Linen Pixel Designs

My cross stitch designs, and occasional other things I come across.

[hekateras]

Prism, from Patterns In My Tea by Linenpixel

Instagram | Twitter | Artstation | Buzzly

[hekateras]

Invisible Inc - Toothless

Fic illustration :D

[arctic-hands]

So I saw a post last night that terrified me but then my app crashed before I could reblog it.

Smart appliances are completely, 100% reliant on wifi? Like if there's an outage or you couldn't pay your internet bill in time and your wifi isn't working, your stove/fridge/door locks etc are rendered completely and totally unusable? The oven doesn't revert back to a basic oven/stove that you just have to operate manually? It's competely inoperable? Is that what you're telling me because that's fucking terrifying. So you're either completely locked in or out of your home if wifi goes down? Who the fuck signs up for this shit?

[thecringeandwincefactory-deacti]

I've read soooo many articles by people in the security industry who are like ALL of my locks at home are manual, always will be. They completely reject this 5G bullshit.

Also, Jesus, if Amazon is cooperating w the police vis-a-vis the doorbell camera bullshit, imagine how they'll cooperate with the police in 20 years by turning off your access to water because you getting shook down to for a donation to a local police charity.

[dirtydirtychai]

P l e a s e stop with the internet of things. Your fridge doesn’t need to be connected to WiFi. Stop inviting unnecessary tools of surveillance into your home.

[systlin]

Listen, I am a Professional Security Person.

And if anyone even mentions any of this smart home bullshit near me I hiss like an offended cat.

A good old fashioned manual lock and deadbolt is the way to go, and I don’t trust my fridge to be talking to my doctor or my washing machine to have internet access, because I promise you that shit will go sideways immediately.

You know who can compromise your wifi enabled baby monitor or security cameras and watch the camera feed of your house? Literally anybody with 10 spare minutes and some freeware.

[nomadicism]

Imagine the possibilities of a hacking a WiFi-enabled gas-powered tankless water heater and telling it to open the valve without igniting the pilot light. This is how the revolution begins.

[afishtrap]

As someone who works in cloud technologies, it’s not the connection to the cloud that gives me nightmares. It’s the storage on the cloud. 

Your rhoomba is a neat and efficient little machine because it stores all of its programmed paths on the cloud: every room it’s been in, the location of every piece of furniture, and most importantly, the hours it’s put to work. Enough data, and we have most (if not all) of your floorplan, the location of every major and minor piece of furniture, the occupants of your household, and a pretty good idea of your household’s usual routines. 

Your cloud-connected thermostat tells us when you’re home, when you wake up and go to sleep, when you’re gone each day, and when you’re away for vacation. If that thermostat also has occupant-sensitivity (ie, turns itself down when no movement is detected), that’s a wonderful datapoint for tracking your routine. And if the thermostat ‘learns’ your patterns enough to make suggestions, that’s because it’s stored everything in the cloud -- along with a million other peoples’ data -- to statistically infer your behavior.  

Your fridge can suggest recipes for the food you’ve stored, which means it’s sending a query into the cloud with the contents of your current fridge. If it’s predictive -- such as, telling you what you normally eat and suggesting a grocery list to resupply -- that means it’s also tracking what you eat, how often, how fast, and when you do most of your shopping. 

With your address, I can get publicly-available data on how much you paid for the house, its age, and its rough square footage. With the rhoomba, I can guess you go out and do things on your days off, that you have at least two animals, that you have a child of walking age, an eat-in kitchen and a formal dining room, hardwood floors, and at least three rugs of high enough quality that you’ve programmed the rhoomba to avoid munching on the rugs’ tassels. The thermostat tells me you’re early to bed and early to rise, that you keep the house relatively cold during hot summers and that your HVAC sometimes struggles to keep up. Your fridge tells me how much you spend, your entertainment patterns, and a fairly good idea of your tastes. 

I now know when, where, what, and how to hit you with advertisements for holiday entertaining, high-end furniture, home remodeling, appliance replacement -- and a fairly well-educated guess on your budget for each. There’s a billion other things I could see, deduce, and use from your collected data. Not every use would be as innocuous as advertisements, either. 

