patrickallencooper

just tech infused philosophical meanderings

[patrickallencooper-blog]

OpenCV Build on Windows

It is such a rarity to actually find instructions on how to properly build libraries in that development earthquake that is Windows. I have to commend the “Learn OpenCV” blog on being one of the only end to end guides on how to successfully build the library on Windows as things currently stand. Just be sure to install VS 2015 with complete with the accompanying C++ tools as shown. In case you’re wondering if VS2017 will work here the answer is a definitive no. I can attest to this.

On top of many other associated issues, the real brunt of the problem there comes from difficulty CUDA 9.1 has with VS 2017. The two are incompatible. I’m hoping Nvidia will do something about this, but since we’ve already seen a patch come and go since the problem was identified in September 2017, this is starting to look more and more unlikely.

I’ve had success using this same build guide with versions 3.3.1 and 3.4.0.

[mathblab]

“An approximate answer to the right problem is worth a good deal more than an exact answer to an approximate problem.”

— John Tukey

Statistician known for: exploratory data analysis, box plot, coining the term ‘bit’

[hana--studies]

cute coffee shops n’ coding

[thingsprogrammersshout]

Things Programmers Shout #831

*code works* I’d like to thank stack overflow. // submitted by @bellaferra

[patrickallencooper-blog]

The honest truth behind our success.

[patrickallencooper-blog]

The Case for Cognitive Dissonance

It is a strange thing that we encourage struggle in athletics, but not in academics. The most mentally capable people I know are those that relish in the difficulties ascribed by mental dissonance. I meet intelligent people all the time who somehow manage to accomplish very little because they race from project to project, meeting resistance on each one until their cognitive dissonance tolerance threshold is reached and they rush off to the next problem without bothering to solve the problem they were working on. The problem isn’t willpower, these people have just as much willpower as anyone else, the problem in question has to do with a very particular kind of discomfort.

It leaves me wondering if we shouldn’t be focusing on a different kind of problem solving in schools, especially at the university level. With the advent of more and more capable automation, the demand for solving trivial problems rapidly continues to diminish. College demands of its students at the undergraduate level that they solve tons of really simple problems, without any real regard for truly difficult multilayered work, this is reserved for the graduate level. This leaves most graduates with the workplace capacity of a bucket of scripts rather than capable information workers.

When I first started working right out of high school I remember being terrified by the amount of time it took me to solve problems and the general uncertainty surrounding the tasks I was assigned. I was working on some really hard computer vision problems. Anyone who knows anything about the current state of the field can relate just how far from straightforward our solutions are right now. This shift caused a culture shock that sent me reeling. When you go from progression that is regimented to the point where each problem is labeled and sequenced, with the each problem’s individual substeps being likewise categorized and sequenced, to a world where the problems themselves are poorly defined it can be devastating. Yet, these real world problems are precisely the problems that are worth solving. The goals of each researcher are different and the tools are all hurriedly built for highly specific tasks with little care for the community they are a part of. This isn’t to say that this kind of work environment is bad. Throughout high school I had a nagging impression that many of the skills I was learning would be utterly useless later in life. My early adolescent self would be proud he was pretty close to the reality of the situation. The detective work involved in these more complex tasks is hugely rewarding, especially for the introverted plotters and schemers of the world, of which I am one. However, for others, those who lean heavily on the social reward grades, teachers, and a supportive framework this shift can be devastating. Current events suggest the millennial generation is highly susceptible to this effect for reasons that can be explained elsewhere.

Cognitive dissonance tolerance can be enhanced through a variety of methods. The best method for developing this tolerance is to assign tasks that are unsolvable and to dedicate oneself to explaining why. This is so seldom done, oftentimes when students are assigned these tasks they are merely asked to develop a poor or approximate solution instead of really understanding why it’s impossible this sort of problem and using this new knowledge to work around the problem. Knocking down the impossible is the first step to establishing a good way to approach a novel task, yet exactly zero time is seriously dedicated to developing this faculty in most students during their K-12 education. It is my belief that mentally ill individuals discovering strong theoretical principles has become such a norm throughout history because of their abnormally high intelligence working in conjunction with their brains which have been acclimatised over the years with large amounts of mental dissonance. This is perhaps why these individuals have been historically less likely as well to fall for dogmas within classical arguments surrounding the rationalist, empiricist divide and other disputes.

[patrickallencooper-blog]

Day 4, Uberman Sleep Schedule

Day 4 is here. Missed a day of posting on account of being quite busy with school and work. I think that overall this sleep schedule is encouraging me to have a generally better state of mind. There’s something oddly nice about knowing you’ve optimized the equation of your life. This is especially true for me as someone who functions very poorly in situations where time is unstructured and poorly formatted. From a purely aesthetic standpoint it makes living a more pleasant experience. In a final sense it lends a sense of authenticity to my life that it was otherwise lacking. Reporting progress at work and school felt sort of like an appeal to some sort of arbitrary amount of time I had to work on that particular task. Now, it feels like I’m offering up genuine effort.

On the downside my eye has started twitching. I need to keep my eye on the matter. 😉

Happiness (Emotional wellbeing): 9 Mental Aptitude (Memory, Speed, etc.): 8 Physical State (Ability to exercise, clarity of vision, etc.): 6.5

[patrickallencooper-blog]

Building Natively in 2018

It has been my experience that building libraries for large scale C++ projects, often with several having to stretch the most generous definition of cross-platform, is quite difficult. It may be useful to compare it to something we’re all usually a bit more familiar with that of factoring. Both processes are more art than science, both require an intuition of the processes and setups to work on, both involve rearranging data to fit the situation and both are easy to do properly, and yet miss out on the important setup that would have assured our success.

