WEBVTT 00:00:00.001 --> 00:00:06.140 Hello and welcome to Python Bytes, where we deliver Python news and headlines directly to your earbuds. 00:00:06.140 --> 00:00:11.100 This is episode 318, recorded January 10th, 2023. 00:00:11.100 --> 00:00:12.320 I'm Michael Kennedy. 00:00:12.320 --> 00:00:13.460 And I'm Brian Okken. 00:00:13.460 --> 00:00:17.580 And this episode is brought to you by Microsoft for Startups Founders Hub. 00:00:17.580 --> 00:00:19.960 Thanks to Microsoft for again supporting the show. 00:00:19.960 --> 00:00:25.200 And I, Brian, just want to let people know real quick at the beginning here that they should join us here on YouTube. 00:00:25.200 --> 00:00:27.220 Like right now, live if they want. 00:00:27.340 --> 00:00:32.940 They could go to pythonbytes.fm/stream slash live or just go to the website and click live stream at the top. 00:00:32.940 --> 00:00:37.080 We typically record on Tuesdays at 11 a.m. at Pacific Time. 00:00:37.080 --> 00:00:39.560 And you can also see the older video versions. 00:00:39.560 --> 00:00:41.060 People want to find them there. 00:00:41.060 --> 00:00:42.900 So it's always nice to have people in the audience. 00:00:42.900 --> 00:00:46.880 And we also put information in there about future episodes. 00:00:46.880 --> 00:00:49.780 So if for some reason it changes, we keep that up to date. 00:00:49.780 --> 00:00:50.260 Or you do. 00:00:50.260 --> 00:00:50.440 Yeah, exactly. 00:00:50.440 --> 00:00:55.420 Usually what I'll do is I'll just schedule whatever time we're going to record a week in advance. 00:00:55.600 --> 00:00:59.620 And YouTube will notify you if you smash the bell, you know, that sort of thing. 00:00:59.620 --> 00:01:00.580 Smash the bell. 00:01:00.580 --> 00:01:00.860 Yeah. 00:01:00.860 --> 00:01:02.560 What do you got to kick us off with today? 00:01:02.560 --> 00:01:06.320 Well, I want to start with something like right out off the presses. 00:01:06.320 --> 00:01:08.960 So we have 703. 00:01:08.960 --> 00:01:11.420 This was suggested by Will Shanks. 00:01:11.420 --> 00:01:13.980 And the author is Sam Gross. 00:01:14.300 --> 00:01:19.200 And it really was the PEP was created just yesterday, January 9th. 00:01:19.200 --> 00:01:20.220 Or at least. 00:01:20.220 --> 00:01:20.640 Yeah. 00:01:20.640 --> 00:01:21.560 So that's pretty cool. 00:01:21.560 --> 00:01:23.340 So very recently. 00:01:23.340 --> 00:01:24.380 That new PEP smell. 00:01:24.380 --> 00:01:25.000 Yeah. 00:01:25.000 --> 00:01:26.500 It has that new PEP smell. 00:01:26.500 --> 00:01:36.860 And what also one of the things that the PEP thing, peps.python.org has a like a dark mode 00:01:36.860 --> 00:01:37.760 and light mode thing. 00:01:37.820 --> 00:01:38.920 I didn't know that. 00:01:38.920 --> 00:01:39.860 I was playing with that today. 00:01:39.860 --> 00:01:40.240 Anyway. 00:01:40.240 --> 00:01:45.780 So the PEP 703 is making the gil optional in CPython. 00:01:45.780 --> 00:01:48.700 I knew that there was talk about this. 00:01:48.700 --> 00:01:50.540 I heard Guido talking about it. 00:01:50.540 --> 00:01:52.600 Possibly going in 312. 00:01:52.600 --> 00:01:55.160 But I didn't know if it really was. 00:01:55.160 --> 00:01:55.760 And it does. 00:01:55.760 --> 00:01:57.260 I mean, it's still in draft stage. 00:01:57.260 --> 00:01:58.980 So it's not it's not promised yet. 00:01:59.100 --> 00:02:04.000 But it's it's targeting 312, which is the one that's being worked on right now. 00:02:04.000 --> 00:02:04.720 It's pretty exciting. 00:02:04.720 --> 00:02:08.640 So, yeah, a couple of things that jump out from reading just the title. 00:02:08.640 --> 00:02:10.540 Optional, not removal. 00:02:10.540 --> 00:02:11.100 Yeah. 00:02:11.100 --> 00:02:13.460 OK, that's interesting. 00:02:13.460 --> 00:02:14.100 I have. 00:02:14.100 --> 00:02:16.240 Like you said, it's still still brand new. 00:02:16.240 --> 00:02:20.920 So I haven't got a chance to research this a lot yet to understand the nuances of what that is. 00:02:20.920 --> 00:02:26.360 But, you know, a lot of the prior talk, we had the Gilectomy and other things like that. 00:02:26.780 --> 00:02:32.800 Eric Snow's subinterpreters, all of those were about how do we remove the gil or effectively remove it. 00:02:32.800 --> 00:02:36.240 And this is like, well, well, not really. 00:02:36.240 --> 00:02:39.500 It is not really optionally selectable while you're running. 00:02:39.500 --> 00:02:42.340 It's a the option is a compile time flag. 00:02:42.340 --> 00:02:49.460 So, yeah, this is and we probably should link to that that that really long interview with Guido. 00:02:49.460 --> 00:02:54.740 But there's a there's an interview with Guido Van Rossum that's that in it, 00:02:54.860 --> 00:02:57.160 talks about some of the guild challenges. 00:02:57.160 --> 00:03:00.180 And it's an interesting bit conversation anyway. 00:03:00.180 --> 00:03:06.980 So there's a flag, a --without Gil flag that's going to be added for compiling CPython. 00:03:06.980 --> 00:03:12.720 So and when I first heard this, I'm like, well, I don't really want to compile CPython. 00:03:12.720 --> 00:03:17.460 I'm I mean, I could, but I'm not that kind of guy that likes to do that in their free time. 00:03:18.100 --> 00:03:19.300 So we have good news. 00:03:19.300 --> 00:03:20.900 I'll share with that with you later. 00:03:20.900 --> 00:03:34.080 But the idea around this is in there's a lot of a lot of interest because there's a lot more scientific computing and machine learning and all this sort of stuff that's running on Python. 00:03:34.080 --> 00:03:38.200 Now that wasn't before or just I don't know what it was doing before. 00:03:38.200 --> 00:03:40.740 But there's a lot of there's a lot of interest in this. 00:03:40.740 --> 00:03:47.140 And scientific computing especially has a lot of concurrency that you can speed things up dramatically. 00:03:47.140 --> 00:03:52.180 If you can run if you've got like these huge vectors that you can all compute at the same time. 00:03:52.180 --> 00:03:54.220 So there's there's ways around it. 00:03:54.220 --> 00:03:56.800 We have some really fast libraries that can do this. 00:03:56.800 --> 00:04:01.000 But these are like engineering feats that people have to do to make this all work. 00:04:01.000 --> 00:04:05.340 And they're usually in C API extensions and you and whatnot. 00:04:05.780 --> 00:04:13.900 So one of some of the motivation is that the there's a there's a lot of obstacles for concurrency because the GIL is in place. 00:04:13.900 --> 00:04:14.600 So what do we do? 00:04:14.600 --> 00:04:22.000 So the I'm going to jump to the distribution part, which I'm pretty excited about distribution. 00:04:22.000 --> 00:04:25.540 The distribution is going to be easier than I thought. 00:04:25.540 --> 00:04:27.740 So I thought people are going to have to build it themselves. 00:04:27.740 --> 00:04:30.280 It doesn't look like that's the case. 00:04:30.360 --> 00:04:40.160 It looks like Sam is going to work with Anaconda to distribute a without GIL version of Python together with compatible packages. 00:04:40.160 --> 00:04:42.400 And I imagine so this will take time. 00:04:42.400 --> 00:04:49.500 One of the things that's is going to be important is to get this out is like quickly so that people can start building packages. 00:04:49.500 --> 00:04:54.400 Because in one of the reasons why I'm glad this is like almost a different distribution. 00:04:54.400 --> 00:04:59.780 I know this is going to be a lot of work, but people can't really use it until the packages that. 00:05:00.120 --> 00:05:05.440 That like allow them to use it, you know, can are there straight pythons. 00:05:05.440 --> 00:05:13.100 The C layer might if you're integrating with C or something, it might assume that there's a GIL and not the safety it needs to do or something. 00:05:13.100 --> 00:05:13.660 Yeah, right. 