Disqus - Latest Comments for divide_by_zero

Re: A Job Is Not Alienating; It Is Empowering

dividebyzero — Mon, 19 Aug 2019 19:15:45 -0000

This article is exceedingly naive. You could at least try to make the case that unhappy workers whose job are "not the greatest thing ever", who are numerous and identify with many of those ideas, might be perhaps not living in a genuine free market, or might be experiencing the effects of forms of tyranny that might be denounced by a classical liberal as much as other political groups too. Or might even be experiencing situations where there may not be any culprit except indirect imperfections in a market that should be free in principle, things such as lack of access to information. I would love to hear some honest and insightful liberal opinions about why many people are unhappy at work instead of seeing this dismissed as phony conspiracy theories from your political rivals.

Re: Why Is Brazil So Happy?

dividebyzero — Sat, 18 May 2019 10:56:05 -0000

We're all hopeful for Brazil, but there is plenty to be worried about. Recent direct meddling of the president on gasoline price is a great example of how Bolsonaro is nowhere close to being a politician with a liberal mindset https://www1.folha.uol.com....

Re: Data Modeling in FP vs OOP

dividebyzero — Sun, 04 Nov 2018 05:21:36 -0000

Trying to look for something in my projects. In the first one it seems easy to get rid of inheritance. These three last derived classes could just be functions that come up with the `x` and `y` and produce a `Sample` object... Why did I not do this at first? No idea. https://gist.github.com/nlw...

The next one I'm not so sure how to refactor. I'm sorry I couldn't make the code simpler, but the idea is we have an algorithm that takes an object with different methods, and the different implementations of the class let us solve different problems with the same algorithm. What would be the alternative to having the base class here?
https://gist.github.com/nlw...

Re: Emacs: Effective Windows Management

dividebyzero — Thu, 30 Aug 2018 14:51:58 -0000

Maybe I can just make something different, I'm redeveloping my workflow, but this is what I'm doing right now. I have emacs on python mode in one window, then C-c C-c and C-c C-z to run the script and open the repl. Then I open another window and change that to the repl, change the first window to the editor buffer, and that's OK. Then what happens is that if I have to kill the session C-c C-k y, if I run the script again again I get a new repl buffer in the first window. This is a screenshot to better explain https://imgur.com/a/097XEqt

I just need to find a way to prevent the repl buffer from opening, maybe _ever_ opening. Of course I can easily kill this new buffer, but I'm trying to make everything as slick as possible

Re: Emacs: Effective Windows Management

dividebyzero — Wed, 29 Aug 2018 06:50:55 -0000

Been trying to follow this advice, but one thing that nags me is that when working with python and an interactive console, for instance, you sometimes end up having a new shell buffer opening up in the same window where my code is. How do I prevent that?

Re: Emacs: Xah Fly Keys Customization

dividebyzero — Mon, 20 Aug 2018 14:58:42 -0000

I'll be definitely trying out mapping capslock! What I mean about the home key, though, is that xah-fly-keys by default maps M-SPC, F8, Menu and Home all to entering command mode, and I would like the Home key not to be binded (only to normal cursor movement behavior). I managed to just comment out the binding line at the xfk source code, and I'm quite happy with that, but it would be definitely better to do this from my config file instead. So I am wondering, what would be the best way to remove the Home binding without having to edit the source code?

Re: Emacs: Xah Fly Keys Customization

dividebyzero — Sat, 18 Aug 2018 10:43:22 -0000

Been starting to use xfk, and I think it feels great. Definitely believe I'll switch to it. I am a big fan of the "Home" key for navigating text, though, and I would quite like to unset it for switching to command mode. What is the best way to do it? Or actually just setting Home to move in the command mode could be enough...

Re: The Big Problem With Self-Driving Cars Is People

dividebyzero — Thu, 27 Jul 2017 14:22:59 -0000

I have deep respect and admiration for professor Brooks, but I can't avoid raising some questions regarding this column.

