Python Bytes - #1 Intro to the show and pip 9 is out!

Episode Date: November 7, 2016

See the full show notes for this episode on the website at pythonbytes.fm/1

This is Python Bytes, episode number one, recorded Monday, November 7th, 2016. Hello, everybody, and welcome to Python Bytes, where we deliver Python headlines directly to your earbuds. This is one of your hosts, Michael Kennedy, along here with Brian Aukin. Hey, Brian, how's it going? It's going great. I'm super excited to be doing this podcast with you. I'm really thrilled to be bringing these little tidbits to everyone each week or biweekly, whatever we decide to go with. Yeah, I'm really excited too. I think this is definitely something that's missing out in the community right now.

Yeah, the goal is you subscribe to Python Bytes. And if you listen every week, you'll basically get a lot of the top headlines that you need to know about. So let's jump right into the headlines. The first one that I want to talk about is PyData, specifically PyData from Washington, D.C. The videos for this conference are out. So if you didn't make it to PyData D.C. and you want to check them out, there are 64 videos, which is about 30, 35 hours worth of content on YouTube right now. And so a couple of the videos that I thought were really great, you know, looking through them, where one of them was Elasticsearch and Redis, how and when to use them. And I don't really know

too much about Elasticsearch. So I'm really excited to watch this video and learn more about it. Another one's the one by Renee Tate. She hosts a podcast and a website called Becoming a Data Scientist. And she did a talk on advice from her podcast guests about becoming a data scientist. So I think that's really cool. One that was also cool was visual diagnostics for more information about machine learning and basically tuning your machine learning. And then one about data reproducibility and scientific computing called You Got Your Engineering to My Data Science Addressing the Reproducibility Crisis. Yeah, I was really impressed with these videos.

And there's a couple of things that I really liked about them. I like the format that shows, you can see both the slides and the presenter on a little kind of a picture in picture sort of thing of the presenter talking as well. I think that was done well. I also really like the – man, these are like 35-minute

or half-an-hour presentations, and they're just packed with data. Yeah, it's not like people are trying to fill an hour, hour and a half. They're like, here's the essence. Let's go. And it's really great. And, you know, I definitely agree with you on the quality. I'm really, really impressed with all the Python conferences about how high quality their AV work is and how good of an effort they make to basically record and share, not just with the attendees, but with everybody, whatever went on there. Yeah, it's pretty impressive. All right.

What do you got first, Brian? Oh, all right. One of the things that came up is another tutorial about EasyArgs. It's a command line library for making arguments better. I actually didn't take too much of a look at it. I just think it's one of the things I wanted to point out was it's kind of interesting that we have yet another way to make command line arguments easier to parse. I've lost track now of how many extra libraries there are. Yeah, let's see. ArgPars, Click, DocOpt, and a bunch more, right?

Yeah, I was just thinking that it seems like a rite of passage now. So if you're serious about being a Python developer and open source and blogging about it you have to do two things you have to write a tutorial about how generators work and you have to make your own command line library that's awesome i think you might be right there i haven't done either of those but um but i i actually do like um i'm making fun of it a little bit but but I'm glad that people are doing it. If everybody's trying to solve the problem, it means there's a lot of people not happy with the current solutions. And I'm one of those people that likes to write command line tools. So I say more power to them to write another. Yeah, absolutely. More power to them. That's awesome. So the next one that I

wanted to talk about is something called safety-db, which is on GitHub. And, of course, we'll put all the links to everything we're talking about in the show notes. You can check them out in your podcast player. But safety-db is interesting because when we write and deploy our applications, especially web applications because they have the biggest surface area and they touch the most people, right? They're publicly out on a server. We don't just deploy our code. We deploy all the packages that we're using in our code and then the packages that those packages depend upon and so on. So, for example, on my training website,

I've got a mailing list and I use MailChimp for that. MailChimp itself depends upon DocOpt, speaking of command line parsers and so on. How do I keep track of all these things and know if there's a security vulnerability? Like what if there's a problem in DocOpt that would let people hack into my server? I didn't install DocOpt and maybe it's not even on my mind. I don't even realize that it got installed. This GitHub set of data keeps that information current so I can check hey does my project have any

vulnerabilities that I may or may not even be aware of through these this sort of hierarchy of dependencies I'm using wow that's what I didn't realize it was that that in full yeah so I'm I'm just starting to get into more of the web development aspect of it but I do I mean just even with what I do now I depend on MailChimp and other services. I didn't even think about that, of looking at their vulnerabilities. Yeah, isn't that crazy? And there was something with one of the libraries that Flask is built upon a while ago that caused some big hoopla. And I'm sure it got patched right away and it was no

