0allena0

There are two kinds of robots in Hollywood: innocuous servants that are one glitch away from murdering their masters, and the bots designed for wholesale slaughter from the start.

Here's a look at the latter, a purpose-built class of killer machines-and the pioneering models that gave new shape to one of our oldest nightmares.

MECHANICAL MONSTERS

The stars of the 1941 Superman animated short presented the most lasting and awkward archetype of deadly robotics-the dreaded humanoid.

ED-209

While there's nothing particularly deadly about the two-legged design of ED-209, which first appeared in 1987's RoboCop, it was iconic enough to remain intact for the 2014 remake.

SENTINELS

The Matrix (1999) provided the most alien take on killer bots: the horrific squid-legged, spider-headed Sentinels, which attack victims with multitudes of razor-sharp tentacles.

This article originally appeared in the July 2013 issue of Popular Science. See the rest of the magazine here.

Find Us On Facebook — Business Insider: Science

Join the conversation about this story »

We take off.

On June 28, NASA flipped the switch for GALEX, shutting down the highly successful astronomical observatory. This was not unexpected; the mission had a fantastic 10-year run, and given NASA’s limited budget, the funding to operate it was needed elsewhere^*. It’s always sad to see a workhorse like this one put down, but it had a long and very successful life.

GALEX—the Galaxy Evolution Explorer—observed the sky in ultraviolet, a high-energy form of light. Young stars still embedded in their birth clouds pour out this kind of light copiously, and the primary goal of GALEX was to observe galaxies in the UV to determine the rate of star formation. By observing galaxies at all distances, from the local Universe to the farthest reaches of the cosmos, it would see them at different ages, allowing scientists to see how galaxies evolve over time. It was the only UV camera in space capable of doing this kind of survey.

Along the way it made some pretty amazing observations. Galaxies are lovely in the UV, but GALEX wasn’t limited to looking just at them. It also took a peek at Mira, a red giant star just 400 light-years from Earth. The star is dying, shedding its outer layers while at the same time plowing at high speed through the thin material that lies between the stars themselves. The result is that the star looks like a comet with an extended tail, except that tail is 130 billion kilometers long. You can even see the bow shock wave, the arc in front (on the right) of the star as it rams through that interstellar material, compressing it.

It also looked at stars long dead: That picture above shows the Cygnus Loop, a vast structure created by the expanding debris of a long-ago exploded star, slamming into and sweeping up the gas around it. Located about 1,400 light-years away, the nebula is huge: 80 light-years across. Even at that tremendous distance, its apparent size is six to seven times that of the full Moon on the sky! The gas is compressed into tendrils and filaments hot enough to emit ultraviolet light, making it appear like an enormous jellyfish.

Lastly, I can’t leave this without putting up that magnificent image of the Andromeda Galaxy, our nearest big spiral galaxy neighbor in the Universe. I wrote all about this image before, so you can get your fill (Phil?) of the science there. But it’s impossible not to be awed by the sprawling complexity of the spiral arms in this gorgeous galaxy, the site of the births of countless billions of stars. Because it’s so close—about 2.5 million light-years away, our back yard as galaxies go—Andromeda is one of the best-studied objects in the sky. In the fall it’s even visible to the naked eye, a faint smudge that belies its profound nature.

I never worked with GALEX directly, but I’ve worked extensively with a different space-based UV camera, and it’s hard to see one of its brethren turned off. But the data GALEX has taken will live on in archives that will be used by astronomers for many years. And I hope that in the future there will be more and more sophisticated observatories like GALEX yet to come.

^*I think the fact that NASA has to make a choice to turn off functioning missions due to budget constraints is a national embarrassment. NASA costs us very, very little and gives us so, so much. See the bottom section of this post for my editorializing on this topic.

File this under “OMG this is so cool why didn’t I think of it?!”: a video of a superconducting magnet sailing up, down, and around a track shaped into a Möbius strip:

I love this so so so much. It’s all explained in the video (starting around two minutes in) but in case you missed it:

1) There exists a kind of material that, if you chill it down to very low temperatures, becomes a superconductor, meaning electrons can flow through it freely without resistance. What that means is that a lot of weird stuff can occur. For one thing, if you induce an electric current in a superconductor, that current will flow forever without resistance.

As an analogy, imagine filling your sink with water. If you give the water a brief swirl with your hand the water will flow around the sink. Eventually, though, friction slows it and the water will be still once again. But if the water were frictionless, it would swirl forever. A superconductor is like a frictionless sink, filled with electrons instead of water. Once electrons start to flow, making a current, they never stop.