First of all, I don't believe there is much doubt about how the cars will behave. Of course in any ambiguous situation they will be designed to err in the side of caution. And that will probably mean that, yes, in some situations that human drivers today drive pretty close to pedestrians that might mean they'll "crawl". But I don't quite believe that cannot be perfected with time. I believe people act in ways that are quite predictable and detectable, and we just need to come up with a suitable model of pedestrian motion.

About taking cues like where people are looking at and how they are gesticulating to decide whether someone might cross the street... That sounds a bit like trying to figure out exactly how our mind works, and then hoping to implement algorithms that reflect this belief of ours about our cognitive systems. Whenever I see a researcher doing this, I immediately think of COG, one of the most notorious projects from the professor. In that project, and others, it was shown that letting the robot learn by itself what to do might be a much better strategy to AI than trying to come up with detailed models that emulate our thinking (or our idea of how our thinking works). Similarly, I believe we should as soon as possible let go from this perhaps anthropocentric attitude regarding autonomous cars. Let's not worry about the car being able to "understand" that some part of the human body that has some name is moving in a certain way, and what the person might "mean" with that... The car has its sensors and its goals, and it is going to learn what to do. The decision process might in the end not look like anything we can feel familiar with, but might still perform as well, or even better than a human driver does. Let the robots learn the best way they can how to predict pedestrian motion and do their jobs.

Would that mean that some autonomous cars would behave like "nagging" drivers? Absolutely. They will also behave like drivers who are less prone to murder other people. Could it not be the case that our supposed ability to "handle ambiguities" is actually a false negative for "careless driving"?

About jerks playing around with cars on the road... Of course this will happen. Not even elevator doors are safe from user abuse. But how would this ill-intended pedestrian not suffer any retaliation in the case I am sitting inside the self-driving car? The passengers in this case are even more free to retaliate, and might even display hand gestures to the abuser with both hands if deemed necessary. There is a big difference in the application scenarios of fully autonomous and passenger-less vehicles, and a vehicle carrying people, and they should be considered separately. Fully autonomous vehicles are subject to even worse kinds of "abuse" problems, such as theft, and the possibility of someone playing around with the vehicle pales in contrast.

Safety from theft is also one of the major obstacles preventing applications such as autonomous trams and trucks. The driver is not there just to drive, they also check if passengers are paying their fees, etc. This requires more work to be automated, so naturally it will take some more time to show up on the streets than just cars that can self-drive with human oversight. But I don't quite believe no-one is looking into self-driving trucks and trams and buses, I am not sure what are the most mature initiatives out there, but the professor makes it seem like this is an unexplored territory, and I would strongly question that. (UPDATE: here's a project https://www.fastcompany.com... )

Finally, about all the use cases where owners might let their cars walking around by themselves, and sending their cars to school lines, etc, I find it interesting the professor firmly holds the idea that people will really continue to be interested in owning their cars. In all those cases any problems would go away when you consider that people may just start using more frequently car-sharing services such as Uber and whatnot. Younger people already own less cars today even without automation. I don't even have a driver's license... There will be many changes in our society regarding transportation when self-driving cars become a reality, and that involves owning cars, dealing with parking, and even the discussed "dance of death" that pedestrians and cars do in some roads.

The debates around AI, robotics and self-driving cars are being overtaken today by people who don't really come from the field. I praise the voice of researchers of such caliber as professor Brooks. It is really a great exercise to ponder how adoption of self-driving cars will work in practice, and I found it interesting I could have very disparate views from someone like him. I guess we all really need to have more and more debates about the subject.

Re: The Neophyte's Guide to Scala Part 12: Type Classes

dividebyzero — Sat, 24 Jun 2017 14:54:23 -0000

Great article. I must protest the GoF bashing a bit, though. I do understand some people were traumatized by it when young, and can't seem to overcome it... But I think Odersky himself once said that languages evolve by turning patterns into language features. Could we say here that Haskell actually did that to the adapter pattern, and Scala is still missing an opportunity of implementing it really well in the language? Also, it's interesting to think how this topic relates to erasure... Would Scala be better without it? Or should it have better tools to let us deal with it? Maybe this is actually a major reason we should learn to do that in Scala, and maybe this also relates to the fact it is so important in Haskell? Thanks for leaving me with more questions than I had before!

