Shared posts
The Revelations of Lady Murderface
Striking Aerial Photos of Hong Kong Shrouded in Fog by Andy Yeung
All photos © Andy Yeung.
For his latest photo series Urban Fog, photographer Andy Yeung launched a DJI Phantom 3 from Victoria Harbor in Hong Kong to capture the city at night while covered in mist. Yeung said he was inspired by a similar project of the city photographed during the day, and was intrigued to see how he could present the city lights at night as they illuminated the fog. The photos are remarkable for their likeness to a thunderstorm with the cool lights of the city glowing inside the fog like flashes of lightning. You can see more of Yeung’s work on his website. (via Designboom)
Photos of Monumental Waves Crashing in Australia by Warren Keelan
Trying to capture a medium that’s in a constant state of flux would seem stressful in any situation, but photographer Warren Keelan works comfortably in a wetsuit amongst crashing waves on the South Coast of New South Wales, Australia, always trying for the perfect shot. Whether working completely submerged or perched precariously on the cusp of a behemoth swell, he’s consistently able to find the right angle and lighting to highlight the monumental power of the constantly moving ocean. He shares about his process:
I’ve always had a fascination with nature, especially the ocean and its ever changing forms, and I am compelled to capture and share what I feel are special and unique moments in the sea. I love the raw, unpredictable nature of water in motion and the way sunlight brings it all to life, from both above and below the surface. For me, the challenge is creating an image that hopefully tells a story or leaves an impression on the viewer.
Keelan has a gallery in his hometown of Wollongong, Australia, and many of his photos are avilable as prints online. You can also follow him on Instagram. (via This Isn’t Happiness)
I Have No Idea What This Startup Does and Nobody Will Tell Me
Drones Rigged with LED Lights Dramatically Illuminate Landscapes at Night
All images provided by Reuben Wu.
Photographer Reuben Wu‘s (previously) latest series attempts to bring the alien mystique of planetary exploration to our own world, creating theatrically-lit compositions with the aid of GPS-enabled drones. “Lux Noctis” is influenced by a confluence of 19th century romantic painting and science fiction which is expressed in the dramatic ways each drone lights the earthen subjects from above.
“My aim is to portray a unique perspective of the planet we live on by illuminating night landscapes with an aerial LED light,” said Wu. “Scenes which show not only the beauty of the landscape but also the versatility and awesomeness of adapting new technology to create art.”
Wu used a prototype AL250 light by Fiilex mounted on a 3DR Solo UAV. Typically known for their ability to capture visuals below rather than light them, Wu’s drones serve as flying light beams which circumvent expensive cranes or helicopters previously used to light scenes.
You can see more of Wu’s dramatically lit work on his Instagram and Facebook. (via PetaPixel, thnx John!)
Maciej Ceglowski’s Heroic and Lazy Stand Against IFTTT
HsiufanI'm most intrigued by the inside baseball knowledge of how IFTTT built their integrations.
Pinboard creator Maciej Ceglowski:
Because many of you rely on IFTTT, and because this email makes it sound like I’m the asshole, I feel I should explain myself.
In a nutshell:
IFTTT wants me to do their job for them for free
They have really squirrely terms of service
JSON document fast lookup with MySQL 5.7
In this blog post, we’ll discuss JSON document fast lookup with MySQL 5.7.
Recently I attended Morgan Tocker’s talk on MySQL 5.7 and JSON at FOSDEM, and I found it awesome.
I learned some great information from the talk. Let me share one of them here: a very useful trick if you plan to store JSON documents in your MySQL database and want to retrieve the documents from some attribute’s values. So let’s look at how to do JSON document fast lookup with MySQL 5.7!
In this short example, I show you how we can speed up this type of search using JSON functions and virtual columns.