2) Another important thing is that if you stick a conductor in a moving magnetic field (or move a magnet over a conductor), it will induce a current in the conductor, which will start electrons flowing. One of the basic laws of electricity and magnetism is that the flowing electron current (which we call electricity) will itself generate a magnetic field, and it’ll do so in a way to resist the original changing magnetic field. It’s a bit like if you feel yourself falling over, and you flail your arms (or grip something quickly) to get your balance back.

3) Still with me? Now combine this. In a superconductor, if it’s set up properly, you can generate a flow of electrons that’ll whiz around inside it. If you then put that superconductor on a track, it will levitate above the track because gravity pulls it down, changing the magnetic field it feels, so it sets up its own magnetic field to resist that. Same thing if you hang it upside down; gravity pulls it away from the track, so it changes its own field to hold on to the track tighter.

4) Finally, if you look at the magnetic field of the track, it changes rapidly near the edges, toward the inside and outside. That acts like a buffer, keeping the superconductor on the track (because, again, the superconductor’s own magnetic field will resist that change, forcing it to the middle of the track). But the field doesn’t change along the track, so the superconductor doesn’t care about moving in that direction. Poke it, and it’ll glide, floating, along the track.

5) Now we can combine everything. Chill a puck made of yttrium barium copper oxide (or as we in the science biz call it, YBCO, to sound—wait for it, wait for it—cool) with liquid nitrogen so that it becomes a superconductor. Put it on the track. The magnetic field lines of the magnets in the track grip the puck invisibly, and it can glide up and down the track. Yay!

All of the above is well-known and most students of physics have seen this demo a zillion times. But the idea of shaping the track into a Möbius strip? That is pure genius! A Möbius strip is a shape that has only one side. Here’s how to make one: Take a strip of paper that is much longer than it is wide. Give one end a half twist, so it’s flipped over, and attach it to the other end. Voila! A Möbius strip. Take a pencil, put the tip anywhere on the strip, and start drawing a line along it. After a moment, you’ll meet up to the point where you started. If you hadn’t twisted the paper, and just attached one end to the other to form a squat cylinder, that won’t happen. There’s an inside face and an outside face. In a Möbius strip there’s only one face, connected to itself.

So in the video, the floating superconductor slides along the track, flips upside down, keeps moving, and eventually comes back to the place it originally started. If you’re impatient, the cool stuff starts at 4:45 in the video.

This is really just a clever way to show an old demo, but I love it. It’s got everything! Exotic materials, liquid nitrogen, superconducting magnets, and a Möbius strip all combined in a live-action Escher drawing. It’s nerdvana.

Hmmm, well, actually, it doesn’t have everything. What it really needs is for the puck to have a little fan on it, a propeller, so it can keep going around and around forever. Or until it warms up and the superconductivity collapses (maybe attach a hose to it somehow to keep it supplied with liquid nitrogen…?). That would make a fun exhibit at a museum, too.

Science! I love this stuff!

Tip o' the Meissner Effect to Philippe Roux on G+.

Mark Showalter, an astronomer at the SETI Institute, was playing with some old Hubble images of Neptune and got a surprise: He found a moon!

I love this story. Neptune has ring fragments, arcs of material orbiting it that are essentially segments of rings. Showalter studies these arcs and went to the Hubble data archives to get a pile of old archived images of the planet taken by the space telescope, finding about 150 in all. He then used some techniques that allowed him to track the motion of the ring segments around Neptune. On a whim, he applied the technique to look for anything outside the ring system, and boom! There was a white dot that persisted in many of the images: Neptune’s 14^th moon.

The moon orbits Neptune at a distance of about 100,000 kilometers (65,000 miles) and is only about 20 kilometers (12 miles) across, which is pretty small. Also, since Neptune is over 4 billion kilometers (2.5 billion miles) from Earth, the moon is incredibly faint. That’s why it took so long to find, and it was almost by accident! Showalter’s method revealed the tiny iceball, and that’s the reason I love this story; he had the data, and he was really just playing around with it to see what he could see. I used to do that all the time with Hubble data, and once I found a dying star. It’s worth sometimes just indulging yourself; you never know what you might find.

The moon has a provisional name of S/2004 N 1 (which means it’s a planetary Satellite first seen in data from 2004, it’s orbiting Neptune, and the first one discovered in that year). It orbits the planet in about 23 hours, just under one Earth day.

Which gives me an idea for a potential name. Moons of Neptune are named for beings associated with the god of the oceans. There was no Roman god of tides per se, but the minor deity Volturnus was the god of waters, which is close (and not being used by any asteroid; always a problem when naming names). And tides on Earth have a period of just over a day, but hey, close enough to this new moon’s orbital period. So why not? Plus, Volturnus is a seriously kickass name.