00:05:13.660 --> 00:05:21.480 And even even if there's no changes that have to be made, they're going to be they're going to be have to these packages are going to have to be recompiled. 00:05:21.480 --> 00:05:25.580 But there probably are changes they have to make, but they're trying to keep it to a minimum. 00:05:25.580 --> 00:05:30.000 But anyway, they'll at least have to get recompiled and they'll be not binary compatible. 00:05:30.000 --> 00:05:31.560 With both versions of python. 00:05:31.560 --> 00:05:35.320 So we'll have this to python world for a little bit, but it's really targeted. 00:05:35.320 --> 00:05:38.660 And I think going through anaconda is a good way to do it. 00:05:38.660 --> 00:05:45.420 It's targeted to people developing these C API extensions so that they can start developing it. 00:05:45.420 --> 00:05:46.720 So that's pretty, pretty neat. 00:05:46.720 --> 00:05:54.260 And this collection of packages that already support the GIL lists python will grow over time. 00:05:54.260 --> 00:05:59.300 And I don't know, I'm okay with having that be coordinated through anaconda. 00:05:59.300 --> 00:06:00.120 It seems reasonable. 00:06:00.120 --> 00:06:00.780 Yeah. 00:06:01.160 --> 00:06:05.240 One of the things I loved about this article is it talked about some of the challenges. 00:06:05.240 --> 00:06:07.080 So zooming really fast. 00:06:07.080 --> 00:06:08.700 Apologies for people watching the video. 00:06:08.700 --> 00:06:19.480 But there's a lot of discussion around the motivation of all the lot of the difficulties of different types of parallelism that you are hard to express with with the GIL in place. 00:06:19.480 --> 00:06:24.000 And it's just an interesting, this part's an interesting read. 00:06:24.000 --> 00:06:27.700 There's effects on Python library usability. 00:06:27.700 --> 00:06:28.840 There's effects. 00:06:28.840 --> 00:06:35.820 You want to be able to GPU heavy workloads require multi-core processing to coordinate the whole thing together. 00:06:35.820 --> 00:06:41.400 There's a lot of stuff that we've done engineering wise to work around the GIL. 00:06:41.680 --> 00:06:47.240 And it's kind of exciting that this is, I didn't think, I actually didn't see this happening this quickly. 00:06:47.240 --> 00:06:48.620 So this is pretty exciting. 00:06:48.620 --> 00:06:50.360 Yeah, this came out of the blue for me too. 00:06:50.360 --> 00:06:55.980 The, one of the things that the discussion, oh cool, I found it. 00:06:55.980 --> 00:06:57.820 The, what's going to change? 00:06:57.820 --> 00:06:59.620 So I had this surprise to me. 00:06:59.620 --> 00:07:07.140 The surprise to me was that the, there's supposedly going to be relatively few changes to the public Python and C APIs. 00:07:07.440 --> 00:07:12.920 So this isn't like a moving from Python two to Python three, you have to rewrite your Python code. 00:07:12.920 --> 00:07:15.440 Hopefully you won't have to change your Python code at all. 00:07:15.440 --> 00:07:17.280 It's going to be some of these extensions. 00:07:17.280 --> 00:07:22.040 And these are, I mean, like the people writing numpy and things like that. 00:07:22.040 --> 00:07:31.760 Pandas, they might, they're probably going to have to do some changes and pillow and, and, you know, types or what, like pie torch and things like that. 00:07:31.760 --> 00:07:37.160 They really care about this and they're going to be working with it, but they, they're the ones that didn't want the GIL there. 00:07:37.220 --> 00:07:37.740 In the first place. 00:07:37.740 --> 00:07:40.160 So I think they're motivated to make these changes. 00:07:40.160 --> 00:07:41.740 It's pretty exciting, but okay. 00:07:41.740 --> 00:07:43.080 So what, what is changing? 00:07:43.080 --> 00:07:44.820 And a lot of the internals of CPython. 00:07:44.820 --> 00:07:53.560 So there's, and I think this is a fascinating thing just as a, like a computer science discussion of, of this is a big problem. 00:07:53.560 --> 00:07:54.600 What are the problems? 00:07:54.600 --> 00:07:57.560 Well, there's, and how do we, what are the changes to solve it? 00:07:57.560 --> 00:08:01.540 And there's reference counting and memory management, container thread safety. 00:08:01.540 --> 00:08:06.380 Like if you've got multiple, multiple threads using the same list, how do you deal with that? 00:08:06.380 --> 00:08:08.400 Locking and atomic APIs. 00:08:08.400 --> 00:08:11.060 And there's a great discussion in here on all those. 00:08:11.060 --> 00:08:18.400 So I really loved this just as a refresher of some of these CS topics and how they're still relevant in, in today's programming. 00:08:18.400 --> 00:08:19.420 So it's pretty cool. 00:08:19.420 --> 00:08:19.820 Yeah. 00:08:19.820 --> 00:08:20.320 Yeah. 00:08:20.320 --> 00:08:27.300 The container one's pretty interesting, especially with Python's extreme dependency upon dictionaries. 00:08:27.300 --> 00:08:27.940 Yes. 00:08:27.940 --> 00:08:30.720 Even being like the foundation of classes and stuff. 00:08:31.380 --> 00:08:35.400 The reference counting is the one that's always been held out as a holdup, right? 00:08:35.400 --> 00:08:41.740 Like, well, we have reference counting and that can be so fast because we don't have to do any locking or any thread safety around it. 00:08:41.740 --> 00:08:45.620 As soon as you remove the gill, all that stuff goes out the window. 00:08:45.620 --> 00:08:52.160 It's, it talks about a couple of building on, on one of the things that was done previously, which is pretty interesting. 00:08:52.160 --> 00:08:57.560 And, you know, Sam Gross behind this and a bunch of other folks have just put together a whole host. 00:08:57.740 --> 00:09:00.580 You know, a lot of this comes out of Cinder, I believe at Meta. 00:09:00.580 --> 00:09:03.760 There's a lot of small building blocks. 00:09:03.760 --> 00:09:05.920 This is not just like, well, here's the change, right? 00:09:05.920 --> 00:09:11.600 And one of the things that I saw in the reference counting section is immortal. 00:09:11.600 --> 00:09:13.040 I think it was immortalizing. 00:09:13.040 --> 00:09:13.960 Oh, yeah. 00:09:13.960 --> 00:09:14.660 Right. 00:09:14.660 --> 00:09:21.680 So one of the things was, can we make it so they basically, we're pretty sure that we don't want to ever remove these. 00:09:21.680 --> 00:09:27.420 So we're just going to take them out of the reference counting checks and just assume that they're going to be here for the life. 00:09:27.420 --> 00:09:28.360 Of the program. 00:09:28.360 --> 00:09:28.880 Yeah. 00:09:28.880 --> 00:09:33.380 Just for performance and memory caching and all that kind of stuff. 00:09:33.380 --> 00:09:39.920 However, you also don't have to lock something that you're not incrementing and decrementing because you've, you've decided not ever to do that. 00:09:39.920 --> 00:09:40.100 Right. 00:09:40.100 --> 00:09:40.160 Right. 00:09:40.160 --> 00:09:43.200 So there's like some interesting piling on effects. 00:09:43.200 --> 00:09:53.600 And also that was talking about, I didn't, I don't remember the word, but something about a certain, a lock free thread, thread safe type of way to do the increment for the. 00:09:53.600 --> 00:09:54.060 Well. 00:09:54.060 --> 00:09:55.500 For the reference counting. 00:09:55.820 --> 00:09:58.920 I mean, I was familiar with just like brute force reference counting. 00:09:58.920 --> 00:10:02.800 Of course, it's a, it's a common way to deal with shared objects. 00:10:02.800 --> 00:10:09.500 But the, one of the things I loved about this is the discussion around biased reference counting and deferred reference counting. 00:10:09.500 --> 00:10:14.900 The notion that like statistically looking at how multi-threaded applications work. 00:10:14.900 --> 00:10:22.740 Most objects really are only like most, most things are only, most objects are only accessed by one thread. 