Gonna be a total contrarian douche and say: there absolutely are stupid questions, its that there's no gain in shaming a person for asking them.

Maybe the stupid questions wouldn't be such a problem (and I'm not sure they are: if anything, the site is overly aggressive in closing good questions as off-topic) if there weren't stupid answers getting upvoted, which is much worse from what I can tell.

I think it depends a lot on what you’re asking. Stack Overflow is fine for error messages and even “why isn’t my code working”, assuming the latter is sufficiently novel and you’re lucky enough not to run afoul of an IRC mod brigade.

I really don’t need that kind of help anymore, though. If I have a question, it is generally nuanced or subjective. Most of my SO search are closed as off-topic or a poor fit for a Q&A format. (Clearly they aren’t either of those things, since I’m Googling for a Q&A about them.) Quite often I’ll also see questions flagged as duplicates, even when they absolutely aren’t - the mods simply aren’t technically skilled enough to distinguish between the questions, and nobody who understands the subject has the power to correct them.

tbf I’m not an active participant there, I’m only a passive consumer via Google search results. But it seems clear from even my passive use that SO has extreme user moderation problems.

SO is a distributed program subroutine in your browser that takes an error message as input and code monkey commands as output.

Of course the good questions are marked as off topic or duplicate. Reid may be right that 'stupid' questions exist, but not on sites where all the questions are stupid on purpose.

And honestly, the site is a godsend as far as I'm concerned.

Also, SO isn't the most annoying thing to come across when googling something. That would be forum threads where OP is told to "just google it".

The site sure is something. If Google stopped indexing them, it would take me a really long time to notice. It’s not something I’d wish on the students and junior developers of the world, but once you have enough experience with a particular stack every single SO result is a waste of space.

Obvious root cause: user moderated website that hosts expert answers attracts mostly non-experts who need those answers => experts are moderated mostly by non-experts

The Physics SE is pretty ****ty, and it's the one I have the most time spent on (for fun, not real use). They're very elitist about the types of questions they like to answer, until an advanced user answers a trivial question and then everyone jumps in and/or upvotes the overkill answer just because a famous user wrote the answer. They're incredibly particular about upvotes, to the point where alternative solutions to a problem that arrive at the answer in a different but perhaps less-novel way are ignored and left with 0 upvotes despite answer the question and answering it differently that the latest answer from QMechanic. This makes it very difficult to gain the rep needed to actually have an impact on the site.

It also makes me not want to answer unless I am both the first person to answer and I know I have the most "enlightening" answer to the problem, because **** putting time in TeXing up a proof or taking the time to really qualitatively explain the physics going on if it's just going to be ignored and sit at 0 votes.

E- I'm not salty about people not upvoting all of my answers, I'm salty about perfectly acceptable answers getting absolutely no credit because they don't come from the right person and making it impossible to move towards being a respected member of the site.

i really hate quora though. how many Indian high school students have to ask what it's like to take *random class* at *prestigious university* and then ask if they can get in with *project they worked on*

No matter how long you study the real numbers, the amount of counter intuitive results you learn about it never seems to end.

Let q_n be an enumeration of the rational numbers, and U_n=(q_n-Ɛ/2^n,q_n+Ɛ/2^n) an open cover of each q_n.

U=∪U_n should be a cover of ℝ, yeah? Given rationals are dense and so forth.


Well **** you it isn't. It's not even close to a cover, its Lebesgue measure is way smaller than the real numbers.

It's hard to imagine but it's possible to show that not every irrational number is in U.

It's amazing how every year a bunch of people ask the exact same questions as the previous year about college admissions. It's like they don't know how to research these things.

Just finished my last day of teaching this year! Now for the final death sprint through mathematics before a week of break!

The real question is whether to spend break coding for money or fun.

Just in case this is a factor, I should warn you that nobody who hires for software companies cares about your GitHub. The only experience that matters is your professional experience, and the only code sample they’ll ever look at is an assignment they give you.