It’s easy to forget just how far away Neptune is; it’s the gatekeeper of the outer solar system, and even with our fastest spaceships it takes many, many years to get out there. In some cases our best views are from telescopes right here on (or above) Earth. Hubble’s been orbiting our planet for more than 23 years now. What other treasures are buried in its data, just waiting to surface?

If you live in the Northern Hemisphere and go outside around midnight in the summer, you’ll see a big W made of five brightish stars in the northeast. That’s Cassiopeia, the constellation of the queen. Due to geometric circumstance, we are looking into the plane of our galaxy when we peer in that direction. That means we’re looking into a region of space that has more stars, more gas, and more dust than average.

Which means it’s a good place to look for interesting objects. And in fact it’s where you will find the gorgeous nebula Sharpless 2-188, the gas cast off by a dying star:

This image was taken by my friend Travis Rector (and Heidi Schweiker) using the Kitt Peak 4-meter telescope. They used two filters; one shows hydrogen gas (orange) and the other oxygen (cyan). I highly suggest you grab the 4,900-by-3,800 pixel version, or at least the 1,200-by-940 pixel image, because wow.

Sh2-188 is what we call a planetary nebula, a bit of a misnomer since the first that were discovered looked small, disk-shaped, and slightly greenish through a telescope. Now we know they’re formed by the winds from dying stars, vast gales of gas expelled as the star sheds its outer layers and exposes its small, hot core. The leftover star, called a white dwarf, is so hot it blasts out ultraviolet light, enough to cause the gas to glow.

Planetaries come in fantastic shapes, mostly due to the way the wind from the star changes over time and interacts with previously blown-out material. But Sh2-188 is different, even by the standards of the planetary nebula menagerie. It’s highly asymmetric. Even though you can see it makes a complete circle (well, oval), one side is far brighter and more interesting-looking than the other.

This is almost certainly due to the culprit star itself, screaming through space at high speed. It’s not 100 percent clear which star is at the heart of this nebula, mostly because there are so many to choose from! But there is a faint star down to the lower left of the inside of the cloud, faint and not at all distinguishable by eye from any of the others, that is most likely the actual “central” star. It’s blue—expected for the hot, dense, cinder needed to light up this gas—and about the right brightness.

Also as expected, measurements over time indicate the star is moving through space toward the lower left, probably at well over 100 kilometers (60 miles) per second. That would explain much of why this nebula is so asymmetric; as the star (and its wind) rams through the gas in space, the stuff in the direction of travel piles up, just like a snowplow pushing through snow. It’s also possible that there’s just more stuff in between the stars in that direction, so there is more material to pile up. But either way, the speed of the star is the major reason this planetary is so unusual.

I found this picture browsing through Travis’ gallery, and even before I read his description I thought to myself, “Huh. That looks familiar!” and sure enough, Travis had the same thought and mentioned it in the caption.

Does it look like the logo for any browser you might be using?

Once you see it, the resemblance is pretty good. I can see ears, a nose, the tail, even the foreleg! And while it may not actually be on fire, the surface temperature of the stars is far hotter than the Sun. Of course, it’s in the constellation of Cassiopeia, the queen, not Vulpecula, the fox, but still. At 60 trillion kilometers in diameter, it makes a heckuva logo.

Mozilla, you may have a new direction for your branding …

Related Posts:

The Flaming Skull Nebula. Seriously.
A Dying Star With the Wind in Its Hair
A Glowing, Bubbly Bauble in Space
A Star on the Edge of a Weird, Lovely Death
The Green Ghost of a Distant Dead Star

Today is the anniversary of the first time humans set foot upon another world. Michael Collins, Buzz Aldrin, and Neil Armstrong lifted away from Earth on July 16, 1969, and on July 20, 1969, Aldrin and Amstrong effectively sliced all of history into two different halves.

Collins represents an interesting paradox. While he was one of the three Apollo 11 astronauts to go to the Moon, he didn’t walk on its surface. His task was to stay in orbit, on board the Command Module, while Armstrong and Aldrin cavorted upon the lunar surface.

Some people feel sorry for him, getting almost all the way to the Moon but denied the opportunity to land there.

But he also had a singular opportunity: He became the farthest man in the Universe. When he circled the far side of the Moon, the nearest people to him were thousands of kilometers away, and ignoring them, the rest of humanity was 400,000 kilometers distant. A quarter million miles.

This happenstance is both amazing and melancholy, an experience he describes in his book, “Carrying the Fire”. Musician Simon Lacey read the book, and became so inspired by Collins’ adventure that he wrote several classical pieces of music about it. A friend of Lacey’s, who is a TV editor, agreed to put together a time-lapse video of the Earth seen from the International Space Station, and together they created this wonderful collaboration:

Lovely! The music is sweet, but still with a hint of loneliness to it. The emotional sway of the music makes the connection to Collins more palpable.