00:10:22.740 --> 00:10:25.300 There's not that many objects that are shared. 00:10:25.480 --> 00:10:41.600 So having like a, each object having a thread owner and the process for what reference counting is for the owner is different than the process for the, or, or different, you know, the non-owner, non-thread owner, you know, people accessing the object. 00:10:41.600 --> 00:10:47.000 And, and all of this is, hopefully invisible to, the people using it. 00:10:47.000 --> 00:10:57.600 I mean, clearly at some point we're going to have to have like some way in Python to be able to utilize this is like, it should, it, it can't be invisible forever. 00:10:57.800 --> 00:10:59.700 so it'll be interesting. 00:10:59.700 --> 00:11:00.180 Yeah. 00:11:00.180 --> 00:11:00.600 Yeah. 00:11:00.600 --> 00:11:02.820 The bias reference counting, that's the word I was looking for. 00:11:02.820 --> 00:11:04.020 That's, that's pretty interesting. 00:11:04.020 --> 00:11:06.440 Like the owner thread might not have to lock it. 00:11:06.440 --> 00:11:07.440 It just does whatever. 00:11:07.440 --> 00:11:09.680 And the others have to care. 00:11:09.680 --> 00:11:18.920 Well, and this also, a lot of this is around, I mean, we do reference counting, but partly because of the reference counting and how our object model is. 00:11:18.920 --> 00:11:30.320 I mean, one of the things I was surprised by when I moved from C++ to Python or picked up Python and with objects and everything is that like in C++, you totally know when the destructor happens. 00:11:30.320 --> 00:11:34.140 But in Python, we don't really talk about destructors too much. 00:11:34.140 --> 00:11:39.960 We have constructors, but we don't really put a ton of work in destructors because you don't really know when it's going to really go away. 00:11:39.960 --> 00:11:46.200 and, I think that's a good thing because now we can say it's actually going to be a little bit longer. 00:11:46.200 --> 00:11:58.780 You thought it was going to, you know, it probably would have disappeared earlier, but we're going to defer that and have a garbage collection phase where it also includes, making sure that the threads are concurrency is saved and stuff like that. 00:11:58.780 --> 00:11:59.040 So. 00:11:59.040 --> 00:11:59.560 Yeah. 00:11:59.560 --> 00:12:00.060 Interesting. 00:12:00.060 --> 00:12:02.500 It's a, it's a very fascinating read. 00:12:02.500 --> 00:12:05.340 I just had a quick read of it this morning, but, uh. 00:12:05.340 --> 00:12:07.640 It looks very clever and I'm, I'm all for it. 00:12:07.640 --> 00:12:08.200 I'm here for it. 00:12:08.200 --> 00:12:08.680 Let's do this. 00:12:08.680 --> 00:12:10.900 I don't want to like paint it too, Rosie. 00:12:11.180 --> 00:12:17.320 And everybody paying attention to this knows that there is a performance hit, for single threaded stuff. 00:12:17.320 --> 00:12:19.660 So, and I thought it was going to be larger. 00:12:19.660 --> 00:12:29.620 the reference in here talks about somewhere between a five and 10% performance hit, but we've already seen these like massive gains in performance. 00:12:29.620 --> 00:12:34.700 So, and the work on improving the performance of Python is going to continue. 00:12:34.700 --> 00:12:36.900 So I personally optimistic. 00:12:36.900 --> 00:12:46.460 I think that the gains in, in performance elsewhere are going to far outweigh the possible slowdown of, you know, single threaded stuff. 00:12:46.460 --> 00:12:51.280 I mean, a lot, a lot of, a lot of people don't really care about 5% difference. 00:12:51.280 --> 00:12:54.360 I mean, the code they're running, that's more simple. 00:12:54.360 --> 00:12:57.740 It, it's probably not performance sensitive a lot of the times. 00:12:57.740 --> 00:13:02.220 And when it is, you, you might be trying to squeeze even more performance out of it. 00:13:02.220 --> 00:13:02.480 Right. 00:13:02.480 --> 00:13:05.640 And, you know, my little mini here has eight cores. 00:13:05.640 --> 00:13:07.660 Python only cares about one of them. 00:13:07.660 --> 00:13:08.020 Yeah. 00:13:08.140 --> 00:13:17.720 If I could use all eight or six or something, I could do way more than 5% better, you know, so I could more than make up for that if I actually leverage this feature. 00:13:17.720 --> 00:13:18.300 Yeah. 00:13:18.300 --> 00:13:21.880 And if I don't care about performance plus or minus 5% probably doesn't matter. 00:13:21.880 --> 00:13:22.360 Yeah. 00:13:22.360 --> 00:13:28.020 I don't, I don't, I don't, I don't, I think they have to bring it up because there's going to be people saying, ah, but it's going to be slower. 00:13:28.020 --> 00:13:29.760 And so I'm glad that. 00:13:29.760 --> 00:13:29.920 Yeah. 00:13:29.920 --> 00:13:38.100 And that's also, that's always been the reason we still have the GIL is it's like, we would like to make changes, but they make a con the most common use case less good. 00:13:38.100 --> 00:13:43.080 So we're not doing it, but maybe this is close enough possible improvements even down the road. 00:13:43.080 --> 00:13:43.800 So yeah. 00:13:43.800 --> 00:13:48.300 I'm just, it's going to be, I'm excited to see how this, how this pans out. 00:13:48.300 --> 00:13:50.120 So yeah, I'm excited as well. 00:13:50.120 --> 00:13:50.920 Very, very cool. 00:13:50.920 --> 00:13:56.620 I would say certainly the biggest news of the show is right here. 00:13:56.620 --> 00:13:58.580 And we've got some good jokes and stuff. 00:13:58.580 --> 00:13:59.680 People can stick around for those. 00:13:59.680 --> 00:14:00.040 They should. 00:14:00.040 --> 00:14:00.320 Yeah. 00:14:00.320 --> 00:14:01.100 Stick around to the end. 00:14:01.100 --> 00:14:02.620 What do you got for us next? 00:14:02.620 --> 00:14:03.720 All right. 00:14:03.720 --> 00:14:06.360 Well, I want to tell you about the ferret. 00:14:06.360 --> 00:14:07.700 You heard of ferret DB? 00:14:08.140 --> 00:14:08.840 I have not. 00:14:08.840 --> 00:14:15.060 Like a mongoose, but it's not quite a, I honestly, if you showed me a mongoose and a ferret and said, which is which? 00:14:15.060 --> 00:14:16.160 I don't know if I could tell. 00:14:16.160 --> 00:14:18.620 Maybe the ferret mongooses are more intense. 00:14:18.620 --> 00:14:19.120 I don't know. 00:14:19.120 --> 00:14:20.740 But ferrets are more pet-like. 00:14:20.740 --> 00:14:21.940 They're pretty similar, right? 00:14:21.940 --> 00:14:23.700 Is Mongo from mongoose? 00:14:23.700 --> 00:14:25.700 No, it's from humongous. 00:14:25.700 --> 00:14:33.560 But there are some frameworks in the JavaScript space called mongoose from the Node.js side. 00:14:33.560 --> 00:14:38.680 Anyway, so ferret DB is this thing that I think someone told me about. 00:14:38.680 --> 00:14:40.980 Let me see if I got some credit. 00:14:40.980 --> 00:14:43.220 Yeah, this comes from John Boltmeyer. 00:14:43.220 --> 00:14:44.280 Boltmeyer, sorry. 00:14:44.280 --> 00:14:45.820 And thanks for sending that in. 00:14:45.820 --> 00:14:47.460 It said, ran across this. 00:14:47.460 --> 00:14:49.900 So ferret DB is pretty interesting. 00:14:49.900 --> 00:14:50.780 We'll see where it goes. 00:14:50.780 --> 00:14:54.300 It's a truly open source mongo DB alternative. 00:14:54.760 --> 00:14:57.660 By the way, I'm going to accept the cookies. 00:14:57.660 --> 00:15:02.540 You know, a cool way you don't have to have cookie policy pop-ups is not to track other people. 00:15:02.540 --> 00:15:03.580 That's even better. 00:15:03.580 --> 00:15:05.420 But I'll go and accept it for them. 00:15:05.420 --> 00:15:08.000 Our site doesn't have any, does it, Brian? 00:15:08.000 --> 00:15:09.320 Because we don't have third-party cookies. 00:15:09.320 --> 00:15:12.840 So 5,000 GitHub stars already. 00:15:12.840 --> 00:15:13.980 That's pretty incredible. 00:15:14.540 --> 00:15:22.880 And these folks are big fans of MongoDB and also really big fans of 100% true open source. 