This our test table:
Table: test_features Create Table: CREATE TABLE `test_features` ( `id` int(11) NOT NULL AUTO_INCREMENT, `feature` json NOT NULL, PRIMARY KEY (`id`) ) ENGINE=InnoDB AUTO_INCREMENT=206561 DEFAULT CHARSET=latin1 mysql> show table status like 'test_features'G *************************** 1. row *************************** Name: test_features Engine: InnoDB Version: 10 Row_format: Dynamic Rows: 171828 Avg_row_length: 1340 Data_length: 230326272 Max_data_length: 0 Index_length: 0 Data_free: 3145728 Auto_increment: 206561 Create_time: 2016-03-01 15:22:34 Update_time: 2016-03-01 15:23:20 Check_time: NULL Collation: latin1_swedish_ci Checksum: NULL Create_options: Comment:
We can see the data length is almost 230M:
+--------------------+--------+-------+-------+-------+------------+---------+ | TABLE | ENGINE | ROWS | DATA | IDX | TOTAL SIZE | IDXFRAC | +--------------------+--------+-------+-------+-------+------------+---------+ | json.test_features | InnoDB | 0.17M | 0.21G | 0.00G | 0.21G | 0.00 | +--------------------+--------+-------+-------+-------+------------+---------+ -rw-r----- 1 mysql mysql 228M Mar 1 15:23 /var/lib/mysql/json/test_features.ibd
As an example here is one record (the data is coming from https://github.com/zemirco/sf-city-lots-json):
{ "type": "Feature", "geometry": { "type": "Polygon", "coordinates": [ [ [ -122.41983177253881, 37.80720512387136, 0 ], ... [ -122.41983177253881, 37.80720512387136, 0 ] ] ] }, "properties": { "TO_ST": "600", "BLKLOT": "0010001", "STREET": "BEACH", "FROM_ST": "600", "LOT_NUM": "001", "ST_TYPE": "ST", "ODD_EVEN": "E", "BLOCK_NUM": "0010", "MAPBLKLOT": "0010001" } }
Now let’s try to find all records where the street is “BEACH”. “Street” is part of the array attribute properties.
mysql> SELECT count(*) FROM test_features WHERE feature->"$.properties.STREET" = 'BEACH'; +----------+ | count(*) | +----------+ | 208 | +----------+ 1 row in set (0.21 sec) mysql> explain SELECT count(*) FROM test_features WHERE feature->"$.properties.STREET" = 'BEACH'G *************************** 1. row *************************** id: 1 select_type: SIMPLE table: test_features partitions: NULL type: ALL possible_keys: NULL key: NULL key_len: NULL ref: NULL rows: 171828 filtered: 100.00 Extra: Using where 1 row in set, 1 warning (0.00 sec)
As you can see, we perform a full table scan to achieve this.
With MySQL, we have the possibility of using virtually generated columns. Let’s create one for the streets:
mysql> ALTER TABLE test_features ADD COLUMN street VARCHAR(30) GENERATED ALWAYS AS (json_unquote(json_extract(`feature`,'$.properties.STREET'))) VIRTUAL;
I use “json_unquote()” to avoid to add the JSON string quotes in the column, and later in the index.
You can verify the size of the table on disk, and you will see this doesn’t increase (as it’s a virtual column).
Even if we can now use the “street” column in the search, that won’t help. We still need to add an index on it:
mysql> ALTER TABLE test_features ADD KEY `street` (`street`);
And now we can see that the size is larger, because we have added the size of the index:
-rw-r----- 1 mysql mysql 232M Mar 1 15:48 /var/lib/mysql/json/test_features.ibd
Now we can try to run the query like this:
mysql> SELECT count(*) FROM test_features WHERE street = 'BEACH'; +----------+ | count(*) | +----------+ | 208 | +----------+ 1 row in set (0.00 sec)
Let’s have a look at the Query Execution Plan:
mysql> explain SELECT count(*) FROM test_features WHERE street = 'BEACH'G *************************** 1. row *************************** id: 1 select_type: SIMPLE table: test_features partitions: NULL type: ref possible_keys: street key: street key_len: 33 ref: const rows: 208 filtered: 100.00 Extra: Using index
And finally we can verify this in the statistics available in sys schema:
mysql> select * from sys.schema_index_statistics where table_name='test_features'G *************************** 1. row *************************** table_schema: json table_name: test_features index_name: street rows_selected: 208 select_latency: 72.59 us rows_inserted: 0 insert_latency: 0 ps rows_updated: 0 update_latency: 0 ps rows_deleted: 0 delete_latency: 0 ps *************************** 2. row *************************** table_schema: json table_name: test_features index_name: PRIMARY rows_selected: 0 select_latency: 0 ps rows_inserted: 0 insert_latency: 0 ps rows_updated: 0 update_latency: 0 ps rows_deleted: 0 delete_latency: 0 ps 2 rows in set (0.00 sec)
As you can see, this is very fast. If you already know how you want to retrieve data out of your JSON document, it’s very easy to add such indexes in MySQL.
