WEBVTT NOTE Opening 00:00:00.588 --> 00:00:12.608 [opening stuff] 00:00:14.070 --> 00:00:16.835 Emacs Chat 26 is Ross A. Baker. 00:00:17.616 --> 00:00:20.181 You can find him at rossabaker.com. 00:00:20.201 --> 00:00:22.364 Today, we've got a lot of interesting things 00:00:22.404 --> 00:00:24.748 to dig into, including his two-part Emacs 00:00:24.788 --> 00:00:25.870 configuration, 00:00:25.850 --> 00:00:29.017 plus of course, all the things that he does. 00:00:29.037 --> 00:00:31.964 The goal with Emacs Chat is to show some of 00:00:32.024 --> 00:00:33.568 the things that aren't obvious from the 00:00:33.628 --> 00:00:35.533 configuration, like how the different pieces 00:00:35.613 --> 00:00:38.199 actually work together or what the workflow feels like. 00:00:38.560 --> 00:00:40.885 But before we dive into that, Ross, can you tell 00:00:40.925 --> 00:00:42.549 us a little bit about your background? 00:00:42.569 --> 00:00:43.391 What's the context here? 00:00:43.411 --> 00:00:44.213 What do you like to do? 00:00:45.307 --> 00:00:45.788 [Ross]: Oh, sure. NOTE What Ross does 00:00:46.369 --> 00:00:48.513 So for work, I'm a backend engineer. 00:00:48.814 --> 00:00:51.679 I'm somewhat well known in the Scala community. 00:00:51.699 --> 00:00:54.504 I've done that for a long time on an open-source basis. 00:00:54.524 --> 00:00:56.709 I've taken an interest in Rust here recently. 00:00:56.769 --> 00:00:59.915 Those are the two primary languages I work in. 00:00:59.935 --> 00:01:02.940 I work for a financial company. 00:01:03.001 --> 00:01:05.866 We maintain the apps for small- and regional-sized 00:01:05.906 --> 00:01:07.990 banks in the United States. 00:01:07.970 --> 00:01:09.391 Fairly fun work there. 00:01:09.412 --> 00:01:11.994 It's kept me quite busy here lately. 00:01:12.234 --> 00:01:14.217 For other things I like to do for fun, like to go 00:01:14.277 --> 00:01:16.118 for a jog, like to push away from the computer 00:01:16.179 --> 00:01:16.699 now and then. 00:01:16.839 --> 00:01:19.002 So I go jogging every day, try to run about nine 00:01:19.042 --> 00:01:19.882 miles every day. 00:01:22.745 --> 00:01:27.911 My two big de-stressors are the running 00:01:28.091 --> 00:01:30.373 and the fiddling with my Emacs configuration, 00:01:30.574 --> 00:01:32.115 kind of opposite sides of the coin. 00:01:32.135 --> 00:01:34.838 But I need them both to stay happy, I think. 00:01:35.847 --> 00:01:37.029 That makes sense. 00:01:37.649 --> 00:01:39.171 Also a soccer coach. 00:01:39.452 --> 00:01:41.274 My kid is in high school and coached his 00:01:41.314 --> 00:01:42.336 recreational team. 00:01:42.676 --> 00:01:44.118 They have a lot of fun with that as well. 00:01:45.139 --> 00:01:45.840 [Sacha]: Yeah. 00:01:45.900 --> 00:01:47.603 One of your blog posts mentions that you've been 00:01:47.623 --> 00:01:50.967 publishing your youth soccer webpage in ox-hugo. 00:01:52.029 --> 00:01:52.469 [Ross]: That's right. 00:01:53.310 --> 00:01:54.472 [Sacha]: Emacs can be used for everything. 00:01:55.413 --> 00:01:56.054 [Ross]: It can, yeah. 00:01:56.254 --> 00:01:57.636 That's right. 00:01:57.903 --> 00:02:01.708 [Sacha]: We're also very curious about how 00:02:01.768 --> 00:02:04.491 people get into Emacs and what makes them stick 00:02:04.531 --> 00:02:05.192 with Emacs. NOTE How Ross got into Emacs, used other editors, and then came back to Emacs 00:02:06.593 --> 00:02:10.518 In your story for the Emacs Carnival in Take 2, 00:02:10.878 --> 00:02:12.681 you shared how you got into Emacs. 00:02:12.741 --> 00:02:16.425 Well, you were introduced to Emacs in 1997, but 00:02:16.465 --> 00:02:17.366 you actually left. 00:02:17.567 --> 00:02:20.350 You used other editors for a while because of work, 00:02:20.718 --> 00:02:22.440 and then you came back to Emacs. 00:02:22.460 --> 00:02:25.244 Was it primarily for Haskell and Clojure, or were 00:02:25.284 --> 00:02:30.411 there other reasons that got you interested in it? 00:02:30.431 --> 00:02:33.039 [Ross]: Yeah, when I was at the Haskell shop, 00:02:33.040 --> 00:02:35.477 it was something that a lot... When 00:02:35.497 --> 00:02:37.360 you get out into the workplace, there's very few 00:02:37.500 --> 00:02:39.763 Emacs users, relatively speaking, in most 00:02:39.783 --> 00:02:40.344 workplaces. 00:02:40.364 --> 00:02:41.746 But when you're in a Haskell shop, that's what 00:02:41.806 --> 00:02:42.947 almost everybody uses. 00:02:42.967 --> 00:02:45.531 So being surrounded by that, that was a good 00:02:45.851 --> 00:02:47.113 reason to get back into it. 00:02:47.173 --> 00:02:48.694 I'd been dabbling in it a little bit 00:02:48.695 --> 00:02:50.337 before that, even. 00:02:50.317 --> 00:02:53.199 I had a really good time with it originally. 00:02:53.200 --> 00:02:55.563 I got into it like so many people did back then. 00:02:55.583 --> 00:02:57.506 I got into it in college, showed up, and that's 00:02:57.546 --> 00:02:58.867 what the professor was using. 00:02:58.887 --> 00:03:00.509 The first language that we learned there was 00:03:00.569 --> 00:03:02.872 Scheme, which is a Lisp dialect, so it's a 00:03:02.932 --> 00:03:03.933 natural fit there. 00:03:04.734 --> 00:03:07.938 Back then Emacs and Vim were the two dominant 00:03:08.018 --> 00:03:09.881 editors at that point, so it was pretty much one 00:03:09.921 --> 00:03:10.421 or the other. 00:03:10.441 --> 00:03:12.083 If you were learning a Lisp dialect in 00:03:12.123 --> 00:03:13.905 school, the professor was going to guide you 00:03:13.945 --> 00:03:15.227 toward Emacs. 00:03:15.207 --> 00:03:16.828 I got that initial shove 00:03:16.829 --> 00:03:18.171 and fell in love with it there. 00:03:18.351 --> 00:03:20.134 Then I got out into the workplace, and in the 00:03:20.154 --> 00:03:22.817 workplace back then, it was very rigid. 00:03:22.837 --> 00:03:24.960 You would have to use the commercial editor that 00:03:25.000 --> 00:03:27.363 was integrated with all the IBM software, 00:03:27.364 --> 00:03:30.086 so I had to use these specific things 00:03:30.087 --> 00:03:31.870 and I couldn't use Emacs anymore. 00:03:33.011 --> 00:03:35.875 I did dabble in Vim a little bit as well and I 00:03:35.935 --> 00:03:37.958 liked the efficiency of editing 00:03:37.938 --> 00:03:40.081 that came from both Emacs and Vim. 00:03:40.221 --> 00:03:42.204 Vim was easier to emulate than these other 00:03:42.265 --> 00:03:43.426 editors that I had to use. 00:03:43.446 --> 00:03:46.771 That had me on a Vim path for a little while. 00:03:46.791 --> 00:03:48.895 But then scripting it just wasn't as satisfying 00:03:48.955 --> 00:03:49.896 as Emacs. 00:03:51.198 --> 00:03:52.620 It was always really calling to me. 00:03:52.680 --> 00:03:55.445 I felt like I was in exile while I was outside of 00:03:55.525 --> 00:03:55.965 Emacs. 00:03:55.985 --> 00:03:57.668 Then, when I had a chance to start picking my 00:03:57.728 --> 00:03:59.971 own editor and customizing things again and 00:04:00.072 --> 00:04:01.794 getting up to speed with other people who were 00:04:01.874 --> 00:04:04.017 using Emacs, it was just natural to come home, 00:04:04.078 --> 00:04:05.700 and here I am again. 00:04:06.440 --> 00:04:07.081 [Sacha]: Nice. 00:04:08.623 --> 00:04:11.687 Many people haven't had the experience of working 00:04:11.747 --> 00:04:14.291 with other Emacs users in the same company. 00:04:14.391 --> 00:04:15.613 So what is it like? 00:04:15.813 --> 00:04:20.039 Were you swapping lots of config around, pair 00:04:20.079 --> 00:04:21.380 programming, that sort of thing? 00:04:22.563 --> 00:04:24.264 [Ross]: Yeah, pair programming, 00:04:24.265 --> 00:04:25.567 sharing config. 00:04:25.787 --> 00:04:28.511 I am fortunate at my current workplace, even 00:04:28.571 --> 00:04:30.334 though it's a much smaller percentage than it's 00:04:30.374 --> 00:04:31.635 been in some other workplaces. 00:04:31.655 --> 00:04:33.678 It's also a large company, and we have an active 00:04:33.758 --> 00:04:36.362 Emacs channel there within the company Slack. 00:04:36.663 --> 00:04:38.725 A lot of people are interested in it there, so 00:04:38.785 --> 00:04:41.429 I've got my little inside community in addition 00:04:41.469 --> 00:04:43.271 to the outside Emacs community as well. 00:04:44.572 --> 00:04:46.555 Fortunately, I still get that even to this day, 00:04:46.735 --> 00:04:47.856 so that's pretty fun. 00:04:48.757 --> 00:04:49.677 [Sacha]: You're the second person has described 00:04:49.678 --> 00:04:53.181 the wonders of having a company internet channel 00:04:53.182 --> 00:04:57.928 on Slack, just about Emacs. Yes. Very awesome. NOTE Config focusing on built-ins 00:04:58.280 --> 00:05:04.213 I'm wondering, with your current company, is that 00:05:04.233 --> 00:05:08.862 the reason why you have so many "gotta run with 00:05:10.125 --> 00:05:12.911 just the built-ins", "gotta run even if you've got 00:05:13.011 --> 00:05:15.557 Emacs 28"... Is that the reason for those kinds of 00:05:16.438 --> 00:05:18.583 considerations in your base config? 00:05:19.559 --> 00:05:21.382 [Ross]: That was kind of a premature 00:05:21.522 --> 00:05:23.966 optimization that worked out very well for me. 00:05:23.986 --> 00:05:27.613 I used to run a very extensive config. 00:05:27.633 --> 00:05:29.616 I'd look at all the classic configs that would go 00:05:29.676 --> 00:05:32.320 out there and install all the packages from MELPA. 00:05:33.042 --> 00:05:35.165 I'd try them all out. I'd just accumulate 00:05:35.205 --> 00:05:36.768 these things. I went through some Emacs 00:05:36.808 --> 00:05:37.990 bankruptcies. 00:05:37.970 --> 00:05:39.693 Then people in the community started talking 00:05:39.753 --> 00:05:43.138 about, okay, Emacs 29, it added `use-package` out 00:05:43.178 --> 00:05:43.658 of the box. 00:05:43.758 --> 00:05:47.544 It added... I can't remember when project.el came 00:05:47.604 --> 00:05:50.228 along, but it added these things that started... 00:05:50.308 --> 00:05:52.852 You used to have to get them from outside, or you 00:05:52.872 --> 00:05:55.456 used to have these other libraries, and Emacs got 00:05:55.496 --> 00:05:57.058 a lot better out of the box. 00:05:57.138 --> 00:05:58.560 I started to say, okay, 00:05:58.540 --> 00:06:00.043 I'm going to do one more bankruptcy. 00:06:00.083 --> 00:06:01.967 I'm going to see what I can do with built-ins. 00:06:02.027 --> 00:06:04.832 I started experimenting down those lines, even 00:06:04.853 --> 00:06:07.137 though I was always using that extended config. 00:06:07.157 --> 00:06:09.241 I was challenging myself. How much can I get from 00:06:09.301 --> 00:06:12.147 a built-in config and then just have a little bit 00:06:12.207 --> 00:06:14.872 of extra for those gaps? Because there are gaps. 00:06:15.173 --> 00:06:17.918 I do love Emacs built-ins, but I am more 00:06:17.979 --> 00:06:19.722 comfortable with the regular one. 00:06:19.853 --> 00:06:21.976 But I challenge myself that way. 00:06:22.256 --> 00:06:24.199 This month, I got put on a project where 00:06:24.239 --> 00:06:26.022 I have to work in a sandbox environment. 00:06:26.042 --> 00:06:28.646 So it's not regular operating procedure, but I 00:06:28.686 --> 00:06:30.709 have to do all my work in the sandbox environment. 00:06:30.729 --> 00:06:32.531 There's no network egress. 00:06:32.551 --> 00:06:36.116 I have to say, okay, if I want packages in here, 00:06:36.217 --> 00:06:38.179 I have to talk to the admins of that system and 00:06:38.239 --> 00:06:40.242 say, I want these packages. 00:06:40.262 --> 00:06:41.644 I lobbied for Emacs on it. 00:06:41.725 --> 00:06:43.247 They thought, well, that's kind of weird, but 00:06:43.307 --> 00:06:45.029 it's just one more line of apt-get in the 00:06:45.049 --> 00:06:45.931 Docker container, 00:06:45.971 --> 00:06:47.633 so they're fine with it. 00:06:47.613 --> 00:06:51.133 It was an old version of Debian Bookworm. 00:06:51.134 --> 00:06:53.267 I was running Emacs 28. I'm like, 00:06:53.268 --> 00:06:56.167 okay, well, I've got this built-in configuration. 00:06:56.168 --> 00:06:57.567 This is its moment to shine. 00:06:57.568 --> 00:06:59.967 I tried to run it on there. 00:06:59.968 --> 00:07:03.300 I was able to copy the file up there. I tried 00:07:03.301 --> 00:07:05.200 running it on there, and I run into things 00:07:05.201 --> 00:07:07.309 I've been using `setopt`. I've been using 00:07:07.289 --> 00:07:11.876 one of the XDG packages for setting your X desktop 00:07:11.896 --> 00:07:12.817 group directories, 00:07:12.837 --> 00:07:15.922 so to have a standard place to put your config 00:07:15.982 --> 00:07:17.743 files versus your cache files 00:07:17.744 --> 00:07:19.527 versus your state files. 