This is the first of several pieces by Lacey, and I’m looking forward to seeing the next.

As for Collins, a little while back a picture circulated the ‘net (shown at the top of this post). It shows the Lunar Module carrying Aldrin and Armstrong returning to orbit, about to rendezvous with the Command Module carrying Collins.

You’ve probably seen it; I have looked at this picture countless times. But it still startled me when I saw the caption someone had dreamed up for it: “Michael Collins is the only human being, living or dead, not in the frame of this picture.”

Perhaps that won’t always be true; there will come a time when a picture of the Earth doesn’t encapsulate all of humanity. But for now it does, and while we can’t all experience the feelings Collins had at that time, perhaps, on this day, one of greatest of all anniversaries, the music by Lacey can help us understand them.

Фото: Норвежский Лесной

Фото: Норвежский Лесной

Wolfram Alpha is probably the most useful site on the internet.

It's not a search engine, it's not an encyclopedia, and it's not a calculator, but it's a little bit of all of that. It's really the only member of its field. 

Originally developed as an online version of Stephen Wolfram's Mathematica software, its basic functionality is that of a math equation solver.

Over the years, however, it's grown substantially, and has really matured as a site to become one of the coolest and most informative sites online. 

Here are some of the coolest things you can do with it. 

1. You can find the nutrition facts for literally any food, as customized as you'd like it.

2. Use Wolfram Alpha to find out more than you ever wanted to know about why the weather is unbearable today.

3. The site makes traveling by plane a cinch.

Masao Yoshida, the manager of the Fukushima Daiichi power in Japan during its meltdown in  2011, has died of cancer, the New York Times reports. He was 58 years old.

The plant’s operator, Tokyo Electric Power, said that Yoshida died of esophageal cancer. Experts have denied a link to the nuclear meltdown due to the speed of his illness.

Yoshida is best known for a series of key decisions made after the plant began melting down after a powerful earthquake and tsunami crippled the plant's cooling systems, including using seawater to cool overheating fuel rods.

Importantly, the first day after the earthquake, Yoshida ignored a requests from both Prime Minister Naoto Kan and senior Tepco officials to stop injecting seawater into the reactors. While there were fears that the seawater might cause fission chain reaction, Yoshida's plan proved successful and he was never reprimanded for his disobedience.

In a video released last year, Yoshida is shown disagreeing with an unnamed company official at Tepco headquarters two days after the quake. The official wants to use fresh water to cool the remaining fuel rods, hoping that they might be reused. Yoshida responded, "We don't have the option to use fresh water. That will cause further delays."

While Yoshida has faced some criticism for failing to prepare the plant for the risk of tsunami, his solemn response to handling the disaster led to a reputation as a hero. He reportedly asked staff left at the plant to write their names on a blackboard to keep a record of who was there, and had to be dissuaded from leading a "suicide mission" to try and pump more water in to one of the reactors himself.

“If Yoshida wasn’t there, the disaster could have been much worse,” Reiko Hachisuka, the head of a business group in Okuma town and one of a panel that investigated the accident, told Bloomberg today. Former Prime Minister Naoto Kan also tweeted a tribute, saying, “I bow in respect for his leadership and decision-making.”

Join the conversation about this story »

The next highly-mobile rover to land on Mars will be designed to return samples from the Red Planet to Earth, according to a 160-page report [PDF] prepared by NASA's Mars 2023 Science Definition Team (SDT) released on Tuesday, July 9.

The team of scientists and other Mars experts was assembled in January to outline a mission concept for the successor to NASA's Curiosity rover, which has been rolling around the Red Planet and searching for signs of microbial life since it landed on Aug. 6, 2012.

Scientists have learned from Curiosity that Mars once had an environment that was potentially habitable for life, John Grunsfeld, NASA's associate administrator for science, said in a press conference on Tuesday. What is still not known about our neighboring planet is "the context of when that was in Mars' history or compared to the Earth's history," he added.

For the next large mission, the SDT recommends in situ exploration, meaning scientists would select one key site on Mars that has been confirmed to have had potentially habitable conditions. Specifically, the next rover will be looking for evidence of biosignature preservation, or marks potentially left by ancient forms of life.

The next rover is modeled after Curiosity. It would be car-sized with slight modifications to accommodate a new suite of science instruments, including the ability to collect Martian samples to send back to Earth for further analysis.

"The ability to collect and cache scientifically compelling, well-documented samples from in situ rock outcrops is unprecedented in Mars exploration and is the necessary first step in a systematic plan to search for life," the report said. 

