Programming Throwdown - Funky Languages

Episode Date: October 15, 2014

This show covers Funky Languages: esoteric languages that are created mostly for fun. Tools of the show: Jason: Open Broadcaster Software Patrick: Plex. Books of the show: Jason: Impro for Storytellers: http://amzn.to/1sMohKv Patrick: The Android’s Dream (John Scalzi): http://amzn.to/1wHkOL0 ★ Support this podcast on Patreon ★

Hosting provided by Host Tornado. They offer website hosting packages, dedicated servers, and VPS solutions. HostT.net. Programming Throwdown, Episode 37, Funky Languages. Take it away, Jason. Hey, everybody. So, yeah, we got an interesting question from many interesting questions yeah we got several interesting yeah definitely several interesting

questions um i try to answer them on uh you know google plus slash facebook slash twitter wherever they come slash email but uh this one was particularly interesting, so we wanted to discuss it on the show. So Will wrote in and said, basically, what do you guys think about moving to Silicon Valley? I live on the East Coast. He lives in a small town in Georgia. I'm assuming he's doing tech, developing mobile applications. And he just wants to know, should uh you know make the trip over to silicon valley so i don't know what do you think i'll i'll uh i'll let you have the first shot of this there are many factors which should influence one's decision to relocate oh god first

uh so no no but seriously like um if you've never visited and you haven't like if you're just thinking like should i interview the answer is yes you should probably interview well that makes sense more generally you should interview uh part of that is you'll probably get a chance to come out here because we are in the silicon valley so that's why i say here right so you get a chance to come out and visit see what it's like but i mean everybody will tell you that property is really expensive expensive to buy a place to live in or even just to rent a place at least in the really uh kind of main part of silicon valley but there is a reason why everybody wants to live here that is that there's a lot of good places to work and you know the area is beautiful the

weather is great um and people love paying taxes. People love paying taxes. I think you make a really good point. If you interview, provided you make it past the phone screen, you basically get a free trip to Silicon Valley. That's for most big companies. Every company is a little different, but most big companies are that way. Even startups, actually. is a little different but most big companies are that way even startups actually like uh when i was looking around when i was still on the east coast um uh startups would also pay for you to for you

to come okay yeah at the time i was interviewing at tesla and uh at the time they were like really small company and they they uh they totally flew me over there or offered to fly me over there ended up picking a different company. But, but yeah, so, so, you know, you can interview, get a free trip,

make sure you, if possible interview on, you know, like a Monday or Friday. So you can come the weekend or at least, that's true, you know, take some days off or something.

And be upfront with the recruiter, right? Because their job is to be very nice to you. And I was, I'm always, I give this advice, but I know I was always very timid and asking for things or accommodation basically but you know if you have a significant other a spouse or something don't take advantage of it in my opinion but like you know if they're debating whether or not they would also like to move you should say

hey like as a as a couple we're determining this like is it possible for my wife to join me on this trip or whatever yeah that's that's a really good point i mean i i didn't do that and i kind of regretted it i mean it it worked out fine but uh it was a little nerve-wracking for my wife to uh like literally her first day in california was the day we moved and it's like, there's really no going back. And so, yeah,

if I had to do it over again, I would have asked, you know, if you think about it, if you interview at like a big company, you know, uh, you know,

two or $300 plane ticket. I mean, they, you know, they're not gonna, it's not gonna set them back and it could really do. And having another person isn't really that big a deal because like, you know, they don't have to pay for another hotel room or really anything um so there i did ask it

from one recruiter um not for the bay area but um for another area and i did ask and they said that i could i didn't end up taking them up on it because i ended up taking another offer um but i they were very you know sure of course you know like that's no problem and i had actually already flown there and interviewed and i was asking to fly me and my wife together out again to do kind of like an exploration trip and they were okay with that as well yeah right so um that totally makes sense so getting back to the question oh yeah um yeah no no i mean so you're right that you should definitely interview and all that so we didn't waste time or anything okay but but the uh should you move Silicon Valley so um here's my take on it so it's a little bit of a different question but it'll come back to the main point somebody asked me how come we can't just recreate Silicon Valley you know in Florida

or or somewhere else right and there's been many attempts. There's a place in India that's supposed to be like a technology triangle that's trying to be like an incubator. There's a lot of places all over the world that are like kind of tech incubators. And a lot of them get traction. And of course, they're successful to some degree, but none of them are close to the success of Silicon Valley. Right. And what I think about this is that you have sort of this pool of talent and a lot of the information you need to be a successful tech company is sort of latent. you know it's it's lessons learned it's not something you can really find in school especially now because you know there's so much disconnect like people in silicon valley aren't going back to university and teaching or anything like that so there's just so many lessons to be learned and so much um sort of what's the word for it where it's like uh you just kind of have to you know be grandfathered into it there's

not really a way to learn other than yeah it's like this this language that's evolved over time that basically you know to be a successful company you need a couple of say you know silicon valley veterans and so if there was some kind of migration out of the valley then we could get these incubators and other places but at the moment you know the best place and one of the valley then we could get these incubators and other places but at the moment you know the best place and one of the only places to get sort of veterans who have been doing you know big data or or you know practical machine learning or whatever um is is here and those people aren't moving so um so there are other but there are other places like i i read stuff a lot um about new york right it's not to the same degree of startups but like new york city has a lot of

tech jobs and and yeah that's true and it somewhat i think is partly just the you need a certain density of people and then a certain percentage of those people to be you know technologically minded advanced computer science people like that kind of stuff and the areas like if you take uh you know jason and i are both from florida if you take the areas in florida where we were even if there may be a lot of people the density is relatively low and uh the number of people who are from a technology background is also very low right and there's a chicken and an egg problem because let's say you're a brand new company. Like you're some startup, you just have an idea.