00:07:19.547 --> 00:07:21.910 In Emacs 28, they had most of those variables, 00:07:21.950 --> 00:07:24.134 but one of those variables didn't exist yet. 00:07:24.174 --> 00:07:26.137 So I felt a few paper cuts that way. 00:07:26.777 --> 00:07:29.121 Up near the top of my config, I've got a 00:07:29.161 --> 00:07:30.563 compatibility layer. 00:07:31.015 --> 00:07:32.978 The compatibility layer, there's a nice 00:07:33.018 --> 00:07:34.339 compat package out there. 00:07:34.379 --> 00:07:34.960 I can't remember. 00:07:35.281 --> 00:07:37.143 I think Omar, who you interviewed recently, I 00:07:37.263 --> 00:07:39.246 believe he's one of the people behind it. 00:07:39.266 --> 00:07:41.669 And tarsius, I think he's involved in that too. 00:07:41.809 --> 00:07:43.131 I could be misquoting on that. 00:07:43.772 --> 00:07:45.294 Maybe I'm misattributing that. 00:07:45.414 --> 00:07:47.697 But anyway, there's this wonderful compat package 00:07:47.777 --> 00:07:48.138 out there. 00:07:48.598 --> 00:07:50.180 If you're a package author and you want to have 00:07:50.220 --> 00:07:52.243 your Emacs package running on older versions, 00:07:52.824 --> 00:07:53.665 it's great for that. 00:07:53.645 --> 00:07:55.869 But the whole point of this is I don't have any 00:07:55.929 --> 00:07:56.951 external packages. 00:07:56.971 --> 00:07:58.954 So where I need this compatibility layer the 00:07:58.994 --> 00:08:01.799 most, I can't use this lovely compat package out 00:08:01.839 --> 00:08:02.360 of the box. 00:08:02.840 --> 00:08:05.705 So I had to reinvent a few shims. 00:08:05.725 --> 00:08:07.068 I think I could bring that up. 00:08:07.088 --> 00:08:08.250 Am I sharing my screen? 00:08:08.390 --> 00:08:09.592 [Sacha]: Yes, you're sharing your screen. 00:08:10.473 --> 00:08:12.677 That was one of the things that immediately 00:08:12.717 --> 00:08:14.099 struck me about your config. 00:08:14.139 --> 00:08:16.303 You're so hardcore about not using 00:08:16.323 --> 00:08:18.066 external packages that even your 00:08:18.333 --> 00:08:19.935 adapting to older versions, 00:08:19.955 --> 00:08:23.299 you're re-implementing things yourself in order 00:08:23.339 --> 00:08:26.644 to be able to stick with your constraints. 00:08:26.664 --> 00:08:28.326 I think that's hilarious, by the way. 00:08:28.366 --> 00:08:30.268 You started it off with it as a personal 00:08:30.328 --> 00:08:32.211 challenge because, of course, people are allowed 00:08:32.251 --> 00:08:34.933 to set arbitrary challenges for themselves, 00:08:34.995 --> 00:08:37.798 and then it turned out to be surprisingly useful 00:08:37.857 --> 00:08:40.542 for you in this limited environment. 00:08:40.802 --> 00:08:42.164 [Ross]: Yeah, that's one that really saved 00:08:42.203 --> 00:08:42.683 my bacon. 00:08:43.005 --> 00:08:44.647 [Sacha]: Yeah, this is great. 00:08:44.667 --> 00:08:46.990 So even things like `setopt`... Then you have 00:08:47.155 --> 00:08:52.333 `keymap-set`. You've got some replacements 00:08:52.334 --> 00:08:55.633 for these modern niceties to make it still work 00:08:55.634 --> 00:08:58.233 in Emacs 28. 00:08:58.234 --> 00:09:01.500 [Ross]: Yeah, I leaned into the new keybindings. 00:09:01.501 --> 00:09:03.655 There's `keymap-set`, `keymap-global-set`, 00:09:03.675 --> 00:09:07.467 `keymap-global-unset`. Those are roughly the same as 00:09:07.468 --> 00:09:09.367 what you had in older Emacs versions 00:09:09.368 --> 00:09:11.667 but you don't need to pass the keyboard macro 00:09:11.668 --> 00:09:14.648 around all the things, so it's a little bit more concise. 00:09:14.881 --> 00:09:16.645 There were a few more advantages to it and I'd 00:09:16.685 --> 00:09:19.029 already been on Emacs 29 or Emacs 30 00:09:19.030 --> 00:09:19.993 everywhere else. 00:09:20.013 --> 00:09:22.137 I'd already leaned into those, and then I got 00:09:22.198 --> 00:09:23.039 into this environment. 00:09:24.342 --> 00:09:26.307 But as you can see, they're fairly easy. 00:09:26.367 --> 00:09:28.692 This is not a 100% full fidelity. 00:09:28.712 --> 00:09:31.558 There's a few nuances that get missed in doing this, 00:09:31.559 --> 00:09:33.463 but for the most part, it works pretty well. 00:09:35.536 --> 00:09:37.199 I'm not aiming for perfect here. 00:09:37.299 --> 00:09:40.003 I'm aiming for "this works well enough at Emacs 28 00:09:40.264 --> 00:09:42.066 for the life of this project, which should last 00:09:42.127 --> 00:09:42.687 about a month." 00:09:42.788 --> 00:09:45.592 And otherwise, I've got the nice full glory Emacs 00:09:45.652 --> 00:09:46.714 for my daily driver. 00:09:47.535 --> 00:09:50.239 It's straddling that divide fairly well. 00:09:50.259 --> 00:09:54.566 [Sacha]: I just looked up when Emacs 28 was released. 00:09:54.606 --> 00:09:57.291 This is 2022, so four years ago. 00:09:57.912 --> 00:09:58.172 [Ross]: Yes. 00:09:58.653 --> 00:10:00.556 [Sacha]: A lot of Emacs has changed since then. 00:10:00.736 --> 00:10:02.539 But of course, Emacs being Emacs, 00:10:03.363 --> 00:10:05.747 there's a lot of effort put into making sure the 00:10:05.847 --> 00:10:09.232 old stuff keeps working, which is handy for these cases. 00:10:09.993 --> 00:10:10.574 That's great. NOTE simple-orderless #+SCREENSHOT: emacs-chat-26-ross-a-baker-00-11-23-700.jpg #+SCREENSHOT_TIME: 00:11:23.700 00:10:12.276 --> 00:10:14.479 So you have a lot of these shims, 00:10:14.480 --> 00:10:16.483 and one of the things that you've ended up 00:10:16.843 --> 00:10:21.530 re-implementing along these ways is a version of 00:10:21.650 --> 00:10:23.894 orderless that works without having to take the 00:10:23.954 --> 00:10:25.356 entire orderless package in there. 00:10:25.676 --> 00:10:27.639 Can you tell us a little bit more about what you 00:10:27.679 --> 00:10:28.881 like about your setup? 00:10:30.447 --> 00:10:33.912 [Ross]: I wasn't happy with any of the 00:10:34.513 --> 00:10:36.496 built-in completion styles. 00:10:36.516 --> 00:10:39.220 Well, some of them are okay, but the overall... 00:10:39.221 --> 00:10:41.945 I had used orderless in my more extensive config. 00:10:42.466 --> 00:10:44.889 Orderless, for people who don't know, is a way 00:10:44.949 --> 00:10:45.831 of completing things. 00:10:45.931 --> 00:10:47.153 Let's see if I can show it off. 00:10:47.293 --> 00:10:55.286 So I can do... If I do "file" and "find"... 00:10:55.306 --> 00:10:57.208 Well, I thought it would work. 00:10:57.829 --> 00:10:58.851 [Sacha]: Do you need a space? 00:11:00.772 --> 00:11:01.884 [Ross]: Yeah, I don't know why the space 00:11:01.925 --> 00:11:02.734 wasn't completing. 00:11:02.754 --> 00:11:04.453 So I can do "file". 00:11:04.720 --> 00:11:05.922 "find file". 00:11:06.002 --> 00:11:06.743 I could do the space. 00:11:06.823 --> 00:11:09.026 I could do it in any order as well. 00:11:09.046 --> 00:11:11.148 Or I saw there was an "ido" in there. 00:11:11.228 --> 00:11:13.151 If I do it in reverse order, that's the idea of 00:11:13.251 --> 00:11:15.634 orderless is if I'm using the space, I got thrown 00:11:15.714 --> 00:11:15.974 off there. 00:11:16.015 --> 00:11:18.117 It was auto-completing a hyphen on me I didn't 00:11:18.157 --> 00:11:18.518 expect. 00:11:19.139 --> 00:11:21.081 Anyway, if I remember some tokens and things, 00:11:21.101 --> 00:11:23.664 let's say I'm looking for `ido-find-file`, but I 00:11:23.704 --> 00:11:25.266 type "find" first and then I do... 00:11:25.387 --> 00:11:29.211 What orderless does is it takes the input and it 00:11:29.252 --> 00:11:30.272 splits it by space 00:11:30.273 --> 00:11:31.855 and it will find them in any order. 00:11:32.716 --> 00:11:35.300 It's a nicer completion style than I think any of 00:11:35.340 --> 00:11:36.743 the ones that come out of the box are. 00:11:36.903 --> 00:11:38.325 I got used to it. 00:11:38.345 --> 00:11:40.167 But if I'm in this lean config, 00:11:40.168 --> 00:11:41.210 orderless is something... 00:11:41.230 --> 00:11:42.252 It is not a built-in. 00:11:42.292 --> 00:11:44.095 That's something that I was leaving behind. 00:11:44.115 --> 00:11:46.879 That was one that was really starting to itch. 00:11:46.899 --> 00:11:49.363 Then I found this blog post by James Dyer 00:11:49.404 --> 00:11:51.968 where he'd implemented a very slimmed-down 00:11:52.008 --> 00:11:53.931 version of orderless. 00:11:53.911 --> 00:11:55.912 It wasn't the full features that you get, 00:11:55.913 --> 00:11:57.614 but it was something that was good enough 00:11:57.615 --> 00:12:00.819 that captured the basic principles of it. 00:12:00.839 --> 00:12:01.560 I took that. 00:12:01.620 --> 00:12:02.841 I found a few bugs in it, 00:12:02.842 --> 00:12:04.043 and I found a few more things. 00:12:04.163 --> 00:12:06.807 It was a very, very simple version of orderless, 00:12:06.827 --> 00:12:08.649 and I wanted a couple more bells and whistles. 00:12:08.669 --> 00:12:10.851 The completions weren't quite right to my tastes. 00:12:11.372 --> 00:12:12.513 I took that blog post. 00:12:12.553 --> 00:12:13.655 I ended up with a middle ground. 00:12:13.695 --> 00:12:15.836 It's not as good as the full orderless, 00:12:15.837 --> 00:12:17.438 and it's a little bit more extensive 00:12:17.439 --> 00:12:19.682 than the blog post that I found. 00:12:19.662 --> 00:12:21.324 But I'm able to run that, 00:12:21.325 --> 00:12:22.867 and I don't really miss 00:12:22.868 --> 00:12:24.851 the full features of orderless that much. 00:12:25.252 --> 00:12:27.316 There are all sorts of cool things that it does 00:12:27.376 --> 00:12:30.301 where you can say, okay, these are regexp tokens. 00:12:30.361 --> 00:12:31.703 These are literal tokens. 00:12:31.804 --> 00:12:34.028 These are NOT filters. 00:12:34.048 --> 00:12:36.151 You could do all sorts of combinatorial logic 00:12:36.211 --> 00:12:37.754 with it with various syntax. 00:12:37.774 --> 00:12:39.898 You can customize the syntax for it. 00:12:40.334 --> 00:12:42.858 It's all great, but I wasn't using those extended 00:12:42.898 --> 00:12:45.222 features of that package that much. I am happy 00:12:45.262 --> 00:12:47.205 with this lighter-weight version. That way 00:12:47.245 --> 00:12:49.789 I'm able to use it even when I'm off here in the 00:12:49.850 --> 00:12:50.931 security sandbox. 00:12:50.971 --> 00:12:52.233 I'm able to do it there too. 00:12:53.095 --> 00:12:55.177 It exposes the underlying tension 00:12:55.178 --> 00:12:56.841 of this setup. 00:12:57.023 --> 00:12:59.366 I would recommend 100% use orderless. 00:12:59.426 --> 00:13:00.367 It's better maintained. 00:13:00.387 --> 00:13:01.949 The maintainer is a lovely person. 00:13:02.409 --> 00:13:03.511 It's more full-featured. 00:13:03.531 --> 00:13:04.712 You can customize it. 00:13:04.732 --> 00:13:06.593 I ran into a little bit of a hiccup here 00:13:06.594 --> 00:13:07.836 and it's my own config. 00:13:08.697 --> 00:13:11.079 This is something, it's going to be more 00:13:11.119 --> 00:13:12.481 professionally maintained if you're using 00:13:12.561 --> 00:13:12.962 orderless. 00:13:12.982 --> 00:13:14.322 So if you're not adapting 00:13:14.323 --> 00:13:16.885 to my built-ins only method, 00:13:16.886 --> 00:13:20.230 I would 100% recommend using orderless instead. 00:13:20.210 --> 00:13:21.972 But because I have this constraint, 00:13:21.973 --> 00:13:22.894 I'm building this up. 00:13:23.395 --> 00:13:24.657 What I built up is kind of nice. 00:13:24.697 --> 00:13:26.018 It's one of the more interesting parts 00:13:26.019 --> 00:13:28.143 of my package, or of my config. 00:13:28.303 --> 00:13:30.125 I could spin it off as its own package, 00:13:30.126 --> 00:13:32.008 but if I spun it off as its own package, 00:13:32.009 --> 00:13:34.312 it's just a strictly worse version of orderless 00:13:34.313 --> 00:13:37.338 at that point, so what's the point? 00:13:37.358 --> 00:13:38.820 That's one of the tensions that you have in these 00:13:38.860 --> 00:13:41.144 built-ins, is anytime that you build something 00:13:41.204 --> 00:13:43.227 that is really interesting, you want to spin it 00:13:43.328 --> 00:13:45.030 off and share it, but at that point, you've 00:13:45.070 --> 00:13:47.314 defeated the purpose of being built-ins only. 00:13:49.117 --> 00:13:54.905 [Sacha]: I think the idea then of looking at 00:13:54.925 --> 00:13:57.008 the packages that you really like and if you find 00:13:57.048 --> 00:13:58.570 that you're only using a subset of their 00:13:58.590 --> 00:14:01.534 functionality, being able to strip it down or 00:14:01.634 --> 00:14:03.276 take advantage of somebody else's work and 00:14:03.316 --> 00:14:05.740 stripping it down to the subset that you use 00:14:06.240 --> 00:14:08.403 makes it a lot easier for you to put it all 00:14:08.423 --> 00:14:11.187 together in your batteries included, built-in 00:14:11.908 --> 00:14:13.050 only configuration. 