Is What I See What I Get? — Math & Optics in Type Design
Every so often, I’m preoccupied by the question, “Is what I see also what I get?” I catch myself thinking that there’s a discrepancy between what I “saw” and the information my eye actually took in.
The differences between reality, what I see, and what I think I see are especially important to consider when dealing with type. A simple yet fascinating example of a “misperceived” reality can be found in the optical perception of horizontal and vertical bars of the same width. In Detail in Typography, Jost Hochuli states:
For a given weight of line, a horizontal line appears heavier than a vertical line. To achieve optically balanced verticals and horizontals, which appear to be of the same weight, the horizontal must be somewhat narrower. This applies not only to straight lines but also to curves.1
It may come as a surprise that most people can only perceive this phenomenon when they are made overtly aware of the discrepancy. Helvetica’s ‘H’ and ‘S’ offer another example. The crossbar of the ‘H’ is set slightly higher than the center in order for it to appear to be in the middle of the letter. Similarly, the lower curve of the ‘S’ is slightly larger than the upper one, so that the two appear equal in size. If you rotate both characters 180 degrees, the differences become apparent immediately.
By rotating glyphs, such as these from Helvetica, one can immediately see the optical corrections that make top and bottom halves appear equal.
In both of these optically corrected examples, observers do not “see” reality. Instead, they perceive something close to symmetrical letterforms. Observers will most likely only notice the discrepancy if it is measured; when an immediate comparison is given; or, as in the example of the letters ‘H’ and ‘S’ above, with the help of inversion. Otherwise, they are convinced that what they see is mathematically correct. Even if you have many years of experience with optical illusions in the context of type design, you may still stumble if a perceived reality is challenged. In his book Kursiv, Hendrik Weber notes a great example. When Gerard Unger produced his typeface Hollander Kursiv, he slanted the individual letterforms at different angles so they would appear parallel.2
Even if you are aware of this fact, it is impossible to see through this optical illusion without helper lines that trace the degrees of incline. The discrepancy between the angles can be demonstrated easily using Adobe Garamond with the addition of such helper lines.
There are many examples of this kind that are either consciously or unconsciously manipulated by designers, especially for fonts that are intended for body text. For that reason, it’s always wise to subjugate the mathematical reality to the “seen” or “optical” reality.
Consider the following situation: a painting is to be hung on a wall. Unfortunately, the floor and the ceiling are not parallel to each other. A discussion starts: should the painting be hung with the aid of a tape measure or the naked eye? If it is hung using a tape measure, one may feel satisfied that it is hung “correctly”. However, the painting will most definitely appear crooked. If it is hung using only the naked eye, it can be hard to shake the feeling that it is “off ” somehow. Luckily, this is only a dilemma for the people hanging the painting. Unknowing observers are unlikely to notice that anything is off. The painting will hang “true” in their perceived reality. This is exactly the intent of fonts created for reading, and not for “being seen”. New technologies have made it easier than ever to create fonts that are mathematically precise, tempting people to trust the “tape measure” rather than the eye. It takes courage and knowledge to recognize that the perceived exactitude of the eye is more accurate than the measured “truth”.
The human eye is an organ that has evolved over millennia — to a level of perfection that even Charles Darwin, despite his theory of evolution, had trouble coming to terms with: “To suppose that the eye, with all its inimitable contrivances … could have been formed by natural selection, seems, I freely confess, absurd in the highest possible degree.”3 Vision is one of the most powerful senses we possess. We should learn to rely on the eye, even though the tape measure may offer a different opinion.