You know, you're Ubar and you want to be, you know, a phone app where people can request bars that they show up, bartenders and they show up. You know, you're going to have like a little bit of funding to get started and you need to get some very talented people or at least one very talented person to build, you know, version one of the system. That means you're going to have to find someone who's willing to quit their job and take a huge risk to join your company. And that means you have to go through a lot of people like

it might be a one in 10,000 chance that you find someone who's in that position. And so if you go to, say, Florida, where there might only be 9,000 people who have that kind of skill set, now you're really in trouble, right? And so because of that, you don't have this sort of snowball effect and so silicon valley is like one of the few places where you can um where you can go and find some very senior people to work on something you know that's just an idea right yeah so so i would say should i move it's obviously subject to a lot of personal decisions and it is expensive to live here but i mean there are two things that are good is one is jason kind of talked about like those talented people it's also good for you as an employee because there's a lot of people to learn from uh and a lot of really smart people typically um right yeah there are not smart people

too i guess but there are a lot of smart people and if you can find them you can learn a lot not even at your own company but just like everywhere right it's all around you yeah definitely the other thing is like, for me, it's much better to move on my terms. Like I decided like I wanted to move and moved out here. And then now if something happened and I was forced into needing to get a different job or there was a downturn in my company, right?

Then I would be forced to relocate. And if I'm here, the chance of me having to relocate away is uh low compared to where i was before yeah that's true now there's still the problem of like what happens like what happened in 2000s with the tech burst tech bubble burst um but i'm already in the tech industry so like in general it's going to be a bad time for me regardless of where i am yeah definitely definitely so uh yeah actually it's kind of it's quite interesting when the tech bubble burst i don't know if there was necessarily like uh like people out of a job actually no i take it back yeah i read stories about people like i think

it got pretty bad yeah yeah that's true but uh to be honest i find it very unlikely i do think that there is a little bit of a bubble now, but I think if it bursts, there's such a lack of talent that the companies that are established will just swoop up all of the talent. I don't feel like that well will run dry.

So basically, the long answer is it's up to you. I think it's fair to speak for for both of speak for both of us I say that you know we love it here but but yeah you should definitely bring you know your significant other if you have one and and and give it a good evaluation you know before you come so yeah so speaking of Silicon Valley and startups how does a a startup work? Yeah. So, you know, I think, you know, we've been in Silicon Valley, you're completely inundated

with startups, like reading about startups, friends who go to work for startups, things like that. But, you know, I was thinking about this the other day that when I was in Florida, I really had no idea sort of what a startup is or anything. And, you know, we're both tech savvy people, but it's just not a, it's really just a part of Silicon Valley culture. And so I felt like I wanted to try to share that with everyone. So, you know, neither Patrick nor I have ever worked in a startup.

We both have friends who are at startups or have started startups. And yeah, basically, so this is kind of how it works in a nutshell. And Patrick, please correct me because I'm sure I'm going to mess stuff up. But you typically try to start by contacting venture capitalists. So you make a very quick prototype. So some common examples are the Zappos, that Zappos shoe sales website. The guy actually started by going to Payless, like a real brick and mortar, like a Payless shoe store, taking pictures of the shoes there and then going on his computer and making

an HTML page selling those shoes for I think it was like a 10 or 15% markup online. And then literally, if someone would buy the shoes, he would walk down the street to the Payless, buy the shoes, mail them himself, right? So it's like super, super low tech. But it was enough to go to a venture capitalist and say, look, I've been doing this for a month and I got, you know, 2000 people ordering shoes. And then what will happen is another way to do it is they have these. I don't know if I'd really call them competitions. I guess maybe like incubators. like Stardex and some of these other ones where you sort of enter a competition and you pitch an idea, you know, with varying degrees of completion. And then they choose, you know, some winners and

those winners, you know, get connected with venture capitalists. But your goal is to get connected to a VC or a venture capitalist. What they will do is they will give you some money, like maybe $5 million or $10 million. But they don't give it to you. They give it to the company. And then as part of giving the company this money, they also sort of set up the company and they set up all the infrastructure you need to run a company. So you and the venture capitalists

together figure out for example how much you should get paid so like however much you should get paid that's money that isn't going into the company and so you know also like how much of the company you own like if you want a really big salary then the venture capitalist is going to say okay but now you're only going to own 10% of your company instead of owning like, you know, 50% of your company. So it's sort of like how much do you want to risk and things like that. Once you've sort of set all that up, then now you have this money. You can use the money and your prototype to try to get, you know, some of your friends or other people in the valley to join your company and and you can pay them and it's the same deal for them like you can give them money and you can give them

you know a piece of the company and they can sort of decide how much of either they want to take and then uh you're off you're up and running and you're a startup um then you go through some rounds um so the deal is you know eventually you'll burn through your say five million dollars You're up and running and you're a startup. Then you go through some rounds. So the deal is, you know, eventually you'll burn through your, say, $5 million. Like you'll spend it all. And then you'll say, okay, well, we have a good company, but, you know, we're not making enough money to cover our expenses yet. So then you go back to these venture capitalists and you say, look, I took your five million dollars and I built up this company and the company last year made let's say half a million dollars so what I do need now is I need ten million dollars so that I can like continue to expand the company and this this and this and this and this reason

is why you know I'm gonna turn a profit and all that stuff and so they give you some more money to build a bigger company and uh you know at some point you become very big and you become sustainable so for example like look at uh uber for example so uber is still a startup technically because you know they haven't ipo'd um or anything like that um but uber is as far as i know is actually i don't know for a fact but i believe they just closed a new round all right oh they closed i think it's one of the probably not this is careful right so like there's always a trade-off between how fast you grow and you know funding your growth from your profits so i i believe i'm correct in saying that i think uber has continued to raise more uh investment capital uh to be able

to expand even faster to have a first mover advantage so basically they're trying to get to as many stories as possible as many cities as possible and so they could slow down with the cities they are and be essentially profitable but it costs a lot of money for them to roll out in a new city and so they'd rather keep bringing on new cities rather than getting profitable and they're like one of the most well-funded startups in history so like they have been given like their investors agree with them basically and and that's what they've chosen to do so yeah what patrick is saying like another part of that is you know and groupon has this as well is is um or had this is uh it's very hard to sort of i mean you hear a lot about patent trolls and patent lawsuits and things like that but in general if

you have a business idea it's pretty easy to copy and it's also very hard to shut down clones especially clones in other countries and things like that so a company like uber they have a good idea they need to get their idea in as many cities as possible so that there isn't like a detroit version of uber you know run by a different company a houston version of uber so on and so forth so that they can own all of the cities right and so to patrick's point like it might be more important to own all of the cities so there's only one uber even if it's like going to take a ton of money and then they can make that money up later um so yeah so that part is pretty complicated. At some point, your startup becomes pretty big. And then, actually, let me step back a bit. So there's multiple ways for your startup to not become a startup.