Martian dust and rocks collected over the mission would be preserved in a sample tube and placed into a sample cache. The cache can hold up to 31 samples which would be jettisoned back to Earth. 

Back on Earth, scientists would perform sophisticated laboratory measurements — beyond the capability of what could be done on Mars — to further examine the samples.

The mission is expected to cost $1.5 billion not counting the cost of the launch vehicle. The total cost to build, design, and launch Curiosity was $2.5 billion.  

PHOTOS: This Big Ship Is Powered Entirely By The Sun

SEE ALSO: Scientists Spent Weeks In The Desert Pretending They Were On Mars

Join the conversation about this story »

More than 800 firefighters are battling two massive wildfires moving through mountain areas near Las Vegas and Reno, Nevada, NBC News reports.

The fires 35 miles northwest of Las Vegas are only 15% contained as of Tuesday evening, and span almost 31 square miles. So far, there have been no injuries or reported property damage.

More than 500 residents and another 98 teens at a youth correctional camp have been evacuated. Several local casinos have said they would make their rooms available to the displaced, Casper Star-Tribune reports.

Time has more:

It’s burning in pinion, juniper and bristlecone pine forest in rugged territory covering more than 30 square miles and continues to threaten about 400 homes. More than 500 residents and another 98 teenagers at a youth correctional camp have been evacuated since the weekend.

“It’s dry,” Nichols said. “We’ve got torching trees and spotting fire. We’re being extremely careful and monitoring the safety of firefighters and the public.”

The caution is justified — after the death of 19 firefighters who were battling a blaze in Arizona less than 10 days ago.

In addition to hundreds of personnel, there are 44 engines, 10 water tenders, nine helicopters, and a DC-10 tanker plane that can drop 11,700 gallons of fire retardant on the scene, according to NBC News.

The Vegas-area fires are just two of 21 uncontained large fires currently burning across the country.

Join the conversation about this story »

AccuWeather.com Meteorologist and Social Media Coordinator Jesse Ferrell writes in his blog: "When I saw this photo this morning on Social Media, I couldn't believe it was real."

Ferrell has vetted the photo and says that it is one of dozens of photos of the photogenic waterspout that came on shore north of Tampa, Fla., Monday night (additional photos on Twitter and Facebook from Denis Philips (ABC) and Paul Dellegatto (FOX)).

The photo (and video below) was taken by Joey Mole, of Safety Harbor, Fla.

Joey says he had never seen a waterspout before, and neither had his neighbors, who had lived there for 40 years. It made landfall only 250 feet from Joey's dock and went over his neighbor's home, two houses down. Joey says minor damage occurred, including missing shingles and downed trees.

Find Us On Facebook — Business Insider: Science

Join the conversation about this story »

Как уже обсуждалось в предыдущем посте из серии "ЛаТеХ для продвинутых", одним из самых универсальных средств для контроля над расположением картинок, таблиц и подписей к ним в пределах плавaющего объекта является пакет floatrow.

В предыдущей части мы рассмотрели всевозможные способы взаимного расположения картинки и подписи (снизу, сверху, сбоку), выравнивания их по горизонтали и т.п. Этот же пост будет посвящён расположению нескольких объектов (рисунков, таблиц) в ряд. Обратите внимание, что имеется в виду именно расположение нескольких объектов типа figure и/или table, а не нескольких картинок/таблиц (обычно помеченных как а, б, в, ...) в пределах одного объекта, как это делают пакеты subcaptionили subfig.

Все примеры из данного поста могут быть найдены здесь и здесь в виде tex-файлов готовых к компиляции. Документация к пакету floatrow доступна на русском и английском языках.

Краткий анонс

Сначала мы познакомимся с командами \ffigbox{}{} и \ttabbox{}{}, которые создают бокс для картинки/таблицы и соответствующей подписи.

Затем будут освещены средства, необходимые для контроля над горизонтальным выравниванием:

• Окружение floatrow для расположения иллюстраций и таблиц в ряд.
• Длины FBwidth (ширина картинки) и Xhsize (ширина остатка страницы).
• Параметр floatrowsep, определяющий расстояние между картинками в ряду.

И, наконец, будут разобраны способы выравнивания объектов по вертикали, как то

• Команды \TopFloatBoxes, \BottomFloatBoxes, \CenterFloatBoxes, выравнивающие весь float.
• Параметры heightadjust и valign для выравнивания только картинки/таблицы.
• Выравнивание подписей.