00:14:14.131 --> 00:14:14.652 [Ross]: Right. 00:14:14.800 --> 00:14:15.824 [Sacha]: It's not super long. 00:14:16.105 --> 00:14:18.353 It can be understood if you go through the 00:14:18.393 --> 00:14:18.774 function. 00:14:19.236 --> 00:14:20.460 So that's great. 00:14:21.825 --> 00:14:23.813 Are there other things like that that you find 00:14:23.853 --> 00:14:27.004 yourself wanting to extract from other packages? NOTE Size indicator 00:14:29.245 --> 00:14:30.967 [Ross]: I did this really weird little 00:14:31.028 --> 00:14:33.171 thing called size indicator mode. 00:14:33.391 --> 00:14:34.573 I'd already spun that off. 00:14:34.673 --> 00:14:36.375 I don't know if it was worthwhile or not. 00:14:37.076 --> 00:14:40.021 One thing is all of your list commands, they run 00:14:40.121 --> 00:14:43.025 things in binary bytes. 00:14:43.185 --> 00:14:45.569 When we say a kilobyte, what does a kilobyte mean? 00:14:45.589 --> 00:14:49.775 Does that mean 1,000 bytes or 1,024 bytes? 00:14:49.895 --> 00:14:51.577 Most of the commands that you run, they're going 00:14:51.617 --> 00:14:55.523 to be in these binary units. 00:14:56.752 --> 00:14:59.459 What you get out of the box in Emacs is 00:14:59.460 --> 00:15:03.067 the decimal unit. Their definition is 1000 instead 00:15:03.068 --> 00:15:06.700 of 1024. I thought, well, that's kind of weird. 00:15:06.701 --> 00:15:09.667 I built this size-indicator-mode where it goes 00:15:09.668 --> 00:15:12.300 and it translates things so it's unified across 00:15:12.301 --> 00:15:15.133 the way the coreutils work. It's the 1024 base 00:15:15.134 --> 00:15:16.400 instead of the 1000 base. 00:15:16.401 --> 00:15:18.933 Trivial little package, but it was a lot of fun 00:15:18.934 --> 00:15:20.767 I threw it out there. It's up on my Codeberg 00:15:20.768 --> 00:15:22.422 if anybody wants to pull that in. 00:15:23.718 --> 00:15:25.861 [Sacha]: Yeah, so it just reduces the 00:15:25.921 --> 00:15:28.666 friction of having to think, okay, is this in 00:15:28.746 --> 00:15:30.448 this unit or is that unit? 00:15:30.468 --> 00:15:32.411 It'll always be in the units that you like to use 00:15:32.451 --> 00:15:34.595 when you're talking about files. 00:15:34.615 --> 00:15:35.516 [Ross]: Exactly, yeah. 00:15:35.556 --> 00:15:37.179 It keeps things consistent that way. 00:15:37.199 --> 00:15:38.761 That's a package that I spun off. 00:15:38.821 --> 00:15:39.822 Is that something that... 00:15:39.823 --> 00:15:41.505 It's a very lightweight configuration. 00:15:41.525 --> 00:15:42.846 It probably could have stayed here 00:15:42.847 --> 00:15:43.669 in my built-ins. 00:15:44.129 --> 00:15:45.751 I doubt anybody else has that itch, 00:15:45.752 --> 00:15:46.512 but I shared it out there 00:15:46.513 --> 00:15:47.975 and a lot of people thought, ooh, neat. 00:15:47.995 --> 00:15:49.598 I think a few people did install it. 00:15:49.618 --> 00:15:51.801 I'm glad I spun that one off, I suppose. 00:15:52.253 --> 00:15:53.756 There's always that fuzzy line. 00:15:53.797 --> 00:15:55.460 Where does it grow into something that you think 00:15:55.521 --> 00:15:57.686 other people are going to be interested in, versus 00:15:57.726 --> 00:15:59.169 what are the ones that you should just keep 00:15:59.289 --> 00:16:01.414 in-house and they can always copy it out of your 00:16:01.434 --> 00:16:02.577 config if they're interested? 00:16:03.419 --> 00:16:05.203 It's a hard line to walk in this style. 00:16:05.223 --> 00:16:07.749 I haven't exactly found the balance yet, I don't 00:16:07.789 --> 00:16:08.170 think. 00:16:08.892 --> 00:16:10.614 [Sacha]: A lot of your config isn't so much 00:16:12.256 --> 00:16:14.517 new functionality as it is 00:16:14.518 --> 00:16:15.919 just filing off the corners 00:16:15.920 --> 00:16:18.603 or making something a little bit nicer to use. 00:16:19.624 --> 00:16:21.326 For example, when I was going through your 00:16:21.987 --> 00:16:24.329 config, both the base config as well as the 00:16:24.369 --> 00:16:26.131 extended one where you have more packages and 00:16:26.151 --> 00:16:29.275 things set up, I was like, oh yeah, okay. 00:16:29.836 --> 00:16:32.639 You fiddle with your visual-bell, but you also, 00:16:33.299 --> 00:16:34.540 in your extended config, 00:16:34.541 --> 00:16:36.163 set it to mode-line-bell instead. 00:16:36.323 --> 00:16:38.375 That sounds interesting. 00:16:39.357 --> 00:16:39.637 [Ross]: Right. NOTE Graceful degradation 00:16:40.719 --> 00:16:44.344 I go for graceful degradation. When I was doing 00:16:44.424 --> 00:16:47.128 web pages as far back as the 90s, 00:16:47.168 --> 00:16:48.408 when I was first into Emacs, 00:16:48.409 --> 00:16:50.191 everything was about graceful degradation 00:16:50.192 --> 00:16:52.515 as you had these new browser features. 00:16:52.535 --> 00:16:54.758 You're trying to use these new shiny things. 00:16:54.959 --> 00:16:55.459 I'm old enough. 00:16:55.499 --> 00:16:57.462 I've been doing web pages since before there were 00:16:57.502 --> 00:16:58.744 cascading style sheets. 00:16:58.764 --> 00:17:01.167 But anyway, as we were starting to do CSS and as 00:17:01.207 --> 00:17:02.810 we were starting to do JavaScript, it was all 00:17:02.850 --> 00:17:04.531 about progressive enhancement: 00:17:04.512 --> 00:17:06.616 do the shiny things for the browsers that can 00:17:06.675 --> 00:17:09.320 support it, but also have this nice baseline of 00:17:09.361 --> 00:17:11.965 functionality for those that can't. I've tried 00:17:11.986 --> 00:17:15.292 to do that with this base level config. 00:17:15.473 --> 00:17:17.958 I've got the visual-bell that's built into Emacs 00:17:18.018 --> 00:17:19.681 and it flashes the screen. It puts up this 00:17:19.800 --> 00:17:21.864 annoying triangle sign. 00:17:21.865 --> 00:17:23.568 I don't like that that much. 00:17:23.569 --> 00:17:26.875 This one, if I can make the bell ring... 00:17:27.412 --> 00:17:28.574 Oh, this would make the bell ring. 00:17:29.496 --> 00:17:30.137 Why is it not? 00:17:30.498 --> 00:17:32.282 [Sacha]: I guess `C-g` doesn't just 00:17:32.322 --> 00:17:33.384 automatically make it... 00:17:33.745 --> 00:17:36.150 Oh, `C-g` works for me, so I don't know. 00:17:36.230 --> 00:17:38.234 [Ross]: It usually works for me, too. 00:17:38.314 --> 00:17:41.079 I don't know why it's not triggering a bell here. 00:17:43.284 --> 00:17:43.805 [Sacha]: That's OK. 00:17:44.807 --> 00:17:45.669 [Ross]: Curse of the demo. 00:17:46.130 --> 00:17:46.350 [Sacha]: Yeah, yeah. 00:17:46.370 --> 00:17:47.833 Anyone interested can go look it up. NOTE emacs-lock-mode #+SCREENSHOT: emacs-chat-26-ross-a-baker-00-19-01-733.jpg #+SCREENSHOT_TIME: 00:19:01.733 00:17:48.860 --> 00:17:51.006 Then you have a couple of other things 00:17:51.067 --> 00:17:54.256 around, for example, making it easier to not kill 00:17:54.316 --> 00:17:57.565 a buffer. I hadn't known about 00:17:57.566 --> 00:17:58.870 `emacs-lock-mode` either. 00:17:59.070 --> 00:18:00.996 [Ross]: Oh yes, Emacs lock. 00:18:05.785 --> 00:18:08.532 Yeah, so there's this `emacs-lock-mode` 00:18:08.533 --> 00:18:09.557 and it is not the same. 00:18:09.657 --> 00:18:12.084 It's easy to confuse this with the lock files. 00:18:12.105 --> 00:18:15.395 You know those files that you get the hash signs, 00:18:15.415 --> 00:18:19.086 like you get ".#name-of-the-file#". 00:18:19.353 --> 00:18:21.816 That's something from when Emacs was used more on 00:18:21.896 --> 00:18:23.218 multi-user environments. 00:18:23.258 --> 00:18:26.202 Whenever you had a modified file, it would create 00:18:26.242 --> 00:18:28.425 these lock files, and it would prevent two users 00:18:28.465 --> 00:18:30.608 who were editing the same file on their own Emacs 00:18:31.128 --> 00:18:32.350 from stepping over each other. 00:18:32.911 --> 00:18:34.994 How often in modern times are we working on a 00:18:35.034 --> 00:18:36.055 multi-user system? 00:18:36.075 --> 00:18:38.318 If it is a multi-user system, are they both 00:18:38.418 --> 00:18:40.000 using Emacs on the same file? 00:18:40.080 --> 00:18:42.063 That's something that was very important in the 00:18:42.123 --> 00:18:44.466 late 90s and is not very important now. 00:18:44.868 --> 00:18:46.430 This is not those file locks. 00:18:46.510 --> 00:18:48.632 This is a buffer locking, 00:18:48.633 --> 00:18:50.376 and with the scratch buffer... 00:18:50.636 --> 00:18:52.920 I've got this turned on. 00:18:52.940 --> 00:18:53.561 Evaluate it. 00:18:53.841 --> 00:18:55.443 Make sure it worked. 00:18:55.463 --> 00:18:57.146 Let's go to our scratch buffer. 00:18:57.446 --> 00:18:58.628 Here's my scratch buffer. 00:18:58.708 --> 00:19:00.711 I don't want to accidentally kill it. 00:19:00.731 --> 00:19:02.713 If I try to kill it, it's going to say 00:19:02.714 --> 00:19:04.857 "Buffer *scratch​* is locked and cannot be killed." 00:19:05.618 --> 00:19:09.103 I'll use this for notes all over the place. 00:19:10.771 --> 00:19:13.454 I've accidentally killed things too many times. 00:19:13.494 --> 00:19:16.176 The scratch buffer is not typically backed by a 00:19:16.256 --> 00:19:18.999 file, and I've lost too many things. 00:19:19.019 --> 00:19:22.122 This way, I know this scratch buffer is always 00:19:22.182 --> 00:19:23.924 going to be a persistent place. 00:19:24.404 --> 00:19:25.365 There are other approaches to it. 00:19:25.385 --> 00:19:27.147 I think there's a `persistent-scratch` package 00:19:27.167 --> 00:19:27.667 that's out there. 00:19:27.687 --> 00:19:29.829 But again, that gets me out of the built-in zone. 00:19:29.949 --> 00:19:31.531 That might be a progressive enhancement, 00:19:31.551 --> 00:19:33.513 something that takes your scratch buffer and 00:19:33.553 --> 00:19:34.514 backs it with a file. 00:19:34.534 --> 00:19:35.875 That would probably be a good thing to do. 00:19:35.975 --> 00:19:37.557 So that's one of the packages I'll be looking at 00:19:37.657 --> 00:19:38.938 adding in the other config. 00:19:39.121 --> 00:19:41.087 But emacs-lock is built in. 00:19:41.107 --> 00:19:42.793 It's one of those things that's so obscure you 00:19:42.833 --> 00:19:44.699 can't even find it in the Emacs manual. 00:19:45.301 --> 00:19:46.064 It's just out there. 00:19:46.084 --> 00:19:47.107 I don't remember where I picked it up. 00:19:47.127 --> 00:19:49.154 I think just scanning somebody else's config I 00:19:49.174 --> 00:19:50.900 picked it up somewhere, but I couldn't find it in 00:19:50.940 --> 00:19:51.983 the manual at all. NOTE exiting Emacs: yes-or-no-p 00:19:52.334 --> 00:19:54.177 [Sacha]: I was going to ask you where you found it. 00:19:54.217 --> 00:19:56.381 You occasionally read other people's 00:19:56.441 --> 00:19:58.926 configurations for interesting ideas. 00:20:00.529 --> 00:20:05.177 And then you have a bunch of niceties. 00:20:06.279 --> 00:20:07.400 Actually, before I go into that, 00:20:07.401 --> 00:20:08.924 @Ray-On-Emacs has a question. 00:20:08.944 --> 00:20:10.447 "How do you exit Emacs then? 00:20:10.507 --> 00:20:14.293 Or do you never exit Emacs?" 00:20:14.754 --> 00:20:16.317 [Ross]: I almost never exit it. 00:20:16.803 --> 00:20:19.108 [Sacha]: I think the `emacs-lock-mode` for the 00:20:19.128 --> 00:20:22.135 Scratch doesn't kick in if you're closing Emacs 00:20:22.376 --> 00:20:22.877 as a whole? 00:20:23.940 --> 00:20:24.721 [Ross]: Correct. 00:20:24.741 --> 00:20:26.145 [Sacha]: If you're closing Emacs as a whole, it will 00:20:26.686 --> 00:20:29.312 happily still discard your Scratch buffer, but as 00:20:29.372 --> 00:20:32.820 Ross says, you could also not exit Emacs. 00:20:34.066 --> 00:20:37.352 [Ross]: Yeah, I've got the flippant line in 00:20:37.392 --> 00:20:38.575 here: ¨Exit Emacs? 00:20:38.595 --> 00:20:40.499 Surely there's been a terrible misunderstanding,¨ 00:20:40.519 --> 00:20:42.723 But I do configure the way that I do it. 00:20:42.883 --> 00:20:45.689 I want to make sure that the confirmation of it, 00:20:45.749 --> 00:20:46.671 it's the yes or no. NOTE yes-or-no-p 00:20:46.791 --> 00:20:48.072 I don't want to accidentally hit that 00:20:48.073 --> 00:20:52.041 key binding, so I will use the full fledged yes-or-no-p`. 00:20:52.021 --> 00:20:54.164 That's one area that I diverge from a lot of 00:20:54.204 --> 00:20:56.628 people's common things. One of the very early 00:20:56.688 --> 00:20:58.551 things a lot of people set in their config is 00:20:58.591 --> 00:21:00.773 they'll rebind this `yes-or-no-p` 00:21:00.774 --> 00:21:04.079 to the `y-or-n-p` so that way, they only have to type 00:21:04.119 --> 00:21:05.661 the one character. 