The most common is it dissolves. So I think it's like over 90% of startups fail. And the number one reason they fail is actually in fighting. So the people who work at the startup, you know, clearly the people work at the startup believe in the startup to varying degrees, and are invested in the startup to varying degrees, you know, some of them are getting a bigger salary, but are getting less, you know, portion of the company as a result. And so that creates a lot of tension.

And so actually the number one reason why startups fail is that the people, there's social problems and the startup falls apart. Of the startups that succeed, there's two major categories. One is, and this is for more of like startups that are more technology and less of like a product um is that the startup gets acquired by a bigger company so let's say you know i make a startup where i do statistics for people where i built like some cool statistics system um that isn't really a company like i couldn't go to patrick and sell them statistics like it's really like a business service well and so it is a company and you can make money but jason's talking a lot about venture capital and it's hard for them they want to recoup a lot of

money quickly and so they're taking a lot of bets and they don't want to bet on someone who says over the next 10 years i'm going to make a million dollars each year but not grow i'm just going to keep making a million dollars like i don't have a business plan i don't want to make a billion dollars i just want to keep making a million dollars every year forever yeah that's true that's true they're not really interested in gaining just a percentage of profits each year they want you to have an exit they want you to as jason's saying sell out to another company or he'll talk about the other option right right like most of the time I mean there are exceptions obviously but most of the time if I if I if a company gets acquired it's because it has some kind

of service and you know usually if a company IPOs, it's usually like a product. It could be like a product that serves businesses, but sort of a very well-defined product. But if it's like a service, like we do machine learning services to other companies, that usually turns into an acquisition. So let me explain both of them. So an acquisition just means, you know, a bigger company says, hey, look, here's the startup. It is like, you know, say eight people. They're all really talented and they all work together and they're not fighting. They haven't destroyed each other. So, you know, it'd be great if we could get eight people who are all very talented to work at our big company. Like, it'd be awesome. We could just put them in a team together and we know there's not going to be any problems. So they'll say, you know, let's buy the

startup. And so they might make a pretty big investment. Like they might spend like, say, 20 or even a hundred million dollars and they buy the startup. But what they're really buying is they're buying those people. Like they know that some of the people on that startup will stay at the big company. Some of them will leave because you can't force someone to work for you. But some of them will stay. And they have a very cohesive team.

And they'll usually work something in writing where everyone has to stay for at least a couple of years and things like that. That's an acquisition. Another way that a startup can not be a startup is through what's called an IPO. And that's where a startup becomes a publicly traded company. So do you want to explain this? You'll probably be better than me.

Sure. IPO is an initial public offering. And that basically says that you're going to divide up the company into certain percentage shares and you're going to offer those to the public and there's all sorts of rules about how you can do this and if you look at an example like facebook is kind of the most recent famous one but tesla as well so these are often traded to a small number of people initially on what's called a secondary market, a closed market, where you have to meet certain criteria in order to be able to trade. In other words, not everyone can buy Facebook shares. You have to be an investor, basically.

Then they decide to IPO, which means that there's no company. They either just want to or there's no company big enough to buy them or no company that's interested in paying on their terms, whatever the reason may be. They decide that they were just going to sell some portion of the percentage to the public. And so they do what's called initial public offering. And you work through an investment bank to handle all sorts of complication and going to various retirement funds and getting them to agree to buy portions of facebook and uh you know all this stuff but basically what's going to happen is that uh it'll be regulated it'll be known you're going to declare uh here's all my earnings now because before you kept them private and you agree to all this stuff and say now the

public can buy and sell you know a share of share of Facebook. And for all those, say you sell, I'm going to sell 20% of the company and, you know, 20 shares. So basically each share is 1% of the company and Facebook keeps 80 of them and sells 20 of them to the public. And they raise money by the people buying them at whatever the initial offer price is. Right, Exactly. So, and then after, from then on, when it's traded,

like if you go into your brokerage account and you buy Facebook right now, Facebook doesn't get any money from that. Unless you happen to be, unless it happens to be that a person at Facebook is selling you that share, but most likely, you know, like Jason could sell me that share of Facebook. It's not coming from that company more. So they don't continue to make money off of those shares right so what

happens if like if you buy all the shares and you just own the company like no because they don't offer the whole company so like in the case of facebook mark zuckerberg now has the equivalent since after the initial public offering he owns we could look it up because it has to be public but like let's say he still owns 30 of the shares um but then his shares may be special shares where he gets like two times the number of votes so he may get like you know essentially whatever that would be works out to be but he would get you know more votes for his 30 of the shares uh shares than the rest of people. Or some combination of people basically hold. So in the simplest case where you don't do vote multiplying,

it's that Mark Zuckerberg and whoever his other co-founders are own 51% of the shares, and they only ever release 49% to be sold. Gotcha. Gotcha. So that's how they maintain ownership and then there's all sorts of clever ways to kind of keep ownership without having to own all the stock keep control i should say gotcha that makes sense so yeah so um that's how so yeah so that's startups in nutshell i mean it's it's it's it's known for being a lot of work for very little pay. But you're looking at the total expected value.

So you're hoping that if there's some kind of acquisition or something like that, that you can, like what they call cash out. So you might work for, you know, one fourth of what you could make as a regular employee at another company, but you also own, let's say, 1% of the company. And so then if the company becomes as big as Facebook, now you own 1% of $40 billion, which is what? $400 million. Which is a lot of money. So you're taking a chance. You're sacrificing your pay for a chance to do really well. And, I mean, statistically speaking, you won't do as well as a Facebook or Google,

which is highly unlikely. But, again, if it's something you really believe will be successful and it will be like a acquire or something like that, uh, uh, acquisition or something like that, then, uh, um, then, you know, you could do really well. I mean, if you knew for sure that you would get acquired or IPO, if that was a guarantee that then you, then you would work there. Like, like we would all be working at startups. No, no, no, no, no.