00:15:22.880 --> 00:15:26.040 And that's why they decided to make this thing called ferret DB. 00:15:26.040 --> 00:15:27.800 So here I'll read from their page a bit. 00:15:27.800 --> 00:15:35.140 MongoDB is a life-changing technology for many developers, empowering them to build applications faster than using relational databases. 00:15:35.540 --> 00:15:44.620 However, MongoDB abandoned its open source roots, changing the license to SSPL, making it unusable for many open source and commercial projects. 00:15:44.620 --> 00:15:48.800 Now, MongoDB is still open source, but this is like, I'm pretty sure this is a license. 00:15:48.800 --> 00:15:53.520 So basically, you can't make our thing as a service and then sell it without us. 00:15:53.520 --> 00:15:56.760 Like AWS had a MongoDB as a service. 00:15:56.760 --> 00:16:02.240 And Mongo is like, hello, we want to be, we didn't expect it to be used in this way. 00:16:02.240 --> 00:16:07.340 So I don't think it's, I mean, it's not like a huge bash against Mongo, I don't really think. 00:16:07.340 --> 00:16:12.380 But if you really wanted it to be unencumbered entirely, well, MongoDB is not it, right? 00:16:12.380 --> 00:16:12.800 Right. 00:16:12.800 --> 00:16:14.340 So here's the idea. 00:16:14.340 --> 00:16:22.380 So ferret DB is built on top of Postgres SQL, but it is wire compatible with MongoDB. 00:16:22.380 --> 00:16:31.820 That is, you can use the different libraries or drivers from different languages, you know, Python, JavaScript, .NET, whatever. 00:16:31.820 --> 00:16:36.800 You should be able to just change the connection string over to the ferret DB and off it goes. 00:16:36.800 --> 00:16:37.280 Wow. 00:16:37.280 --> 00:16:44.440 And it exchanges the binary JSON encoding across the wire, just like MongoDB does. 00:16:44.440 --> 00:16:51.180 I'm guessing they're using the JSON column stuff in Postgres SQL to match. 00:16:51.180 --> 00:16:54.380 I couldn't imagine they're trying to do a relational thing over there. 00:16:54.380 --> 00:16:56.060 They're probably just like stashing the blobs. 00:16:56.060 --> 00:16:58.260 But yeah, it's pretty neat. 00:16:58.260 --> 00:16:58.940 What do you think? 00:16:59.360 --> 00:17:08.680 That's actually, the architecture of that sounds great for an open source project to not try to do the entire database backend, but built on top of Postgres. 00:17:08.680 --> 00:17:09.360 Yeah. 00:17:09.360 --> 00:17:12.040 I mean, there's Edge, there's EdgeDB, right? 00:17:12.040 --> 00:17:15.260 Which is quite popular in the Python space. 00:17:15.260 --> 00:17:17.220 And I believe they're doing basically the same thing. 00:17:17.220 --> 00:17:18.460 Like build something. 00:17:18.460 --> 00:17:18.780 Yeah. 00:17:18.780 --> 00:17:19.840 Powered by Postgres. 00:17:20.340 --> 00:17:22.080 Maybe just the query engine. 00:17:22.080 --> 00:17:25.200 I'm not sure if the database engine as well for the EdgeDB folks. 00:17:25.200 --> 00:17:27.960 I interviewed them, but still pretty, pretty neat. 00:17:27.960 --> 00:17:28.920 I can get ahead with the joke. 00:17:28.920 --> 00:17:29.220 Yeah. 00:17:29.220 --> 00:17:40.360 So, FairDB, formerly MangoDB, is an open source proxy that translates MongoDB wire protocol queries to SQL with Postgres as a Postgres SQL as the database engine. 00:17:40.360 --> 00:17:41.540 MangoDB. 00:17:41.540 --> 00:17:43.100 That's a little bit too close, I think. 00:17:43.260 --> 00:17:44.080 Yeah, a little bit. 00:17:44.080 --> 00:17:45.020 Mango. 00:17:45.020 --> 00:17:47.080 Yeah, that's maybe pushing it. 00:17:47.080 --> 00:17:53.540 So, yeah, you pretty much write the regular MongoDB code that you would write, but it apparently works. 00:17:53.540 --> 00:17:56.760 Now, somewhere, I don't know where the roadmap is. 00:17:56.760 --> 00:17:57.640 Let's see. 00:17:57.640 --> 00:17:59.120 They did a blog post. 00:17:59.120 --> 00:18:00.260 Yeah, maybe that's where I should be looking. 00:18:00.640 --> 00:18:04.340 And on the blog post, they have something about their roadmap. 00:18:04.340 --> 00:18:04.940 I don't know. 00:18:04.940 --> 00:18:06.900 I'll link to it if I can find it. 00:18:06.900 --> 00:18:16.580 But there's a FairDB roadmap, and there's one really notable thing that people are going to need to be holding out for, and that's indexes. 00:18:16.580 --> 00:18:17.740 Oh, okay. 00:18:17.740 --> 00:18:24.080 And so if you don't have indexes, you have a real hard time doing fast database stuff. 00:18:24.080 --> 00:18:28.940 So I think they said they're trying to get it kind of all finished up by the end of 2023. 00:18:29.240 --> 00:18:34.280 So this is not like a, hey, you should just jump right on this and run with it. 00:18:34.280 --> 00:18:37.240 Because I don't think it's far enough along. 00:18:37.240 --> 00:18:39.540 It's written in Go, for what that's worth. 00:18:39.540 --> 00:18:45.620 MongoDB itself is written in C, so it's not like it makes it more or less friendly to Python folks. 00:18:45.620 --> 00:18:47.240 So, oh, here's the public roadmap. 00:18:47.240 --> 00:18:47.660 Perfect. 00:18:47.660 --> 00:18:49.660 So it's linked off their GitHub pages. 00:18:49.660 --> 00:18:54.700 And so you can see the end of alpha is supposed to be in the past, and it does look like they're all done. 00:18:54.700 --> 00:18:57.260 You can see their beta stuff they're working on. 00:18:57.260 --> 00:19:01.640 But then for general availability, this index stuff that's supposed to come. 00:19:01.640 --> 00:19:03.620 Actually, I said 2023. 00:19:03.620 --> 00:19:08.320 This is supposed to be done January 2023, which felt like a long time ago. 00:19:08.320 --> 00:19:10.240 Not long ago, but that's actually now. 00:19:10.500 --> 00:19:15.560 I don't know how far along it is, but support for issues, for indexes. 00:19:15.560 --> 00:19:22.260 Look, this issue here looks, I don't see a lot of traffic on it, but still, it would be basically required, right? 00:19:22.260 --> 00:19:27.280 If you've got many, many gigs of data, then you have to have indexes. 00:19:27.280 --> 00:19:27.840 Right. 00:19:27.840 --> 00:19:42.480 However, the devil's advocate here, development work on a project that utilizes this can proceed before you get tons of users and stuff, knowing that it's coming soon, I guess. 00:19:42.480 --> 00:19:43.380 Yeah, that's a good point. 00:19:43.380 --> 00:19:44.220 If you trust him. 00:19:44.460 --> 00:19:50.160 Yeah, although because it's wire protocol, you could just run MongoDB and then flip it theoretically, right? 00:19:50.160 --> 00:19:50.880 You don't have to. 00:19:50.880 --> 00:19:54.160 It's not like a complete new creation you've got to build upon, right? 00:19:54.160 --> 00:19:56.540 It's meant to mirror something that exists. 00:19:56.540 --> 00:19:57.440 So you could do that. 00:19:58.200 --> 00:20:05.720 I think we've got like five, six, seven gigs of data in the database that backs Python bytes. 00:20:05.720 --> 00:20:12.400 Like if you do queries against, you know, give me the 20 things out of this, these millions and millions of records that match this without an index. 00:20:12.400 --> 00:20:17.260 Actually, I think MongoDB may refuse to do it. 00:20:17.260 --> 00:20:23.820 There are certain situations where it's like too much data, no index, not going to be able to answer that question. 00:20:23.820 --> 00:20:28.980 So it might be really slow or it might actually error out and say this, you need to create an index. 00:20:28.980 --> 00:20:33.080 So anyway, long story short, like this is a big deal until this comes in. 00:20:33.080 --> 00:20:37.720 I would not touch it in terms of a, I'm going to try to run on top of this. 00:20:37.720 --> 00:20:38.000 Yeah. 00:20:38.000 --> 00:20:49.520 But if that, if that comes through, then you already have Postgres, but you want some kind of document like experience for part of your app or one of your apps, but you don't want to switch to Mongo. 