Команда \ffigbox

Основной рабочей лошадкой во всех примерах ниже является команда \ffigbox предоставляемая пакетом floatrow. Её синтаксис прост:

01:  \ffigbox[ширина][высота][вертикальное положение]{подпись}{картинка}

01:  \usepackage{floatrow}
02:  \usepackage{graphicx}
03:  ...
04:  \begin{figure}[!h]
05:   \ffigbox{\caption{My Caption}\label{fig:ffb}}%
06:           {\includegraphics{roman_a}}
07:  \end{figure}

Для таблиц определена аналогичная команда \ttabbox: далее мы сосредоточимся на картинках, но всё без исключения применимо и к таблицам.

Расположить несколько рисунков в ряд нам поможет окружение floatrow:

01:  \begin{figure}
02:  \begin{floatrow}
03:   \ffigbox{\caption{My Caption left}\label{...}}%
04:           {\includegraphics{...}}
05:   \ffigbox{\caption{My Caption right}\label{...}}%
06:           {\includegraphics{...}}         
07:  \end{floatrow}
08:  \end{figure}

Горизонтальное выравнивание

Волшебная длина \FBwidth

Предыдущий рисунок выглядит очень неуравновешенно, так как ширина отводящаяся на подпись не соответствует ширине картинки. Гораздо более симпатичного результата можно добиться вот так:

01:  \usepackage{floatrow,graphicx,calc}
02:  % Создаёем новый разделитель
03:  \DeclareFloatSeparators{mysep}{\hspace{3cm}}
04:  ...
05:  % Настраиваем значение разделителя для объектов 
06:  \thisfloatsetup{floatrowsep=mysep}
07:  % А вот и сам плавающий объект
08:  \begin{figure}
09:   \begin{floatrow}
10:    \ffigbox[\FBwidth+1cm]{\caption{...}}%
11:            {\includegraphics{...}}
12:    \ffigbox[\FBwidth+1cm]{\caption{...}}%
13:            {\includegraphics{...}}         
14:   \end{floatrow}
15:  \end{figure}

По порядку

1. В строках 10 и 12 использован необязательный аргумент в \ffigbox, который задаёт ширину плавающего объекта. В этом примере мы задаём ширину на 1см больше, чем ширина соответствующего изображения: т.е. подпись будет выдаваться на 0.5см с каждой стороны. Заметьте, что ширина самого изображения доступна через \FBwidth, a для того, чтобы можно было к ней прибавить 1см, надо загрузить пакет calc в строке 1.
2. Расстояние между двумя объектами (Figure 3 и Figure 4) задано в строке 6 с помощью параметра floatrowsep. Обратите внимание, что нельзя просто написать floatrowsep=3cm, вместо этого надо определить новый разделитель (назовём его mysep) в строке 3, а затем использовать в floatrowsep=mysep. Пакет предопределяет следующие разделители: none, quad (=1em), qquad (=2em), columnsep (равен длине \columnsep, которую ЛаТеХ использует для разделения двух колонок текста).

Волшебная длина \Xhsize

Помимо очень полезной длины \FBwidth, пакет предопределяет длину (точнее, ширину) \Xhsize, равную оставшейся ширине страницы. Эта длина может быть использована точно так же, как и \FBwidth, в первом необязательном аргументе \ffigbox.

Например, в случае, когда одна из подписей очень длинная, под неё может быть желательно выделить побольше места:

01:  \begin{figure}
02:   \begin{floatrow}
03:    \ffigbox[\FBwidth+1cm]{\caption{My Caption left}}%
04:            {\includegraphics{...}}
05:    \ffigbox[\Xhsize]{\caption{My Caption right very .... very long}}%
06:            {\includegraphics{...}}         
07:   \end{floatrow}
08:  \end{figure}

Несколько картинок в ряд

До сих пор мы рассматривали расположение только лишь двух рисунков в ряд. Однако, ситуацию легко обобщить для произвольного числа рисунков. Для этого небходимо задать необязательный аргумент к окружению floatrow равным числу рисунков, как проиллюстрированно в третьей строке нижеследующего примера.

01:  \thisfloatsetup{floatrowsep=qquad}
02:  \begin{figure}
03:  \begin{floatrow}[3] % три рисунка в ряд!
04:   \ffigbox[\FBwidth+0.5cm]{\caption{...}}{\includegraphics{...}}
05:   \ffigbox[\Xhsize/3*2]{\caption{...}}{\includegraphics{...}}  
06:   \ffigbox[\Xhsize]{\caption{...}}{\includegraphics{...}} 
07:  \end{floatrow}
08:  \end{figure}

Из этого же примера очевидна гибкость использования \FBwidthи \Xhsize: к ним можно прибавлять длины, их можно умножать на скаляры. Кроме того, обратите внимание, что параметр \Xhsize в строках 5 и 6 имеет разные значения: он каждый раз означает ширину, оставшуюся на странице на момент вызова \ffigbox!