Sources:
This essay was originally published in m0ire, Issue 3. Translation from the German by Matthias Suchert and original copyediting by Lauren D. Weinberg.
Making the measured truth a visual one is the alpha and omega of type design. Some other well-illustrated meditations on this include Ian Moore’s “Making Geometric Type Work” and Tobias Frere-Jones’ recent series, “Typeface Mechanics”.
Use rems for global sizing, use ems for local sizing
The following is an unedited extract from my [forthcoming book](http://book.webtypography.net/). I chose this chapter because I felt it would be useful standing on its own. I should point out that this is a very technically-focussed extract, while the other chapters I’ve written contain a much higher proportion of typographic theory.
## The amazing em (and friends)
The web has always been a type-based medium. Cascading Style Sheets joined HTML as a part of the web in 1996, and did so with typographic foundations from their inception. The principal underpinning to those foundations is the `em` unit.
The beauty of the `em` in CSS is that it provides a unit of length which relates directly to font size. Ems remain fundamental to modern web design. As you will see throughout this book, ems enable you to design truly scalable web pages, which is why we are introducing you to them before we go any further.
Ems have a long-standing tradition in typography, where they have typically been used for horizontal measurements. Ems are named after the letter ‘M’ (and pronounced as you would the letter), but are not directly related to the character itself. It is more likely the em was derived from the width of the metal ‘sort’ which held a capital M in moveable type printing presses.
In CSS, the `em` is a general unit of length related directly to font size. We use it to define widths, heights, padding, margins and other measurements in both horizontal and vertical directions – something typographers with a traditional background may find surprising. The `em` unit can also be used to define the font size itself.
The relationship between ems and font size is the same on the web as it is in traditional printed typography. It’s very simple: one `em` is a distance equal to the font size. So if the font size of an element (``, `
`, `
`, etc) is 16 pixels then one `em` is equivalent to 16 pixels; in an element with 36 pixel type one `em` equals 36 pixels; and for 120 pixel type one `em` is 120 pixels.
Just as in traditional typography, a crucial aspect of the relationship between ems and font size is that it is independent of the font’s design. One em equals the font size whether the typeface is an elaborate calligraphic script, Japanese kanji, or a plain sans-serif. To illustrate the relationship between length and font size, consider these styles:
#box1 {
font-size: 36px;
width: 1em;
height: 1em;
}
#box2 {
font-size: 120px; width: 1em; height: 1em;}
These styles will render like:
and
Both boxes have a height and width of 1 em but because they have different font sizes, one box is bigger than the other. Box 1 has a font-size of 36 px so its width and height is also 36 px; the text of box 2 is set to 120 px and so its width and height are 120 px.
The fundamental advantage of using ems is that distances, lengths and spaces will all scale proportionately with text size. If your reader adjusts their default text size to better suit their requirements, everything sized in ems adjusts itself accordingly. This effect is particularly useful when relationships between elements on page are tied to type size, for example margins and padding.
As we saw in the prior examples, lengths set in ems are relative to the font size of the selected element. However when setting `font-size` in ems, we do so relative to the inherited font-size, that is to say the font-size of the element’s parent. Consider the following mark-up:
…
…
with the following styles applied:
#d1 { font-size: 16px }
#d2 { font-size: 32px }
p { font-size: 1em }
These will render as:
Even though both paragraphs have `font-size:1em` they display different sized text because their parent elements have text sized differently.
## Rem units
The `rem` unit was introduced to CSS3 in 2005. It is very like the `em` in that it lets you set lengths relative to font size. The key difference is that rems are not relative to a selected element’s font size, they are always relative to the `` element. By extension, this means that rems are always directly related to the browser’s default text size, which can be adjusted by the reader. This gives us typographers additional precision and ease of use, while still ceding ultimate control to our reader.