00:20:49.520 --> 00:20:50.660 That would be pretty cool. 00:20:50.660 --> 00:20:57.220 So one of the things notable in the menu bar of FairDB is careers. 00:20:57.220 --> 00:21:05.920 So this is not, this is intended to be something commercial at some point, unless they just feel like hiring people for the fun of it. 00:21:05.920 --> 00:21:07.620 Yeah, that is interesting. 00:21:07.620 --> 00:21:09.760 So where are they going with it? 00:21:09.760 --> 00:21:09.900 Right. 00:21:09.980 --> 00:21:17.280 So I would imagine they have a FairDB as a service in mind out there, but you're right. 00:21:17.280 --> 00:21:20.020 It's certainly, it has an ink and. 00:21:20.020 --> 00:21:22.100 I mean, that's not necessarily bad. 00:21:22.100 --> 00:21:25.600 That's probably a good thing that there's money behind it, but just interesting. 00:21:25.600 --> 00:21:26.480 It is interesting. 00:21:26.480 --> 00:21:26.780 Yes. 00:21:26.780 --> 00:21:27.500 Very interesting. 00:21:27.500 --> 00:21:27.980 All right. 00:21:27.980 --> 00:21:36.120 Well, FairDB, something I put out there for people to keep their eye on, not necessarily to jump on right away, but quite interesting. 00:21:36.440 --> 00:21:36.640 Yep. 00:21:36.640 --> 00:21:42.400 Oh, by the way, we have a couple of excited folks out in the audience. 00:21:42.400 --> 00:21:44.820 A GIL optional would be amazing, says Ciro. 00:21:44.820 --> 00:21:49.300 And I know that Ciro is working on some real high performance web framework stuff. 00:21:49.300 --> 00:21:50.760 So that would be cool. 00:21:50.760 --> 00:21:55.100 And how about a Docker container prebuilt with CPython with no gill? 00:21:55.100 --> 00:21:57.120 Pick your Docker container and run that one. 00:21:57.120 --> 00:21:57.580 Sure. 00:21:57.580 --> 00:21:58.200 That'd be cool. 00:21:58.440 --> 00:21:59.480 Yeah, that's not a bad idea. 00:21:59.480 --> 00:22:00.020 Yeah. 00:22:00.020 --> 00:22:04.320 And Seth out there says, good thing we invested in faster CPython earlier. 00:22:04.320 --> 00:22:05.760 So now we can roll back a little. 00:22:05.760 --> 00:22:07.080 Yeah. 00:22:07.080 --> 00:22:08.580 I mean, kind of, we knew that was coming. 00:22:08.580 --> 00:22:12.600 So that's one of the reasons for the speed ups in 3.11 and planned for future. 00:22:12.600 --> 00:22:12.980 So. 00:22:12.980 --> 00:22:13.940 Yeah, indeed. 00:22:13.940 --> 00:22:14.440 All right. 00:22:14.440 --> 00:22:20.180 Let's switch over to real quickly talk about our sponsor before we move on here. 00:22:20.400 --> 00:22:23.920 As I said at the beginning, this episode is brought to you by Microsoft for Startups. 00:22:23.920 --> 00:22:25.140 We just got an email. 00:22:25.140 --> 00:22:28.740 I don't know if you saw, Brian, from somebody who signed up for the Microsoft for Startups 00:22:28.740 --> 00:22:30.420 program and said, this is a really great. 00:22:30.420 --> 00:22:34.580 I was a little bit skeptical, but I signed up and this is a really great project, a really 00:22:34.580 --> 00:22:37.600 great service and offering for them. 00:22:37.600 --> 00:22:39.600 And that they actually, it's working really well. 00:22:39.600 --> 00:22:39.880 So. 00:22:39.880 --> 00:22:40.620 That's great. 00:22:40.620 --> 00:22:41.200 Yeah. 00:22:41.200 --> 00:22:42.180 It was really good to hear. 00:22:42.180 --> 00:22:46.280 So, you know, Microsoft knows that starting a business is hard. 00:22:46.280 --> 00:22:51.620 They want to help support startups that can then later, you know, be customers of theirs, 00:22:51.620 --> 00:22:52.060 of course. 00:22:52.060 --> 00:22:57.920 So they put together this Microsoft for Startups program to help them understand what startups 00:22:57.920 --> 00:23:02.660 need to help them be more successful and to create a digital platform to help them overcome 00:23:02.660 --> 00:23:03.280 those challenges. 00:23:03.280 --> 00:23:05.940 And they call that Microsoft for Startups Founders Hub. 00:23:05.940 --> 00:23:11.480 And it provides all founders at any stage with free resources to solve startup challenges, 00:23:11.720 --> 00:23:17.940 provides tech benefits, access to expert guidance and skilling resources, mentorship, network 00:23:17.940 --> 00:23:19.020 connections, and much more. 00:23:19.020 --> 00:23:22.560 It doesn't require the startups to be investor backed or third party validated. 00:23:22.560 --> 00:23:23.520 You just apply. 00:23:23.520 --> 00:23:26.740 Microsoft says it looks like you would be a good fit for this program. 00:23:26.740 --> 00:23:27.400 You're in. 00:23:27.400 --> 00:23:33.340 And with that, you get access to a bunch of cloud resources, GitHub and Azure, Office 365. 00:23:33.340 --> 00:23:40.040 I think you also get access to open AI, which, you know, chat GTP, pretty hopping these days 00:23:40.040 --> 00:23:41.920 and might be something cool to leverage. 00:23:41.920 --> 00:23:47.020 And through Microsoft for Startup Founders Hub, becoming a founder is also no longer about who 00:23:47.020 --> 00:23:47.380 you know. 00:23:47.380 --> 00:23:51.620 You have access to their mentorship network, which I think is one of the biggest values you 00:23:51.620 --> 00:23:52.220 probably get. 00:23:52.220 --> 00:23:57.820 Talk to people about idea validation, fundraising, management, coaching, sales and marketing, and 00:23:57.820 --> 00:23:58.360 all those things. 00:23:58.360 --> 00:24:03.120 And you get to book one-on-one meetings with those mentors to make your idea a reality 00:24:03.120 --> 00:24:06.380 today with the critical support you'll get from Microsoft for Startups Founders Hub. 00:24:06.380 --> 00:24:11.000 To join the program, just visit pythonbytes.fm/foundershub 2022. 00:24:11.000 --> 00:24:12.280 The link's in your show notes. 00:24:12.280 --> 00:24:14.440 So if you're interested, give it a click. 00:24:14.440 --> 00:24:16.140 All right. 00:24:16.140 --> 00:24:16.600 Over to you, Brian. 00:24:16.600 --> 00:24:25.080 Well, I was, I want to guess transition is difficult for me apparently right now, as are 00:24:25.080 --> 00:24:25.460 words. 00:24:26.060 --> 00:24:31.920 Anyway, the next article I want to talk about is four tips for structuring your research 00:24:31.920 --> 00:24:33.380 group's Python package. 00:24:33.380 --> 00:24:40.500 So I think this is just a Python package, but research group, I think this is from David Aaron 00:24:40.500 --> 00:24:40.980 Nicholson. 00:24:40.980 --> 00:24:48.540 And I think the idea around it is a lot of people with scientific computing and various things 00:24:48.540 --> 00:24:53.360 like that might not be, might be struggling with lots of Python code. 00:24:53.360 --> 00:24:58.620 So they're putting it into subdirectories and stuff like that. 00:24:58.620 --> 00:25:04.820 So this is talking about packages in that sense of it's a directory with a __init__() and you 00:25:04.820 --> 00:25:08.080 can reference it by importing it through normal import. 00:25:08.080 --> 00:25:13.320 The import mechanism, it's not about packaging for like IPI or something like that. 00:25:13.320 --> 00:25:15.740 It's still some really great information. 00:25:15.960 --> 00:25:22.620 So the four tips are one, give your packages and modules terse single word names whenever 00:25:22.620 --> 00:25:23.200 possible. 00:25:23.200 --> 00:25:26.340 And there's some great examples of what not to do later. 00:25:26.340 --> 00:25:31.760 Number two, import modules internally instead of importing everything from modules. 00:25:31.760 --> 00:25:34.040 Make use of sub packages. 00:25:34.560 --> 00:25:41.040 And the last one is prefer modules with specific names instead of things like utils, helpers 00:25:41.040 --> 00:25:41.740 and support. 