00:21:05.681 --> 00:21:08.025 Emacs is fairly thoughtful about which ones that 00:21:08.065 --> 00:21:10.969 it makes you do "yes" or "no" versus the "y" or "n". The 00:21:11.009 --> 00:21:13.172 things that are quick and inconsequential, those 00:21:13.213 --> 00:21:14.815 are the ones where it gives you the prompt that 00:21:14.835 --> 00:21:16.898 it's just a "y" or "n" answer. 00:21:16.878 --> 00:21:18.540 The ones that are going to be more 00:21:18.600 --> 00:21:20.682 destructive and you really want to stop and slow 00:21:20.722 --> 00:21:22.002 down and think about, that's what the 00:21:22.003 --> 00:21:22.984 `yes-or-no-p` is for. 00:21:23.445 --> 00:21:25.767 People stomp all over that in their configs, 00:21:25.787 --> 00:21:27.589 and I think that's one area where I'm a little 00:21:27.649 --> 00:21:29.151 bit out of the mainstream. 00:21:29.291 --> 00:21:31.333 I like the defaults on Emacs there. 00:21:31.834 --> 00:21:33.676 It is thoughtful about which ones it prompts you 00:21:33.716 --> 00:21:34.316 for that on. 00:21:34.897 --> 00:21:37.079 But if I remember right, `confirm-kill-emacs` was 00:21:37.199 --> 00:21:39.181 one that was just a "y" or "n". That's one of the 00:21:39.221 --> 00:21:40.743 more dramatic ones that's out there. 00:21:41.484 --> 00:21:44.086 So I did customize that to make sure, yes, I 00:21:44.147 --> 00:21:45.568 really mean that. NOTE Processes 00:21:45.548 --> 00:21:47.451 Then there's some annoyances where 00:21:47.452 --> 00:21:49.374 if you've got a bunch of inferior processes 00:21:49.375 --> 00:21:50.857 running after I've already confirmed 00:21:50.858 --> 00:21:52.040 that I want to kill Emacs, 00:21:52.060 --> 00:21:53.763 well I just said I want to kill Emacs, 00:21:53.764 --> 00:21:56.087 so I didn't want to be prompted about the processes. 00:21:56.088 --> 00:21:59.434 I do quit Emacs enough I've customized how it shuts 00:21:59.495 --> 00:22:02.960 down. I don't do it that often. 00:22:02.961 --> 00:22:05.025 [Sacha]: I imagine if you really want to stop yourself from 00:22:05.377 --> 00:22:08.122 closing Emacs, you might just 00:22:09.003 --> 00:22:13.531 change the key binding and make you M-x it 00:22:14.052 --> 00:22:15.554 instead in order to exit it, 00:22:15.574 --> 00:22:17.858 so you really absolutely don't run into that. NOTE Moving keymaps 00:22:18.359 --> 00:22:21.264 In fact, your config has some of these lambdas 00:22:21.364 --> 00:22:24.710 where you are telling yourself, okay, this 00:22:24.770 --> 00:22:26.473 keyboard shortcut has changed. 00:22:26.909 --> 00:22:29.115 You have this thing in Magit, I think, where 00:22:29.156 --> 00:22:32.345 you're like, I moved the key map and now 00:22:32.445 --> 00:22:34.411 it's somewhere else. 00:22:34.431 --> 00:22:35.554 I thought that was interesting too. 00:22:35.574 --> 00:22:39.265 [Ross]: I'm trying to remember what that one was called. 00:22:39.717 --> 00:22:41.440 [Sacha]: There you go. 00:22:42.201 --> 00:22:43.262 [Ross]: Yeah, the `crab-juice--moved-to`. 00:22:43.503 --> 00:22:46.547 I'll bind my old key map as I'm trying 00:22:46.647 --> 00:22:48.009 to move things around. 00:22:48.851 --> 00:22:51.094 I used to use Spacemacs for a while. 00:22:51.174 --> 00:22:52.797 I used Doom Emacs for a while. 00:22:52.837 --> 00:22:54.639 I've looked at a lot of the other configs and 00:22:54.679 --> 00:22:57.343 I've said, okay, this architecture makes a little 00:22:57.444 --> 00:22:58.024 bit more sense. 00:22:58.085 --> 00:23:01.049 I want to rebind my prefixes that way, but still, 00:23:01.169 --> 00:23:02.712 for the ones that are deeply ingrained, 00:23:02.932 --> 00:23:04.735 it's sort of like a deprecation warning, except 00:23:04.775 --> 00:23:05.956 for key bindings. 00:23:08.202 --> 00:23:18.879 It looks like I used to run on `C-c g`. 00:23:18.979 --> 00:23:21.503 So if I try to do that, it's going to say, nope, 00:23:21.624 --> 00:23:22.665 moved to `C-c v c`. 00:23:22.886 --> 00:23:23.747 And now I can do that. 00:23:23.927 --> 00:23:27.493 `C-c v c`, and there's my Magit command. 00:23:28.368 --> 00:23:30.731 [Sacha]: Yeah, I like that you actually 00:23:30.791 --> 00:23:33.114 ended up mapping the whole keyboard shortcut 00:23:33.274 --> 00:23:35.697 instead of just the `C-c g` one because, of 00:23:35.737 --> 00:23:40.023 course, your muscle memory is getting you to put 00:23:40.143 --> 00:23:43.127 in the whole `C-c g` and then something. 00:23:44.008 --> 00:23:49.835 This one lets you have the full keyboard 00:23:49.875 --> 00:23:52.138 shortcut before it tells you the message. 00:23:54.389 --> 00:23:55.991 [Ross]: This is using user error. 00:23:56.011 --> 00:23:57.633 So I thought when I hit that one, I thought I 00:23:57.654 --> 00:23:59.396 would flash my mode line, but the mode line is 00:23:59.436 --> 00:24:00.117 not flashing. 00:24:00.217 --> 00:24:01.999 I know it flashes on my other machine. 00:24:02.881 --> 00:24:04.062 It flashes on my work machine. 00:24:04.082 --> 00:24:05.364 It doesn't flash on this one. 00:24:07.908 --> 00:24:09.330 [Sacha]: The challenge is getting a config 00:24:09.370 --> 00:24:11.292 to work in many different environments. 00:24:11.352 --> 00:24:12.774 You've got your work machine, you've got this 00:24:12.814 --> 00:24:15.899 personal machine, and then you've got the sandbox 00:24:16.039 --> 00:24:18.542 one that's got a really old version of Emacs on it. 00:24:19.617 --> 00:24:21.721 Okay, so a couple of other things that you've 00:24:21.782 --> 00:24:22.303 mentioned. NOTE Writing in Org Mode and Markdown 00:24:23.726 --> 00:24:27.313 You do a lot of writing in Org Mode as well, right? 00:24:27.353 --> 00:24:29.197 You've mentioned using the literal programming 00:24:29.237 --> 00:24:30.901 part of it a lot. 00:24:30.921 --> 00:24:32.524 You do a lot of work with Markdown. 00:24:33.547 --> 00:24:35.812 Do you have any interesting workflow tips for 00:24:35.872 --> 00:24:37.816 people who are finding themselves writing with 00:24:37.896 --> 00:24:39.820 either Org Mode or Markdown in a work environment? 00:24:41.588 --> 00:24:44.092 [Ross]: That's always a tension as well. 00:24:44.212 --> 00:24:46.296 I would rather be in Org Mode. 00:24:46.356 --> 00:24:48.279 Org Mode has more bells and whistles. 00:24:48.359 --> 00:24:50.963 You can program more things in it. 00:24:51.905 --> 00:24:54.709 It's a programmable Markdown in a sense. 00:24:55.891 --> 00:24:57.794 I'm much more comfortable in Org Mode in general. 00:24:57.934 --> 00:25:01.139 Now, I don't use Org as a personal organizer as much. 00:25:01.220 --> 00:25:01.881 I would like to. 00:25:01.981 --> 00:25:03.343 I've dabbled in it a little bit. 00:25:03.323 --> 00:25:05.046 I've seen some of your previous interviews and 00:25:05.066 --> 00:25:06.750 people doing mind-blowing things with that. 00:25:06.790 --> 00:25:07.992 I'm jealous of what they're doing. 00:25:08.012 --> 00:25:09.455 I would love to get to that level. 00:25:09.956 --> 00:25:12.721 I'm still a relative newcomer to Org versus how 00:25:13.583 --> 00:25:14.705 old I am at Emacs. 00:25:14.745 --> 00:25:16.429 I'm a relative newcomer to Org. 00:25:16.869 --> 00:25:18.913 So I haven't embraced that yet, but I do use it 00:25:18.974 --> 00:25:21.318 as my general, if I'm writing a document by 00:25:21.378 --> 00:25:23.863 myself, it will 100% be in Org. 00:25:23.843 --> 00:25:26.586 This document that we're in, it's my entire website. 00:25:26.606 --> 00:25:28.488 My Emacs config is in here. 00:25:28.688 --> 00:25:31.471 My Nix configuration for how my systems work is 00:25:31.571 --> 00:25:31.992 in here. 00:25:32.733 --> 00:25:34.074 I like to do barbecue. 00:25:34.114 --> 00:25:36.317 Some of my barbecue recipes are in here. 00:25:36.997 --> 00:25:38.979 Just general blog posts about things that have 00:25:39.040 --> 00:25:41.282 nothing to do with technology are in here. 00:25:42.043 --> 00:25:43.805 It's all one big Org file. 00:25:44.145 --> 00:25:45.967 That style is emphasized. 00:25:46.047 --> 00:25:49.351 I'm using ox-hugo to take this document and export 00:25:49.391 --> 00:25:50.492 it to the website. 00:25:50.860 --> 00:25:54.826 They encourage having one big large file for that. 00:25:54.906 --> 00:25:56.388 I followed suit on that. 00:25:56.428 --> 00:25:59.893 That's why I have just this one giant document. 00:26:00.033 --> 00:26:02.717 In the work context, if I'm starting a small 00:26:02.777 --> 00:26:04.540 document, I'll start it in Org Mode. 00:26:04.600 --> 00:26:06.583 But then when I started in Org Mode, if other 00:26:06.623 --> 00:26:09.147 people want to edit it, we got a small Emacs community, 00:26:09.167 --> 00:26:11.430 so depending who I'm sharing it with, that might be fine. 00:26:11.470 --> 00:26:13.393 There's a couple people I can pair with on that. 00:26:13.433 --> 00:26:15.255 If I'm working with them, everything is great. 00:26:15.356 --> 00:26:17.779 But that's not something... 00:26:17.928 --> 00:26:20.451 There are Org implementations that work in other 00:26:20.531 --> 00:26:22.895 editors, but it isn't as good anywhere else as it 00:26:22.975 --> 00:26:24.056 is in Emacs. 00:26:24.116 --> 00:26:26.439 That's what everybody's going to typically use. 00:26:26.459 --> 00:26:28.202 And if they're not using Emacs, they're not going 00:26:28.222 --> 00:26:30.605 to be that warm to Org. 00:26:30.905 --> 00:26:33.669 I hate to say the dreaded two-letter acronym, but 00:26:33.709 --> 00:26:35.992 with AI, you're seeing a lot more Markdown. 00:26:36.132 --> 00:26:37.954 People are starting to program in Markdown. 00:26:37.994 --> 00:26:39.897 Markdown is bigger than it's ever been before. 00:26:39.977 --> 00:26:42.180 It's just unavoidable at this point. 00:26:42.160 --> 00:26:44.542 So what I'll do is I'll start a document in Org 00:26:44.622 --> 00:26:46.564 and then I can export the Org document to 00:26:46.604 --> 00:26:49.767 Markdown if I'm going to put it into GitHub or 00:26:50.008 --> 00:26:51.189 sometimes I'll just use Google. 00:26:51.289 --> 00:26:53.551 It really depends on how the collaboration model is. 00:26:53.591 --> 00:26:55.213 If I'm going to be collaborating with a lot of 00:26:55.253 --> 00:26:57.093 people up front and iterating real quickly, 00:26:57.094 --> 00:26:59.817 I'll do that in a Google Doc and I'll export that to Markdown. 00:27:00.077 --> 00:27:01.798 If I'm writing it myself as a first draft, 00:27:01.799 --> 00:27:03.801 I'll start it in Org and export that to Markdown. 00:27:04.221 --> 00:27:06.063 Markdown ends up being the target in the 00:27:06.083 --> 00:27:08.185 workplace for anything that's permanent though. 00:27:08.317 --> 00:27:08.958 I don't love it. 00:27:09.038 --> 00:27:11.724 I wish Org had won. All my personal projects, 00:27:11.744 --> 00:27:13.647 they'll have a README.org on them rather than 00:27:13.687 --> 00:27:16.633 README.md, but it is what it is and you've got to 00:27:16.653 --> 00:27:18.056 play nice with your co-workers too. 00:27:19.298 --> 00:27:21.142 [Sacha]: Yeah, I find myself having a hard 00:27:21.182 --> 00:27:23.385 time remembering to switch the link syntax. 00:27:23.386 --> 00:27:27.213 I have a "do what I mean", just insert the link. 00:27:27.214 --> 00:27:28.415 I don't care what syntax it's in, 00:27:28.416 --> 00:27:29.017 just make it work. 00:27:30.019 --> 00:27:30.460 [Ross]: Nice. 00:27:32.245 --> 00:27:34.088 [Sacha]: You mentioned sometimes 00:27:34.188 --> 00:27:36.952 in your Zoom calls when you need to exchange 00:27:37.032 --> 00:27:38.775 notes with people faster, then you have to use 00:27:38.815 --> 00:27:40.297 Google Docs for that one because that's what 00:27:40.317 --> 00:27:41.138 they're familiar with. 00:27:41.659 --> 00:27:44.023 But there might be some easy ways for you to ship 00:27:44.063 --> 00:27:49.010 that back and forth between Org and Google Docs. NOTE Ethersync? 00:27:49.170 --> 00:27:52.596 [Ross]: One thing I'm really looking 00:27:52.676 --> 00:27:54.959 forward to is there's this project. 00:27:54.979 --> 00:27:56.261 I can't remember what it's called. 00:27:56.341 --> 00:27:59.586 Ethersync. Yes, thank you. 00:28:00.460 --> 00:28:02.683 I've seen some various attempts at that and 00:28:02.723 --> 00:28:04.406 I had a little bit of success with them, but none 00:28:04.426 --> 00:28:05.227 of them have lasted. 00:28:05.267 --> 00:28:06.609 I'm really rooting for that team. 00:28:07.129 --> 00:28:08.972 I feel like Emacs got rejuvenated. 