained there anymore.
Military suffers from that too (Score:2)
A lot of military projects were looking at (and in some cases, actually built) giant war machines, from huge yield nuclear bombs to 200-ton tanks. In the aircraft division, take a look at the Antonov An-225 Mriya, which I had the pleasure of seeing live a couple months ago (landing and take-off). In the armored division, there were many projects, such as the rather obscure Japanese O-I project, or the widely known German Maus and "E" series. The Germans really liked their big armored vehicles.
The USSR was a
Re: (Score:2)
The M1A1 tank is 67.6 tons. I think a 200 ton tank would need a nuclear powered engine, either that or it goes once around the block before needing a fill up.
Re: (Score:2)
Not necessarily.
https://en.wikipedia.org/wiki/... [wikipedia.org]
Sure, blame the engineers and designers (Score:3)
Itâ(TM)s not the ego of totalitarian rulers and endless appetites of the economy, right?
And all this time (Score:4, Interesting)
I thought gigantism was an disease caused by an excess of growth hormone during childhood [nih.gov].
Big good. Bigger ? (Score:2)
Depending on where you stand you might see all of this either as an admirable quest for new horizons (a quintessential human striving) or irrational and wasteful overreach (a quintessential human hubris).
You don't know the limits until you pass them!
Some endeavours become limited by unchanging physical constraints that were not obvious until they were reached. Others by economic factors which might start out favourable to a development, but can change over time (just think of Concorde) and sometimes change back again.
However, what factors that provide an incentive to make bigger (or in the case of semiconductors: smaller) structures is reflected as much in the animal kingdom and plants, To anthropomorph
Searching for limits (Score:2)
Diminishing returns (Score:2)
Most of the examples are a case of diminishing returns. Take shipping, getting bigger and bigger improves the efficiency of the vessel, in terms of tons carried per dollar of cost in fuel and crew. But it doesn't linearly scale, at some point the efficiency gain from going bigger isn't that big and you have penalties like restrictions on use of canals, restrictions on ports that can take the ship and a therefore a limited number of destinations that make sense.
Same for the A380, the 747 was the biggest plan
Or ... (Score:2)
Only mediocre engineers (Score:2)
Or more likely actually "management" supported by mediocre engineers that can see the detail and get it right, but do not understand the overall picture. You find those unfortunately everywhere (good engineers are rare). What they do often results in designs that work somewhat but that people like me feel a strong urge to fix because they are just not what they could be.
Good engineering lives by KISS and by pragmatism and requires realistic-self-assessment and some humility on the side of the designer. All
Re: (Score:2)
Economies of Scale (Score:2)
It's most economically efficient to build something as big as it can be when you have large fixed costs and low marginal costs.
The land under a skyscraper in NYC is very rare and expensive so they make the building as tall as is economical because one story in a tower rents for almost as much as a one story building (or maybe more with a view).
Supermax tankers require the same size crew as a regular tanker and barely more fuel, etc.
But the pyramids are a bad example because they're tuned to certain resonant
Error is essential (Score:2)
... chill! Never heard of post-Panamax? Optima are best/quickest approached from both sides. Many things are unpredictable, dealing with everything is engineering; science isolates. If there is anything "exceptional" about America, it is tolerence for error.
Conrad Knew This Long Ago (Score:2)
Re: (Score:2)
Of course, the Titanic was tiny in comparison to modern cruise ships. :)
Re: (Score:2)
Not really. Remember that the larger ones come with additional risks that the smaller ones do not have. For example, the long Suez-Canal blockage would not have happened with a smaller ship (container freighter though), because it would a) not have gotten into that steering problem and b) could have been towed free much more easily.
Re: (Score:2)
Nice comment, well stated.
But, I would amend part of your last paragraph. Your examples are good, except for the cruise ship. If you meant an ocean liner, that would be true, but the era of ocean liners as trans-continental transportation died in the era of air travel. Now, cruise ships are nothing but Vegas vacations on the waves, where the destination is a triviality and as such speed and fuel efficiency mean nothing. Instead "packing them in" makes money. The market has changed, and big cruise ships