If a browser’s default text size is 16px then 1 rem is always 16px regardless of where a selected element might sit in the page or what its context might be. If your reader changes their default text size to 21px then 1 rem will always be 21px. Let’s take our previous example, and set the paragraph font size in rems instead of ems:
#d1 { font-size: 16px }
#d2 { font-size: 32px }
p { font-size: 1rem }
These will render as:
### Use rems for global sizing, use ems for local sizing
In rems and ems we have two extremely useful and versatile units which enable us to relate lengths, distances and dimensions to font size. However it is not always obvious when it’s better to use one rather than the other. We’ll be using both rems and ems throughout this book, so hopefully the advantages of each unit will become clear, but as a guideline, use rems to scale something with the page (global sizing) and use ems to scale within a component (local sizing).
Take the example of a pull quote containing two short paragraphs. The spacing between the paragraphs will depend on the size of the paragraph text, so you could feasibly set that spacing in either rems or ems. Should you decide during your design process that the pull quote should be set a bit larger, then the spacing between the paragraphs must also be increased. Therefore the paragraph spacing is directly related to the paragraph text size and therefore be considered local sizing within a component. Hence you should use ems in this case.
As for the paragraph text itself, that does not have any direct relation to another part of the page. It is only related directly to the default text size. Therefore the paragraph text size it can be considered global sizing and thus should be set in rems rather than ems. We will be considering text sizing in much more detail later on.
## Ch units
The `ch` unit was introduced in CSS 3 in 2006. It is equivalent to the width of a character, hence ‘ch’. More specifically 1 ch equals the width of the zero (`0`) character in the current font. This means that, unlike ems and rems which do not change with different type designs, a `ch` changes as the font-family changes. In the cases where it is impossible or impractical for browsers to determine the width of a ‘`0`’ (perhaps because the font doesn’t include a ‘`0`’), browsers will set 1ch to be equal to 0.5em.
The `ch` unit can be useful if you want to set lengths or sizes which relate directly to the width of the font. For example you might want to set the width of a block of text to be wider for an expanded font, and narrower for a condensed font. Using `ch` could achieve this automatically for you, for example the following two text blocks both have a width of `34ch` but because they use fonts of different widths, the blocks are different sizes.
CSS includes many more units of length: you may have noticed we’ve omitted to mention pixels at this point, and there are plenty of others we haven’t yet touched upon. `Em`, `rem` and to a lesser extent their cousin `ch` provide us will the primary tools for designing and developing websites which will bend, squash, stretch and most importantly adapt to the whims of our readers and their web-enabled devices.
Gilt.com – When entering your email address in the sign up form,...
Gilt.com – When entering your email address in the sign up form, the input field auto completes common domains.
Shipwrecks and Deep Ocean Scenes Encapsulated Inside Translucent Whale Sculptures
Image provided by Isana Yamada
Image provided by Isana Yamada
Japanese artist Isana Yamada' s project Samsara is composed of six translucent whales mounted on thin pedestals that give each of the sculptures an illusion of movement. The whales, illuminated from within, provide a window to strange worlds locked inside their resin-coated bodies: churning submarine volcanoes, fluffy white clouds, and even polar bear skeletons that float within. The project, staged at Tokyo University of the Arts, references the circle of existence found in Buddhist traditions with each whale displaying a separate scene. The whale that represents the human dimension contains a sunken sailboat, imagery that symbolizes a difficult voyage or plight.
Yamada’s work will also be shown in an exhibition of sculptural works at the Artcomplex Center of Tokyo from March 1st through 6th. You can see more of his work on his Facebook page here. (via My Modern Met)
Images by @muzintansaki
Image provided by Isana Yamada
Image provided by Isana Yamada
Image provided by Isana Yamada
This Photograph of the NYC Winter Storm Looks Like an Impressionist Painting
Photo © Michele Palazzo. Jonas Blizzard in New York, 2016.
While walking through the Jonas Winter Storm that swept across the East Coast last week, photographer Michele Palazzo captured this incredible shot of the Flatiron Building against a backdrop of swirling snow. With the exception of a few minor details like logos and a food cart, the image looks like an impressionist painting right out of another another century. The cloudy atmosphere and gusty winds create patterns that appear uncannily like brush strokes. You can see more of Palazzo’s shots from the morning of January 23rd on EyeEm.