big deal. But if you don't know that you need to go patch your system that's that's a problem so safety db yeah great cool i'll check it out check it out all right so next up we've got a github project that came to my attention this week um called fast style transfer and it's um, an interesting library that, that uses, um, TensorFlow, which I've never used TensorFlow. Have you used TensorFlow, Mike? No, I'm, I'm just starting to work on a project, um, a data science project that is probably using TensorFlow, but I've only heard good things. I have not yet used it. Okay. Well, this, this one is something that makes me want to give it a try because it uh it

the demo on it looks really cool you you take a um a painting or a picture of some artist style and you can apply that style to another picture like your own picture and i've seen something like that before but what i'm really excited about is this will apply it to videos as well and i've always i guess ever since the Ahas Take On Me video, I've liked that notion of like an artistic style over a short video. It just seems like fun. That sounds really fun.

So I could like take my training videos and turn them into like a Van Gogh type of thing. Yeah, that'd be great. I don't know if people could read your slides but um but they would feel very creative while they're watching it no i i can't i can't think of actually any commercial reason why i would try this but but the artist in me wants to give it a shot no it sounds really cool and a chance to play a tensorflow also cool. So another big piece of news that came out this week is PIP. So PIP has a full new major version.

PIP 9 is out and comes with a couple of new features. One is the ability to check the installed package dependencies to see if everything is set up correctly there. You can use PIP show in a less verbose way. You can also say pip list and give it a not required option, which will show you all the packages that don't have aren't there because of their dependency on a thing. They're sort of top level, which is really cool. There's a ton of fixes that came as part of this release. And as you might expect with any major release of some new thing, four days later, pip 901 was released with five fixes for bugs that were introduced in pip 9 yeah um and actually there's a there's the list of what's in pip 9 is is big it is and i'm

one of the things that is cool is there's there's features in there that i didn't even know existed um so like the pip check is like that's really cool. I'm excited about that. But also they're changing the list format or they're going to. So if you do pip list now on nine, it'll give you a warning that in the future, the default format will be a column format. And the column format you can add now with a dash dash format equals column. And it's really easy to read. So I'm excited about that. Yeah, I'm really excited that this is here as well.

Packaging is such a foundational, important thing in the whole ecosystem. And so anything that makes that better, I'm for it. Yeah, and I use, like most people, I'm using virtual environments all over the place. And keeping track of what's in every one of them is pretty cool. Yeah, it definitely is. Okay, so I'm going to jump to one. keeping track of what's in every one of them is pretty cool. Yeah, it definitely is. Okay. So I'm going to jump to one, the next topic that you brought up, which is a Reddit. Actually, it's a sort of a blog post and a Jupyter project about the new MacBook Pro

and what Reddit users think of the MacBook Pro. And at first, you brought this up and I'm like, this is ridiculous. I know that everybody's upset, like half the people are upset, and half the people are excited, but most of the people I talk to are developers, and they're kind of ticked about all the changes of the MacBook Pro. But this Reddit article, or this article about incorporating Reddit, it's not about, it's not about the

specifics of that. It's using a sentiment. It's talking about sentiment analysis to analyze all the words used in these Reddit threads and, um, and, and try to determine from this, the words, if people are excited or upset about it. And I had never even heard of sentiment analysis. So I think it's neat to show a data analysis project that is topical on something that people are talking about right now. Yeah, because we all have a thing that we care about. Either we work for a company that makes a product or has some sort of public persona, or we have an open source project. And it would be nice to know, are people liking my work? And this ability to apply sentiment analysis to it, it sounds really cool. And I

think the MacBook Pro is a great thing to just grab because it's timely. Yeah. And I think it'd be kind of neat to, it was interesting to apply it to that and probably safer than applying it to something like current politics because you don't want to get in the middle of that right now. No, no, no, no. It's T-minus 28 hours or something. Nobody wants to get in the middle of that. Yeah, cool. So I really thought that was neat as well.