Re: AMD Ryzen 7 1700X CPU Review

dividebyzero — Sun, 04 Jun 2017 07:02:03 -0000

For those disappointed about the lack of support for quad channel, "for scientific applications" to quote the article, don't despair and move to Intel just yet! New 16 cores "Threadripper" processors just announced by AMD support quad.

Re: Have We Forgotten about Geometry in Computer Vision?

dividebyzero — Tue, 09 May 2017 04:35:15 -0000

Congrats, I hope more deep learning people start sharing your insight.

Re: If Haskell is so great, why hasn't it taken over the world? And the curious case of Go.

dividebyzero — Mon, 10 Apr 2017 14:51:32 -0000

It is hard to figure out how could any language take over the world. This is harder than debating economy and war, topics where people actually try to "take over the world"... What does that even really mean for a programming language? Percentage of adoption in new projects? Total projects? Projects making more money to programmers? Percentage of new code produced that is not just another CRUD app?...

One crazy thing about languages is that sometimes you look at what's happening out there in the "real world", and no language really seems capable of really taking it over. There is still Cobol code hanging around in companies. Physicists and others cling stubbornly to their old Fortran code passed on by their academic forefathers. Just the other day I watched a lecture from a great guy who has many admirers, and in the beginning he confessed how he once lost a couple of weeks solving some concurrency problem in an embedded project. The dude coded in assembly, and after that some friend finally thought him into picking up C. Really, in the 21 century...

Tons of programmers seem to be comfortable about learning one or two languages in their whole lives. Then they just stick to their guns. For example, I have an impression that Java only got so popular because it became the first serious language tons of people learned at college, not because there was a huge migration of C++ (or Cobol) coders. If that theory is right, then Haskell adoption doesn't grow more because it would be specially unattractive to beginners.

I really can't decide what I think about taking over _the_ world, but I can see Haskell is taking over _my_ world... After many years doing C and Python, one day I got fed up with run time errors and looked for a new language, compiled, functional, and found Scala. It was great learning it, I must say it even kind of returned to me some joy in programming itself. And I did study a lot in your book, so thanks for that! :)

After a couple of years into Scala, I am finally starting to study Haskell, and I am very well impressed with it. When I picked Scala I did consider Haskell, but it striked me as over complicated and impractical, and documentation wasn't good. I got an impression it would be like working with Prolog (with which I have some experience with).

After learning Scala and giving Haskell a second chance, I found out I was completely wrong. The language is actually very practical. For instance, I though "IO was hard", and today I see how stupid is that myth. For many things it seems more simple and uncompromising than Scala. But one thing I realize is that a lot of what must be learned for someone coming from C and Python like me really has to do with FP in general and not so much about e.g. syntax.

After learning Scala I tried to convince everyone I knew to try it out, and I only managed to convince a couple of friends... But little by little I think FP will gain more and more territory. I don't know about taking over the world, but I am always glad when I read about people taking FP. We are "taking over the world that matters"...

Not to be too off-topic, I am not so sure this issue of interactions between applications is such a natural next step from FP, or something strongly tied to language design. I see this as a very old issue parallel to programming languages, that relates more to things like Corba, UML and REST, etc. It is about domain modeling and I don't know, ontologies or whatever. But here again, like with FP, it is something that programmers only really start to understand its importance after they acquire some practice... So maybe what I am trying to say is that the reason Haskell isn't taking over the world is because FP in general isn't taking over the world, and e.g. DDD isn't taking over the world, well at least not very fast, but it will as programmers start to learn about all these higher-level stuff sooner in their lives...

