such_heights: amy, eleven, rory and river having a picnic (who: ponds)
[personal profile] such_heights
Well, two years after I started this thing, think I'm finally ready to post it. Here goes!

The Aetherion
Doctor Who ; PG-13 ; Amy/Rory, Eleven/River ; 45k
After the Doctor is kidnapped, Amy, Rory and River embark on a long journey to find the only thing that can save him - a mysterious artefact called the Aetherion.

Content notes: deals with depression and recovery from trauma.

Many thanks to [personal profile] purplefringe for a heroic amount of handholding, and to [personal profile] pocketmouse for the beta.

Read at the AO3
wiscon_posts: (Default)
[personal profile] wiscon_posts posting in [community profile] wiscon
As one of the concerns brought up regarding the Frenkel decision was that all available reports were not considered, the subcommittee convened to consider Rose Lemberg's report would like to ask if anyone has additional information about the events publicly discussed by Rose Lemberg [http://rose-lemberg.livejournal.com/354919.html] and Saira Ali and Alex Daly MacFarlane [http://sairaali.livejournal.com/201577.html] that they would like to make WisCon aware of at this time.

We do have Rose Lemberg's report, Saira Ali and Alex Dally MacFarlane's open letter to WisCon, and WisCon's other correspondence with Rose Lemberg, FJ Bergmann, Saira Ali, Alex Dally MacFarlane, and other witnesses. If you have additional statements or information, please email it to safety@wiscon.info by Monday, October 6.

Thank you.

Albert Einstein, Anti-Racist Activist

Sep. 22nd, 2014 06:05 pm
[syndicated profile] thisaintliving_feed

Posted by s.e. smith

Usually, when I delve deeper into the lives of famous figures and days gone past, I end up surfacing with a dirty, uneasy feeling. There’s Margaret Sanger and her support of eugenics, for example, a huge disappointment for me when I realised that the woman I’d heralded as an icon and important figure in the early push for reproductive freedom was involved in the movement for selfish, racist, classist, ableist reasons. She’s not the only historical figure who played a key role in social justice movements but who didn’t necessarily have the most enlightened approach to the people she was claiming to help.

There’s something about seeing our icons fall that’s always upsetting, even though we should know better. And that’s why I’m always tremendously excited to find out something good about someone, in a way that’s almost pathetic: Some indicator that there is, in fact, a shred of hope for humanity and some people did the right thing even if it was difficult. Evidence of goodness also stands as proof, for me, that ‘well, it was just like that at the time’ or ‘everyone thought that way’ are false arguments, because some people didn’t think ‘that’ way and were fighting to change society.

When we say that people with offensive views and ideas were just products of their time, we neatly ignore the people who were alive and working during the same time against the very ideas propagated by the people who ‘couldn’t help themselves.’ One such person was Albert Einstein, who wasn’t just a noted pacifist, but a passionate anti-racist. He identified racism as a white problem, and noted that the white community needed to be responsible for fixing it.

He made anti-racist speeches, visited historically Black schools and colleges, supported local communities of colour, and joined forces with Paul Robeson to fight fascism, segregation, and lynchings. Throughout his lengthy career, he was deeply involved with racial issues, and this is something contemporary media rarely covered, and something I knew nothing about until it flickered across my radar on Twitter, as these things so often do. There is, in fact, an entire book looking at Einstein’s relationship with race and America’s racial problems, which underscores the huge body of material the writers had to work with.

It is perhaps not surprising that Einstein’s contributions to anti-racism were erased at the time. It was easy to focus on the media-friendly physicist who amazed people with his mind, and to quietly skate around details of his personal life. His work can’t have made contemporary media comfortable, either, as he was unafraid when it came to specifically confronting white complicity and talking about what whites needed to do. In an era when fascism was on the rise in the US, eugenics was culturally accepted, and the nation was still firmly enforcing Jim Crow, segregation, and other outright racism, Einstein stood out as a banner in the cultural landscape.