00:28:08.992 --> 00:28:11.456 It used to be, okay, it was hard to use Emacs 00:28:11.496 --> 00:28:13.478 because if you're in a niche language and you're 00:28:13.538 --> 00:28:15.822 using a niche editor, you had to find a community 00:28:15.862 --> 00:28:18.165 that used both of those to have decent support. 00:28:18.145 --> 00:28:19.688 Then LSP came along. 00:28:19.708 --> 00:28:20.990 You have your language servers. 00:28:21.010 --> 00:28:23.475 So now I can write in these niche languages that 00:28:23.515 --> 00:28:25.358 not everybody's writing Emacs, but I still have a 00:28:25.398 --> 00:28:27.863 good experience thanks to Eglot and thanks to the 00:28:27.923 --> 00:28:30.007 language server provided by those communities. 00:28:30.348 --> 00:28:33.093 Emacs became viable again for doing a lot of 00:28:33.173 --> 00:28:34.315 development. 00:28:34.295 --> 00:28:38.201 And I'm hoping that Ethersync kind of becomes 00:28:38.281 --> 00:28:40.844 like the LSP of collaborative editing, where it's 00:28:40.925 --> 00:28:43.148 not something where it's only Emacs users sharing 00:28:43.168 --> 00:28:45.571 with each other, but it works across editors and 00:28:45.631 --> 00:28:47.494 people can collaborate in real time that way. 00:28:47.634 --> 00:28:50.098 That would replace that Google Docs use case 00:28:50.178 --> 00:28:50.519 I have. 00:28:50.959 --> 00:28:52.742 The only reason I use Google Docs is we can see 00:28:52.782 --> 00:28:54.825 each other typing and edit things on the fly 00:28:54.865 --> 00:28:55.265 together. 00:28:55.386 --> 00:28:57.729 So I'm very excited about that Ethersync project. 00:28:57.789 --> 00:28:58.770 I hope that takes off. 00:28:59.632 --> 00:29:02.936 [Sacha]: Yeah, I think that will probably be 00:29:02.997 --> 00:29:05.778 more feasible than hoping people will switch to Emacs 00:29:05.779 --> 00:29:06.840 and then use CRDT. 00:29:11.586 --> 00:29:14.109 There are other interesting 00:29:14.309 --> 00:29:16.151 collaboration things in your config. NOTE Managing Github with Forge #+SCREENSHOT: emacs-chat-26-ross-a-baker-00-30-35-700.jpg #+SCREENSHOT_TIME: 00:30:35.700 00:29:16.171 --> 00:29:19.435 You've mentioned Forge. You have some things 00:29:19.535 --> 00:29:21.878 in here for cloning to different directories 00:29:21.958 --> 00:29:24.180 depending on the repository, committing with 00:29:24.200 --> 00:29:26.383 different addresses, confirming before merging 00:29:26.483 --> 00:29:27.264 pull requests. 00:29:27.664 --> 00:29:29.346 Can you tell us a little bit about these quality 00:29:29.366 --> 00:29:30.728 of life improvements for you? 00:29:31.552 --> 00:29:36.182 [Ross]: Let me see if I can come up with an example here, 00:29:36.242 --> 00:29:36.944 first of all. 00:29:38.367 --> 00:29:41.994 So this is Forge for people who have not seen it. 00:29:42.014 --> 00:29:43.738 Let's do... 00:29:49.642 --> 00:29:50.803 [Sacha]: So when you don't need something 00:29:50.863 --> 00:29:53.566 like real-time collaboration, for the people who 00:29:53.606 --> 00:29:56.910 are watching, Forge lets you manage your Git 00:29:56.950 --> 00:30:01.575 repositories with the pull requests and other 00:30:01.615 --> 00:30:02.957 changes and things like that, right? 00:30:03.337 --> 00:30:04.659 I don't use it yet, so... Yes. 00:30:05.299 --> 00:30:07.902 [Ross]: Yeah, so it's essentially a GitHub client. 00:30:07.922 --> 00:30:11.786 So I was looking... I'm here in my Forge test directory. 00:30:12.167 --> 00:30:14.349 I'm trying not to share any of my work repositories. 00:30:14.389 --> 00:30:15.711 I use this primarily at work. 00:30:15.731 --> 00:30:18.754 I just set it up on this machine last night. 00:30:19.021 --> 00:30:21.183 Still probably a couple of rough edges. 00:30:21.203 --> 00:30:22.744 I can list the topics. 00:30:23.945 --> 00:30:26.387 And if I list the topics, I can see I've got pull 00:30:26.407 --> 00:30:27.969 request 20 open. 00:30:28.549 --> 00:30:31.372 I can hit enter on it and I can see, okay, it's open. 00:30:31.412 --> 00:30:32.673 It's in state pending. 00:30:32.753 --> 00:30:33.934 This is where it's going. 00:30:34.134 --> 00:30:35.095 These are the commits. 00:30:35.696 --> 00:30:38.558 I can look at the commit and see, okay, what is 00:30:38.638 --> 00:30:39.619 this person doing? 00:30:39.799 --> 00:30:43.722 Oh, they're trying to add another line to this file. 00:30:44.323 --> 00:30:46.245 And if I say, okay, that looks good. 00:30:46.565 --> 00:30:48.927 I can do, what is it? 00:30:53.513 --> 00:31:00.043 I can look at this and I can say... This works 00:31:00.103 --> 00:31:03.107 better if it is in a full screen editor. 00:31:04.349 --> 00:31:07.554 Let's do... I thought I was tracking this repo 00:31:07.634 --> 00:31:08.055 already. 00:31:11.640 --> 00:31:12.882 I had this working last night. 00:31:12.942 --> 00:31:14.164 Curse of the demo again. 00:31:14.945 --> 00:31:18.551 Anyway, I'm able to approve pull requests from 00:31:18.591 --> 00:31:18.771 this. 00:31:18.831 --> 00:31:20.854 I'm able to... 00:31:21.998 --> 00:31:23.000 Reject pull requests. 00:31:23.020 --> 00:31:24.763 I'm able to comment on pull requests. 00:31:24.823 --> 00:31:25.784 I can look at the diffs. 00:31:25.805 --> 00:31:29.731 So all of my integration with GitHub is done 00:31:29.791 --> 00:31:31.154 through this. 00:31:31.174 --> 00:31:34.519 I'm in an environment right now, I'm a staff engineer. 00:31:34.660 --> 00:31:36.283 I participate on a lot of teams. 00:31:36.363 --> 00:31:38.847 I get over a thousand GitHub notifications a day. 00:31:39.127 --> 00:31:41.211 The GitHub notifications that are built into 00:31:41.231 --> 00:31:42.994 the browser are just not sufficient. 00:31:43.295 --> 00:31:45.940 That's something I was able to pull it in with 00:31:46.020 --> 00:31:48.585 Forge. I'm able to tag things. I'm able to 00:31:48.685 --> 00:31:49.687 mass mark the things. 00:31:49.787 --> 00:31:51.951 I get a lot of infrastructure things that they're 00:31:51.991 --> 00:31:54.014 important, but they're not important to me. I 00:31:54.054 --> 00:31:56.419 get notified on them all the time through the 00:31:56.459 --> 00:31:58.162 various automations. I'll get them. 00:31:58.242 --> 00:32:00.947 I'll get a hundred of those at a time and I can 00:32:00.987 --> 00:32:02.770 just mark a region of those and knock out a 00:32:02.830 --> 00:32:05.455 hundred and say, okay, that's off my plate. 00:32:05.435 --> 00:32:08.798 Moving it into Emacs, as awkward as this demo is, 00:32:09.139 --> 00:32:10.720 it works great on my work machine. 00:32:11.201 --> 00:32:13.243 I'm able to keep up with things so much better 00:32:13.283 --> 00:32:15.465 than I was before. 00:32:15.485 --> 00:32:15.985 Let's see. 00:32:16.326 --> 00:32:18.908 I should be able to, if I want to look a little 00:32:18.968 --> 00:32:21.190 bit deeper and see things in the GitHub, at least 00:32:21.230 --> 00:32:21.751 this works. NOTE #+SCREENSHOT: emacs-chat-26-ross-a-baker-00-32-21-833.jpg #+SCREENSHOT_TIME: 00:32:21.833 00:32:21.811 --> 00:32:23.893 I can open it and see this directly. 00:32:23.913 --> 00:32:25.795 Then if I want to merge it from here, I can. 00:32:26.616 --> 00:32:31.020 Or I should be able to merge. 00:32:31.777 --> 00:32:33.820 Now I can do a regular merge or a squash 00:32:33.860 --> 00:32:36.084 merge or a rebase just the same as I have this 00:32:36.144 --> 00:32:36.665 button here. 00:32:36.705 --> 00:32:37.967 That's all built in here. 00:32:38.087 --> 00:32:47.543 Let's go ahead and merge this pull request. 00:32:47.563 --> 00:32:47.903 [Sacha]: All right. 00:32:47.923 --> 00:32:50.968 That little detail about you dealing with like a 00:32:51.088 --> 00:32:54.107 thousand notifications a day? 00:32:55.249 --> 00:32:59.195 It makes the guardrails that you added to Forge 00:32:59.216 --> 00:33:01.499 in your config even more interesting because you 00:33:02.060 --> 00:33:04.585 can stop yourself from accidentally automatically 00:33:04.665 --> 00:33:06.668 merging in things that need more review or 00:33:06.708 --> 00:33:09.473 whatever, which of course is difficult to 00:33:09.513 --> 00:33:11.095 demonstrate at the moment because 00:33:11.096 --> 00:33:13.680 work stuff, private. 00:33:13.660 --> 00:33:17.305 But if other people are listening and thinking, 00:33:17.345 --> 00:33:19.426 they're dealing with a similar volume, 00:33:19.427 --> 00:33:23.733 you can modify Emacs to stop you from making mistakes 00:33:24.173 --> 00:33:24.614 like that. 00:33:25.595 --> 00:33:28.499 You can add just enough friction for the cases 00:33:28.559 --> 00:33:31.503 where you need to pay more attention. 00:33:31.523 --> 00:33:34.246 And @PuercoPop, there's a gh-notify 00:33:34.286 --> 00:33:36.669 package specifically for high-volume GitHub 00:33:36.750 --> 00:33:37.571 notifications. 00:33:38.332 --> 00:33:41.035 I'm not sure... Have you come across that one yet? 00:33:42.230 --> 00:33:44.758 [Ross]: I'm curious what that looks. 00:33:44.798 --> 00:33:49.753 The term "notify" concerns me, like if 00:33:49.793 --> 00:33:52.079 it's helping me manage notifications in bulk, 00:33:52.080 --> 00:33:54.087 I'm interested. If it's going to pop up something 00:33:54.147 --> 00:33:56.170 every time I get a new notification, 00:33:56.190 --> 00:33:57.372 I'm very much disinterested. 00:33:57.393 --> 00:34:00.177 I don't know whether to be excited or horrified, 00:34:00.197 --> 00:34:01.800 but I'll take a look. 00:34:01.820 --> 00:34:03.263 [Sacha]: I think because @PuercoPop is 00:34:03.343 --> 00:34:05.206 recommending it specifically in the context of 00:34:05.286 --> 00:34:08.172 high-volume notifications, it might provide you 00:34:08.212 --> 00:34:10.576 that inboxing where you can filter. 00:34:10.636 --> 00:34:14.322 I haven't looked into it myself, but I'm getting 00:34:14.422 --> 00:34:16.505 the sense that this is a problem that many people 00:34:16.887 --> 00:34:18.028 who use Emacs face 00:34:18.029 --> 00:34:20.353 and who solve it a very Emacs-y way. 00:34:21.210 --> 00:34:28.296 [Ross]: Yes. Forge has all these nice things built into it already, 00:34:28.297 --> 00:34:29.940 but I wanted to customize the workflow a 00:34:29.980 --> 00:34:30.400 little bit. 00:34:30.760 --> 00:34:32.601 It's right here in the comments, the things that 00:34:32.641 --> 00:34:32.882 it does. 00:34:32.942 --> 00:34:34.463 It makes sure that it's mergeable, 00:34:34.464 --> 00:34:36.445 there's no merge conflicts, 00:34:36.446 --> 00:34:38.546 makes sure that all the status checks ran, 00:34:38.547 --> 00:34:40.830 makes sure that all the approval is done. 00:34:41.070 --> 00:34:43.671 One thing that the Forge does not do is 00:34:43.672 --> 00:34:47.877 it does not show you comments on your pull request. 00:34:48.667 --> 00:34:50.951 There's a code-review package that's out there. 00:34:51.172 --> 00:34:52.875 It's been through a lot of forks. 00:34:53.014 --> 00:34:54.958 I'm still integrating that into my workflow. 00:34:54.978 --> 00:34:57.644 I've got a fork of a fork of a fork that I have. 00:34:58.125 --> 00:35:00.369 This is life in Emacs, for better or for worse. 00:35:00.790 --> 00:35:02.994 I've got a fork of a fork of a fork of a fork so 00:35:03.034 --> 00:35:04.096 I can see those comments. 00:35:04.116 --> 00:35:05.919 I haven't fully integrated it with Forge yet. 00:35:05.939 --> 00:35:07.502 That's a work in progress. 00:35:07.718 --> 00:35:09.941 But anyway, I didn't want to be merging these 00:35:10.001 --> 00:35:12.405 things and ignoring if somebody's taking the time 00:35:12.425 --> 00:35:13.346 to give me feedback. 00:35:13.386 --> 00:35:15.048 I don't want to merge things because sometimes 00:35:15.089 --> 00:35:16.991 people will approve things, but they'll say, hey, 00:35:17.031 --> 00:35:17.993 just a little nit to pick. 00:35:18.033 --> 00:35:19.034 You got a typo here. 00:35:19.395 --> 00:35:21.257 Don't want to slow down progress, but you might 00:35:21.297 --> 00:35:23.060 want to take a look at this. 00:35:23.080 --> 00:35:25.584 If I'm just sitting here entirely inside of 00:35:25.604 --> 00:35:26.605 this, I'll never see that. 00:35:26.625 --> 00:35:29.389 So this also checks for unresolved threads. 