anon1253

Or get an IDAS-LPS, CLS or UHC filter (either on the lens or as a sensor clip on). Or stack a bunch of images with proper dark substraction. If you stack enough light sub frames you can efficiently reconstruct dynamic range. Of course the ETTR advice is good too, but you can go much further even with consumer equipment.

Yep, thing that's often overlooked is that the Waddenzee is pretty much a lost cause though, with all the one-of-a-kind flora and fauna that lives there. If the sea rises a couple more centimeters the area will no longer be dry during ebb tide. Go "Wadlopen" while you still can!

There's a larger truth hidden in here: experts disagree. It was one of the things that took me by surprise when building Machine Learning systems, but it makes sense if you think about it. For all but the most trivial matters, it's the job of the expert to exercise personal judgement. It's why they're experts, not because they have a solid understanding of the "knowns", but because they have developed methods and intuitions about the "unknowns". Ask 10 different experts a complex question in their field of interest, and you'll likely get widely varying answers. My guess is that it also played a role in the early expert systems AI bust (AI winter). There are interesting papers published about this as well, e.g. http://www.sciencedirect.com/science/article/pii/S0951832007... the money shot is in this chart http://imgur.com/a/shaii but you'll see similar distributions everywhere where expert judgement is warranted.

Between Flickr dying, instagram being Facebook, and 500px being, well the post is still spot on… I'm not sure what a nice place is for looking at and sharing pro/semi-pro photography. Seems like the comodtization of photography is both a blessing and a curse when it comes to the social sphere of (pro) photography

Well, I agree wholeheartedly https://joelkuiper.eu/cloud

Any relation to https://github.com/QuiltProject ?

It's not just that overpopulation is a major problem. It's that I'd rather save them the burden of existence in a world that will be substantially less habitable.

Not quite sure what to make of your comment other than some non-constructive jab at my personality, though. But if you have any thoughts, I would of course like to hear them! We're all just making this up as we go along.

- I eat mostly vegetarian

- I live in a small apartment block that is very insulted and has a lot of natural light

- don't own a car and use bike / public transport

- work 100% remote

- I compensate CO2 when flying, even when for business

- I try to have a minimal set of possessions, and buy second-hand if possible (and am really aggressive about ad blocking, don't watch commercial TV)

- I got a vasectomy to not have children

- I try to write and engage in social media about it, read the primary literature

- Donate to reforestation / tree planting initiatives

But it's all bullshit isn't it? I doesn't change a damn thing. The fixation on the idea that individual consumer behavior will make a noticeable impact is laughable. The only way to make a difference is by changing the companies and business that pollute in the first place, not the people who buy the stuff. And that means taxing, de-investing, subsidizing alternatives, and for the love-of stop subsidizing fossil fuels. But that's never going to happen.

I mean even putting a stupid sail on cargo ships would cut down pollution more than the CO2 anything you and I could ever do. Or if we hike up the prices of transport considerably (so fruits from bloody Africa don't cost 20 cents in Northern Europe) we can change our entire logistics pipeline.

It's silly to think we're unable to change. Nearly all of the CO2 emissions were done in just a generation. Post-world war 2 industrialization, automation and globalization caused the vast majority of CO2. There are still people alive today who lived through WO2. Why are we talking like this is an insurmountable problem, that the momentum is just too big? It literally took /one/ average human life span to get where we are. It could take less than one to go back to pre-industrial levels. And it's not just a matter of better technology, it's mostly a matter of social adaptation. If we're dead set on infinite growth (in terms of consumption, GDP, population) we're going to have a problem with or without climate change. The fix is: a smaller population, less consumption, and less production. But that's contrary to /everything/ capitalism stands for. The problem is thus not so much technology, but rather the socio-economic paradigm we're in. How to change this without going all "you dumb communist beta cuck snowflake"? I don't know. But here are some ideas:

- Incentivize less spending by "anti-ads" (could be interesting to add as an ad blocker), basically ads that signal that you /don't/ need what's advertised. Interesting AI problem that can piggyback off of all the stuff that's being done in the ad space itself

- Implement economic incentives for CO2 scrubbing, by carbon-tax, subsidies or secondary effects (e.g. useful non-fuel biproducts, or things like greener living spaces / habitable deserts)

- Stop subsidizing fossil fuels, agriculture, etc. Will it hike up the prices? Sure, but that's kinda the idea.

- Look at the big picture first. Driving your car less, LED bulbs, sure…they kinda work. It's somewhat of a moot point with cargoships that almost burn raw oil, equivalent to the CO2 output of a small city. This is an optimization problem: try to do the big things first.