Any of the blog posts you’ve read to the contrary were at best mistaken, but more often a deliberate lie. Recruiters sometimes lie about wanting to see your GitHub because they want to attract people who enjoy coding in their free time (i.e. people who will accept atrocious conditions and can be ruthlessly exploited) but nobody actually cares about the code you’ve produced. The target is really people who agree with the sentiment, not the people who actually follow through on it.

So if you’re spending the break coding as a career building exercise, do it for pay. 100% of the time every time.

Isn't Github a social media site? It could make sense to have activity on there if you'd like to build your professional network that way, i.e., by linking to your homepage and having projects with a good README.md that sells the conceptual / technical idea behind the repo. Although most people who professionally write software won't have anything on Github at all, because nothing they've ever written is open source.

I heard that the whole idea of posting junk to Github to demonstrate kind of metric of an intrinsic desire to code would never have worked anyway, because it's too easy to just fork a million projects and change a few lines.

Although this is probably just me talking from perceptions here. There are a lot of flashy projects that you'll run across on Github that do important things, but most people likely won't have a repo of any sizeable complexity worth sharing, unless they have a genuinely good idea they developed because they are in a graduate program or are already an experienced software engineer, or are otherwise are involved with a project that's already sunk a huge number of man hours into it. And at this point you might as well host the thing on your homepage and write a paper about it.

Technically yes I suppose, although I’ve always treated it as a free repository host for bad undergraduate projects I don’t care to back up myself.

But that isn’t how professional networks work - much to the consternation of open source devs who think it should work that way, e.g. “I wrote a library google uses everywhere but they won’t hire me unless I can reverse a linked list on a whiteboard”. Professional networks are a mix of tit-for-tat (reasonably likeable person who can do me favours in return) and genuinely good people who you’d like to work with again, or who will make you look good for recommending them. Throwing code on GitHub (i.e. over a fence) doesn’t accomplish that. In the former case, the favour was already done for free. In the latter case, seeing someone’s code tells you exactly nothing about how they’d be to work with - assuming you even cared to look at the code.

This, too.

Nobody looks at a candidate’s GitHub anyway. I never have, so I honestly couldn’t tell you if people are gaming them that way. It’s just an irrelevant line on a resume that I ignore.

There are a lot of really good open source projects and I don’t post this to poop on the people who work on them. Not at all. I’m just saying that, despite what everybody on HN says, it doesn’t contribute to a hiring decision anywhere. (What people say happens on forums is usually what they believe people should be doing, rather than what is actually done.) If you have a cool project, you want to share it with the world, and you can’t think of any reasonable way of monetizing it, by all means you should make it open source and post it to GitHub. Just don’t expect it will get you hired anywhere.

What can I really learn from reading your open source code? I can learn whether you test your code, get it reviewed, have a consistent style - all things I can enforce through policy and training, so irrelevant. It would also take me a really long time to read your project to understand even that much.

Can I learn how well you can read a spec? Follow instructions? Whether you ask follow up questions? How well you take criticism? How you’d adapt to my problem, instead of a problem you chose? No. And those are the things that matter most.

So given that, let's say that someone wanted to incentivize open source development of something, despite that it gets nobody hired, how do you do that. Are projects like this doomed to depend on idealists who don't care about furthering their carer and delusional amateurs who don't understand how to build a career?

Don't forget that programming today has plenty of roots in academia, if you go back through the decades. The idea of being someone who 'hacks' on software for the good of the community is probably a concept originating at the MIT a.i. lab in the `60's.

In fact, Richard Stallman is the literal embodied manifestation of your idealist, and his whole reason for starting GNU was the commercialization of lisp systems that came from the a.i. lab.

That's definitely true. I guess the reason I said open source development of something, is because I wasn't asking about programming explicitly. Let's say... open source technology to sequester carbon.

Not that it is relevant to this thread...