Update: The folks over at EyeEm have a nice roundup of Jonas storm photos.
Update 2: The snowflakes you see in the window of the Flatiron are part of a paper origami installation by Chelsea Hrynick Browne in the Prow Artspace.
The First Round Capital Holiday Train Wreck
Your annual reminder of the hidden costs of taking venture capital is here — it's the First Round Capital Holiday Video, a yearly cringe-fest of startups parodying the year's biggest pop hits, with lyrics tweaked to reflect the worst of startup culture. (Full lyrics at the end of the post.)
I have a grim fascination with these videos and their ever-increasing production budgets. Every year, I watch them with my hands shielding my eyes, and collect them in a YouTube playlist. (For some reason, 2009's video is only on Vimeo.)
Of course, if you ask First Round about it, and probably most of the founders in the video, they'll say, it's all in fun! We're just blowing off some steam at the end of the year! I know one of the partners at First Round. I know some of the people at the startups in the videos. I don't think their intentions are bad.
But once these startups have taken funding, do they really have a choice?
This year, Crunchbase says First Round invested in 57 startups, a median amount of $8.5M and an average of $18.5 million.
If someone gives you $8.5 million, sits on your board, and owns a significant part of your company, you're going to dance if they say "dance." You're going to sing if they say "sing."
And you'll ask your entire team to do it too, and they'll be captured on video, streamable on YouTube for years after they quit or were laid off.
In a situation like that, you can't really say no. You just put on the costume, smile, and dance.
Continue reading...Famous YouTube stars are barely scraping by
Gaby Dunn, co-star of the YouTube channel Just Between Us, explains the dismal economics of being a mid-list YouTube star.
I’m 27 years old and have been building an online following for 10 years, beginning with a popular Livejournal I wrote in high school. A couple of years ago, after moving to Los Angeles, I made the transition from freelance writing to creating online video. The channel I have with my best friend Allison Raskin, Just Between Us, has more than half a million subscribers and a hungry fan base. We’re a two-person video creation machine. When we’re not producing and starring in a comedy sketch and advice show, we’re writing the episodes, dealing with business contracts and deals, and running our company Gallison, LLC, which we registered officially about a month ago.
And yet, despite this success, we’re just barely scraping by. Allison and I make money from ads that play before our videos, freelance writing and acting gigs, and brand deals on YouTube and Instagram. But it’s not enough to live, and its influx is unpredictable. Our channel exists in that YouTube no-man’s-land: Brands think we’re too small to sponsor, but fans think we’re too big for donations. I’ve never had more than a couple thousand dollars in my bank account at once. My Instagram account has 340,000 followers, but I’ve never made $340,000 in my life collectively.
The high highs and low lows leave me reeling. One week, I was stopped for photos six times while perusing comic books in downtown L.A. The next week, I sat faceless in a room of 40 people vying for a menial courier job. I’ve walked a red carpet with $80 in my bank account. Popular YouTube musician Meghan Tonjes said she performed on Vidcon’s MainStage this year to screaming, crying fans without knowing whether she’d be able to afford groceries.
Touch Arcade on Apple TV's App Store and the state of iOS gaming
Anita Sarkeesian on harassment through impersonation
HsiufanThis is nuts.
A Japanese Ramen Shop Now Has a Michelin Star
This is what Michelin-starred ramen looks like.
Michelin has unveiled its 2016 guide to Tokyo restaurants, and a ramen spot has joined the tire company's coveted ranks of the city's high-end sushi haunts and kaiseki temples. Sugamo's popular Japanese Soba Noodles Tsuta, considered one of the best shops in the city, was awarded a single star this year for its abbreviated menu of ramen or tsukemen in shio soup or shoyu soup, both made with a chicken and seafood stock featuring kelp, clams, and pike. (On Tuesdays, the restaurants transforms into Miso No Jin and serves miso ramen.)
Though this is the first time a ramen specialist has been given a star in any of Michelin's guides, the inclusion of a cheap, no-frills noodle restaurant isn't unprecedented: Tsuta's star parallels that of Hong Kong hole-in-the-wall Tim Ho Wan, the dim sum house that was billed as the world's cheapest Michelin-starred restaurant when it was first inducted. Additionally, earlier this year, Michelin included street vendors in its Hong Kong guide for the first time.