It was a guarantee. work there like like we would all be working at startups no no no no guarantee you're yeah the the entire the incentives would be structured differently and you wouldn't get the same payout you're absolutely right so what i'm saying is like uh yeah if if there was a guarantee that it could ipo and you were getting like like one or two percent of the company or something like that then everyone would do it but yeah then to pat's point, like once a company is like Uber, where everyone knows it's going to be successful, then they know that there isn't a risk. And so then there's no reason. You get directly less percentage of the company at that point. Right. But it can still be really profitable because a lot of times they still really need people and so yeah they're willing to it's it is easier for them to give stock than dollars

right right so you still can like i have no idea but i assume now you can probably still get a pretty good deal going to uber uh and getting stock because they're more easily payable to you in stock than than them but then you're having to hope that by the time you can actually sell that stock, that it'll be worth more than what it is now, which is essentially nothing because you can't sell it. Yeah, right. Yeah. That's another thing.

That's actually good to talk about is it's very hard to know the value of a non-public company. And we've seen this right like with Facebook you know when Facebook had the IPO they IPO it and I'm gonna get this wrong but I mean is it roughly like thirty or forty dollars a share which means you take say thirty dollars and you multiply by the number of shares and that's how much you know they think that Facebook as a company is worth but then like within a month they ended up at I think it was like twelve dollars a share so what that means is that the public thought Facebook was worth one-third of what Facebook thought they were worth and I mean it's more

complicated than that but but in a nutshell what that means is you know the people who were at facebook before they went public thought they were getting a certain amount of like they thought that they were worth something like their shares and everything and then within a month of the ipo they they found out that the shares are actually worth a third of what everyone thought they were worth and so there's a lot of uncertainty but um you know because of that risk um you know startups will give you more uh shares and things like that because they know you're taking on a risk that other people aren't willing to take so good all right i think it's time to move on to the news

okay so i got the first one this came from a reader bentz and he sent us this is it's kind of interesting if you have a github account and someone uh makes a pull request against your repo repository i don't know i'm probably getting the exact wrong because i use too many different tools um but that basically when that email comes in now in gmail it'll actually give you a little button on the side of the subject line that allows you to directly go and view it uh so essentially there's now integration between gmail and github uh for viewing uh updates and stuff that they come in that's awesome is that like uh did they have to work with the gmail team or is there like an api or something i was actually just wondering that and i clicked on i don and I don't know.

Oh, okay. Well, they're probably... I think it's like a way for the GitHub embed something in their email basically that tells like, hey, there's this thing, Gmail, this is how you render it or whatever. Gotcha. That makes sense. But because Gmail is so big, I guess it's worth it and so many github users use gmail but like you would probably

have to do something different like there's no standard i don't think so like they wouldn't now also get something in yahoo yeah right right yeah um cool so yeah my news is on rust so actually i we should cover this. This is a brand new, well, sorry, it's a brand new language to me. But I read about it when I saw this news article. So, the news article is Rust 0.12.0 is released. And so, I was looking into Rust, and I'm not going to say too much because we should dedicate a show to it. But basically, it's a language that's intended to be very fast for system software-y stuff, like C or C++.

But the compiler is very, very smart. And they've reduced sort of what you can do a little bit although it's it's it's almost as fully functional as fully featured as you know crc plus plus they just they just handicap you a tiny bit but in exchange the compiler is actually extremely clever and uh i think it's pretty cool i'm going to do more research into it and we'll have a whole show dedicated to it but uh but yeah you know take some time and check it out if you're into doing low level you know c c++ tasks um this could be uh really cool so and it compiles to like machine code and all of that so if that's important yeah so the next one is about uh intel processors underestimating the error bounds of sine. I think it might just be trigonometric functions in general by 1.3 quintillion.

Yeah, a little bit of sensationalism. Yeah. So what this is, there will be a link or you may have already seen or you'll be able to search for it. But so interestingly, you design a processor out of silicone silicon and um they you know it's kind of there forever it just stuck with whatever is there and intel has some had some famous kind of problems and what happens when things go wrong or what inputs and this one's a little more subtle and it's apparently been around for several years and the documentation around the various functions are incorrect and one of it is that i believe it's actually in the processor as far as i could tell when i was reading it now now

i'm thinking like i wonder because intel actually also makes a compiler um for their processors so i hope i hope i'm right in saying this is actually the processor because i read it with an assumption okay good you're right because i read it with an assumption and now i was like oh no maybe i okay anyways uh and so basically the interesting part of this is that when you do a sign of a number um that number could be pretty much you know the input could be anything so in in double or flow whatever the range of inputs is as big as you can support in doubles and obviously you really only want to define sine as a function of just like the first repeating interval or some repeating interval so

you want to take whatever the input is and essentially wrap it around because sine is periodic right so sine keeps repeating itself so you want to reduce whatever the input is and essentially wrap it around because sine is periodic, right? So sine keeps repeating itself. So you want to reduce the input to something that's in a very specific range and then give a really precise answer in that range. Else you're forced trying to represent the infinite part of sine out and it gets really complicated. And this is actually a really common technique for a lot of things where you have a number that doesn't overflow or have a limit but actually just keeps wrapping around and defining that wraparound function and it turns out that there was a problem when you give a number that is very close to sign or to pi very close to pi in the input and they had not as many bits of precision as they thought they had and so what it causes like a

whole like the output basically is far less precise than it should have been and it's actually in the silicon because of the way they're reducing that input the way they're trying to wrap it around yeah exactly yeah you end up with what's called catastrophic cancellation which is an awesome name. But basically, it just means, if you try to subtract two numbers, that so in other words, so just a quick recap on floating point, right? The way floating point works is a little more complicated than this, but let's just simplify it. Imagine if you had, say, a number, like,, 0.12345, right? So you could just represent that in floating point by saying 1, 2, 3, 4, 5 times 10 to the negative, you know, sixth or whatever it is, right? So that's sort of what they do. So they end up with

like a regular number and then they store an exponent. Now in their case, they do times 2 to the whatever, but let's just pretend it's 10 to make life easy. So if you have like 1, 2, 3, 4, 5, 6, and then another number that's like 1, 2, 3, 4, 5, 7, then when you subtract those two, all of those first numbers are going to go to zero and the difference is going to be one right so that's going to cause like your exponent to change a lot too right like you can think in your head like think you have these two big numbers but now

you subtracted them and you end up with this little number and so actually the little number you kind of have a lot of room to the right like if there were other numbers it was one two three four five six dot like point seven eight nine ten eleven or seventy nine nine nine nine nine you know those things after the dot become important like before it would they were small because you're this huge number and then dot and this tiny piece onto it, right? But as soon as you subtract two huge numbers, then the tiny pieces after the dot, all of a sudden they're important. But if you got rid of them because the number was big and they weren't important, so you chopped them off,

it's like you chopped off something that wasn't important, but it became important later and that's called catastrophic cancellation and so and that's really bad right and so that's exactly what happens when you take a sign of a number close to pi so yeah and so the reason you get such a large thing is because when you're dealing with 64-bit numbers it's very easy to quickly escalate things to the point where you have this quintillion counts of error or whatever yeah yep yeah exactly um math cool hard especially point stuff yeah floating point stuff is just total nightmare it's just tricky there's a lot of corner cases. It's really hard to test very well. Yeah, yeah, definitely.