The best way is if the software product is an artifact of e.g. a professional services organization, and not the core competency/main revenue source of the company. Then the company has an incentive to both open source it (make it auditable and crowdsourced labor) and still fund its development (because having the artifact is important for delivering its primary product).

Some of these cases are more obvious than others. Like RHEL, it’s free (“free”) and funded by Red Hat’s support services. Having a Linux distribution that’s both weirdly well known to Red Hat’s support staff and also weirdly different from all other distributions is an important thing for them. It makes Red Hat irreplaceable for the companies that license RHEL.

Then you have ducknuts like rethinkdb though. Oh my goodness. They wanted to do open source product+support/consult but, well, if your dB is bad enough that you need to hire a consultant out the gate then nobody’s going to use it. No need to say what happened to rethinkdb. AFAIK the company failure also destroyed the open source project but actually I don’t care enough to find out.

So, while this is pretty much the only way to properly incentivize open source under our current economic model, it won’t work for every product, and you need to think long and hard before trying it.

Well, the #1 priority is gonna be that stupid food and rent b.s. that most engineers like spending money on. Basically any employed engineer is gonna be working under a strict IP assignment / antimoonlighting agreement, so unless you are paying for their food and rent, they aren’t gonna be able to work for you even if they want to.

That means you need to make money somehow. If the tech is open, you won’t make the money that way, so you’ll need to make it some other way that uses the tech - leasing, installation? Idk, seems like the real money is a traditional invention/patent/licensing scheme,based on a gamble that the incentives will eventually align behind carbon sequestration.

Also Jonesie, I would say Stallman might be somewhat of an outlier, but that also might just be because I am outside of the developer community and he is so notable.

This is definitely the problem I have been running into, even to get people to work on things as trivial as open source machine tools variations or a solar concentrator and steam engine to run them. Of course, most of the people I talk to want to put ****ing computers in everything instead of using appropriate technology, and I'm at a loss as to what the cultural solution for that malady is other than finding some people who are into this stuff and making a reality series that makes them look badass so everyone else feels left out.

This is also one of my main interests in blockchain type stuff. I figure the narratives around it are so world changey and ideologically similar (even if most of the participants are really naive) that if I can make it simple to send my non profit crypto donations that are tax write offs I should be able to get enough revenue to pay a few people. Especially if I can get a system of machines that people see as community scale carbon neutral (if I can develop the idea that it is carbon negative so much the better, but I doubt that's actually do-able) then some sort of carbon credit system on the blockchain could fund a few salaries. If we could actually be using this stuff to grow food that would probably supplement it somewhat, and allow me to make pretty compelling media about it, but I know few engineers who are content with not climbing some ladder or able to do that and pay off their loans.

But if people are going to be doing the blockchain ****, there's money to be had for me (hah) to put to use developing objectively useful stuff and getting the word out about it so that when climate refugees start showing up they can meet their needs out of old cars and ****, at least in my area. Or when I am a climate refugee I know what to do I guess. I don't know, we're so ****ed I guess.

Stallman was a loonie even in his heyday. His personal philosophy solidified around his own selfish ambitions on MIT’s computer resources - he wanted root on every computer, the operators wanted to keep idiot undergrads from ruining everything. There are noble consequences to his beliefs (e.g. companies should not be able to make computers that their owners cannot fully access and control) but those are incidental to what he really believed (total information freedom and privileged access to all computers by all people).

Computers and software was engineered even back then, and people did pay money for it. GNU/FSF also isn’t opposed to collecting money (you are allowed to charge for discs) but there’s an unaddressed tension there, between the freedom to do whatever you want (which necessitates open source) and the ability to charge for your programming work. This is probably why RMS squat in an MIT office as a physics MSc student instead of inventing an open source coding job.