That made him all the braver than modern-day white anti-racists, who are simply building on the groundwork that other people have established (often without credit). Einstein worked side by side with the Black community, and he spoke openly to people of all races about his desire for an end to racism and racialised violence — but he was careful not to pin the blame on communities of colour. Instead of positioning himself as a charitable do-gooder helping to bridge the gaps or bring people of colour into modern society, he saw himself much as the opposite, it would seem: He spoke as a man who wanted to lead white people to the path of behaving responsibly and with respect towards their fellow human beings.

Yet, this can’t have made his media followers very comfortable at the time, as he was challenging the status quo in a truly radical way, along with other early white anti-racists who were ready to take responsibility for their community. As the years progressed, therefore, this part of his history was steadily and quietly erased, making it harder and harder to find information about it. It’s a great pity, because even as Einstein was revered and beloved for his work in physics, he should by right also have been recognised as a civil rights icon, and should have been a figure of much interest to people working on and discussing early civil rights pushes.

Today, the fact that we still don’t know about this fascinating and utterly engaging aspect of Einstein’s personality and life is nothing short of shameful, and it’s an indictment of our own society. Decades later, we still can’t face up to our race problems as a society, and that means that we can’t honour the contributions of early civil rights crusaders. To admit that Einstein was involved in anti-racist work, after all, is to admit that this is an issue which has been going on for decades, and that far from being a radical and new thing, civil rights and equal access to society have always been important causes — important to abolitionists in the 19th century, important to Einstein and people like him in the early 20th, important today as racism becomes slippery, sliding between the cracks and leaving its stain on society.

The true story of Einstein makes me wonder how many other fascinating stories I’m missing, how many other people in different eras were fighting the good fight without recognition. Trailblazers are always there, speaking up and pushing for change before the rest of society is ready to acknowledge them, and that means that all of us are accountable for our own bigoted views: We can’t claim to be products of our time, not when the people of our time are actively rejecting the busted ways of thinking we carry with us.

Image: Albert Einstein painted portrait, thierry ehrmann, Flickr.

turlough: Mikey & Gerard Way & Ray Toro & Frank Iero in van, c. 2003 ((mcr) we were so young)
[personal profile] turlough
Brian was quiet for a minute, watching Bob finish patching up his hand. "What are we going to do with them?" he asked. "I mean..." he flailed his other arm around, "they're kids. What do you even do with kids? What if they don't get fixed? I got off the road for them! I paid for this tour out of my own pocket! If they're stuck like this..."

Bob clamped a hand over his mouth. "Did you not hear the 'shut up'? Just... go to sleep, ok? Maybe they'll be back to normal in the morning."

"And if they're not?"

"If they're not, we'll deal with it then." Bob pushed him back onto the bed, then got up to turn off the light. Brian listened to him shuffling around in the dark, then crawling into the other bed.

"You know," Bob said a few minutes later into the silence, "when I offered to come to Europe with you to help babysit the band, this wasn't exactly what I had in mind."


- [archiveofourown.org profile] lordessrenegade's Five Tired Boys and a Broken Down Manager
umadoshi: (to-do list (totaldevotion))
[personal profile] umadoshi
Posting in hopes of getting at least some of these out of my head!

--It probably says something that when I found the bedroom fan in the bathtub this morning, my first thought was "How the hell did the kittens get that in there?" (What it says is probably that I wasn't awake yet, but I'm more entertained by the thought that it says how highly I think of my kittens' capabilities.)

--When I hit a certain point in my rewrite script, I can stop work for the day and get back to rereading Untold. Yes, I'm aware of how sad it is that my work is going slowly because I'm too distracted by thoughts of the thing I want to do and can't until I get my work done. *facepalm*

--I now have beta notes back on two fics from [personal profile] wildpear (first notes on one, second notes on the other), and I wonder what the odds are that I can get one of them posted before going back to the office on Thursday.