There’s something peculiar about building libraries natively in a world with Docker and Anaconda. In a way it is one of the least rewarding, most difficult, and yet simultaneously most mind-numbing tasks a computer scientist can dedicate himself to. It seems an endless slog of libraries which at some magical point in the past may have worked together but now are an endless series of deprecated features and classes which have to be individually isolated and brought to the forefront if they are going to be used productively. My intent here isn’t to complain, the benefits of C++ are apparent, even today… if not especially today. The importance of building natively and generally not allowing abstraction to creep too much into our functional designs cannot be overstated. So here are a few tips from the trenches:

1. Trust the library designers! It is tempting after building a few libraries, shuffling some environment variables, and unpacking a couple header libraries to imagine that you have the full story on every situation where builds are required. You do not. Always look for documentation and procedures. 2. Sometimes, close enough may actually work. Indulgences like this aren’t common in computer science or in technical fields in general, but often times you need to go into building a library with a firm idea of what you intend to use it for. If all you need are the binaries, don’t build it. That being said if you build a library, but all of it. Trying to isolate only what you want from a framework of references is a disaster waiting to happen unless the library is very simple. 3. Ask yourself often what platform is best for your use case and build only for that system. If the project is multiplatform, get ready to make concessions… large ones. You may have to alter one version greatly, and often expect to resort to entirely separate libraries to get the build working properly. 4. Keep a journal of exactly what you’re doing. Think of this as commenting for your builds since you won’t have the usual paper trail code provides.

[patrickallencooper-blog]

The Journey Continues...

It’s now day two of my uberman adventure and things are going quite well. So far no grogginess, though I have noticed it is slightly harder for me to coordinate myself as well as I otherwise might. This seems fairly minor, stuff like messier handwriting than I usually have and a slightly lower degree of control over vocal intonations, however, so far nothing groundbreaking. I am still able to function at work and at school as well as I otherwise might have been able to. Today: Happiness (Emotional wellbeing): 7 Mental Aptitude (Memory, Speed, etc.): 8 Physical State (Ability to exercise, clarity of vision, etc.): 7

[patrickallencooper-blog]

Starting the Uberman Sleep Schedule

I’ve decided to go cold turkey into the wonderful world of polyphasic sleep. It may seem a bit aggressive, but here I’ll be chronicling my experiences on the uberman sleep schedule and reviewing generally how my state of health and overall performance in school and elsewhere has been affected by the shift. I’ve begun by making some necessary preparations. This includes a large selection of teas for regulating my mood at night as well as learning a large enough amount about meditation to allow myself to relax through periods of mental disruption. And last, but not least, a neck pillow for allowing me to sleep in my car outside of work when the nap roles around for every four-hour interval. With regards to exactly what the uberman sleep schedule is, it is a series of 6, 20 minute naps interspersed throughout the day at equal intervals. This translates into a nap session every 4 hours for a grand total of two hours spent sleeping per day, though it is likely the inconvenience of locating somewhere to sleep throughout the day will cost an additional 45 minutes of productive time at least. Another important feature specific to me is that I have OCD, while I have the illness nearly totally under control lately, I should mention that sleep deprivation has been known to exasperate symptoms. I’ll be keeping a particularly close eye on this moving forward. Finally, the most common question people have asked me when I decided to go on this sleep diet, so to speak, is, “how are you going to work and attend classes with such a strict nap schedule?” The answer is fortunately for me that my work schedule and sleep schedule are flexible enough that I can make time for resting between classes, along with before and after work, with one nap sandwiched in during a break in the car outside. This is hardly ideal, but in my unique position it is possible to have a relatively uninterrupted day life while maintaining what the uberman prescribes. Each day I’ll be assessing my overall mental state. I’ll do my very best to maintain the schedule for two weeks. If at that point the data I’ve collected shows a decline in happiness and performance, then I will stop immediately. If not I may have found a new productive alternative to losing a third of my life Today: Happiness (Emotional wellbeing): 8 Mental Aptitude (Memory, Speed, etc.): 8.5 Physical State (Ability to exercise, clarity of vision, etc.): 8 Okay, let’s see how this goes! 😊

[patrickallencooper-blog]

Some things about the Google Pixel 2

Things that are great: The squeeze feature. Didn’t expect to love this, but I do. The speakers. Dual front facing… enough said. The buttons. The small stuff matters, the buttons have great longevity and feel consistently crisp. The camera, especially for portraits. Google has implemented some great DOF approximation for such a small device.

Things that are okay: The screen. No burn-in yet, hoping it won’t be a problem given mine isn’t the XL version. The battery. Great for a full day of intense use and not a second more.

Things that are bad: The gigantic bezels. What’s up with the webbing in front of the killer speakers? Feels cheap and I’m not expecting it to stay appealing for the duration of my phone’s life.

[patrickallencooper-blog]

Are blogs useful?

A few thoughts about what blogs are really useful for. “When tempted to do anything in secret, ask yourself if you would do it in public.” – Thomas Jefferson There is some sort of crazy notion that blogs are inherently self-interested. This might have some merit, but what do you have to do with it? The idea of publicity is more important here than the reality of publicity. It’s interesting for the heavily introspective to externalize their thoughts, feelings, emotions, etc. This notion not only should be discarded by the irrelevant, but also it seems likely to benefit a freethinking ecosystem. Many knowledge workers today are unfortunately trapped within internal dialogues for too long and for too often this creates a sort of internal and self-referential paradox where no kind of motivation can be provided. This might be remedied by simply inviting the creation of a narrative. That seems reasonable now doesn’t it?