Re: True fullscreen in xmonad

dividebyzero — Tue, 10 Jan 2017 06:46:38 -0000

Thanks for that, I use a bright red 1-pixel border, and it was impossible to watch Netflix! :)

Re: OpenVPN Configuration Guide for Linux/Ubuntu Version 16

dividebyzero — Tue, 03 Jan 2017 18:13:13 -0000

I tried using OpenVPN on my Ubuntu 15.04 (in Netherlands), but although the connection would be set up OK, I could not load any webpages. PPTP worked just fine, though!... UPDATE: it stopped working! =P

Re: Why Jessica and Trish Getting Together on Jessica Jones Would Be a Bad Idea

dividebyzero — Fri, 23 Sep 2016 04:38:22 -0000

Yeah, I didn't mean it like that, and I apologise... The under-representation of LGBT is definitely a pressing issue I didn't have in mind. I hope the idea of "extending" the test can make us talk more about who are these characters from movies that pass the test. If it is actually still difficult to escape hetero-normativity, or of it turns out that it's even more difficult than in other movies, then it's a big problem. I pay a lot of attention to the Bechtel test, but never thought about it much in the light of LGBT representation. Sure will try to do it in the future.

Re: Why Jessica and Trish Getting Together on Jessica Jones Would Be a Bad Idea

dividebyzero — Thu, 22 Sep 2016 18:21:47 -0000

Sometimes it is a slow process to realize you are taking crap.

I mean a better world where there is a lot of diversity on screen and nobody feels underrepresented.

Re: Why Jessica and Trish Getting Together on Jessica Jones Would Be a Bad Idea

dividebyzero — Thu, 22 Sep 2016 18:13:08 -0000

Yeah, I dunno... You're obviously right, and I'm sorry... I'm almost turning into a shipper now :]

Re: Why Jessica and Trish Getting Together on Jessica Jones Would Be a Bad Idea

dividebyzero — Thu, 22 Sep 2016 18:03:32 -0000

Yeah I don't know, I will be paying more attention, I'm just talking crap... But this doesn't look a like a very nice world to me, we can do much better than any of these options.

Re: Why Jessica and Trish Getting Together on Jessica Jones Would Be a Bad Idea

dividebyzero — Thu, 22 Sep 2016 17:54:12 -0000

You totally have a point on the shoehorning thing.

Re: Why Jessica and Trish Getting Together on Jessica Jones Would Be a Bad Idea

dividebyzero — Thu, 22 Sep 2016 17:44:30 -0000

Must have been a white straight male!

Re: Why Jessica and Trish Getting Together on Jessica Jones Would Be a Bad Idea

dividebyzero — Thu, 22 Sep 2016 17:39:33 -0000

Thing is, passing the Bechdel test and having a lesbian couple on screen is bound to have some correlation, because it might be easier to pass the test once you already have a lesbian couple. To pass the test with heterosexual characters would be trickier. But this is all just a stupid nerdy discussion, it is pointless to argue about any of this once you realize there are so few samples of any of these categories to make any tests! So I'm just happy for Jessica Jones whatever happens, and hope the shippers find a way to keep some hope alive. It's not good to lose hope.

Re: Why Jessica and Trish Getting Together on Jessica Jones Would Be a Bad Idea

dividebyzero — Thu, 22 Sep 2016 17:22:50 -0000

OK, it does sound like I may be removing the original test spirit of criticizing male-centrism, and that was not my intention. It's just that sometimes it looks like the only escape for a woman not to be overly worried about a man would be homosexuality. A hetero woman would have no choice except to be constantly serving their male masters. If a woman does not worry about some man in some time, there must be something "wrong" with her. "Independent" woman are destined to be homosexual, and then there would be a strong correlation between successfully passing the Bechdel test, and not being straight. But of course, now it looks like I'm an a homophobic crusade, trying to deprive people from a chance of seeing non-heteronormative relationships on screen, what is not my intention too. Pity there is so little screen time that might fit some diversity, that people have to dispute. But whatever, I don't really care at all if they have something in the story so I don't even know what I'm complaining about. It would be some major fan service, so why not? Maybe at least a dream scene some day, or an alternate universe / holodeck episode. That is totally happening, I can see it already.