--It would be really good if the BPAL Weenie update went up tonight. (It's now late enough that I'm worried something's actually wrong with Beth or her family, even though I expect that if something were that wrong, she or someone on staff would have mentioned it by now.)

--I'm thinking of growing my hair out for a while, which in practice would likely mean "until I get sick of it", since I've essentially given up on the idea of ever having actual long hair again. (Wow, it's been over ten years since I cut it off.) Being friends with the Elf has transformed my notion of "long hair" to the point that I no longer feel my hair's maximum growth length really counts, anyway. (Hers is down to her knees. Mine maxes out at something like elbow length.)

----->I miss being able to put my hair up, despite knowing that when it's long enough to do so I mostly get frustrated by my inability to do anything interesting with it. ("Put it up" usually means "twist it off my neck and into a clip. Messily.")

----->I want to be able to put BPAL in it. At chin length it's long enough that I can do that, of course (at any length it's long enough, really), but I like pulling it back and then having all the lovely smell around my head when I take my hair down for bed.

----->It's been two years now since I buzz-cut my hair. O_o And the haircut I had before that was a pixie cut, albeit briefly, so I can barely remember how long it's been since the last time I had my hair longer than a bob.

--I remember having notions of catching up on Agents of SHIELD (read: seeing past the series pilot) before going back to Casual Job. My past self is hilarious.

--I don't plan to make any Yuletide nominations (IIRC you can even if you don't intend to participate officially, but I feel weird about the thought of doing so), but I'm watching the unofficial spreadsheet with interest. And hoping Newsflesh gets nominated again; surely it will? *fidgets* October Daye has been nominated already, which is all to the good. ^_^

Because I care: Project Persephone!

Sep. 22nd, 2014 01:02 pm
james_davis_nicoll: (Default)
[personal profile] james_davis_nicoll
This project has direct relevance to the challenges of the 21-century where our megacities & urban environments will grow at astonishing rates. Yet the building industry, utilities and energy companies necessarily lag behind the physical demands of a growing city and where inflexible infrastructures become inadequate or inappropriate then urban decay sets in with crime, homelessness, waste & resource management issues, traffic congestion etc.


IN THE FACE OF ETC ONLY A STARSHIP WILL SAVE US!

And the nice thing is they can recycle images from old L5 articles.
james_davis_nicoll: (Default)
[personal profile] james_davis_nicoll
Because I will be dealing with two works I will describe as "tomes".
[syndicated profile] icanhascheezburger_feed
It's National Sea Otter Day, Let's Have A Caption Contest!

We all know sea otters are awesome, but did you know they're helping to fight climate change? The good people at Take Part have the scoop!

In celebration of these hand holding eco-heroes, we want you to caption this photo! Click and use our builder to add your best caption before 12pm PT Friday, 9/26. The top entries will be published by Saturday morning for you to vote and decide whose was the best!

Even if you don't enter your own submission, click here to see all the captioned submissions.

By the way, if your more of a fan of captioning marsupials, stay tuned for another contest this Wednesday (9/24).





Submitted by: (via Getty Images)

(no subject)

Sep. 22nd, 2014 12:26 pm
synecdochic: speech bubble: "You are a TERRIBLE PATIENT." (terrible patient)
[personal profile] synecdochic
I have reached the stage in my post-surgical recovery when I am clearly healed just enough for my Houdini instincts to kick in. Three days running I woke up to find I had woken up and not only unwrapped the Ace bandages being used for elbow compression, but rerolled them as I went. (In the proper direction, even -- they're Velcro-close, so you have to wrap them correctly as you take them off.) I've been leaving the arm unwrapped during the day a lot of the time because my skin is so irritated from being all wrapped up and I'm sick of my hand swelling up so much, but I really didn't want to just skip the compression completely, so I switched back to sleeping in the light compression sleeve I was using during the day pre-surgery.