I think Stallman deserves a good deal of credit for his vision, understood literally as a reaction to taking the MIT a.i. lab stuff commercial. Symbolics may have been the pinnacle of Lisp computing, but what good is commercial software and hardware if it just get's slaughtered in the market by cheaper workstations and ultimately PC's, resulting in all that awesome software just getting locked up in a cabinet somewhere? Making GNU into a virus means that today you're going to have far more people using Elisp than Common Lisp, even if it's inferior. Ditto for Linux w.r.t. BSD and whatever other stuff AT&T lawyers didn't want us to have.

That said, there's more to the world of software than what's visibly open source. I'm sure Symbolics' customers did lots of cool things with those expensive machines, even if the platform failed to thrive in the long run in the way that open source did.

That said, a lot of open source is still just happenstance. Netscape -> Mozilla was just a big F.U. to Microsoft on their way to the graveyard.

GPL isn’t a virus*, not really. And people use elisp because they use emacs, not because elisp is a good/open lisp. I think if RMS deserves any credit, it’s because he tapped into the zeitgeist well enough to organize people to create and share some pretty rad versions of the Unix userland tools - even if, individually, those tools are quite trivial.

Gcc was also super important since there weren’t any non-commercial c compilers at the time. So gcc was also more or less a prerequisite for all of the BSDs to bootstrap themselves as well. RMS deserves credit for getting it off the ground, even if he was responsible for GCC almost fading into irrelevance... twice? Three times?


* RMS/FSF make some pretty wild claims for what GPL2/3 effectively prohibit, per IP attorneys I’ve spoken with over the years. It’s always best to play it safe, but when it comes down to it the question is about derivative works vs interoperability, and no, judges do not understand or give a **** about whether the end users operating system happens to load a GPL library into the same address space as your program.

This is why e.g. nmap has a custom GPL that also restricts what you’re allowed to do with the output, versus the GPL which expressly authorizes you to distribute GPL software with proprietary software as long as they aren’t “too tightly integrated”, whatever that means.

Sorry, I got my computing history from the wrong disgruntled Lisp hacker. It was actually Unix and C that are the ultimate computer viruses. Which I guess is what GNU piggybacked onto.

Okay, that more or less lines up with how I was viewing him, but with more specific terminology.

I also love when the Stallman copypasta gets adapted to completely unrelated contexts:

I'd just like to interject for moment. What you're refering to as Linux, is in fact, GNU/Linux, or as I've recently taken to calling it, GNU plus Linux. Linux is not an operating system unto itself, but rather another free component of a fully functioning GNU system made useful by the GNU corelibs, shell utilities and vital system components comprising a full OS as defined by POSIX.

Many computer users run a modified version of the GNU system every day, without realizing it. Through a peculiar turn of events, the version of GNU which is widely used today is often called Linux, and many of its users are not aware that it is basically the GNU system, developed by the GNU Project.

There really is a Linux, and these people are using it, but it is just a part of the system they use. Linux is the kernel: the program in the system that allocates the machine's resources to the other programs that you run. The kernel is an essential part of an operating system, but useless by itself; it can only function in the context of a complete operating system. Linux is normally used in combination with the GNU operating system: the whole system is basically GNU with Linux added, or GNU/Linux. All the so-called Linux distributions are really distributions of GNU/Linux!

As well as the modified navy seal copypasta:

What the **** did you just ****ing say about me, you proprietary slave? I’ll have you know I graduated top of my class at Harvard, and I’ve been involved in numerous free software projects, and I have contributed to over 300 core-utils for GNU. I am skilled in Lisp and I’m St. IGNU-cius, saint of the Church of Emacs. You are nothing to me but just another unethical non-free software advocate. I will distribute the **** out of your source code with freedom the likes of which has never been seen before on this Earth, mark my ****ing words. You think you can get away with saying that **** about me and the GPL on the Internet? Think again, ****er. As we speak I am contacting my colleagues at FSF and your binaries are being reversed engineered right now so you better prepare for the storm, maggot. The storm that wipes out the pathetic little thing you call your copyright. You're ****ing dead, kid. Free software can be anywhere, anytime, and it can ensure your freedom in over four ways, and that’s just with the GPLv2. Not only am I extensively skilled at hacking, but I have access to the source of the entire GNU userland and core-utils and I will use it to its full extent to wipe your miserable proprietary code off the face of the continent, you little ****. If only you could have known what ethical retribution your little “clever” program was about to bring down upon you, maybe you would have ensured your users' freedom. But you couldn’t, you didn’t, and now you’re paying the price, you goddamn idiot. I will **** free as in freedom all over you and you will drown in it. You’re ****ing dead, kiddo.

