Quill: I’m Peter Quill. People call me Star Lord.
Peter: wait Peter Quill?
Quill: yeah? what?
Peter *who watched a buzzfeed unsolved about his disappearance*: oh my god.
Almost everyone who encountered Peter Mayhew claimed he was one of the kindest people you’d ever meet. In the same realm of Tom Hanks, Mr. Rogers, and Bob Ross.
Just a bit taller.
He loved being a part of Star Wars. He was famous for using his special custom lightsaber cane.
Despite being in a lot of pain, he went to as many conventions as possible to meet people. You could tell his love for Star Wars fans was genuine.
Mark Hammil wrote a touching tribute to him.
“He was the gentlest of giants. A big man with an even bigger heart who never failed to make me smile & a loyal friend who I loved dearly. I’m grateful for the memories we shared & I’m a better man for just having known him. Thanks Pete.”
Rest in Piece, Mr. Mayhew.
If you are in some sort of pleasant afterlife, I hope they finally gave you the medal you totally deserved.
May the Fourth Be With You.
@casualCosplayKatie@Instagram.com as Cloud City Princess Leia.
May the Fourth Be With You! @CasualCosplayKatie: Easiest #princessleiahair by a long shot. Now with full ‘bound outfit (even if I’m not bound anywhere today)!
I know it seems a bit odd to talk about a Marvel book on a Flash website, but stay with me a moment. There are two series currently running, one at DC and one at Marvel, in which mental illness has been portrayed in a major plot line. For DC, it’s HEROES IN CRISIS, with a tragedy occurring at a mental health facility for superheroes called Sanctuary. For Marvel, it’s THE UNSTOPPABLE WASP, in which the lead character is diagnosed with bipolar disorder. So, how do the two story lines match up? Follow us after the jump!
Amazing costume and a great photo!! The painting is the perfect added detail 😀
It’s hard to tell in this shot, but the painting’s the cover on a messenger bag (she hid the straps to pretend to be stealing the painting), so it’s not just a perfect detail, it’s useful too!On Tumblr
Last fall, I conceded that phones have caught up to casual cameras and I’d have to get a nicer one to get better image quality. Well, I finally bought a mirrorless camera. The kiddo found my old SLR, and we’ve split a few rolls of film (re)discovering how to shoot with it. Then he started asking about a modern digital equivalent. Since it was going to be two of us using it, not just me, I felt like I could justify the expense.
I read a bunch of reviews and asked around for advice, finally settling on a Sony Alpha a6000. It’s a few generations back in their advanced amateur line, making it a bit more affordable.
We brought the new camera to WonderCon, and I made some discoveries:
- It actually handles the light level inside the convention center!
- I’ve gotten waaaay too used to just capturing costumes when shooting cosplayers, instead of composing interesting shots. Most of the photos I took on Saturday had good image quality, but were ultimately just snapshots.
- Because of #1 and #2, I ended up with a lot of busy backgrounds. I tried to cut down on the distraction by adding vignetting to some of the photos afterward.
Originally we planned totrade off who had the good camera, but he ended up wearing his giant Minecraft Spider Jockey costume the entire time, so he didn’t have much opportunity to take photos. In the end, he only took one all weekend…but it was the best-composed shot of the entire day!
I took the lesson from that, and while most of my pictures on Sunday were still utilitarian snaps, I did manage to take a few that I think worked out better, like these three:
My full cosplay gallery is on Flickr, including these four photos and all those snapshots.On Tumblr
This wine could be made…useful. #minecraft
Tonight’s waxing crescent moon, taken with the really long telephoto lens. It’s passing in front of the Hyades star cluster right now, which I’d love to get a photo of, but the moon is just waaay too bright to capture the background stars! Plus I’m still getting used to the controls on this camera.
And here’s another moon shot I took in January, about a week after the lunar eclipse. I used a…
It’s almost as though cruelty and neglect can affect people’s mental health.
Everyone gives Batman shit over the state of Arkham but no one ever talks about Iron Heights.