I woke up at one point last night to pee, and discovered that not only had I taken off the compression sleeve, I'd also taken off my usual wrist braces. Apparently my sleeping self is really sick of wearing supportive gear :P

SHIELD fic rec

Sep. 22nd, 2014 06:07 pm
astridv: (ward 5)
[personal profile] astridv
Off the leash by [archiveofourown.org profile] Matarreyes
Gen (with some Skye/Ward UST); Ward, Skye, Coulson, May, Talbot; casefic; 16600 words

Summary: Ward was last seen by the team being led away by the military. When Skye unexpectedly runs into him on a SHIELD mission, she learns that Colonel Talbot had been loath to waste the specialist' talents and is putting his skillset to a good use. Skye is forced to become Ward´s mission control as Talbot and Coulson decide to work together, and is quick to realise that Ward's sweet parole deal is much darker than he lets everybody know.

Habitat Management Units

Sep. 22nd, 2014 09:00 am
holyoutlaw: (me meh)
[personal profile] holyoutlaw

Green Seattle Partnership splits each park into different zones called “habitat management units” (HMUs). This allows GSP to assign different target forest types and reference ecosystems to the different HMUs, and the forest stewards to use techniques and approaches best suited to each HMU.

North Beach Park is split into 11 HMUs; nine of these are discussed in this document. The other two are only accessible by crossing private property lines.

The HMUs were delineated by Nelson Salisbury and Ella Elman when they mapped North Beach Park for EarthCorps in late summer of 2011. The names of the HMUs were decided by the forest stewards. All of the names are descriptive in some way.

The HMUs in North Beach Park are based on two basic characteristics: slopes and uplands, and wetlands. There is some mixture in that all the wetland areas contain some upland slopes, and the upland areas frequently contain some seeps or wet areas in their lower regions.

Within these two divisions, slopes and uplands are assigned their name based on nearby property (ie, Fletcher’s Slope is below Fletcher’s Village; 91st St. Slope is below 91st St.; 92nd St. Wetlands is below 92nd St.), characteristics (the South Plateau is the largest flat area of the park and 80 feet above the rest of the park), or aspect (South Slope, West Slope, North Slope). The Headwaters Bowl is where the groundwater enters the park and begins to form the stream; the Central Valley is in the middle of the park.

Each of these HMUs received a reference ecosystem at the time of mapping, based on broad category of the plant species seen. There are two reference ecosystems for NBP: “mesic-moist conifer and conifer-deciduous mixed forest” and “riparian forest and shrubland.” These are based on NatureServe classifications.

The table below shows the nine HMUs discussed in this book sorted by size, and listed with their target forest type and reference ecosystem. The target forest types are explained in “Target Forest Types,” next week.

Name Size Target Forest Type Reference ecosystem
Central Valley 1.97 ALRU/RUSP/CAOB-LYAM Riparian forest and shrubland
Headwaters Bowl 1.39 ALRU/RUSP/CAOB-LYAM Riparian forest and shrubland
North Slope 1.14 TSHE-PSME/POMU/DREX Mesic-moist conifer and conifer-deciduous mixed forest
West Slope 0.84 TSHE-PSME/POMU/DREX Mesic-moist conifer and conifer-deciduous mixed forest
South Slope 0.76 TSHE-PSME/POMU/DREX Mesic-moist conifer and conifer-deciduous mixed forest
92nd St. Wetlands 0.69 THPL-TSHE/OPHO/POMU Mesic-moist conifer and conifer-deciduous mixed forest
South Plateau 0.57 TSHE-PSME/POMU/DREX Mesic-moist conifer and conifer-deciduous mixed forest
91st St. Slope 0.54 TSHE-PSME/POMU/DREX Mesic-moist conifer and conifer-deciduous mixed forest
Fletcher’s Slope 0.53 TSHE-THPL-ACMA/ACCI/LYAM Riparian forest and shrubland

The Central Valley, Headwaters Bowl, and 92nd St. Wetlands are all primarily wetlands and are discussed first. The other six HMUs are primarily uplands and slopes and are discussed after the wetlands. Within each category, the HMUs are discussed in the order of greatest amount of restoration effort they have received.