00:35:29.369 --> 00:35:31.753 I've got some embedded GraphQL inside of here 00:35:31.793 --> 00:35:34.578 that makes it all work and we grab that data from 00:35:34.638 --> 00:35:37.303 the GraphQL and then I'm able to just do this 00:35:37.383 --> 00:35:39.807 inside a list and prompt on these things and if 00:35:39.867 --> 00:35:41.570 it finds there's anything that isn't right it 00:35:41.630 --> 00:35:44.936 gives me these safety checks along the way and 00:35:44.956 --> 00:35:48.542 yeah it's really nice. NOTE Committing with work addresses vs personal 00:35:49.501 --> 00:35:52.185 One other issue that I had was we had this 00:35:52.245 --> 00:35:53.987 requirement at work where we needed to start 00:35:54.448 --> 00:35:56.571 committing with our work email addresses rather 00:35:56.611 --> 00:35:57.552 than our personal addresses. 00:35:57.592 --> 00:36:00.596 I use my same GitHub account for work and personal. 00:36:00.656 --> 00:36:02.178 I needed to get that moved over. 00:36:02.699 --> 00:36:05.423 To do that with GitHub, you need to do those 00:36:05.483 --> 00:36:07.285 merges via the API. 00:36:07.305 --> 00:36:08.487 I've got things configured. 00:36:08.567 --> 00:36:10.750 All of my work repositories are 00:36:10.730 --> 00:36:13.895 partitioned off into their own subdirectory of my 00:36:13.935 --> 00:36:14.896 projects directory. 00:36:14.996 --> 00:36:17.320 I've got a git config that sets my email address 00:36:17.340 --> 00:36:17.821 to that. 00:36:17.881 --> 00:36:20.405 So if I'm doing things locally, it works. 00:36:20.425 --> 00:36:22.908 But if I'm doing these API merges that are 00:36:22.989 --> 00:36:25.272 necessary for certain repositories with certain 00:36:25.332 --> 00:36:28.657 safety checks, I needed to be able to thread that along. 00:36:28.697 --> 00:36:30.921 I've got a bug fix in here for that, where I'm 00:36:30.961 --> 00:36:32.944 threading along that email address that reads my 00:36:32.984 --> 00:36:35.808 git config and passes that along to the API 00:36:35.968 --> 00:36:36.930 endpoint. 00:36:36.950 --> 00:36:39.133 Otherwise, it was committed as the wrong identity. 00:36:39.113 --> 00:36:40.514 There's a few things here that I'd like to 00:36:40.574 --> 00:36:41.936 contribute upstream to Forge. 00:36:41.976 --> 00:36:43.777 Some of it is very bespoke to me, 00:36:43.778 --> 00:36:45.920 and some of it I think would be useful to everybody. 00:36:45.940 --> 00:36:49.363 I need to tease that apart still. 00:36:49.383 --> 00:36:51.866 [Sacha]: I like your technique of separating 00:36:52.066 --> 00:36:54.288 things into different subdirectories and so that 00:36:54.909 --> 00:36:56.691 all of your work stuff goes in the work directory 00:36:56.771 --> 00:36:59.033 or all of your personal stuff goes into projects 00:36:59.093 --> 00:36:59.534 or whatever. 00:36:59.974 --> 00:37:01.936 Then you can change your settings based on 00:37:01.976 --> 00:37:03.838 the location of the project. 00:37:03.818 --> 00:37:07.644 Even cloning a project will automatically pick 00:37:08.065 --> 00:37:09.787 the right directory to put it in. 00:37:09.807 --> 00:37:11.009 I saw that in your config. 00:37:11.210 --> 00:37:12.612 That's pretty clever. NOTE Emacs tinkering as stress relief 00:37:13.533 --> 00:37:15.496 I'm curious because a lot of people find it 00:37:15.637 --> 00:37:19.923 challenging to balance the workflow improvement I 00:37:19.963 --> 00:37:22.047 get to tinker with Emacs versus actually getting 00:37:22.087 --> 00:37:24.330 work done, the time balance between those two. 00:37:24.370 --> 00:37:26.113 It's a little hard to figure out sometimes. 00:37:26.093 --> 00:37:27.735 What is this like for you when you're 00:37:27.775 --> 00:37:29.498 sitting down, you're doing some work, and you 00:37:29.518 --> 00:37:31.801 realize "There's probably something 00:37:31.821 --> 00:37:34.225 that I can write to make this 00:37:34.265 --> 00:37:35.826 smoother?" 00:37:35.827 --> 00:37:37.770 [Ross]: I've been doing Emacs off and on 00:37:37.830 --> 00:37:39.893 for about 30 years now, and when I get that 00:37:39.953 --> 00:37:42.836 balance right, I'll let you know. 00:37:42.837 --> 00:37:44.760 I don't have it right yet. 00:37:45.668 --> 00:37:48.472 I know I spend more time tweaking 00:37:48.572 --> 00:37:49.913 things than I probably should. 00:37:49.914 --> 00:37:52.698 It's a stress reliever for me. 00:37:52.718 --> 00:37:56.563 I get tired of doing the other things. 00:37:56.983 --> 00:37:58.946 I still love computers. 00:37:58.986 --> 00:38:01.129 I've loved computers since I was six years old. 00:38:01.189 --> 00:38:04.033 Got my first computer then. 00:38:04.093 --> 00:38:06.597 My aunt got me a book on programming when I was six. 00:38:06.637 --> 00:38:08.900 I started going through that book, and I was off 00:38:08.940 --> 00:38:10.562 to the races on that. 00:38:10.542 --> 00:38:13.747 I relax by doing computers. Some of 00:38:13.767 --> 00:38:16.451 the open source chaos is a lot to deal with, and 00:38:16.772 --> 00:38:18.775 using the same languages that I use at work that 00:38:18.835 --> 00:38:20.678 feels too much like work, so being able to 00:38:20.718 --> 00:38:23.042 withdraw and just tweak things... "Okay, this 00:38:23.162 --> 00:38:28.550 irritated me." It's a combination... From a pure 00:38:28.590 --> 00:38:31.415 productivity perspective, I overdo it for sure, but 00:38:31.695 --> 00:38:33.718 I guess the stress relief if I would burn out 00:38:33.798 --> 00:38:36.102 otherwise... Maybe I'm getting it right after 00:38:36.182 --> 00:38:37.444 all. It's hard to say. 00:38:38.403 --> 00:38:40.590 [Sacha]: Besides, you might also be able 00:38:40.670 --> 00:38:43.900 to say to yourself, well, I do have some Emacs 00:38:43.980 --> 00:38:47.431 co-workers. I'm sure they can benefit from this too. 00:38:47.451 --> 00:38:47.812 [Ross]: I do. 00:38:50.059 --> 00:38:51.222 [Sacha]: You're helping them out too. 00:38:51.242 --> 00:38:54.370 Okay, I want to dig into some of the things that 00:38:54.410 --> 00:38:58.074 you have around improving Emacs' capabilities as 00:38:58.155 --> 00:38:59.436 a self-documenting editor. 00:38:59.456 --> 00:39:02.320 You have some configuration snippets that make 00:39:02.380 --> 00:39:03.561 getting help easier. 00:39:04.202 --> 00:39:06.806 I hadn't come across them before, so I wanted to 00:39:07.266 --> 00:39:08.708 make sure other people knew about them too. 00:39:09.108 --> 00:39:12.533 Like your thing about making apropos look at all 00:39:12.573 --> 00:39:14.936 the things and sort by scores. 00:39:15.396 --> 00:39:16.278 It can sort by scores? 00:39:16.378 --> 00:39:19.041 What kind of scores are we talking about? 00:39:19.021 --> 00:39:30.639 How does this work? 00:39:30.640 --> 00:39:32.080 [Ross]: One thing that I have done, 00:39:32.081 --> 00:39:34.565 a little bit off-topic for the question that you've 00:39:34.825 --> 00:39:37.289 asked, but I have linked everything very 00:39:37.329 --> 00:39:41.836 aggressively to the info manual where I found things. 00:39:41.856 --> 00:39:45.602 Sorry, I got caught in a recursive edit here. 00:39:46.932 --> 00:39:51.100 So let's open this. 00:39:53.044 --> 00:39:58.094 What was that doing? 00:39:58.134 --> 00:40:01.099 I'm also struggling because I've been very busy 00:40:01.140 --> 00:40:03.163 at work lately. I'm used to the Mac key 00:40:03.204 --> 00:40:05.007 bindings again. I'm back here on my Linux 00:40:05.047 --> 00:40:08.113 machine. I haven't adapted to the Ctrl versus the Cmd. 00:40:09.038 --> 00:40:11.561 So yeah, all apropos commands will sort their 00:40:11.621 --> 00:40:13.904 lists or results in alphabetical order. 00:40:13.924 --> 00:40:16.186 If sort by scores is non-null, they try to guess 00:40:16.226 --> 00:40:19.811 the relevance of each result. 00:40:19.831 --> 00:40:21.412 My comment looks like it's a little bit obsolete 00:40:21.472 --> 00:40:23.695 here because it looks like for the documentation, 00:40:23.815 --> 00:40:26.538 I do like the sorting by scores, but for regular 00:40:26.578 --> 00:40:27.860 apropos, I did not. 00:40:27.920 --> 00:40:30.643 I think I tweaked something and didn't update 00:40:30.663 --> 00:40:31.584 the documentation. 00:40:31.604 --> 00:40:33.787 [Sacha]: That's okay. I was just curious about it. 00:40:34.267 --> 00:40:38.252 Then you also have some things like, 00:40:38.654 --> 00:40:43.584 if it's an autoloaded symbol, then go ahead and 00:40:43.624 --> 00:40:46.229 load the library so that you can get help for it. 00:40:46.510 --> 00:40:49.415 If there's a shortdoc example, include that 00:40:49.475 --> 00:40:53.764 as well in the documentation of a function. 00:40:54.165 --> 00:40:57.391 So if people are reading configs for ideas, I 00:40:57.451 --> 00:40:59.315 thought I would call attention to this part 00:40:59.335 --> 00:41:00.898 because it sounds really handy. 00:41:02.566 --> 00:41:03.449 [Ross]: Yeah, I found these. 00:41:03.569 --> 00:41:04.912 I've just been going through... 00:41:04.913 --> 00:41:06.137 Part of my relaxation... NOTE Under-appreciated Emacs built-ins 00:41:06.157 --> 00:41:08.303 This is such a dorky thing to say, but part of 00:41:08.343 --> 00:41:10.430 the relaxation is I'll go through the Emacs 00:41:10.470 --> 00:41:11.774 manual looking for things. 00:41:11.794 --> 00:41:14.362 The overall structure of this document follows 00:41:14.422 --> 00:41:16.508 the Emacs manual fairly closely, and that's 00:41:16.825 --> 00:41:18.087 not by coincidence. 00:41:18.147 --> 00:41:20.191 I've just been going through chapter by chapter. 00:41:20.311 --> 00:41:23.476 Okay, well, I started with Emacs 19 or whatever. 00:41:23.576 --> 00:41:25.600 Emacs 31 is around the corner. 00:41:25.620 --> 00:41:27.243 I better at least catch up to all the goodies 00:41:27.263 --> 00:41:27.924 that are in 30. 00:41:28.024 --> 00:41:30.368 I'll read the manual, and I'm 00:41:30.408 --> 00:41:31.690 finding a lot of these things there. 00:41:31.730 --> 00:41:33.473 That's why I have so many links back to the 00:41:33.533 --> 00:41:36.678 manual inside of my code, just because I'm going 00:41:36.719 --> 00:41:37.320 through that. 00:41:37.340 --> 00:41:39.964 That's where you can find a lot of these hidden gems. 00:41:40.180 --> 00:41:42.423 I'm hosting the Emacs Carnival this month, and 00:41:42.463 --> 00:41:44.546 it's the underappreciated Emacs built-ins. 00:41:44.566 --> 00:41:46.649 I said, if you want to find any of them, go 00:41:46.729 --> 00:41:47.530 in the manual. 00:41:47.570 --> 00:41:49.473 They're there by the bushel. 00:41:49.493 --> 00:41:53.098 [Sacha]: It feels almost like a book club. 00:41:53.118 --> 00:41:54.600 We're going through the manual together and 00:41:54.660 --> 00:41:58.846 finding all these gems that I wouldn't have come 00:41:58.866 --> 00:42:00.929 across otherwise, because Emacs source code is too 00:42:00.989 --> 00:42:02.071 big and all that. 00:42:02.151 --> 00:42:05.035 I've been really appreciating the Emacs Carnival 00:42:05.055 --> 00:42:06.357 submissions thus far. 00:42:07.585 --> 00:42:12.371 Of course, there are lots of other things that 00:42:12.452 --> 00:42:15.436 are not in Emacs that you've also been exploring 00:42:16.176 --> 00:42:18.480 and checking out. NOTE gptel 00:42:20.502 --> 00:42:23.907 Am I allowed to ask about the gptel stuff? 00:42:24.808 --> 00:42:25.128 [Ross]: Of course. 00:42:25.169 --> 00:42:31.457 [Sacha]: I was curious about how the quick lookup stuff 00:42:31.497 --> 00:42:32.518 has been working out for you. 00:42:32.538 --> 00:42:34.521 What kinds of things do you like to use that for? 00:42:36.290 --> 00:42:38.640 [Ross]: That's one that it was pretty 00:42:38.680 --> 00:42:40.689 exciting and I haven't really fully worked it 00:42:40.769 --> 00:42:41.954 into my workflow yet. 00:42:42.055 --> 00:42:43.842 I tried it. I liked it. I haven't 00:42:43.902 --> 00:42:44.726 internalized it. 00:42:44.746 --> 00:42:46.654 I was trying that one night. 00:42:47.562 --> 00:42:50.425 what what is capable of is you can just highlight 00:42:50.485 --> 00:42:53.809 something and do. It's almost like a quick 00:42:53.829 --> 00:42:55.711 Google search, except it'll bounce it off the 00:42:55.751 --> 00:42:58.754 model of your choice. It will pop up things a 00:42:58.794 --> 00:43:00.977 little bit faster that way. For the most part, I'm 00:43:01.017 --> 00:43:03.520 just using stock gptel, where I've just got the 00:43:03.540 --> 00:43:06.022 chatbot window. I'll copy a region and I'll 00:43:06.042 --> 00:43:08.986 put it into gptel. I'll run it off the model, say 00:43:09.026 --> 00:43:11.348 hey, I got a question about this, help me with this 00:43:11.388 --> 00:43:14.671 compile error, help me with that, and so forth. 00:43:14.672 --> 00:43:15.713 I don't... 00:43:17.110 --> 00:43:19.853 Talking about AI is always difficult because it's 00:43:19.893 --> 00:43:23.236 such a controversial thing and to use it at all, 00:43:23.276 --> 00:43:24.457 you offend a lot of people. 