And yes, your collected data is available for purchase by corporations beyond the one who made a single appliance. Anyone who tells you otherwise is trying to sell you something.

Remember, Target was able to use its customer data to set up a predictive system that identified pregnant women before the women were even aware themselves --- and that was just sales data. Imagine what Target, and all its corporate brethren, can learn about you now that you’ve basically put your entire private life on the cloud, right there for the taking. 

[earthstory]

The PETM – Why It Feels Like Deja Vu

The Palaeocene Eocene Thermal Maximum was an event that took place 56 million years ago at the end of the Palaeocene and lasted for 150,000 – 200,000 years. The average global temperature rose by 5-9°C (9-16°F) leading to mass migrations and extinctions of flora and fauna.

The event was discovered when marine sediment cores from Antarctica showed a large and sudden excursion ( a sharp spike in the values) in carbon isotopes, indicating that a large amount of CO2 had been rapidly expelled into the atmosphere. Sounds familiar doesn’t it?

Keep reading

[linenpixel]

The Embroidered Computer uses historic gold embroidery materials to craft a programmable 8 bit computer.

[superlinguo]

When does maluma/takete fail? Two key failures and a meta-analysis suggest that phonology and phonotactics matter: New article in i-Perception

When I was an undergraduate student, one of my lecturers drew these two shapes on the board and asked us which one we would call takete, and which one we would call maluma:

The majority of the class thought that the rounded /u/ vowel, liquid /l/ and nasal /m/ all suited the rounded puff of cloud, while the aspirated voicelessness of /t/ and /k/ and the more forward vowel sound suited the spikier shape. We were a typical survey of participants. You might have participated in this kind of matching test. Often the words bouba and kiki are used instead, to similar effect.

While I was working in Singapore I had the good fortune to meet Suzy Styles, an Assistant Professor in the Department of Psychology at NTU. Suzy is a psycholinguist who is interested in how the brain creates these mappings across the senses (in this instance sounds and shapes), and what role language and its acquisition play in this. Suzy’s work also includes tone languages, such as Mandarin, and it was hearing about this work that made me wonder what speakers of Syuba, with its two tones, would make of a test like this.

Suzy and I decided to test this out. Actually, we had a whole lot of other questions we wanted to ask, but thought we should run the traditional maluma/takete test first just to get a baseline.

We didn’t expect it to fail.

There are very few failures reported for maluma/takete, and the main one (Rogers & Ross) was in the same journal as our paper, but way back in 1975. In that paper the participants all spoke Songe, a language of Papua New Guinea. In order to understand why the test failed, we looked at all the examples where the test worked. This paper is therefore the most up-to-date survey of maluma/takete tests done with ‘normal’ populations. First things first, there are actually very few papers reduplicating this effect, given how much cross-sensory literature appears to be built on it. Secondly, the groups this phenomenon has been tested with are waaaayyyy too boring - lots of English speakers, with some other major languages like French and Italian.

We put forward the argument that the failures possibly happened in Syuba and Songe because the words used have sounds that those languages don’t have, or wouldn’t use in that order. In Syuba kiki and bouba don’t fit the pattern of words. It would be like someone asking you to do the test with words like lrisg or ngoopr. We didn’t even get onto looking how tone comes into play (that’s a story for another paper!).

This test needs to be done in more languages.

As usual, too much of what is assumed to be true is based on a small number of the world’s languages. We want to see this type of test run in many more of the world’s languages. We have put all of our materials and methods up in an Open Science Framework repository for others to use. If you’re doing fieldwork you can use our sounds and shape templates (you can even 3D print them!) and help to broaden the range of languages we have results for. There are also a few more tests in the kit that are the foundation for some upcoming work we’re now writing. If you do have a failed maluma/takete test that you left languishing in a drawer, we invite you to ‘bring out your dead’ and help us figure out what is really going on with these sounds and shapes.