Вертикальное выравнивание

Весь объект

До сих пор во всех примерах использовались картинки приблизительно одинаковой высоты, т.е. особой нужды в форматировании их вертикального положения не возникало. Если же рисунки отличаются по высоте, то встаёт вопрос об их вертикальном положении относительно друг друга.

Эта проблема решается с помощью одной из команд
\BottomFloatBoxes (выравнивание по низу),
\TopFloatBoxes (выравнивание по верху),
\CenterFloatBoxes (выравнивание по центру),
которая должна следовать сразу после \begin{figure}.

01:  \begin{figure}[!h]\TopFloatBoxes % or \BottomFloatBoxes or \CenterFloatBoxes
02:   \begin{floatrow}
03:     \ffigbox{...}{...}
04:     \ffigbox{...}{...}
05:   \end{floatrow}
06:  \end{figure}

Только картинки

В большинстве случаев, однако, необходимо выровнять только картинки, а подписи к ним оставить на одном уровне. Без проблем:

01:  \begin{figure}[!h]
02:   \floatsetup{heightadjust=object,valign=c}
03:   \begin{floatrow}
04:    \ffigbox{\caption{...}}%
05:            {\includegraphics{...}}
06:    \ffigbox{\caption{...}}%
07:            {\includegraphics{...}}         
08:   \end{floatrow}
09:  \end{figure}

Здесь использованы два параметра:

heightadjust: какие части float в ряду должны быть одинаковой высоты. Соответственно, его значения

1. object: место, отводящееся под саму картинку, выполнить одинаковой высоты для всех картинок;
2. caption: то же, но для места, отводящегося под подпись;
3. all: первое и второе вместе;
4. noobject, nocaption, none: отрицание, соответственно, первого, второго и третьего

valign: "тип" вертикального выравнивания. Может быть t (по верху), b (по низу), c (по центру).

Подписи

Для рисунков, когда подписи размещены под иллюстрацией, нет необходимости их дополнительно "выравнивать": подписи и так выровнены "по верху". Однако у таблиц, когда подписи размещены сверху, так как использована директива

01:  \floatsetup[table]{style=plaintop}

01:  \floatsetup[table]{capposition=top}

Чтобы вырoвнять их "по верху" используем

01:  \floatsetup[table]{style=Plaintop}

01:  \floatsetup[table]{capposition=TOP}

Полностью код примера выше: открыть

01:  % используем стиль, в котором подписи таблиц 
02:  % сверху и выравнены "по верху"
03:  \floatsetup[table]{style=Plaintop,floatrowsep=qquad}
04:  %
05:  % Теперь сами таблицы в ряд
06:  \begin{table}[!h]
07:  \begin{floatrow}
08:   % первая таблица
09:   \ttabbox[\FBwidth]{\caption{My Caption left}}%
10:           {\begin{tabular}{c|cc}
11:             Column 1 & Column 2 & Column 3 \\ \hline
12:               a      &  b       &  c       \\
13:               a      &  b       &  c       \\
14:               a      &  b       &  c       \\
15:               a      &  b       &  c       \\
16:               a      &  b       &  c
17:            \end{tabular}
18:    }
19:   % вторая таблица
20:   \ttabbox[\FBwidth]{\caption{My Caption right, let us 
21:                             make it somewhat longer}}%
22:           {\begin{tabular}{c|cc}
23:                      Column 1 & Column 2 & Column 3 \\ 
24:                      \hline
25:                        a      &  b       &  c       \\
26:                        a      &  b       &  c       \\
27:                        a      &  b       &  c       
28:            \end{tabular}
29:           }
30:  \end{floatrow}
31:  \end{table}

Картинка рядом с таблицей

01:  \begin{figure}[!h]\CenterFloatBoxes
02:   \begin{floatrow}
03:     \ffigbox[...]{\caption{...}}{...}   
04:    %
05:    \killfloatstyle
06:    %
07:     \ttabbox[...]{\caption{...}}{...}
08:   \end{floatrow}
09:  \end{figure} 

И всё-таки subfloats

Хотя во введении и было обещано не рассматривать вопросы, связанные с расположением "подрисунков" (subfigures, subtables), стоит отметить, что все вышеприведённые рецепты действуют и в их случае с той лишь разницей, что надо использовать окружение subfloatrow вместо floatrow.