Yeah, it's like deceptively simple. Like you think it's an easy idea. Like, okay, we'll just take the first few digits and then we'll put 10 to the whatever. Like we've been doing scientific notation since we were in sixth grade or whatever. So you think it's going to be easy, but then the problem is computers are like

doing so much math on the same number that uh even if you make like there's a tiny little approximation it can burn you that's what happens here and like you point out there's all sorts of cases where like in actual math the order of operations doesn't matter but in floating point math it does matter so like if you multiply a very large number by a very large number by a very tiny number like the order the associative rule of how you actually do that will change your answer yep yep and you have to be aware of that yep and even people who are trying to do it really well still end up doing it wrong sometimes yeah yeah definitely um cool so uh yes it's other news it's uh this is only for students unfortunately but if you are a student at high school university i think any any institution

is fine um then you can get the github student developer pack which looks pretty freaking amazing um it is just a huge laundry list of things that you get for free um so yeah you get like crowdsourcing from crowdflower uh you know you get a you get a pro github account um you get the unreal engine um you get send grid to like programmatically send emails. You just get a ton of amazing stuff for completely free for no cost as long as you can prove you're a student.

Pretty freaking awesome. If you are a student and you're listening to us, you're high school, college, whatever, get this. Sign up for this. It is incredible. Another thing, a little bit of a side note, but Amazon has something similar, and get that one too. You get, like, free, you know, computers, like, free virtual machines on Amazon.

Yeah, AWS, whatever. Yeah, yeah. So check that out. Cool. All right. It's time for Book of the Show. Book of the Show. So my book is called Impro for Storytellers.

It's pretty cool. It's this guy. He's a famous, or I guess not famous, but he's a very experienced improvisational instructor. And so he does a lot of like corporate events and things like that he also you know teaches uh you know people who have gone on to work at saturday night live or write for saturday night live or things like that um and he wrote this book and he kind of intended it for everyone he has a similar book that's kind of meant for people who want to get into improvisational comedy which is for me not that interesting but but uh this book is really cool it's kind of meant for anyone and it just talks about sort of how to be improvisational

and how to think creatively and um how to sort of think on your feet and um it seems pretty cool i mean i've i haven't got into it yet i've only ordered it so um oh just like a preview yeah i'll have to give a recap but uh you know i i read through uh some of the preview of it and it's uh it looks pretty awesome and so yeah cool mine is of course not helpful it is just a science fiction book the androids dream which uh is by john scalzi a reference also to the philip k dick do androids dream of electric sheep i initially thought there was something weird going on there but is i think it's just an homage to the other work and then you make reference of it in the book uh and so um but it's not at all related as far as like the storylines uh and this is the first book

by john scalzi i think i've read i may have read one other book by him um but he writes self-described by him as you know introductory science fiction so this is um much closer to just normal fiction but with science-y settings so i I often recommend books which are considered more hardcore science fiction, like very deep, all sorts of pseudoscience-y explanations for things and even real science or physics involved. This is really plausible. This is more just of a fun tale.

I don't want to talk too much about it because I never want to spoil anything. But basically, this is just more of a fun tale. It's in a sort of a science fiction setting um and it was a pretty it was a it was a very i guess light read but it was very good i enjoyed it cool cool cool so if you've tried to get through some of the other very large tomes that i've recommended and you haven't uh try this one instead cool sounds good yeah give it a shot tool of the show tool of the show my tool of the show is what i'm using to stream to twitch right now which is called open broadcaster

software um so twitch actually recommended it you know i went to twitch trying to figure out how to stream um uh just to stream the show just kind of for kicks and uh they said hey use this program oh do we have any viewers of our stream we have one person oh yes hello person so calin444 uh major props to you for checking out the twitch stream you know i only gave people an hour's notice. I wasn't expecting a huge turnout or anything like that. The thing that I really think is cool about Twitch is the chat, actually.

Actually, I was inspired to stream this because of our user question. Twitch plays Pokemon. Oh, God. A user question. The person said, what's it like to be in Silicon Valley Made me think we really should try to get A bit of a community going Where people who like the show

And listen to the show can meet each other Literally zero people And we have One person So Callan444 props to you Sorry you couldn't meet anyone But you can meet us uh for what it's worth um and uh yeah i mean we'll do this maybe you know a couple more shows and uh uh see see sort

of what kind of turnout we get but uh but yeah so i'm using this open broadcaster software and uh um it's pretty cool it's actually pretty straightforward to set up. The only trouble it gave me was I made some changes to the settings and I had to shut down the program and restart it for the changes to take effect but I didn't know that I, you know, they didn't tell me to restart the program. So just a minor nit. But you know, other than that it pretty much worked out of the box and uh uh yeah pretty cool pretty cool tool um also they

have a version for mac and a version for linux is coming so yeah cool mine is also video related sort of and that is plex uh which is at plex.tv or i think you can just search plex and this is a combination of things uh which is i guess somewhat controversial but i enjoy it which is a server that you run on i think they have windows and linux at least i run both the windows and the linux ones servers and um a client has a web interface, or also it has Android apps, iOS apps, whatever you want. It works on I have a Chromecast. And that's how I use a lot. It works on the Chromecast. And what it does, it will serve you all of your media, it does audio pictures, videos, you kind of run the server, you point it at where your media is stored. And if it's, for instance, music, it'll be able to access IMDB, properly populate it,