Abstract

Eighty seven years ago Köhler reported (1929:1947) that the majority of students picked the same answer in a quiz: Which novel word-form (‘maluma’ or ‘takete’) went best with which abstract line drawing (one curved, one angular). Others have consistently shown the effect in a variety of contexts, with only one reported failure (Rogers & Ross, 1975). In the spirit of transparency, we report our own failure in the same journal. In our study, speakers of Syuba, from the Himalaya in Nepal, do not show a preference when matching word- forms ‘kiki’ and ‘bubu’ to spiky versus curvy shapes. We conducted a meta- analysis of previous studies to investigate the relationship between pseudoword legality and task effects. Our combined analyses suggest a common source for both of the failures: ‘wordiness’ – We believe these tests fail when the test words do not to behave according to the sound structure of the target language.

See also

Brain, Language and Intersensory Perception (BLIP) lab at NTU

Chart of the most common speech sounds across languages

The Open Science Framework repository for the paper

Reference

Styles, Suzy & Lauren Gawne. 2017. When does maluma/takete fail? Two key failures and a meta-analysis suggest that phonology and phonotactics matter. i-Perception 8(4). 1-17. [Open Access online article]

[linenpixel]

Locating your convention centre/stadium’s coat check on the far side of a labyrinthine partially-submerged trail though a marsh is not a good idea. Admittedly this is not likely to happen anywhere outside of the dream I had last night, but just for the record.

[linenpixel]

Mark Watches “A Single Pale Rose”

I took more screenshots than this, but decided to just go with this one.

[linenpixel]

The Woman Who Gave the Macintosh a Smile

Susan Kare designed the suite of icons that made the Macintosh revolutionary—a computer that you could communicate with in pictures.

[linenpixel]

And this one is just extra.

[linenpixel]

Though actually I’ve used most of the decorating slots, and I don’t usually play with the one in the last post. This one is for people who like eating in a greenhouse, or possibly a garden centre.

[linenpixel]

My surrealist cyberpunk Cook Serve Delicious 2 restaurant.

(Hey, it fits the game’s setting, if you read the emails.)

To be clear, that white cloud is painted on the window.

[linenpixel]

Central: Wait, so my younger self went into a mission without any items because I’d just been rescued, was told to stick close to another agent, went off in my own direction anyway, acquired an EMP pack and an accelerator chip from two separate safes, and then, by the time N-Umi had caught up to me, had snuck past a stationary guard into the nanofab room and bought myself a better neural disruptor than the ones the other agents had? Yeah, that sounds completely in-character.

(Okay, fine, Central does seem to emphasize caution, but maybe she was more reckless when she was younger.)

(Just ignore the timeline issues.)

[linenpixel]

Continued from here.

On the second level, for just about everything after the executive suite I was just exploring the level (using a lot of rewinds) to use as a setting and then making up the action afterwards.

Though as you can see in the second-to-last screenshot, I did get Decker into that situation with the hologram projector.

And I changed the layout slightly in my descriptions in the story, so that I didn’t have to try to explain that pointless and implausible laser barrier.

Xu didn’t actually have a flash grenade; that final guard wasn’t actually a flack guard ... 

Also I didn’t even have Prism there, though I did deliberately get Shalem shot at one point as a stand-in for Prism to see how the guards would react.

But the executive having a medgel in his pocket was genuinely something that happened in-game.

[linenpixel]

Screenshots from the playthrough (set up deliberately with custom settings) I used in this fanfic.

Putting this on my main blog rather than invisiblecaptions because I’ve decided my AO3 fanfic is in a separate continuity.

On this first level, the action in the story corresponds almost exactly to how I was actually playing, except that some of Xu’s positioning is different (and if you look closely, Xu has a tag pistol because I forgot to get rid of it.)

Second level here.

[allthingslinguistic]

Roses are red  Violets are blue  Ideas are green and colourless 

[allthingslinguistic]

More #linguentines on twitter. Check them out and/or contribute! 

[linenpixel]

Mark Watches “Jungle Moon”.