Mirrored from Nature Intrudes. Please comment over there.

ICFP 2014 Notes, Day 2

Sep. 22nd, 2014 08:05 am
tim: A person with multicolored hair holding a sign that says "Binaries Are For Computers" with rainbow-colored letters (binaries)
[personal profile] tim
These notes are about Tuesday, September 2.

I caught the end of Robby Findler's invited talk on behavioral software contracts. That was enough to catch a point that I found thought-provoking: that contracts aren't a subset of types, because contracts can express protocol-based properties (similarly to how session types do), which fundamentally involve assignment. I'm still mulling it over, and I should probably just watch the whole talk, but it might be the answer to a question that has plagued me for years, which is: "are contracts just type signatures that you put in a comment?" (Not meaning to participate in a holy war here -- I assume the problem is my lack of understanding.)

If that's true, it reminds me of typestate in Rust, which I implemented for my intern project and which was later removed from the language. Or, maybe, Rust's typestate wasn't as powerful as contracts are, and that's why people didn't find it useful in practice. I do remember always being really confused about the interaction between typestate and assignment -- we went back and forth between thinking that typestate predicates should only be able to refer to immutable variables, and thinking that we'd take the YOLO approach and leave it as a proof obligation for the user that mutation can't cause unsoundness. So maybe if I had understood contracts at the time, the whole thing would have gone better. In any case, I'd like to read more so that I can articulate the difference between typestate and contracts.

I caught slightly more of David Van Horn's talk on soft contract verification, though I missed part of that talk too. The principle here is to allow assigning blame when an assertion fails at runtime: then, you can write your code so as to have strong enough contracts so that your code is to blame as infrequently as possible when something goes wrong (if I understood correctly, anyway). ("Blame" is a technical term introduced by Dimoulas, Findler, Flanagan, and Felleisen, at least if I have the correct initial reference.) As in soft typing, "soft" here means that the contract checker never rejects a program -- it just introduces runtime checks when it can't convince itself of a particular safety property at compile time. This also recalls Rust typestate for me, which had a very similar approach of falling back to runtime verification (actually, in Rust, all typestate assertions were verified at runtime; we thought that would be a simpler approach, and if the feature had persisted, we might have implemented some sort of analysis pass to eliminate some of the dynamic checks). In my copious free time, I'd love to revisit Rust typestate and compare and contrast it with the work presented in these two talks, as well as gradual typing and effect systems, maybe even as a paper or experience report. (Which, of course, would involve me learning about all of those things.) I want to say that Rust typestate did have an analogous notion to blame: it was all about forcing each function to declare its precondition, so that if that precondition was violated at runtime, we knew it was the caller's fault, not the callee's. But I'd like to read the paper to see how much of a role inference plays.

As a much more trivial aside, I really liked that Van Horn used ⚖ as an operator, at least in the slides (as in, C ⚖ M). There should be more Unicode operators in papers! It's 2014; we don't need to limit ourselves to what was built into a 1990s-era version of LaTeX anymore.

In any case, the parts of Van Horn's and Findler's talks I heard made me think "this is the right way to do what we were trying to do with typestate". I want to be sure I believe that, though. I say this because their approach to handling mutation is to statically check any contracts that don't involve assignment -- other contracts revert to runtime checks, but the checks always happen, either statically or dynamically. My memory is hazy, but in the context of Rust, I think we talked about introducing additional precondition checks at each use of a variable involved in a typestate predicate, but quickly decided that would be inefficient. In any case, those two talks made me want to revisit that work, for the first time in a while!