give you all the kind of what we come to expect ability to kind of search from it. At least I think that works well. And then if you have like your tablet or your phone and you have the app, you know, it's able to stream from your network and stream it across. You say, well well there's like a thousand ways to done this i've talked about various ways before upnp ways dlna ways um and plex i think actually does support uh any dlna client or maybe even upnp client i don't know um but the one nice things it does is it does will handle the transcoding and it's the first

one i found that actually just works so it it is mostly free um they there are some features like casting to a chromecast they actually you have to pay to get that feature um and i kind of feel about how does that work like where do they how do they make you pay so if you have in the app that actually does the casting and there's probably ways around it but i i didn't bother with it because i wanted to support them uh because it actually works uh and when you it won't if you haven't uh signed in with a essentially a paid account it won't give you the option to cast either or if you try to cast i think it tells you like oh you need to upgrade or whatever how is it a monthly fee or yeah so i actually just uh this is kind of i bought like the lifetime subscription

i've never had tivo and we don't have cable so i guess this is like my one of my cable replacement expenses um thing is like it might have recently gone up i saw emails about that but i think i paid like 75 once wow that's great but it's kind of pricey like it's 75 dollars and there are open source alternatives oh okay that's actually a lot dang um but but like i said it just works uh there's probably other ways of doing it and i like looked online um and you know tried to find a non-paying way of doing it like every good software engineer person yeah right not no i don't mean like i try to steal it but i just mean like i tried to like look at open source ones you know just like more native linux ones and there were but you know they don't really work on chromecast they don't really handle like

transcoding on the fly they're picky about you know what media you can use whatever this one just works and it works really well for me um and good this gets into one of those uh uh i'm not a lawyer i can't advise you you probably shouldn't. But I have a friend who, you know, has kids and wanted to be able to stop having to go pull out like DVDs of cartoons to stick on for them. And it turns out if you kind of record those DVDs on your computer and if there's a legal way to do that and you did that and you use this to serve them up to your TV, it makes great because uh it would make it great because you could then not have to go get the dvds and stick them in for your children when they wanted something you could actually just show them the pictures of all the covers and have them choose

which one they wanted to watch yeah this is awesome i'm definitely gonna have to try this out do they have like a free wait what does the premium get you again so i think that's just if you only if you want to so you can install the server, see that it's correctly picking up your stuff and streaming it over the web interface. And then it's just, if you want to Chromecast it, I think is the only thing I use it for. Oh, I see. It also offers you the ability to like do remote syncing. So like now that I have the subscription, I can, my phone can upload my pictures to

my server as backup, automatically uh and some other features like that gotcha gotcha cool that's awesome yeah i'm actually gonna install this tonight all right yeah well let me know like i've recommended even other ones similar to this on the show but this is the first one that i've stuck with because it actually seems to work really well yeah yeah that sounds awesome yeah i have a linux box that's plugged in the tv um that's running ubuntu that I'll definitely install this on. All right, yeah. A little bit of a tangent.

I installed FreeBSD on a virtual box, like virtual machine, just to see. Because it keeps coming up. Like this Plex, they have a binary for FreeBSD. And it came up at work and things like that. And so I was like, look, let me just look at this. Because I never quite understood what FreeBSD really was. And again, like, I played with it for maybe, you know, an hour.

So I still don't really know, you know. I think it's like the big difference between it and Linux is the license. So in other words, you can use FreeBSD commercially and it's okay. You could even modify the source code and not contribute back. I don't know. So there's sort of like political reasons why you'd want to use FreeBSD. But it seems like just not as good as Linux. It could be just because I'm used to using Linux.

But, like, for example, the first thing I did was I tried to, you know, sudo a command so that I didn't have to be root. And sudo, like, doesn't exist in FreeBSD. There's some other methodology for doing it, yeah? Yeah, either that or you have to log in. So I used to be into, like, Linux distributions and stuff, and I gave up, and now I'm just on ubuntu uh same way and it's just because like there's literally only one reason it isn't philosophical political nothing is because when i have a problem and i write in

ubuntu and then my problem i'm most likely to get my answer than any other distribution yeah that's that's like it sounds horrible and i like it's just enough people use it now that like i can typically find an answer to my problem i don't have to post my own thing i don't have to like hack on it for an hour um if there's some like oh you know like you were talking earlier about open broadcaster software is going to have a linux i know they'll most likely have a debian or ubuntu version uh because it's so popular and i can just install it and it's almost guaranteed to work i remember like when ubuntu first came out and uh like right off the bat it was just so superior to all the other distributions in terms of its user friendliness

and i remember like immediately saying like i was using i think gen 2 and ubuntu came out and i was like okay i'm done like they i immediately converted um and you know they just keep getting better and better and you know ubuntu is uh um has a huge like uh um like investor it's actually made in i think south africa or it's at least a South African investor who owns the Ubuntu company. And so it's just very well funded. And I think that's the reason why it's so polished. But anyways, a little digression. All right, we're on to funky languages.

Funky languages. languages so everyone get out your uh you know i don't know uh bell bottoms and your you know afro wigs and uh you're going back to the 70s going back to the 70s so uh uh so yeah so one of them we'll just kind of go through them and and and uh and talk about them briefly. So my favorite is, one of my favorites is Whitespace. And that's exactly how it sounds. It's all tabs. Oh, first of all, props to whoever suggested this. I don't remember who it is, but somebody wrote in and told us to do. You know who you are.

Yeah. So props to you. I'll see if there's a little break, I can look you up. But this is a great idea for a show. So whitespace is all tabs, spaces, and new lines. So literally, that is the language. So, you know, like one tab means something, two consecutive tabs means something else, you know, so on and so forth. So, you know, obviously you'll have to use an editor which can tell you where a tab is

versus four spaces. So you have to have one that sort of highlights the tabs. But other than that, you know, all of these languages are fully functional. Like you could write anything to them. They're Turing complete, but they're not necessarily anyone would ever attempt to write actual programs in them. Any, anything other than fun. Like we don't think these are productive or efficient uses of your time in any way. Yeah. Yeah. Yeah. Like don't bring these up in an interview. Oh, well, so, so this white space one, I literally had a conversation today at work where, uh, at, at the job I work at, when we do interviews, we do them on whiteboards or chalkboards,