There was a bit of criticism for Iron Heights within the Flash book, such as when Ashley Zolomon called the prison “the Rogue Factory” (which she said was a widespread nickname) and accused Wolfe of having no compassion for its prisoners. It seems quite likely that the mistreatment there has made some of the prisoners worse, with I think Roscoe as one of the prime examples. And as awful as Arkham is, at least they make some attempt to treat their inmates; we’ve seen that Iron Heights leaves theirs barefoot in straitjackets and isolated in filthy cells. So it’s no surprise that they end up even more mentally ill and anti-social, which is almost certainly what Ashley was alluding to.
Wally and Wolfe clashed on a number of occasions over the treatment of prisoners there, once Wally found out about it. But Wally had no authority there, he was just a vigilante super-hero, and Wolfe kept running things the way he wanted to. The one win I remember was that Wally was able to get Fallout an actual, comfortable room that absorbed his radiation instead of leaving him hooked up to tubes like he was before.On Tumblr
The other day I trained a neural net to generate the names of cookies, based on about 1,000 existing recipes. The resulting names (Quitterbread Bars, Hand Buttersacks, Low Fuzzy Feats, and more) were both delightfully weird and strangely plausible. People even invented delicious recipes for them. But given that I’ve trained neural networks to generate entire recipes before, why not have the neural network generate the entire thing, not just the title?
Well, this is why.
The first neural network I tried is textgenrnn, which I’ve used to generate things like new species of snakes, names for creepy shopping malls, and terrifying robotics teams. Given 1000 cookie recipes from David Shields’s site, textgenrnn could do a recognizable recipe – but its titles and directions were a bit suspect.
Now, granted, it’s confused about other things, too. A memory approximately 40 characters long means that it doesn’t know how many times it has already added sugar (apparently its favorite ingredient). (Other algorithms, like GPT-2, have ways to zoom out.)
I decided to see if textgenrnn would figure out recipe titles if it trained for longer. It generated the recipe above after it had seen the example recipes 3 times each. Below is what it did after another 3 looks at each dataset (6 in total). The title is… different. The recipe is more chaotic. It has at least moved on from its obsession with sugar.
After 3 more looks at the data (9 in total), things have gotten even worse, though according to the loss (the number it uses to track how closely it matches the training data), it thinks it’s doing better than ever. It seems to be freaking out in particular about the repeated + signs that some recipes use as decorative borders.
There are terms for these kinds of training disasters that sound more like warp engine malfunctions: “mode collapse” usually applies to image-generating algorithms, though, and “exploding gradients” usually is signaled by the neural net thinking it’s doing worse and worse, not better and better.
So I moved back to another algorithm I’ve used for a long time, char-rnn, which seems to do well at long texts like recipes.
The recipes are… better. I’ll give it that much.
Some of its ingredients are questionable, and its directions certainly are. But at least (despite a memory only 50 characters long) it has managed to do the beginning, middle, and end of a long recipe. It’s often fuzzy about how to end a recipe, since my rather messy dataset has recipes ending with sources, urls, and even ISBNs. Recipe notes are also highly variable and therefore confusing.
So what happened with textgenrnn? I mentioned my difficulties with textgenrnn to the algorithm’s creator, Max Woolf, and he urged me to try again. There’s some randomness to the training process. Sometimes textgenrnn degenerates into chaos during training, and even then sometimes it pulls itself together again. When it did well, its instructions even start to make sense. You could make the Butterstrange Bars below (almost). Given this amount of randomness, it’s nice when researchers report the aggregate results of several training runs.
Bonus recipe! Sign up here and you can get an Excellent (not excellent) recipe for “Mrs. Filled Cookies” along with (optionally) bonus material every time I post.
Super camera telephoto lens is out of focus
Even though the tripod’s set, the blur is really bogus
If you use the automated setting hocus-pocus
Still the telephoto lens is slightly out of focus!
(The background is supposed to be blurry, but I had a lot of trouble focusing on the foreground with this lens. I guess it’s just going to take practice.)