00:25:41.740 --> 00:25:46.920 And this is, this is everyone struggling with this, including me, even though I've been preaching 00:25:46.920 --> 00:25:47.800 this for a long time. 00:25:47.800 --> 00:25:48.720 It's hard. 00:25:48.720 --> 00:25:52.760 There are some great examples of what not to do. 00:25:52.880 --> 00:25:54.640 So the, the, what was this? 00:25:54.640 --> 00:25:57.880 The give your modules terse single word names whenever possible. 00:25:57.880 --> 00:26:00.680 And I guess there's a couple examples of what not to do. 00:26:00.680 --> 00:26:07.120 So there's a really long directory name called electromyography toolkit. 00:26:07.120 --> 00:26:08.540 Yeah, that's a bit long. 00:26:08.540 --> 00:26:13.180 And then prefixing everything with EMG afterwards within it. 00:26:13.180 --> 00:26:14.220 Yeah, don't do that. 00:26:14.220 --> 00:26:15.220 You don't need to. 00:26:15.220 --> 00:26:18.060 It's just some great examples of better versions. 00:26:18.620 --> 00:26:18.980 Right. 00:26:18.980 --> 00:26:23.360 It already has the top level part of the package as the namespace. 00:26:23.360 --> 00:26:24.000 Yeah. 00:26:24.000 --> 00:26:28.160 If you want to be explicit, just, you know, package dot sub package. 00:26:28.160 --> 00:26:34.220 And so they're talking about how, how, what it looks like when you're importing it is different. 00:26:34.220 --> 00:26:37.840 So I'm going to scroll down of like importing stuff. 00:26:37.840 --> 00:26:40.500 I guess I'll just drop to one. 00:26:40.500 --> 00:26:43.120 A better thing would be like EMG toolkit. 00:26:43.120 --> 00:26:46.920 And I would say even like, is toolkit really adding any information? 00:26:47.140 --> 00:26:49.100 Maybe just EMG or something. 00:26:49.100 --> 00:26:56.300 And then having, instead of having like solvers and a whole bunch of prefixes, just put solvers 00:26:56.300 --> 00:26:59.080 in a directory name and then put the types of solvers underneath. 00:26:59.080 --> 00:27:03.400 You don't have to have EMGs under everything or yeah. 00:27:03.400 --> 00:27:04.260 EMG underscore. 00:27:04.260 --> 00:27:08.160 Or a brute force solver and a, a star solver and all that. 00:27:08.160 --> 00:27:11.200 It's just solvers dot a star solvers dot brute force or something. 00:27:11.200 --> 00:27:11.380 Right. 00:27:11.380 --> 00:27:11.780 Yeah. 00:27:11.780 --> 00:27:12.040 Right. 00:27:12.160 --> 00:27:16.860 And I'd argue that you could even just probably drop the toolkit and it'd be EMG dot solvers dot 00:27:16.860 --> 00:27:17.620 brute force. 00:27:17.620 --> 00:27:18.860 It'd be a better name. 00:27:18.860 --> 00:27:25.440 But I think, so I'm not going to go through this, this entire thing, but import modules is 00:27:25.440 --> 00:27:26.320 something I've seen. 00:27:26.320 --> 00:27:31.720 A lot of people will do like a from package solver import and then list all the things that 00:27:31.720 --> 00:27:32.380 they're importing. 00:27:32.600 --> 00:27:38.600 And I think, and I've been guilty of this to doing this before also, especially if I'm 00:27:38.600 --> 00:27:39.540 only importing one. 00:27:39.540 --> 00:27:45.260 One of the things that people will do this for is because the entire name is too long. 00:27:45.260 --> 00:27:47.300 Well then, you know, make the name shorter. 00:27:47.300 --> 00:27:52.800 So instead of doing that, just import the package solver and then you dot reference everything 00:27:52.800 --> 00:27:53.280 else. 00:27:53.280 --> 00:27:53.620 Right. 00:27:54.340 --> 00:27:56.500 Or do what the data scientists do. 00:27:56.500 --> 00:28:00.520 If you don't control the name, you know, import numpy as NP. 00:28:00.520 --> 00:28:01.200 Yeah. 00:28:01.200 --> 00:28:05.700 Then NP dot that you can shorten the name for you, even if you can't shorten it for others 00:28:05.700 --> 00:28:07.360 at the source. 00:28:07.360 --> 00:28:07.900 Right. 00:28:07.900 --> 00:28:12.020 So one of the reminders is you can do from, if you're in the same directory, you can do 00:28:12.020 --> 00:28:17.640 from dot import solvers, and then you can reference like solvers and entropy solver and things 00:28:17.640 --> 00:28:18.080 like that. 00:28:18.320 --> 00:28:25.700 Anyway, some great, even discussion of sub packages of, and this is something that I 00:28:25.700 --> 00:28:26.980 always forget that is difficult. 00:28:26.980 --> 00:28:30.780 It's a, it's a mind shift to say, oh, I'm going to throw things into a directory and then 00:28:30.780 --> 00:28:33.360 have a package, but you can, you can do that again. 00:28:33.360 --> 00:28:37.860 It can, it can keep going down and, and some people forget to do that. 00:28:37.860 --> 00:28:39.980 So the reference is that as well. 00:28:39.980 --> 00:28:40.760 So it's good. 00:28:40.760 --> 00:28:43.360 You can have packages inside of packages inside of packages. 00:28:43.360 --> 00:28:43.780 Nice. 00:28:44.120 --> 00:28:48.740 I think Robert out in the audience, that's a great recommendation to simplify the code 00:28:48.740 --> 00:28:49.340 we got to write. 00:28:49.340 --> 00:28:50.380 Yeah. 00:28:50.380 --> 00:28:54.660 Just how about we just extend it a little bit from star import star and like you just have 00:28:54.660 --> 00:28:55.480 access to everything. 00:28:55.480 --> 00:28:58.040 There's sure the startup time is 10 minutes, but it's fine. 00:28:58.040 --> 00:29:00.880 Yes. 00:29:00.880 --> 00:29:05.840 And then Pampho Roy is reminding us that, you know, one of the Zen of pythons is about these 00:29:05.840 --> 00:29:06.380 namespaces. 00:29:06.380 --> 00:29:06.860 Yeah. 00:29:06.860 --> 00:29:07.340 Yeah. 00:29:07.340 --> 00:29:09.860 And I think it's just also be nice to people. 00:29:10.800 --> 00:29:14.360 Just, you know, people trying to write, write their code in Python. 00:29:14.360 --> 00:29:17.600 They're not doing weird stuff because they want to do weird stuff. 00:29:17.600 --> 00:29:20.300 They just don't know what the normal way to do it is. 00:29:20.300 --> 00:29:22.460 So gentle reminders like this post are good. 00:29:22.460 --> 00:29:22.980 Indeed. 00:29:22.980 --> 00:29:24.180 What you got for us next? 00:29:24.180 --> 00:29:28.620 I don't want to really quibble about it, but let's talk about it anyway. 00:29:28.620 --> 00:29:29.260 Quibbler. 00:29:29.260 --> 00:29:34.120 So quibbler actually is, is interesting keeping with the data science theme. 00:29:34.120 --> 00:29:41.180 So this one here is a way to take matplotlib and make it more interactive. 00:29:41.180 --> 00:29:48.420 So you probably know matplotlib will make a nice, like printable, publishable, high quality 00:29:48.420 --> 00:29:53.260 image, but they're generally not as interactive as like bokeh, some of these other frameworks. 00:29:53.260 --> 00:29:56.600 So quibbler is like your data, but interactive. 00:29:56.600 --> 00:29:59.740 Let me just play a little video for you here and I'll just describe it. 00:29:59.840 --> 00:30:03.260 But this way you'll kind of know what this is about if you haven't checked this out. 00:30:03.260 --> 00:30:09.000 So the idea is you could go and plot something with matplotlib and you have some variable 00:30:09.000 --> 00:30:10.520 you would like to make interactive. 00:30:10.520 --> 00:30:18.240 You just wrap it in an iQuib and all of a sudden the entire UI becomes completely interactive, 00:30:18.240 --> 00:30:19.160 which is insane. 00:30:19.160 --> 00:30:26.060 So here's like a triangle with squares around each, based off each of the edge of the triangle. 00:30:26.320 --> 00:30:31.200 And then if you just make the numbers, the inputs, this iQuib type, then all of a sudden 00:30:31.200 --> 00:30:33.460 you can drag them around and interact with them. 00:30:33.460 --> 00:30:38.920 And even, even works on like cells and zooms and other pictures. 00:30:38.920 --> 00:30:39.380 What do you think? 00:30:39.380 --> 00:30:40.600 This is amazing. 