00:43:24.598 --> 00:43:26.479 I'll say, I don't really buy into the 00:43:26.539 --> 00:43:27.500 agentic workflow. 00:43:27.560 --> 00:43:29.723 I'm happy just using things as a chatbot and 00:43:30.163 --> 00:43:31.945 little snippets here and there where I'm still 00:43:32.005 --> 00:43:32.746 firmly driving. 00:43:32.786 --> 00:43:34.828 So the people who are fully bought into it, they 00:43:34.888 --> 00:43:36.369 don't think I'm using it right. 00:43:36.890 --> 00:43:39.733 Then there are a lot of ethical concerns that 00:43:39.793 --> 00:43:40.253 I share. 00:43:40.453 --> 00:43:42.055 I don't have the gptel on this. 00:43:42.355 --> 00:43:43.216 I use it for work. 00:43:43.276 --> 00:43:45.058 I don't use it personally. 00:43:45.038 --> 00:43:47.422 That's just the way I've been able to 00:43:47.442 --> 00:43:49.465 compartmentalize all my concerns with it. 00:43:49.565 --> 00:43:51.989 But just being a technologist, being a principal 00:43:52.049 --> 00:43:54.192 staff engineer, I do need to get up to speed with 00:43:54.212 --> 00:43:54.733 these tools. 00:43:54.753 --> 00:43:57.958 So I have had to get used to them and keep up 00:43:57.978 --> 00:43:59.981 with what they're capable of and what they're not. 00:44:00.001 --> 00:44:03.626 At least for my... Setting aside all the 00:44:03.706 --> 00:44:07.733 concerns around environment and whatnot, all the 00:44:07.773 --> 00:44:09.375 other concerns that we have with it , 00:44:09.355 --> 00:44:11.738 I find that GPTel is that sweet spot where I've 00:44:11.758 --> 00:44:14.020 just got this pair programmer where I can run 00:44:14.080 --> 00:44:16.804 things off, bounce a few questions off it, get 00:44:16.864 --> 00:44:19.387 things, but it's not getting write access to my 00:44:19.447 --> 00:44:20.268 file system. 00:44:20.828 --> 00:44:22.390 I have a security background. Worked at a 00:44:22.410 --> 00:44:24.292 security company for several years. 00:44:24.572 --> 00:44:27.115 Just the access that people are giving to these 00:44:27.236 --> 00:44:29.758 agents, I find absolutely terrifying. 00:44:29.778 --> 00:44:32.021 There's a lot of work that goes into using that 00:44:32.081 --> 00:44:33.743 properly, I think. 00:44:33.723 --> 00:44:36.187 I get most of the benefit. I feel like I'm 00:44:36.268 --> 00:44:39.133 de-skilling myself a little bit less than I do 00:44:39.213 --> 00:44:41.577 when I try the agentic workflows elsewhere. 00:44:41.597 --> 00:44:43.019 So I love gptel. 00:44:43.861 --> 00:44:46.626 Now gptel, it does have this interesting agent mode. 00:44:46.666 --> 00:44:47.968 I haven't tried it yet. 00:44:48.489 --> 00:44:51.675 Again, I have some security concerns around that. 00:44:51.695 --> 00:44:53.518 I need to figure out how to make sure that 00:44:53.558 --> 00:44:55.782 whatever I give it write access to... 00:44:55.762 --> 00:44:57.685 It's not getting too much write access to 00:44:57.785 --> 00:45:00.009 anything. But that is a project that's on my 00:45:00.049 --> 00:45:02.533 radar as well, to try to do things that really do 00:45:02.593 --> 00:45:04.857 benefit from this unattended work. 00:45:05.658 --> 00:45:06.640 I'll dabble in that as well. 00:45:06.680 --> 00:45:08.543 There's a whole gptel ecosystem. 00:45:08.964 --> 00:45:13.511 Karthik does an amazing job with that. 00:45:13.531 --> 00:45:15.434 [Sacha]: I like that you have a very 00:45:15.474 --> 00:45:16.676 considerate approach to it. 00:45:16.776 --> 00:45:19.281 I also like the humor with which you describe 00:45:19.641 --> 00:45:22.005 this and other things in your config. 00:45:23.892 --> 00:45:29.347 You aren't buying into the hype. You're like, 00:45:29.407 --> 00:45:32.014 this is your least unfavorite way 00:45:32.055 --> 00:45:32.957 of doing things. 00:45:33.494 --> 00:45:37.899 [Ross]: Yes, exactly. There's some commentary 00:45:37.999 --> 00:45:40.843 out on one of the nuclear test sites about "No 00:45:40.943 --> 00:45:43.045 deed of honor is performed here." I've quoted 00:45:43.085 --> 00:45:44.826 some of that in this part of the config too, 00:45:44.827 --> 00:45:47.830 just kind of my protest of doing this. 00:45:47.831 --> 00:45:50.273 I've got to do this and there are certain ways... 00:45:50.274 --> 00:45:54.558 If I have to do this, these are the ways that work for me. 00:45:54.559 --> 00:45:56.120 I'm getting some benefits out of this, 00:45:56.121 --> 00:45:59.383 but also I don't entirely feel great about this either. 00:45:59.384 --> 00:46:00.446 That's kind of how I've 00:46:01.146 --> 00:46:02.488 compartmentalized that. 00:46:03.295 --> 00:46:04.977 [Sacha]: Yeah, that makes sense. NOTE Getting older 00:46:04.997 --> 00:46:07.179 One of the other things that you joke about a 00:46:07.339 --> 00:46:09.681 couple of times in your config is getting older. 00:46:10.662 --> 00:46:12.864 So with things like your `repeat-exit-timeout` and 00:46:12.884 --> 00:46:14.806 your `mode-line-bell-flash-time`, you're like, 00:46:14.826 --> 00:46:18.209 okay, might change it to track my senescence. 00:46:22.272 --> 00:46:23.454 I appreciate your sense of humor. 00:46:24.615 --> 00:46:26.016 Are you finding that there are things that you 00:46:26.036 --> 00:46:29.559 like to tweak about Emacs in anticipation of 00:46:29.799 --> 00:46:31.581 change as you get older? 00:46:32.776 --> 00:46:35.867 [Ross]: Well, my font size is definitely bigger than it 00:46:35.987 --> 00:46:36.509 used to be. 00:46:36.650 --> 00:46:39.681 I guess that's something that has changed over 00:46:39.701 --> 00:46:40.002 the years. 00:46:40.042 --> 00:46:42.109 I used to be able to operate on a tiny font, no 00:46:42.149 --> 00:46:44.558 longer on that. NOTE Lindy's Law and tool longevity 00:46:46.799 --> 00:46:50.423 I guess Emacs and aging just feels like I made a 00:46:50.523 --> 00:46:52.665 really good... I wish I'd stuck with it for the 00:46:52.725 --> 00:46:53.486 full 30 years. 00:46:53.506 --> 00:46:55.848 I'd be fumbling around a little bit less in this demo. 00:46:55.868 --> 00:46:57.650 I'm not as smooth as somebody who'd been on it 00:46:57.690 --> 00:46:59.932 for the full 30 years, so I regret that time. 00:47:00.473 --> 00:47:02.795 I've just seen so many editors come and go. 00:47:02.815 --> 00:47:06.239 And when I talk to junior developers, I do a lot 00:47:06.279 --> 00:47:06.999 of mentoring. 00:47:07.600 --> 00:47:09.802 Like, I'm not necessarily going to say that you 00:47:09.842 --> 00:47:12.485 should go down this road, but if you do go down 00:47:12.525 --> 00:47:14.427 this road, it is something. 00:47:14.778 --> 00:47:17.533 that is probably going to be richly rewarding 00:47:17.534 --> 00:47:18.967 for you for a long time. 00:47:18.968 --> 00:47:20.800 There's this thing... I think it's called 00:47:20.801 --> 00:47:23.833 Lindy's Law. I'm sure I've quoted it here somewhere. 00:47:23.834 --> 00:47:25.851 The expected lifetime of something is 00:47:25.912 --> 00:47:28.720 proportional to how long it's already existed. 00:47:28.852 --> 00:47:31.433 And with Emacs... Emacs has been around... 00:47:31.434 --> 00:47:34.167 I'm very old. I'm 47. But Emacs is a little bit 00:47:34.168 --> 00:47:36.133 older than I am, depending on how you start counting. 00:47:36.134 --> 00:47:39.333 Emacs has been around for a long time. 00:47:39.334 --> 00:47:42.621 It doesn't have the market share that it used to, 00:47:42.741 --> 00:47:44.284 but it's something that still has a viable 00:47:44.304 --> 00:47:45.527 community. 00:47:46.081 --> 00:47:47.463 You do the Emacs News every week. 00:47:47.523 --> 00:47:49.326 You know probably better than anybody else how 00:47:49.386 --> 00:47:50.748 viable the community is right now. 00:47:50.808 --> 00:47:52.271 It's still an exciting place to be. 00:47:52.371 --> 00:47:55.095 I don't see that changing anytime soon, whereas 00:47:55.135 --> 00:47:57.759 I've seen so much hype over... Everybody wanted to 00:47:57.800 --> 00:48:01.065 get on Atom, and now Atom is a basically dead editor. 00:48:01.165 --> 00:48:03.589 Sublime had this rise and fall. 00:48:03.609 --> 00:48:04.650 So many other editors. 00:48:04.751 --> 00:48:07.515 Now I see people... VS Code was dominant for a 00:48:07.575 --> 00:48:09.418 while, and now I see a lot of people abandoning 00:48:09.478 --> 00:48:10.139 VS Code. 00:48:10.239 --> 00:48:12.082 It's still the dominant editor, but you can start 00:48:12.122 --> 00:48:13.945 to see people moving away from that. 00:48:13.925 --> 00:48:16.107 Emacs is always going to be here. 00:48:16.127 --> 00:48:19.731 So just as I've aged, I appreciate that Emacs, 00:48:20.312 --> 00:48:22.274 that same thing that I was using back when I was 00:48:22.314 --> 00:48:24.696 18, a freshman in college, it's still working for 00:48:24.736 --> 00:48:25.257 me at 47. 00:48:25.677 --> 00:48:28.160 And by the time I hang up my keyboard, pretty 00:48:28.220 --> 00:48:30.162 sure I'm still going to be using it. 00:48:30.402 --> 00:48:32.884 It's done a remarkable job aging gracefully. 00:48:33.305 --> 00:48:35.447 Would I want to use the Emacs 19 today? 00:48:35.467 --> 00:48:36.448 Absolutely not. 00:48:36.808 --> 00:48:39.451 Am I fully comfortable using Emacs 28 in the 00:48:39.511 --> 00:48:40.993 sandbox environment? 00:48:41.013 --> 00:48:43.375 Yeah, there's certain things that I miss. 00:48:43.355 --> 00:48:45.297 But yeah, Emacs 30 is great. 00:48:45.377 --> 00:48:46.979 Emacs 31 is around the corner. 00:48:47.579 --> 00:48:49.261 It's kept up with the times. 00:48:49.622 --> 00:48:50.903 We just talked about AI. 00:48:51.423 --> 00:48:54.006 Emacs and AI, for all my concerns about AI, Emacs 00:48:54.026 --> 00:48:56.068 and AI are a great fit for each other because so 00:48:56.108 --> 00:48:57.810 much of the AI stuff that we're doing is 00:48:57.870 --> 00:48:58.991 text-based. 00:48:59.031 --> 00:49:01.394 Emacs is the best text processing platform that 00:49:01.414 --> 00:49:02.115 there's ever been. 00:49:02.155 --> 00:49:05.058 They fit each other so well. 00:49:05.078 --> 00:49:05.478 [Sacha]: Yeah, yeah. 00:49:05.498 --> 00:49:09.242 And I find it very amusing that with you at 47 00:49:09.282 --> 00:49:12.045 years old and I'm 43, assuming math is right, 00:49:12.346 --> 00:49:15.738 There are people in the Emacs community who would 00:49:15.778 --> 00:49:18.247 consider us still young whippersnappers. 00:49:18.267 --> 00:49:19.572 [Ross]: That's right, yes. 00:49:20.362 --> 00:49:22.464 Different demographics around here, that's for sure. 00:49:23.285 --> 00:49:25.408 [Sacha]: As well as, of course, the other 00:49:25.548 --> 00:49:28.592 end where you have high school students and fresh 00:49:28.632 --> 00:49:31.215 grads still enjoying it, still exploring it. 00:49:32.236 --> 00:49:35.199 And so it's great that we can customize this with 00:49:35.360 --> 00:49:37.903 larger font sizes or different key bindings or 00:49:38.383 --> 00:49:42.468 other ways to catch us from making careless 00:49:42.508 --> 00:49:47.554 mistakes to adapt to us and the things that we're 00:49:47.574 --> 00:49:48.495 working on. 00:49:49.218 --> 00:49:50.219 So all this is very interesting. NOTE Nix is good at managing package versions and customizing them; Matthew Bauer (Bauer IDE) #+SCREENSHOT: emacs-chat-26-ross-a-baker-00-53-01-300.jpg #+SCREENSHOT_TIME: 00:53:01.300 00:49:51.000 --> 00:49:54.726 Oh, actually, on a different tangent, one of the 00:49:54.746 --> 00:49:56.788 things I was curious about is a lot of your 00:49:56.849 --> 00:49:59.773 configuration is set up to work well with Nix. 00:50:00.534 --> 00:50:02.737 Since a lot of people in Emacs are curious 00:50:02.777 --> 00:50:05.501 about reproducible configurations and systems, 00:50:05.521 --> 00:50:07.083 could you tell us more about Nix and your 00:50:07.123 --> 00:50:08.745 experience with it in Emacs? 00:50:09.535 --> 00:50:13.423 [Ross]: Yes, so I'm using that as some 00:50:13.463 --> 00:50:16.630 people use straight.el to pin their packages. 00:50:16.650 --> 00:50:18.434 Some people, they don't bother pinning their 00:50:18.474 --> 00:50:19.616 packages at all. 00:50:19.636 --> 00:50:23.344 I'm using Nix for people who are unfamiliar. 00:50:23.384 --> 00:50:24.827 Nix is... 00:50:25.820 --> 00:50:27.662 Boy, the elevator pitch for Nix. 00:50:29.203 --> 00:50:30.945 It's a package manager, but it's more than an 00:50:31.065 --> 00:50:32.366 Emacs package manager. 