- Slow down. We define the pace at which society moves. Do you /really really really/ need that stuff in 24hours door-to-door? Do you /really really/ need to commute 8 hours each day? These are human choices. The natural world is kinda reluctant about our endeavors. There's no reason why everything must be done /yesterday/ … helps with mental health issues too.

- Incentivize a culture where "moon-shot" ideas are given a chance. There are many ways we can make a big dent, but often costing billions to build. We have no problem building 250 million dollar warplanes (and buy hundreds of them), no problems building half a billion dollar coal plants. Really, no problem with billion dollar sports matches. In fact, we seem to applaud these things. But when talking about rooftop solar, heliostat installations, CO2 scrubbing, planting trees it's all "cost cost cost". We need to rethink that.

And be kind to all living things. We're in for a tough ride either way. Just try to show some compassion and help those who need it more than you do.

Or that's what I try to tell myself when I'm utterly depressed about it https://joelkuiper.eu/change

most implementations are not arbitrary length, but instead rely on padding to make it a fix length sequence.

Very interesting! paredit + agressive-indent made my life so much easier. Now I don't have to worry about parens OR indentation. Unless you work with people who have different preferences of course, but for my own stuff I generally stopped trying to fight agressive-indent (at least it's consistent).

For those who like to know a conceptually a bit more about RDF/Sparql: https://joelkuiper.eu/semantic-web

Bonus: my SPARQL Clojure wrapper https://github.com/joelkuiper/yesparql

We use quite a bit of SPARQL to reason over medical ontologies. It's not exactly "semantic web", but ontological knowledge (such as the Gene Database, UMLS, etc) are naturally serialized in graphs, which makes them an excellent fit for RDF triplestores, hence SPARQL. We regularly query millions of concepts with billions of relations in milliseconds. True, RDF under the hood is just a EAV model, but thinking about the identity between triples, graphs, and logic programming makes a lot of sense for us.

https://joelkuiper.eu/semantic-web

I use Anaconda, which supports different concurrent installs

Fairness is a rather interesting problem with many definitions, see for example [1,2,3], I was hoping for a more extensive set of implementations; still good work!

[1] https://www.amazon.com/Cake-Cutting-Algorithms-Fair-You-Can/...

[2] https://www.cs.cmu.edu/~arielpro/15896s15/docs/paper10.pdf

[3] https://en.wikipedia.org/wiki/Fair_cake-cutting

Additionally if you're really paranoid

- Use LittleSnitch or similar to block traffic per process

- Use ad blockers in your browser and make sure Flash, Java and Silverlight are off by default

- Do update regularly

If you care about security in case your device gets stolen

- Use full disk encryption

- Do /not/ forget to also set a firmware password

You might be interested in "Ghost In The Shell: Identity in Space" https://www.youtube.com/watch?v=gXTnl1FVFBw

Enhanced color isn't exactly right. Most nebula images are done with Narrow Band filters that only let a specific wavelength through. Typically those of Hydrogen alpha, sulfur II and oxygen III emissions. This gives three monochrome channels of data that you can combine however you please to RGB. Or supplement with even more data from broadband R, G, B or luminance filters. Not sure about the Horsehead image, could be a HaRGB or even Near infrared composite. In the case of Juno there are 4 filters: R, G, B and a methane filter (~840nm narrow band). The data you can map to RGB arbitrairly, usually with a linear combination. It's just as much art as science

Here's a silly trick we used for a similar problem: use zlib with a custom compression dictionary. Our application had tons of interlinks (think linked data) that we wanted to expose to the end user, but putting urls in urls is kinda ugly. So we ended up with our own custom compression dictionary and pushing them through zlib. Works like a charm.

Being a huge fan of the genre of "interactive narratives", I have to disagree. Games like Gone Home, Life is Strange, Mass Effect, Skyrim, Fallout, Deus Ex, and others have had a profound impact on me. The author doesn't seem to make a compelling point why they would be better as books or movies. I haven't really been able to derive a more coherent argument from the article than "they're worse because they're worse, so they should not be games". I think that is a bit harsh on the incredible achievements of the games that painstakingly build an interactive narrative on a medium that is both technically and artistically extremely challenging. Take Mass Effect for example, to me the whole franchise has been one of the most successful science-fiction stories ever. It rivals Star Trek, Star War, and The Expanse. Especially since Mass Effect came out during a period where Hollywood had pretty much given up on the genre as a whole. Mass Effect constructs this beautifully detailed and coherent universe set in a future where mankind discovers advanced technology from an ancient race. This propels mankind into a new era of exploration, conflict and intrigue. Te reason why Mass Effect "works for me", besides the "world building", is that you have the ability to construct a secondary narrative on top of the one presented by the game itself. You are allowed, and encouraged, to make decisions and reflect upon them. While it's all arguably superficial and closed-world, the moment you immerse yourself /in the character/ it offers something very few books and movies can: the ability to guide the story, and become a "person" in a complex world. Then again, I'm biased…but so is that author it seems.