Personally, I have a MacBook Pro I'm holding out. I'm not buying this new one for a while. I'm going to see where this whole thing shakes out. Yeah, I guess I didn't want to get too much into it, but I was curious about that, if you were going to get a new one or not. No, I was ready to go customize it, turn all the knobs to 11 and say, ship it now. But it's just, I think there's too many things that are kind of weird and there's not enough of a benefit. So I'm just going to stick with what I got. Yeah, I guess. And I'm, maybe I'm just not doing anything

complicated because I'm on a couple year old MacBook Air. I just don't need very much power, I guess. So yeah, that's good news. Okay. All right. Well, that rounds out our news items for the week. But you know, what else what do you got going on? I know you've you've got the test of code podcast going, you're working on a book you just did some uh presentations at a conference why don't you tell people like what's what's the news around your life yeah well i'm i'm very grateful for november to roll around because um there's i know that um i i know a lot of people in startups and i guess i've heard that people are working in san francisco and stuff work like crazy hours but i'm not used to that I'm I'm a older engineer and I

liked I like my 40-hour work weeks and but October it has been insane I've been doing our projects at work which aren't I can't talk about and most people listening wouldn't care about but they're they've we've been doing like 60-hour weeks lately and on top of that I was doing the um it's Pacific Northwest Software Quality Conference. And I got to talk about, it was a couple weeks ago, I got to talk at that conference. And that was a 90-minute talk. And that's actually when I was looking at the PyData videos, I was impressed with all the information people got out in 30 minutes. Because it's like three times the amount of content that I got out in like 90 minutes.

But it was a good experience to talk. I'm actually kind of glad that my first conference experience was a long one, because I sure got my jitters out of the way, and I'm excited to do a future one. And the book's coming along good um i'm work i'm working with an editor and this this upcoming book is on pi uh focused on pi test uh alone and yeah that's excellent what's the title so people know what to look for we don't have a title yet um so the book about pi test yeah the book about pi test so um but i will definitely share more information with this podcast as it becomes available. Yeah, beautiful.

Hey, how about you? What's going on with you? Well, I've been on an absolute terror of recording TalkPython to me episodes. I recorded six weeks worth last week. So that's given that I spend about 10 hours of research and prep time on each. That meant a very long week. But I'm really, really happy with what I got out. And the next three that I have coming, you know, we talked about pip.

And there's actually a bit of a crisis in the Python ecosystem's core open source infrastructure. And so I'm doing a panel session, including the guy who works on PyPI and piped on stuff. And that's called, Are We Failing to Fund Python's Core Infrastructure? So that comes out this week. And also did something which I love the title, Parsing Horrible Things with Python with Eric Rose. And then Martin Peters, who is either the top or one of the top guys

at Stack Overflow as a user, a contributor, an answerer, has over half a million reputation in Python. And I talked to him about a bunch of the things going on with questions and answers and hard problems that they found on Stack Overflow. And that's just super revealing. I've learned a ton by doing that. Wow. I can't wait to talk about these.

Now, is Donald still on PIP? Donald is still on PIP, but the short the short short short version is he was working at hp hp went through a huge bunch of layoffs cut his whole division and by way of doing that basically defunded pi pi wow so so there's other companies like rackspace and um a couple of others i'm sorry i'm the names, that contribute tons of resources in terms of computing and server and bandwidth. Like the bandwidth and infrastructure charges for PyPI are $40,000 a month. But there's no funding for people to keep that infrastructure running. So anyway, that's what that show is about. I can't wait for that. Yeah, yeah. Check it out if you guys are

interested. All right. Oh, yeah. I forgot to mention, I did talk with Raphael Pierzina. I'm probably getting that wrong, but that's a podcast coming up for testing code focused on PyTest and Cookie Cutter. So that's coming up soon. Absolutely. All right. Well, thank you everybody for listening to this very first episode of Python Bytes. Brian, thank you for launching a new podcast with me. I think it's going to be a lot of fun. I hope people enjoy it. I think it'll be fun too. Thank you, Michael.

Yeah, absolutely. So if you are out there listening and you're like, oh, I have a great news item I want to send you guys, just visit pythonbytes.fm. There's a way right in the menu bar to click and actually send us your news. So if you find something that's cool and you want us to cover it in the next episode,

be sure to send it our way so that it's on our radar. All right. Thank you, everyone. Talk to you later. Bye, Brian. Bye. Thank you for listening to Python Bytes. Follow the show on Twitter via at Python Bytes.

That's Python Bytes as in B-Y-T-E-S. And get the full show notes at pythonbytes.fm. If you have a news item you want featured, just visit pythonbytes.fm. If you have a news item you want featured, just visit pythonbytes.fm and send it our way. We're always on the lookout for sharing something cool. On behalf of myself and Brian Auchin, this is Michael Kennedy. Thank you for listening and sharing this podcast

with your friends and colleagues.