Простой пример ниже демонстрирует это в деталях

01:  \usepackage{subcaption,floatrow,graphicx,calc}
02:  \floatsetup{floatrowsep=qquad}
03:  ...
04:  \begin{figure}[!h]
05:  \ffigbox{
06:    % объявляем, что будут 3 картинки в ряд
07:    \begin{subfloatrow}[3]
08:       \ffigbox[\FBwidth+1cm]{\caption{My Caption left}}{\includegraphics{...}}
09:       \ffigbox[\FBwidth+1cm]{\caption{My Caption mid...}}{\includegraphics{...}} 
10:       \ffigbox[\FBwidth+1cm]{\caption{My Caption right...}}{\includegraphics{...}} 
11:    \end{subfloatrow}
12:  }
13:  { % подпись ко всему float 
14:    \caption{Overall caption.}
15:  }
16:  \end{figure}

1. надо загрузить дополнительно пакет subcaption;
2. вместо окружения floatrow используем subfloatrow;
3. причём помещаем его внутрь первого аргумента \ffigbox (строка 5);
4. а во втором аргументе \ffigbox помещаем подпись ко всему ряду рисунков (строка 14).

Абсолютно аналогично можно это применить и к таблицам, заменив \ffigbox на \ttabbox

Заключительный сложный пример для любознательных

Помимо рассмотренных выше трюков, floatrow позволяет создавать собственные типы float. Кроме того, пакет иеально функционирует в тандеме с другим пакетом caption, который позволяет должным образом оформить подпись.

В следующем примере использованны эти возможности:

1. Определен новый тип float под названием Photo (с собственным счётчиком и другими атрибутами), подпись которого набирается справа от картинки.
2. Для таблиц и др. объектов, подпись отфoрмaтирована специальным образом.

01:  \usepackage{caption,floatrow}
02:  %------------------------------------------------
03:  % Форматируем подписи к таблицам, благодаря пакету caption
04:  \DeclareCaptionLabelFormat{rightline}{\rightline{\bothIfFirst{#1}{ }#2}}
05:  % метка Table справа, затем новая строка, метка наклонным, сама подпись жирным малым шрифтом
06:  \captionsetup[table]{labelformat=rightline,labelsep=newline,%
07:                       labelfont={md,sl},textfont={bf,small}}
08:  % Подпись размещаем вверху от таблицы
09:  \floatsetup[table]{style=Plaintop}
10:  %-------------------------------------------------
11:  % Создаём новый тип float: photo
12:  % Команда \DeclareNewFloatType{тип}{опции} из пакета floatrow
13:  \DeclareNewFloatType{photo}{name={Photo},placement=tbp}
14:  % Определяем аналог \ffigbox для photo
15:  % с помощью \newfloatcommand{команда}{тип}[преамбула][ширина]   
16:  \newfloatcommand{photobox}{photo}%
17:    [{\capbeside
18:      \thisfloatsetup{capbesideposition={right,bottom},% подпись справа внизу
19:                      capbesidewidth=3cm}% ширина подписи 3см
20:    }]%
21:    [\FBwidth]
22:  % форматируем подпись с помощью пакета caption  
23:  \captionsetup[photo]{labelformat=rightline,labelsep=newline,%
24:                       labelfont={md,sl},textfont={bf,small}}%
25:  %--------------------------------------------------
26:  % для обычного "Figure"
27:  \floatsetup[figure]{capposition=bottom}
28:  \captionsetup[figure]{labelfont={md,sl},textfont={bf,small}}%
29:  %---------------------------------------------------
30:  ...
31:  \begin{table}[!h]\CenterFloatBoxes
32:  \begin{floatrow}
33:   \ttabbox[\FBwidth]{\caption{My Caption left}\label{tab:1}}%
34:           {\begin{tabular}{c|c}
35:             ...   
36:     \end{tabular}
37:           }
38:   %  
39:   \killfloatstyle 
40:   % 
41:   \photobox{\caption{Here goes a photo}\label{ph:1}}{\includegraphics{...}}
42:  \end{floatrow}
43:  \end{table}
44:  
45:  See Table~\ref{tab:1} and Photo~\ref{ph:1}. Also a ``normal'' Figure~\ref{fig:1}.
46:  
47:  \begin{figure}[!h]
48:   \ffigbox{\caption{Graph of a function}\label{fig:1}}{\includegraphics{...}}
49:  \end{figure}

Подведение итогов

Пакет floatrow позволяет не только автоматизировать процесс расположения подписи (сверху, снизу, сбоку от картинки), но и предоставляет мощные средства для контроля над расположением объектов в ряд. Предоставляемый им интерфейс облегчает форматирование визуального материала и позволяет добиться желаемого результата без больших затрат времени и сложных комбинаций с minipages.

Все примеры из данного поста могут быть найдены здесь и здесь в виде tex-файлов готовых к компиляции.
Документация к пакету floatrow доступна на русском и английском языках.