Re: Why Jessica and Trish Getting Together on Jessica Jones Would Be a Bad Idea

dividebyzero — Thu, 22 Sep 2016 16:37:37 -0000

It's actually pretty cool they were able to create the feeling of "they might might not have a thing", it's maybe the coolest similar thing since Mulder+Scully. Whatever we might think or whatever might happens, fact is these two characters have a "forbidden love" thing going on, making people write about it, and this is pretty cool, so kudos to the writers for that...

Re: Why Jessica and Trish Getting Together on Jessica Jones Would Be a Bad Idea

dividebyzero — Thu, 22 Sep 2016 16:32:18 -0000

It just seems to me you might be looking for an even stricter test: two woman have a casual non-sexual conversation _despite_ they being lovers. Interesting.

I apologize if I seem like some kind of jerk, I just think there is a lot of fetishism in women/women relationship on screen. I am worried with the idea two woman might only talk if they are lovers, pretty much like a man and a woman too. You don't seem specially worried about that as I am, and that's OK by me. I just love L7 and other female bands, and I am tired of hearing people assume they must be lesbians just because they are a bunch of woman doing something together. It feels to me that Trishica shipping has a bit of that.

There are plenty of 3rd and 4th wave feminists out there who denounce lesbian fetishism on screen by the way, I didn't invent this. It's great to have more female and homosexual representation, but it is less cool when it is due exploitation, more specifically to the pleasure of male viewers (what might actually be inevitable, it's kind of a brutal reality from the nature of the medium, thought by some to be inherently male-centric, what even makes female movie-makers inherently transgressive feminists, it's crazy)

Hey, I don't have a problem if they have a thing, some of my favorite female character couples are lesbians, and some of my favorite movies don't even pass the original Bechdel test, unfortunately, not to say the jerky extended one. I found Jessica Jones pretty cool, even better than Daredevil, even if Matt and Foggy had a thing. Peace out.

Re: Disney Thankfully Pulled This Racist Moana Costume From Their Store

dividebyzero — Thu, 22 Sep 2016 15:59:39 -0000

There's a theory "cultural appropriation" exists so there can be "pure races" and cultures, so there can be "affirmative action" to support such needy cultures, so people may feel the need to keep politicians who fight for that in power. Just a theory!

Disqus - Latest Comments for divide_by_zero

Re: A Job Is Not Alienating; It Is Empowering

Re: Why Is Brazil So Happy?

Re: Data Modeling in FP vs OOP

Re: Emacs: Effective Windows Management

Re: Emacs: Effective Windows Management

Re: Emacs: Xah Fly Keys Customization

Re: Emacs: Xah Fly Keys Customization

Re: The Big Problem With Self-Driving Cars Is People

Re: The Neophyte's Guide to Scala Part 12: Type Classes

Re: AMD Ryzen 7 1700X CPU Review

Re: Have We Forgotten about Geometry in Computer Vision?

Re: If Haskell is so great, why hasn&apos;t it taken over the world? And the curious case of Go.

Re: True fullscreen in xmonad

Re: OpenVPN Configuration Guide for Linux/Ubuntu Version 16

Re: Why Jessica and Trish Getting Together on Jessica Jones Would Be a Bad Idea

Re: Why Jessica and Trish Getting Together on Jessica Jones Would Be a Bad Idea

Re: Why Jessica and Trish Getting Together on Jessica Jones Would Be a Bad Idea

Re: Why Jessica and Trish Getting Together on Jessica Jones Would Be a Bad Idea

Re: Why Jessica and Trish Getting Together on Jessica Jones Would Be a Bad Idea

Re: Why Jessica and Trish Getting Together on Jessica Jones Would Be a Bad Idea

Re: Why Jessica and Trish Getting Together on Jessica Jones Would Be a Bad Idea

Re: Why Jessica and Trish Getting Together on Jessica Jones Would Be a Bad Idea

Re: Why Jessica and Trish Getting Together on Jessica Jones Would Be a Bad Idea

Re: Why Jessica and Trish Getting Together on Jessica Jones Would Be a Bad Idea

Re: Disney Thankfully Pulled This Racist Moana Costume From Their Store

Re: If Haskell is so great, why hasn't it taken over the world? And the curious case of Go.