00:30:40.600 --> 00:30:44.700 Also amazing use of video to show people how easy it is to switch. 00:30:44.700 --> 00:30:45.040 Yes. 00:30:45.040 --> 00:30:46.220 Thank you for doing that. 00:30:46.220 --> 00:30:47.720 Thank you for doing that people. 00:30:47.720 --> 00:30:50.660 But yeah, this is a super cool project. 00:30:50.820 --> 00:30:54.960 I don't, I don't have a ton to say about it actually, but you know, if you're working 00:30:54.960 --> 00:30:58.120 with matplotlib already and you're like, you know, what'd be really nice is if we could, 00:30:58.120 --> 00:31:02.420 you know, move some of these things around and create interactive graphics a little bit 00:31:02.420 --> 00:31:08.560 more than, you know, just wrapping the independent variables in iQuib interactive bits. 00:31:08.560 --> 00:31:12.320 And it turns the whole, the whole graph into something that's, that's pretty excellent. 00:31:12.320 --> 00:31:14.780 I, you know, I think this is, it looks great. 00:31:14.780 --> 00:31:17.000 I don't see why you wouldn't, wouldn't want to use it. 00:31:17.280 --> 00:31:17.400 Yeah. 00:31:17.400 --> 00:31:18.940 That's, that's, that looks fun. 00:31:18.940 --> 00:31:19.340 Yeah. 00:31:19.340 --> 00:31:19.600 Cool. 00:31:19.600 --> 00:31:21.860 And someone out of the audience says it loves Quib. 00:31:21.860 --> 00:31:24.040 So they must have experience with as well. 00:31:24.040 --> 00:31:24.420 All right. 00:31:24.420 --> 00:31:28.360 So if you were doing matplotlib and you want to make it more interactive, check this out. 00:31:28.360 --> 00:31:28.920 Awesome. 00:31:28.920 --> 00:31:30.800 It's all of our items, isn't it, Brian? 00:31:30.800 --> 00:31:33.620 Even though somehow I seem to have like a screen full of tabs still. 00:31:33.620 --> 00:31:34.360 I don't know how that works. 00:31:34.360 --> 00:31:34.920 Yeah. 00:31:34.920 --> 00:31:37.260 So you've got some interesting Python related stuff. 00:31:37.260 --> 00:31:43.960 And I thought I would go with the tradition of Python and say, and now for something completely 00:31:43.960 --> 00:31:44.360 different. 00:31:44.360 --> 00:31:47.640 I just ran across an article that I thought I'd share just quickly. 00:31:47.640 --> 00:31:49.080 I'm not going to talk about it a lot. 00:31:49.080 --> 00:31:55.360 But there was an article in Washington post about how turtle moms talk to their eggs before 00:31:55.360 --> 00:31:55.940 they hatch. 00:31:55.940 --> 00:31:58.280 And I just thought it was a fascinating read. 00:31:58.280 --> 00:32:00.360 Plus it's got pictures of turtles. 00:32:00.360 --> 00:32:02.340 So turtles are awesome. 00:32:02.340 --> 00:32:06.000 I wasn't going to talk about it too much, but one of the things I thought was really cool 00:32:06.000 --> 00:32:11.900 was somebody went out and thought, oh, we don't, we, there's the turtle moms hang out in 00:32:11.900 --> 00:32:18.300 the rivers or somewhere near the hatch, the eggs on the sand. 00:32:18.300 --> 00:32:19.800 And they just wait for him or something. 00:32:19.800 --> 00:32:21.100 Well, are they talking? 00:32:21.100 --> 00:32:23.680 It probably not, but they went out and recorded anyway. 00:32:23.680 --> 00:32:28.000 And it's lower frequencies, frequencies that we don't normally hear. 00:32:28.360 --> 00:32:31.520 And, that's also slower timeframes. 00:32:31.520 --> 00:32:35.320 Like a conversation can be like hours for a reply. 00:32:35.320 --> 00:32:38.040 So it takes a while to go back. 00:32:38.040 --> 00:32:41.460 And then also like some of the eggs talk to each other before they hatch. 00:32:41.460 --> 00:32:46.480 they can, they can talk to each other to help possibly coordinate when to hatch. 00:32:46.580 --> 00:32:51.520 Because if there's predators around, more of them will survive if they all hatch at the 00:32:51.520 --> 00:32:51.860 same time. 00:32:51.860 --> 00:32:53.280 So anyway, just cool article. 00:32:53.280 --> 00:32:54.020 Check it out. 00:32:54.020 --> 00:32:55.260 If you guys want to look at turtles. 00:32:55.260 --> 00:32:55.760 Yeah. 00:32:55.760 --> 00:32:56.880 Science is neat, isn't it? 00:32:56.880 --> 00:32:57.760 Yeah. 00:32:57.760 --> 00:32:58.360 Yeah. 00:32:58.360 --> 00:32:58.980 That's really cool. 00:32:58.980 --> 00:33:05.340 So, kind of not one of my extras, but I'll go through this in is the next coming talk 00:33:05.340 --> 00:33:11.820 Python episode is, around the story of imaging black holes with, Dr. 00:33:11.820 --> 00:33:12.180 Sarah. 00:33:12.180 --> 00:33:16.340 I, I said, yeah, so that's going to be really, really awesome. 00:33:16.340 --> 00:33:20.780 And then there's a good chance, not a hundred percent chance, but a good chance that I also 00:33:20.780 --> 00:33:27.880 will have a science show on the few, the fusion, national ignition breakthrough with Python 00:33:27.880 --> 00:33:28.820 as well coming up. 00:33:28.820 --> 00:33:29.820 So very cool. 00:33:29.820 --> 00:33:30.660 Very cool stuff. 00:33:30.660 --> 00:33:31.340 Very neat. 00:33:31.340 --> 00:33:33.700 You get to talk to so many cool people through that. 00:33:33.700 --> 00:33:34.160 That's neat. 00:33:34.160 --> 00:33:37.140 I know it's, it's, I would never get to talk to these people. 00:33:37.140 --> 00:33:38.300 They're like, Hey, I have a podcast. 00:33:38.300 --> 00:33:39.400 Let's talk about it. 00:33:39.400 --> 00:33:39.760 Amazing. 00:33:39.760 --> 00:33:40.640 All right. 00:33:40.640 --> 00:33:42.380 let's do your extras. 00:33:42.380 --> 00:33:42.900 Yeah. 00:33:42.900 --> 00:33:43.520 Let's do my extras. 00:33:43.520 --> 00:33:46.060 So here we go. 00:33:46.100 --> 00:33:51.640 So I've published the thing on RSS a little while ago and talked about a lot about RSS. 00:33:51.640 --> 00:33:59.080 So I have more RSS stuff because I said reader R E E D E R is a really cool thing, but it's 00:33:59.080 --> 00:33:59.560 Mac only. 00:33:59.560 --> 00:34:01.360 Other people are like, Hey, guess what else I use? 00:34:01.360 --> 00:34:01.920 This is neat. 00:34:01.920 --> 00:34:07.020 So there's a fresh RSS, which is like your own hosted RSS web app. 00:34:07.020 --> 00:34:10.020 I think it also does some read later type of things. 00:34:10.660 --> 00:34:10.900 Feedly. 00:34:10.900 --> 00:34:14.980 I knew about Feedly, but I thought it was just, I don't know, just some sort of web 00:34:14.980 --> 00:34:16.280 app, type of thing. 00:34:16.280 --> 00:34:20.880 But apparently it has like a cool AI engine to help kind of recommend if you get like 00:34:20.880 --> 00:34:23.400 overwhelmed with all the stuff you subscribe to. 00:34:23.400 --> 00:34:24.380 You're like, well, now I can't read it. 00:34:24.380 --> 00:34:25.140 It's just insane. 00:34:25.140 --> 00:34:30.140 You know, it has a kind of a, your own personal algorithmic feed, which is, I know algorithmic 00:34:30.140 --> 00:34:32.940 feeds often get a bad name, but there are benefits too, you know? 00:34:32.940 --> 00:34:33.400 Yeah. 00:34:33.660 --> 00:34:34.680 Have you used Feedly lately? 00:34:34.680 --> 00:34:35.320 I have not. 00:34:35.320 --> 00:34:37.000 I've been using it for years. 00:34:37.000 --> 00:34:41.600 So I didn't know about the AI bit, but I, yeah, I use it a lot. 00:34:41.600 --> 00:34:42.000 Yeah. 00:34:42.000 --> 00:34:43.580 I thought that was the one you had said you were using. 00:34:43.580 --> 00:34:46.760 So, but yeah, the AI engine sounds at least interesting. 00:34:46.760 --> 00:34:49.780 Flim on, Android. 00:34:49.780 --> 00:34:50.900 People care about that. 00:34:50.900 --> 00:34:54.340 And read wise, we talked about, but I've been playing the read wise and it continues to 00:34:54.340 --> 00:34:54.960 be interesting. 00:34:54.960 --> 00:34:55.440 Cool. 00:34:55.440 --> 00:35:00.820 Can, not only when I first thought of that, oh, RSS and, you know, maybe read later, but 00:35:00.820 --> 00:35:05.760 you can throw in like EPUB books, you can do YouTube videos and like, it'll let you 00:35:05.760 --> 00:35:08.060 annotate the transcripts that it makes. 