[via Reuters]
Read more posts by Chris Crowley
Filed Under: michelin, japan, lists, michelin guide 2016, tokyo, tsuta
The Crooked Forest: A Mysterious Grove of 400 Oddly Bent Pine Trees in Poland
This stand of bent pine trees known as the Crooked Forest is easily one of the strangest places in Central Europe. Located outside of Nowe Czarnowo, West Pomerania, Poland, the nearly 400 trees are widely agreed to have been shaped by human hands sometime in the 1930s, but for what purposes is still up for debate. Each tree is bent near the base at 90 degrees, a form that could possibly be helpful in boat or furniture making. Strangely enough, every tree is bent in exactly the same direction: due North. A quick search online reveals a host of conspiracy theories ranging from witchcraft to energy fields.
Whatever the reason, we’re glad photographer Kilian Schönberger (previously) stopped by to capture these photos. You can see more from the series on Behance.
Update: Thank you all for your many, many suggestions about the trees. We’ve heard everything from floods to furniture to fire. There still doesn’t seem to be a consensus.
How anime avatars became a warning
…for practical purposes, if you are besieged by trolls and are also okay with blocking people who might be extremely intelligent and engaging and also fervent otaku, all you need to know is that the “anime avatar” is a mostly though not entirely reliable indicator of trolldom, and block anyone you see with one. But I would suggest you not.
See also this bit 'o fake news from 2014: Study: Anime Avatar Makes Opinions Up To 90% Less Relevant
Robin Sloan on the dystopian vibes of Uber-for-food startups
Google Inbox – When snoozing a webshop order confirmation you...
Google Inbox – When snoozing a webshop order confirmation you can snooze it untill the day of delivery.
Segment: Rebuilding Our Infrastructure with Docker, ECS, and Terraform
This is a guest repost from Calvin French-Owen, CTO/Co-Founder of Segment.
In Segment’s early days, our infrastructure was pretty hacked together. We provisioned instances through the AWS UI, had a graveyard of unused AMIs, and configuration was implemented three different ways.
As the business started taking off, we grew the size of the eng team and the complexity of our architecture. But working with production was still limited to a handful of folks who knew the arcane gotchas. We’d been improving the process incrementally, but we needed to give our infrastructure a deeper overhaul to keep moving quickly.
So a few months ago, we sat down and asked ourselves: “What would an infrastructure setup look like if we designed it today?”
Over the course of 10 weeks, we completely re-worked our infrastructure. We retired nearly every single instance and old config, moved our services to run in Docker containers, and switched over to use fresh AWS accounts.
We spent a lot of time thinking about how we could make a production setup that’s auditable, simple, and easy to use–while still allowing for the flexibility to scale and grow.
Here’s our solution.
Separate AWS Accounts
Samurai Slices A 100 Mph Baseball in Half
Master swordsman Isao Machii is back! This time he’s here to cut a fastball in two.
Apple watch - the time goes from proportional font to monospaced...
Apple watch - the time goes from proportional font to monospaced font when you enter time travel, to stop it jumping around.
Historic flooding in South Carolina
Flooding around Aberdeen Country Club, on Oct. 6 in Longs. S.C.. The Carolinas saw sunshine Tuesday after days of inundation, but it could take weeks to recover from being pummeled by a historic rainstorm that caused widespread flooding and multiple deaths. (Janet Blackmon Morgan/The Sun News via AP)
Unmasking a pseudonymous Twitter user with machine learning
Off-Topic Tuesday -
It's "Off-Topic Tuesday" time!
In the comments to these weekly posts (and only these posts), it's your chance to go as off topic as you like.
Talk about non-comics stuff, thread derail, and just generally chat amongst yourselves.
- The passing of Jack Larson, TV's first Jimmy Olsen.
- The UK Government slapping a Carbon Tax on renewable energy methods which don't generate carbon.
Let the games begin!
comments