but mostly whiteboards. And I said that I'm going to scare the next interview candidate by saying we're going to do coding in white space on the whiteboard. You could have a like, sort of like a Fight Club moment. Like at the end of Fight Club, the guy says, you know, he realizes he's well are you spoiler don't spoil um but but but uh that could totally backfire because you could say you have to write it in white space and the person could say i've just wrote every program possible or like i already did there's the perfect answer it's on the board that was a joke but yes yeah so yeah so so whitespace uh one kind of funny thing about it another funny thing about it is um um it it ignores any real characters and so languages like you know java and c++ ignore whitespace so um so you can actually write a c++ program that has a whitespace program embedded inside of it and like if you compiled it

with the whitespace compiler you'd get one program that worked and if you compiled it with the c++ compiler you'd get a different program that also worked but uh you know potentially did something different they could also do the same actually that would be pretty cool if they have a white space like they do both programs do identical things yeah if they have a c++ to white space transpiler but but one that respected the fact that like there's some spaces that have to be there you know like complicated not doing this in python oh yeah python this wouldn't work but it'd be pretty cool to write a program which took your c++ program and figured out and transpiled it to white space and then put the white space in the program but anyways that's crazy talk um okay so that's white space uh i'll be back with the next one i don't't have very many Arnold Schwarzenegger quotes.

But the next one is Arnold C. And most of the keywords and function calls are replaced with Arnold Schwarzenegger movie quotes. I actually don't know where some of these are from. And I probably will not do any more impersonation voices. I need to read Jason's acting book that he recommended. Yeah, right. So it's showtime talk

to the hand with hello world and you have been terminated that's not arnold that was terrible all right you want to try jason um yeah i mean so no hello world okay go ahead what's that go ahead yeah do it in the right the hello world goes uh it's showtime oh no no it's like uh how does he go like talk through the hand nope i work with some guy who's austrian he's austrian right arnold schwarzenegger um i actually work with a guy who's austrian and he does sound like arnold schwarzenegger which is do you tell him that because i'm sure no one's ever told him that i'm sure everyone yeah everyone tells him that um yeah and so when when the arnold c program exits instead of just you know exit with parentheses it goes you have been terminated yeah so um this one's only awesome if you say all of them in the Arnold Schwarzenegger impersonation.

Yeah, that's right. So real quick, the guy who had the idea was Ryan. So props to you, Ryan. He didn't give us a last name or where he's from or anything. But props to you. This is a great idea for a show, Ryan. Thanks. Okay.

So next one is involving more of your senses. That's right of just your eyes you should also involve your ears and hear the music of your code that's right so the next one is Valito velato but it's a so the input to the compiler is a MIDI file so in other words you you know, you play a song on your keyboard, you give that song to the Valeto compiler and out comes a computer program. So it's pretty wild. You could actually hear your source code before you build it. So this one was actually interesting because when I was reading up about it, it's actually you need to know if some music theory to even understand what it's saying so for instance to uh to have an assignment you know an assignment a variable assignment

then you have to play your second note as a minor third over the first note you don't need a third note and then the variable is a single note then then the expression. I don't even know what that means. I have no idea what you just said. An if statement is a third note being the perfect fifth. Okay. Do they have, like, transpiler? Like, can I take, you know, like, some code at work and turn it into Valeto and then play it?

I don't know. That would be cool. And then just like listen to what your existing one sounds like? Yeah, yeah, yeah. Be like, wow, this is what, you know, this library sounds like. And it even has allowances. So the command note, which is your root note, and these other intervals are on top of those to, you know,

say basically if or whatever, can kind of move around to allow for less repetitive like actually make it sound musical oh i see uh yeah so this one's interesting it's just beautiful to look at the program because it just looks like crazy like that that music that i could never play because it's not just one note at a time and uh they really should have a thing in wikipedia i mean not wikipedia wikipedia has this but they should have a thing on this particular page where you could hear the program uh yeah yeah wikipedia will pronounce something oh actually they have it it's right here okay hang

on i'm gonna play it oops hang on we're gonna get a dnca takedown here's hello world okay it actually there's like two-thirds of it left but yeah that's that's hello okay i heard nothing i was gonna say it sounds a lot like my programs just empty well you didn't hear anything it's okay though well uh hopefully we found it or this will be well i i definitely played it hopefully the your audio audience heard it all right all right cool cool maybe yeah maybe we'll check it later don't worry about it in the editing you should just you should just mux it in

okay yeah anyways it will have been muxed in yeah conversation yeah the I was gonna say the Wikipedia page has a link to the mp3 if you guys want to listen to it cool all right so um the next one is piet you want to talk about piet so uh our previous one was constructing musical programs this one is uh using your drawing artist ability to construct programs.

And this is actually, so Piet, I believe, is the name of, is it Mondrian? Mondrian Piet is the guy who created those kind of like abstract square paintings where like there's like a red rectangle and a blue rectangle. Maybe I'm completely. You're right. Piet Mondrian. Oh, I said it wrong. I think I said Mondrian and Piet. Anyways, Piet Mondrian.

So if you look at his pictures, there's like a classical, like I didn't know specifically, but I have seen his pictures before, like primary colored squares with like black outlines. It almost looked like a stained glass window in some ways. And he's kind of famous for those paintings. And so someone made

a programming language which uses that same similar style to actually be the program so it uses the colors and the pixel locations to create what the program is uh and of all the pages this one actually seems like the most filled out with like here's the compiler and here's like the documentation here's alternative ones and here's some gallery of uh you know collect ones this one seems pretty well put together yeah i love it and if you look at the if you go to the piet website and you look at the sample programs they have like hello world um they have one that prints like fibonacci numbers but then they have one it says towers of hanoi this program solves the towers of hanoi problem

don't ask me how it works i have no idea that's what it says you know you have a good um funky programming language when nobody knows even the person who owns the language website doesn't know how how your program works yeah so again i i have no idea that you could be at all productive in this especially since like it appears that some of the colors are fairly close like different shades of red so a if you're colorblind i don't i i think you're just in trouble oh yeah and then even if you're just like there's like different shades of purple represent different things and maybe if you're like a designer and you're like really into various colors like this would work but so somebody wrote a piet um transpiler like a piet assembler they're

calling it okay and uh um and so they took a video game a text adventure game, and compiled it to Piet. And it is, as you would expect, just like a gigantic. So if you're watching the stream, you can actually see what I'm seeing, which is just a gigantic image. And supposedly, if you compile this image with theX compiler, you'd get a text adventure game. Ooh. Fun. Pretty cool, yeah. But did you ever use, what was that?