00:35:08.060 --> 00:35:09.680 And it's, it's pretty wild. 00:35:09.680 --> 00:35:10.120 Wow. 00:35:10.120 --> 00:35:11.960 and then, okay. 00:35:11.960 --> 00:35:19.780 Going for more RSS here is, just published all the talk Python courses as an RSS feed. 00:35:19.780 --> 00:35:25.040 So if you want to subscribe to that, then soon as we publish a new Talk Python Training course, 00:35:25.040 --> 00:35:29.800 like if we go down here, there's a, pie test one somewhere getting started with pie tests. 00:35:29.800 --> 00:35:30.020 Right. 00:35:30.060 --> 00:35:31.920 So that came out a couple months ago. 00:35:31.920 --> 00:35:35.640 And so people, if they're subscribed to the RSS feed, another one comes out, they'll get 00:35:35.640 --> 00:35:36.200 it straight away. 00:35:36.200 --> 00:35:36.700 It's cool. 00:35:36.700 --> 00:35:36.860 Right. 00:35:36.860 --> 00:35:37.660 That is neat. 00:35:37.660 --> 00:35:38.060 Yeah. 00:35:38.060 --> 00:35:42.240 RSS seeing more of the things over at talk Python, the podcast, not the courses. 00:35:42.240 --> 00:35:44.700 I, we were talking about this right before the show. 00:35:44.700 --> 00:35:50.920 I did a ton of stuff for sort of helping people find when were, was a particular person, a guest 00:35:50.920 --> 00:35:51.460 on the show. 00:35:51.460 --> 00:35:52.940 When were they on the show? 00:35:52.940 --> 00:35:56.220 If they're on one episode, what other episodes did they do? 00:35:56.360 --> 00:36:02.160 So here's a great, huge list of tons of people who have been guests on the show. 00:36:02.160 --> 00:36:04.020 And Brian, who's in number one right now? 00:36:04.020 --> 00:36:07.220 I am with Brett as a close second. 00:36:07.220 --> 00:36:08.400 So a close second. 00:36:08.400 --> 00:36:08.580 Yeah. 00:36:08.580 --> 00:36:10.080 So Brian's been on 10 times. 00:36:10.080 --> 00:36:11.220 Brett Cannon, nine. 00:36:11.220 --> 00:36:12.700 Anthony Shaw, nine. 00:36:12.700 --> 00:36:13.900 Dan Bader, six. 00:36:13.900 --> 00:36:15.780 Lucas Lenga, five. 00:36:15.780 --> 00:36:18.000 Cecil down there at five. 00:36:18.000 --> 00:36:19.100 It's Cecil Phillip as well. 00:36:19.100 --> 00:36:22.460 Anyway, if this is a new resource on the podcast. 00:36:22.460 --> 00:36:26.780 So go, if you want to explore the data for the last seven years in a slightly different 00:36:26.780 --> 00:36:28.100 way, people can check that out. 00:36:28.100 --> 00:36:30.400 And congrats, Brian, for leading the pack there. 00:36:30.400 --> 00:36:31.100 Yes. 00:36:31.260 --> 00:36:38.400 And last, I talked about how to do traveling around, just taking an iPad and still doing 00:36:38.400 --> 00:36:39.080 development work. 00:36:39.080 --> 00:36:40.800 And we talked about prompt and all those things. 00:36:40.800 --> 00:36:44.840 So I basically wrote up an article on that discussion I had there. 00:36:44.840 --> 00:36:45.720 So I'll link to that. 00:36:45.720 --> 00:36:47.280 People can check that out if they want to. 00:36:47.280 --> 00:36:52.880 So there's the written, more formal version called Dev on the Road on my website. 00:36:52.880 --> 00:36:53.360 Cool. 00:36:53.660 --> 00:36:53.920 All right. 00:36:53.920 --> 00:36:56.900 Ready for a joke? 00:36:56.900 --> 00:37:01.740 Oh, by the way, Robert out there is suggesting that we might need to update your photo. 00:37:01.740 --> 00:37:03.800 Now you got your cool hair. 00:37:03.800 --> 00:37:05.840 We might need to update you on that guest list there. 00:37:05.840 --> 00:37:08.240 That was a seven-year-old photo or something. 00:37:08.240 --> 00:37:08.480 Yeah. 00:37:08.480 --> 00:37:10.900 That's from when you were on the show the first time, maybe six years ago. 00:37:10.900 --> 00:37:12.900 No, I really do look like that other photo. 00:37:12.900 --> 00:37:15.200 What you're seeing right now is an AI-generated bot. 00:37:15.200 --> 00:37:15.820 Yeah. 00:37:15.820 --> 00:37:16.900 This is like a... 00:37:16.900 --> 00:37:20.340 Actually, Brian hasn't been here for a few years. 00:37:20.340 --> 00:37:22.080 This is a deep fake and I invented him. 00:37:22.080 --> 00:37:23.160 No, just kidding. 00:37:23.160 --> 00:37:25.060 All right. 00:37:25.060 --> 00:37:25.780 Speaking of jokes. 00:37:25.780 --> 00:37:29.340 The real Brian is still embroiled in a lawsuit to try to get this to stop. 00:37:29.340 --> 00:37:33.200 The real Brian will never defeat the AI. 00:37:33.200 --> 00:37:33.560 Okay. 00:37:33.560 --> 00:37:39.820 So I thought of you actually, Brian, when I saw this one. 00:37:39.820 --> 00:37:42.340 This is on Program and Humor. 00:37:42.340 --> 00:37:51.300 And it says it has this dog-looking thing kind of very deep in thought projected to the sky for reasons I don't understand. 00:37:51.600 --> 00:37:56.020 But it says, am I testing this program or is it testing me? 00:37:56.020 --> 00:38:00.580 I suspect a lot of people can relate to having that go both ways. 00:38:00.580 --> 00:38:01.380 Deep thoughts. 00:38:01.380 --> 00:38:02.340 Deep thoughts. 00:38:02.340 --> 00:38:04.480 And we've got a name for the AI. 00:38:04.480 --> 00:38:05.640 It's called Brian GPT. 00:38:08.220 --> 00:38:08.720 All right. 00:38:08.720 --> 00:38:10.280 I know you got a joke as well for us. 00:38:10.280 --> 00:38:12.900 Maybe more involved than the floating dog. 00:38:12.900 --> 00:38:13.440 Okay. 00:38:13.440 --> 00:38:17.320 I can't remember where I got this, but every cloud architecture. 00:38:17.320 --> 00:38:18.720 It's just, I don't know. 00:38:18.720 --> 00:38:22.200 It's an image with some stuff that I really... 00:38:22.200 --> 00:38:22.600 There's a lot. 00:38:22.660 --> 00:38:27.380 I'm not going to cover everything, but one of my favorite bits of this is the managed services, 00:38:27.380 --> 00:38:31.420 the mismanaged services, and the unmanageable services. 00:38:31.420 --> 00:38:35.260 They're held together by duct tape and like some kind of like super glue. 00:38:35.260 --> 00:38:35.580 Yeah. 00:38:35.580 --> 00:38:37.400 Containers everywhere. 00:38:37.840 --> 00:38:44.260 I do like the cool databases that drive resumes and the real database that's actually doing 00:38:44.260 --> 00:38:45.340 all the actual work. 00:38:45.340 --> 00:38:47.860 The plumbing. 00:38:47.860 --> 00:38:50.020 There's a huge bunch of plumbing around it. 00:38:50.020 --> 00:38:50.880 It just has... 00:38:50.880 --> 00:38:52.700 The data engineers got very excited here. 00:38:52.700 --> 00:38:53.540 Yeah. 00:38:53.540 --> 00:38:54.780 It empties into a data lake. 00:38:54.780 --> 00:38:57.200 Which also drips into a data swamp. 00:38:57.880 --> 00:38:59.880 And there's a blame radius. 00:38:59.880 --> 00:39:02.400 Services that stop stuff. 00:39:02.400 --> 00:39:06.040 There's one tiny cron job that keeps everything from falling apart. 00:39:06.040 --> 00:39:07.160 It's good. 00:39:07.160 --> 00:39:08.360 VPN of sadness. 00:39:08.360 --> 00:39:14.560 And of course, it's all backed by the freaking slick reverse proxy. 00:39:14.560 --> 00:39:15.320 Yeah. 00:39:15.320 --> 00:39:22.820 So this is found on Good Tech Things, which actually has a whole bunch of cool, like funny images. 00:39:22.820 --> 00:39:24.780 So I might steal some more of these from... 00:39:24.780 --> 00:39:27.080 Oh, we might hear from Good Tech Things again. 00:39:27.080 --> 00:39:28.440 This is pretty awesome. 00:39:28.440 --> 00:39:29.520 All right. 00:39:29.520 --> 00:39:30.960 Well, is that everything for us? 00:39:30.960 --> 00:39:31.920 That is. 00:39:31.920 --> 00:39:32.740 That is all of it. 00:39:32.740 --> 00:39:33.680 So... 00:39:33.680 --> 00:39:34.040 Nice. 00:39:34.040 --> 00:39:35.080 Thank you everyone for listening. 00:39:35.080 --> 00:39:36.920 Brian, thanks as always. 00:39:36.920 --> 00:39:37.520 Bye.