I have also had some good fun talking about him as if he and Rick Rubin are the same person in front of CS majors.

I'd just like to interject for moment. What you're refering to as glibc, is in fact, Linux/glibc, or as I've recently taken to calling it, Linux plus glibc. Glibc is not a C standard library unto itself, but rather another free component of a fully functioning C runtime system made useful by the Linux kernel, syscalls and vital system components comprising a full C runtime as defined by ISO/IEC 9899.

Many computer users run a modified version of Linux system calls every day, without realizing it. Through a peculiar turn of events, the version of Linux syscalls which is widely used today is often called glibc, and many of its users are not aware that it is basically the Linux kernel, developed by the Linux Kernel Organization.

There really is a glibc, and these people are using it, but it is just a part of the system they use. Glibc is the wrapper: the library in the system that translates C standard function calls to syscalls that you run. The wrapper is an essential part of a C runtime environment, but useless by itself; it can only function in the context of a complete C runtime. Glibc is normally used in combination with the Linux operating system: the whole system is basically Linux with GNU added, or Linux/glibc. All the so-called glibc distributions are really distributions of Linux/glibc!

I'm sure that Rick Rubin has much better personal hygiene.

100%

Well idk, I don't think selling a bot is a marketable skill and the court history is against bot makers so I doubt I'll try selling it.

I'll probably just do money work because it's money, and money is awesome.

On a random note: default lua behavior is for tostring() of a table to return the memory address of the table. Apparently the ESO devs never overwrote this function so you can just print in the game console memory addresses to basically all UI elements.

I didn't realize this behavior until today, and it's slightly hilarious

https://xkcd.com/1988/

I am a thread contributor

Thanks for the contribution. I'll play. Our "devops" team recommended we use docker for a new project. We complied despite huge misgivings (not really related to docker, but rather their complete lack of ability to sanely build or support anything, ever) and it's been a constant struggle. We've run into bug after bug with docker itself. They are documented and they do eventually get fixed but it's weird to me because docker has been around for so long now and I figured major bugs had been dealt with by now. It really feels like I'm dealing with an alpha-quality product at best. And these are major, show-stopping bugs, where all our instances will just freeze due to underlying filesystem-related issues.

I've also come to completely hate systemd and journalctl.

All our previous components have been written for full virtual machines running on top of vmware ESX hosts, like tons of them. I've been at this company for so long and we've been using ESX since 3.x so yes, I've run into my fair share of vmware bugs as well, but they usually weren't "killed production" type bugs and they usually didn't kill everything at once. And really after we finally upgraded to 5.x some number of years ago things have been fine; upgrades after that have just introduced features, we haven't needed any bug fixes. Obviously there is a bunch more overhead for full VMs (hardware and operational) but we had already invested in both. So yeah, I'm not super impressed with docker :-|

That’s resume driven development for ya.

Do continuous strictly increasing functions send measure 0 sets to measure 0 sets? Taking votes.

Cantor Vitali need not apply, ****'s only monotone.

Turns out yes, and I scored 95% on the analysis final cuz of it. So **** yeah...

Congrats!

Thank you! This semester has been one of the most difficult times for me, the work load was insane. And real analysis is one of the most difficult subjects out there.

Aren't you mostly interested in algebra? I've heard analysts say the exact same thing but about algebra.