Why is the name escaping me now? The one with the little turtles and you can move the turtles. Oh, yeah, Logo. Logo, yes. It's kind of like I feel like these programs should be animated, like actually do stuff when they run, like produce pictures as the output as well yeah like if like it seems anticlimactic that you draw this beautiful program of music or whatever and then the output is like an actual

executable that just runs and does normal stuff yeah yeah totally there should be like a memory visualizer or something oh there you go yeah yeah yeah i wonder if that you should get on that jason all your free time yeah i'll get right on my free time yeah oh man um cool all right yeah that's so there's a ton of these funky languages many more um yeah yeah definitely check them out we covered some you know most of the major ones the major ones i see you did it that was a music reference yeah oh man um so uh uh yeah we covered most of them but uh there's there's plenty more and uh you especially the ones that have you know the assemblers or transpilers where you can take you know your college project and turn it into um valeto or piet without any um you know work on your part that would be pretty hilarious to to submit that you know what i mean it's like i know no yeah well

i don't want to get anyone in trouble but like if you if you could somehow submit a piet drawing instead of your homework but then maybe like also submit your homework then uh programming interviews do not do programming interviews in velado and like drawing notes on the boards and bass clefs and no don't do that yeah could you imagine if someone said, yeah, I only write in Piat. I'm going to need a marker of each type, of each color. Oh, no. No. Okay. This is devolving quickly.

All right. Well, if we missed your favorite funky language, you can write us and let us know. Maybe we'll have to do funky languages part two. Yeah. Yeah, this is a lot of fun um thank you guys for your continued comments of uh that we're doing good that you don't like something that uh you have questions for us it's all really good recently i i was browsing uh the tech category of podcasts in itunes and noticed that of all tech podcasts so that includes things like leo reports like this week in tech uh you

know some really big major brand uh podcasts we were number 74 so that was pretty good i have no idea what arbitrary rating they're using but uh we were number 74 so i was happy to see our little icon in the list with all these other like i know what that podcast is. And Hey, like those people are big. Yeah. Yeah. So, I mean, thanks a lot guys for your, for your support.

I mean, you're the reason why, uh, why, uh, you know, we get, uh,

we get the downloads and the, and the ratings, things like that. So we definitely appreciate it. Um, you know, we're, we're doing,

we're going to do this Twitch stream for at least a couple more shows. So, uh, feel free to jump on. It looks like we're up to like three people. yeah so that's current viewers but how many like over the course have viewed our stream oh i haven't been looking uh but uh i was gonna i was gonna make fun of us and then i briefly went to twitch and noticed the one on their main like their home page that they were featuring was people sitting around a table playing uh pen and paper dungeons and

dragons so i'm not sure that we're that far off from that so maybe this isn't such a bad idea no offense to dungeons and dragons or wonderful people playing it i'm sure it was i've played pen and paper before it's been fun but uh um yeah so so uh you know thanks a lot for your support it's been awesome um you know uh next time I'll give people more notice. And you can jump on and talk to some of the other people who are fans of the show. What else? Hopefully we answered.

Oh, one other thing is, you know, last episode i talked about rejection sampling um so i briefly talked about this idea of like distributions and sampling easy distributions you know to get more difficult distributions and anyways um you know so so check back on the last episode if you want to recap of that but somebody was sort of inspired from that and then kind of did some research on Wikipedia and stuff like that. And actually implemented rejection sampling. And let me actually put it in the show notes. So check this link out. This guy actually implemented it in IPython notebook.

And so you can actually run it in the browser and uh um and see the result and stuff like that even as like visual talk about viewer feedback yeah it's pretty freaking awesome so uh so props to you for for doing that um and uh yeah and it actually made me check out ipython notebook like give it another look and actually i think it's pretty awesome um definitely a big fan so i don't think i said this last time but um i have a little bit of a confession to make so i think okay when uh when i was in what is this probably like uh fifth grade sixth grade science fair i actually essentially did this for my science fair project oh really you made like a notebook kind of thing no no so i made a random number generator out of

like electrical components that i'd found you know like some uh like in radio shack or whatever and you would push the button and it would give you a random number and then i collected a whole bunch of random numbers then like uh used i think it was like mathematic at the time to like plot them all you know with like a quarter quadrant of a circle and like determine how many were in or out of the circle and basically calculated pi using my random number generator as my science fair project wow did you win that sounds like an amazing project so no oh yeah my science fair project was was terrible it actually you're gonna laugh my science fair project literally was a solar powered flashlight

literally like i i felt like if you were like trying to look into an area but the flashlight could still have i didn't get it i didn't get it at first i was like why are you being so down on yourself that sounds fine and then i'm like oh wait like it only worked for like a minute because like the capacitor was tiny right it just is terrible but uh so sebastian uh rashka i'm sorry if i butchered your name but sebastian's the one who um actually wrote this ipython notebook and uh it's very and i wasn't putting it down by saying i did that like as a science fair project a long time ago i was just admitting how nerdy i mean if i was the judge you would have won first place no but i

think the judges didn't get it so i think like someone who did like do trees do better with rock music or not one oh yeah yeah so you were like a savant like no no it wasn't that at all it was just very like it's out there right people like i don't get it like yeah yeah anyways it was more math than science i blame myself yeah you ended up like two positions below the rock music but you did better than the guy who made facebook for the science fair project the judges are just that out of touch no oh no okay all right they were like y combinator judges okay wrapping this up till next time all right yeah we're getting in trouble so uh thanks a lot for the support you guys uh community is awesome we love the feedback the emails the questions uh people who take what we say and actually implement it. Totally props

to you guys. The next time we go on Twitch, we'll give you a couple days or a week notice so we can get more community. There's some people chatting in there, but hopefully we'll get more people on there and you can give us some

instantaneous feedback if we mess up or something like that. Alright. Sounds good. more people on there and you can give us some instantaneous feedback if we mess up or something like that all right cool sounds good all right have a good one guys see you later the intro music is axo by biner pilot programming throwdown is distributed under a creative commons attribution share alike 2.0 license you're're free to share, copy, distribute, transmit the work, to remix, adapt the work, but you must provide attribution to Patrick and I and sharealike in kind.