How does Kobo do it? I was under the impression that the lighting there isn't edge-lit, but rather backlit?

Depends on how much the overlayed LCD screen would impact legibility. I'm not sure. Transparent OLED might work [1]. The backlit would basically be for evenings, where the ambient light might not be enough. I just love the readability and aesthetics of eInk. Practically, it's probably worse at everything...so I can understand the limited uptake.

[1] http://www.mmt.io/transparent-oled-touch-screen-display/

This seems really interesting. But what I've been more interested in is a LCD/e-Ink hybrid display for my workstation. LCDs are basically transparent, so I think you could overlay them in front of a back-illuminated e-Ink display. Say 23"? For things like programming this would be ideal for me. You can put the high-refresh rate things on the overlayed LCD, and the slow documentation, typing, stuff on the e-Ink display. I don't even need a pen, those never work for me (too much input latency).

Another idea would be to pair a large e-ink display with a Macbook Pro style "magic bar" (or whatever they call it). So you can do the typing on the e-ink, with a Submlime text-like minimap for scrolling and context. Anyway, I don't have the chops to pull it off…but it would be great I think!

Oh I totally agree there. I never quite grokked the obsession with JSON these days. But if you go to trouble of writing a query language that /specifically/ targets JSON output ... it seems like it would at least need to be comprehensive. Or come up with another output serializer/deserializer ... but that would put considerable strain on the consumers.

Aren't you afraid you might be reinventing SPARQL ;-) ?

We tried to use GraphQL, but it in our opinion it has a major design flaw: not everything that can be expressed in JSON can be expressed in GraphQL. Since our application makes heavy use of Semantic Web (if you haven't already, checkout SPARQL and RDF which made similar promises). But this means we often end up with responses that contain URIs as map keys. This is not allowed as a filter in GraphQL, even though it is valid JSON. Other parts of our stack contain Clojure(Script), and we make use of namespaced keywords that serialize to "foo/bar" as map keys. Again something that is not allowed in the language, and would break our application. I find it a particularly odd design choice to not make GraphQL spec match the JSON spec. As there are now many things that are impossible to specify as a query, which would otherwise be valid JSON.

It's not the only option either. Combine that solution on a massive scale with re-forestation and (even though eventually it's CO2 neutral) it might actually make a dent. My go-to model was always:

- Plant heliostat molten salt reactors in desert coastal areas [1]

- Use generated heat to desalinate water

- Use desalinated water to provide irrigation

- Plant (potentially GMO'ed) bootstrapping organisms to create a soil (greening)

- Plant trees

- Rinse repeat.

However, your idea makes more sense if the only goal is CO2 scrubbing. But adding the heliostat molten salt reactors in a desert might be quite feasible, they generate tons of heat, which can be translated into energy or simply mechanical force (water pressure)…and there is no shortage of silicates in the desert.

[1] http://www.solarreserve.com/en/technology/molten-salt-energy...

I'm betting on the higher regions of Scandinavia or Iceland.

That really depends on the language I think. Java/C# etc I wouldn't dare to touch without an IDE. Most of my Clojure/Lisp work is done in Emacs (with Evil for Vim bindings, without any syntax highlights). People /still/ think I'm goofy for preferring that. But, I find that if you think of programming as "creative writing" instead of engineering you enter a different, more flow-like, mindset. The IDE and all its billions of buttons just get in the way for me, while just writing s-expressions as trees (using paredit with agressive-indent) never breaks my train of thought.

Don't drop critical thinking and "hammock time" in favor of Agile. Some problems require a bit of personal reflection, without it you can easily end up building something atrocious, even while following all the "best practices". I spend a couple of years at a SCRUM shop and all we did was "finishing stories" without any regard for architecture. The technical debt was truly stunning. These days I easily spend a day just thinking about the problem before writing some code, which then still happens in a REPL-driven style to allow for experimentation and feedback. Best of both worlds I think. Also, most programming is actually fairly easy. Focus on the (business) problem you want to tackle instead of the code. Code is ephemeral, the only thing that matters is the business value and the user experience for getting there.