I missed most of Norman Ramsey's talk "On Teaching How to Design Programs as well, but the paper seems worth reading too. Two things I did catch: Norman saying "Purity has all sorts of wonderful effects" (I think in terms of helping students discipline their thinking and avoid just banging on the keyboard until something works, though I don't recall the context), and him making the point that the HTDP approach makes it easier to grade assignments based on how systematic the student's design is, rather than a laundry list of point deductions.

Next, I went to Richard Eisenberg's talk "Safe Zero-Cost Coercions for Haskell". I have painful memories of this line of work dating back to 2007 and 2008, when I was reviving the GHC External Core front-end and had to figure out how to adapt External Core to represent the new System FC type system features, which (to me at the time) seemed to make the Core type system twice as complicated for unclear benefit. (External Core was removed from GHC some years later, alas.) I'm willing to say at least provisionally, though, that the work Eisenberg presented cleans up the coercion story quite a bit. I also appreciated the motivation he gave for introducing coercions into the type system at all, which I hadn't heard formulated quite like this before: you can typecheck System F just using algebraic reasoning, but when you want to introduce coercions (which you do because of GADTs and newtypes), contravariance ruins everything. I think a way to summarize the problem is that you get overlapping instances, only with type families rather than just type classes.

To solve the problem, Eisenberg and colleagues introduce two different equality relations: nominal ~, and structural ~~. This allows the type system to incorporate coercions based both on nominal type equality, and structural type equality, without having to pick just one. Then, each type parameter gets a "role", which can be either "structural" or "nominal". This allows coercion kinds (my nemesis from the External Core days) to just go away -- although to me, it seems like rather than actually taking coercions out of the kind system, this approach just introduces a second kind system that's orthogonal to the traditional one (albeit a very simple kind system). I guess it's possible that separating out concerns into two different kind systems makes the implementation and/or reasoning simpler; also, as far as I can tell, roles are more restricted than kinds in that there's no role polymorphism. (I'm not sure if there's kind polymorphism, either, although there certainly was in GHC at least at some point.) Actually, there are three roles: "nominal" (meaning "this parameter name matters and is not interchangeable with structurally equal types"), "representational" (for a type that is interchangeable with any others that are structurally equal to it), and "phantom" (type parameters that are unused on the right-hand side of a definition). I wrote in my notes "I wonder if this sheds any light on Rust traits", but right now I'm not going to elaborate on that query!

The implications of the work are that generalized newtype deriving now has a safe implementation; the type system makes it possible to only allow unwrapping when the newtype constructor is in scope. (This, too, reminds me of a Rust bug that persisted for a while having to do with "newtype dereferencing".) The results were that the new role system uncovered three legitimate bugs in libraries on Hackage, so that's pretty cool. Also, Phil Wadler asked a question at the end that began with something like, "...here's how Miranda did it..." (Not something one hears a lot!)

Finally, I stayed for François Pottier's talk "Hindley-Milner Elaboration in Applicative Style", which I understood more than I expected to! He began by saying something that I noticed long ago, but originally chalked up to my own stupidity: Algorithm W in its original presentation, was "imperative, and messy". We want a simpler, more declarative formulation of type inference. Pottier claims that conjunctions and equations are simpler than compositions and substitutions -- I agree, but I'm not sure if that's based on something objective or if that's just what works well for my brain. He defines a constraint language that looks like λ-calculus with existential types, which allows constraint solving to be specified based on rewriting. On paper, it's a declarative specification, but the implementation of it is still imperative (for performance reasons). It sounds like it might be fun to prove that the imperative implementation implements the declarative specification, though perhaps he is already doing that!

Stay tuned for my notes on day 3, when I get around to editing them.

Profile

sasha_feather: Retro-style poster of skier on pluto.   (Default)
sasha_feather

September 2014

S M T W T F S
 1234 5 6
7 8910 111213
1415 1617181920
21222324252627
282930    

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags
Page generated Sep. 22nd, 2014 10:16 pm
Powered by Dreamwidth Studios