00:50:32.526 --> 00:50:35.208 It can manage all of your packages for you. 00:50:35.228 --> 00:50:37.370 So I'm building Emacs with Nix. 00:50:37.450 --> 00:50:39.773 I'm declaring all of my packages that way. 00:50:39.973 --> 00:50:42.955 Then that's specifying... I have a very 00:50:43.015 --> 00:50:46.459 repeatable build of Emacs and it has this version 00:50:46.579 --> 00:50:48.000 of these packages. 00:50:48.020 --> 00:50:50.182 One thing that Nix is very good at is it's very 00:50:50.242 --> 00:50:52.444 good at customizing certain things. 00:50:52.464 --> 00:50:54.486 So if you want to run a patched version of a 00:50:54.546 --> 00:50:55.827 particular package, 00:50:55.807 --> 00:50:59.473 you're able to specify, okay, I want this 00:50:59.533 --> 00:51:02.137 package, I want to use this package definition, 00:51:02.177 --> 00:51:05.302 but I want to overlay these files on top of it. 00:51:05.322 --> 00:51:06.343 Nix excels at that. 00:51:06.424 --> 00:51:08.407 I'm running a lot of customized versions of 00:51:08.447 --> 00:51:10.169 various packages, things that I've tried to 00:51:10.209 --> 00:51:12.613 submit upstream, and the Emacs package might be 00:51:12.653 --> 00:51:15.578 abandoned, or they might be a slow maintainer. 00:51:15.678 --> 00:51:18.022 I need to use it today, and they'll get around to 00:51:18.062 --> 00:51:19.143 it, and they'll merge it in a month. 00:51:19.163 --> 00:51:21.046 So on a temporary basis, I'll want to run a 00:51:21.086 --> 00:51:22.168 patched version. 00:51:22.300 --> 00:51:25.206 Nix makes that workflow very smooth for me where 00:51:25.246 --> 00:51:27.330 I'm able to run patched versions of certain 00:51:27.350 --> 00:51:29.234 packages, like I'm running a patched version of 00:51:29.274 --> 00:51:30.838 Forge like I showed off. 00:51:30.858 --> 00:51:32.641 Some of the things I want to contribute upstream, 00:51:32.682 --> 00:51:33.664 I've got those pulled out. 00:51:33.724 --> 00:51:36.329 I just need to sit down and actually submit them upstream. 00:51:36.970 --> 00:51:39.175 But I've got those ready to go and I'm able to 00:51:39.255 --> 00:51:41.800 run those out of the box. 00:51:41.801 --> 00:51:44.567 It's able to run... It's not just managing Emacs. 00:51:44.568 --> 00:51:46.067 That's the real magic of it. 00:51:46.068 --> 00:51:49.833 If I need certain binaries to support my Emacs config, 00:51:49.834 --> 00:51:52.800 I can declare those, and that all comes together 00:51:52.801 --> 00:51:55.067 and if I pull down Emacs packages, 00:51:55.068 --> 00:51:57.767 those are all going to be there together. 00:51:57.768 --> 00:52:00.883 There's a really fantastic example of this. 00:52:01.243 --> 00:52:04.187 Matthew Bauer has something out there. 00:52:06.361 --> 00:52:08.023 It's Bauer, I-D-E. 00:52:08.364 --> 00:52:10.908 B-A-U-E-R is his last name. 00:52:11.128 --> 00:52:13.451 He has one of the very inspiring configurations 00:52:13.511 --> 00:52:16.055 to me where he builds everything in Nix and he 00:52:16.135 --> 00:52:18.759 bundles all of his binaries that he needs in Nix 00:52:18.839 --> 00:52:22.504 and then he configures his Emacs variables to 00:52:22.545 --> 00:52:24.868 point, not at the git that happens to be on the 00:52:24.928 --> 00:52:27.652 file system, but to the git that he installs via Nix. 00:52:27.692 --> 00:52:31.117 So he knows that the entire thing is 00:52:31.177 --> 00:52:31.958 self-contained. 00:52:32.098 --> 00:52:34.562 That's something I've taken a lot of inspiration from. 00:52:35.015 --> 00:52:39.502 If I'm on a machine that has Nix, I'm able to run this. 00:52:39.522 --> 00:52:42.487 If I want to run my standalone config, I can run 00:52:42.527 --> 00:52:45.051 this command and I can run this command on any 00:52:45.111 --> 00:52:46.873 machine that I have Nix installed on. 00:52:47.595 --> 00:52:49.658 It will run very fast here. 00:52:50.239 --> 00:52:51.741 Should run very fast here. 00:52:52.835 --> 00:52:56.083 and this popped up and this is my base config and 00:52:56.144 --> 00:53:00.836 I could have run this on any machine or the my 00:53:01.277 --> 00:53:02.079 packages are named. 00:53:02.159 --> 00:53:04.967 It's an old Simpsons joke when Homer is stuck in 00:53:05.027 --> 00:53:05.428 New York. 00:53:06.090 --> 00:53:08.937 So that's the khlav kalash and crab juice. 00:53:08.917 --> 00:53:10.581 I think I've got a link to the YouTube video 00:53:10.621 --> 00:53:12.185 somewhere in my config. 00:53:12.205 --> 00:53:14.390 And there's the fully built in one. 00:53:14.410 --> 00:53:16.716 That's something I can get on a new machine. 00:53:17.157 --> 00:53:18.019 I install Nix. 00:53:18.199 --> 00:53:19.843 All of my dot files are there. 00:53:19.863 --> 00:53:22.068 If I want to just run Emacs or if somebody 00:53:22.128 --> 00:53:24.213 else wants to run my Emacs, anybody can take that 00:53:24.253 --> 00:53:25.356 config and run it. NOTE Custom fonts 00:53:25.336 --> 00:53:27.501 Now if you do that, a word of warning. 00:53:27.521 --> 00:53:30.146 One thing that I've done is this font that you 00:53:30.186 --> 00:53:31.709 see is a custom font. 00:53:32.010 --> 00:53:35.918 I took the Iosevka font and I customized a lot of 00:53:35.938 --> 00:53:36.720 the glyphs on it. 00:53:36.820 --> 00:53:38.824 When I do that, I build that in Nix as well. 00:53:38.985 --> 00:53:40.628 You can use Nix to build your fonts. 00:53:40.608 --> 00:53:42.552 That's part of my derivation. 00:53:42.572 --> 00:53:45.558 My Emacs config depends on my fonts. 00:53:45.979 --> 00:53:48.304 If you try to run these, you're going to have 00:53:48.324 --> 00:53:49.406 to compile the fonts. 00:53:49.427 --> 00:53:51.751 I compile a lot of variants of these fonts. 00:53:51.812 --> 00:53:53.635 I compile a lot of alphabets. 00:53:53.655 --> 00:53:55.058 Linguistics is one of my hobbies. 00:53:55.078 --> 00:53:57.183 I love to work in these other alphabets and 00:53:57.243 --> 00:53:57.944 things like that. 00:53:57.964 --> 00:54:00.169 I'm very opinionated on these things. 00:54:00.149 --> 00:54:02.552 Even in languages that I don't speak, I like the 00:54:02.732 --> 00:54:03.813 text to look a certain way. 00:54:03.853 --> 00:54:05.956 Just one of the other ways I blow off steam. 00:54:05.976 --> 00:54:08.459 But anyway, I've got this custom font that I use 00:54:08.519 --> 00:54:08.819 everywhere. 00:54:08.839 --> 00:54:10.121 It's the font on my website. 00:54:10.161 --> 00:54:12.163 It's the font that you see in my editor and my 00:54:12.203 --> 00:54:12.804 terminal. 00:54:13.345 --> 00:54:15.307 If you pull this package, you'll get my font and 00:54:15.327 --> 00:54:17.270 you'll have to compile that and you're going to 00:54:17.310 --> 00:54:19.532 heat the room with your CPU for about 30 minutes 00:54:19.612 --> 00:54:20.353 if you do so. 00:54:20.373 --> 00:54:21.715 So beware on that. 00:54:22.396 --> 00:54:24.799 Otherwise, if you want to run my config, it's all 00:54:24.859 --> 00:54:26.781 out there if you have Nix. 00:54:27.824 --> 00:54:29.609 I don't know how practical that is, but it makes 00:54:29.650 --> 00:54:30.191 me feel good. NOTE Starter kits versus configs 00:54:30.913 --> 00:54:32.037 [Sacha]: Now I'm curious, have you heard 00:54:32.077 --> 00:54:34.464 from people using your base config or your 00:54:34.524 --> 00:54:35.186 extended one? 00:54:38.737 --> 00:54:40.382 [Ross]: I've heard of people 00:54:40.666 --> 00:54:41.567 copying out of it. 00:54:41.607 --> 00:54:43.149 I haven't heard of people depending on it 00:54:43.189 --> 00:54:43.670 directly. 00:54:43.690 --> 00:54:45.112 I would generally discourage people. 00:54:45.232 --> 00:54:47.896 That's kind of the difference between a starter kit and a config. 00:54:47.996 --> 00:54:52.442 You did an excellent series here on starter kits recently. 00:54:52.462 --> 00:54:54.785 I don't see this necessarily as a starter kit. 00:54:54.805 --> 00:54:57.049 This is very opinionated to the way I do things. 00:54:57.109 --> 00:54:58.030 I hope people look at it. 00:54:58.090 --> 00:55:00.393 I hope people draw inspiration from it, copy from it. 00:55:00.513 --> 00:55:01.154 Absolutely. 00:55:01.635 --> 00:55:03.237 But I'm going to change things on a whim. 00:55:03.257 --> 00:55:05.100 If you're using this as a base layer, you're 00:55:05.140 --> 00:55:07.122 going to find my opinions thrust upon you. 00:55:07.163 --> 00:55:08.985 You're probably not going to like half of them. 00:55:10.180 --> 00:55:12.543 [Sacha]: Well, your config is very nicely 00:55:12.603 --> 00:55:15.106 documented and rather enjoyable to read. 00:55:15.206 --> 00:55:19.111 So I can imagine people will get a lot by reading 00:55:19.191 --> 00:55:21.074 it for the things they can copy and paste to 00:55:21.094 --> 00:55:21.975 their config. 00:55:21.995 --> 00:55:23.997 I also particularly like how even your code 00:55:24.017 --> 00:55:27.742 snippets, you've customized it to include the license. 00:55:28.263 --> 00:55:31.627 So it's like every little bit of that is easy for 00:55:31.687 --> 00:55:34.451 people to copy from confidently. NOTE Nix vs Guix; Mac 00:55:34.711 --> 00:55:35.912 Trevok has a question. 00:55:35.992 --> 00:55:38.095 Hopefully without starting a philosophical war, 00:55:38.155 --> 00:55:39.597 why Nix over Guix? 00:55:40.943 --> 00:55:43.647 [Ross]: I would love to run Guix, but I 00:55:43.727 --> 00:55:46.210 have a Mac for work and Guix doesn't have a good 00:55:46.251 --> 00:55:48.434 Mac story. I'm sharing a lot, not just my 00:55:48.574 --> 00:55:50.296 Emacs configuration, but a lot of other 00:55:50.356 --> 00:55:54.042 configurations, because Nix is more broadly scoped. 00:55:54.122 --> 00:55:56.004 I'm able to share a bunch of things between those 00:55:56.044 --> 00:55:57.727 two operating systems. 00:55:57.747 --> 00:56:00.090 So the server that I run, the rossabaker.com, 00:56:00.110 --> 00:56:01.272 that runs on NixOS. 00:56:01.352 --> 00:56:03.635 This laptop that I'm on runs on NixOS, and then I 00:56:03.675 --> 00:56:04.977 have the Darwin machine. 00:56:04.957 --> 00:56:07.121 There are certain things that are compatible 00:56:07.181 --> 00:56:07.962 between Linux. 00:56:07.982 --> 00:56:09.384 There are certain things that are compatible 00:56:09.424 --> 00:56:10.807 between the workstations. 00:56:10.827 --> 00:56:12.810 And then Darwin has its own weird area. 00:56:13.211 --> 00:56:14.753 I'm able to manage all of that with Nix. 00:56:14.813 --> 00:56:16.516 I couldn't do that with Guix. 00:56:16.536 --> 00:56:18.780 If Guix gets a good Darwin story, I will switch 00:56:18.840 --> 00:56:19.902 tomorrow, 00:56:19.922 --> 00:56:21.444 because I would love to be configuring with 00:56:21.464 --> 00:56:21.805 Scheme. 00:56:22.887 --> 00:56:26.593 [Sacha]: Totally different tangent yet again. NOTE Non-work interests: Org for documenting; ox-hugo and multiple languages 00:56:26.894 --> 00:56:30.137 Do you use Emacs for your other interests? 00:56:30.177 --> 00:56:32.579 You've mentioned things like you use it to 00:56:32.659 --> 00:56:36.203 publish your youth soccer homepage. 00:56:36.923 --> 00:56:39.025 What about for your aquarium or linguistics or 00:56:39.105 --> 00:56:40.206 other things you're interested in? 00:56:42.108 --> 00:56:44.650 Tell us about the non-work uses of Emacs in your life. 00:56:46.212 --> 00:56:47.373 [Ross]: That would mostly be just 00:56:47.473 --> 00:56:50.275 documenting things in Org Mode. 00:56:50.916 --> 00:56:52.998 Like I mentioned, it's a big Org doc, and if you 00:56:53.038 --> 00:56:54.699 read this doc, you'll see that it's a very 00:56:54.760 --> 00:56:56.361 sprawling thing. 00:56:56.476 --> 00:57:01.203 For the linguistics, one thing with ox-hugo is 00:57:01.243 --> 00:57:02.425 that integrates nicely. 00:57:02.505 --> 00:57:04.648 I know you've been learning French. 00:57:05.009 --> 00:57:06.812 I think that's cool how you're doing that. 00:57:06.852 --> 00:57:08.574 Spanish and German are my two hobbies. 00:57:08.614 --> 00:57:10.197 I've actually got a trilingual site. 00:57:10.698 --> 00:57:13.262 The English is way more populated than Spanish 00:57:13.302 --> 00:57:13.822 and the German. 00:57:13.842 --> 00:57:15.565 Those are afterthoughts. 00:57:15.764 --> 00:57:18.347 I learned Spanish in college and I learned German 00:57:18.387 --> 00:57:20.389 in high school. I just try to keep current 00:57:20.429 --> 00:57:22.531 with that, but I'm not particularly fluent in 00:57:22.631 --> 00:57:23.692 either. 00:57:23.712 --> 00:57:26.035 There was a wonderful IndieWeb Carnival that 00:57:26.075 --> 00:57:26.455 came up. 00:57:26.515 --> 00:57:28.798 It was called Multilingualism on the Web and 00:57:28.818 --> 00:57:30.880 that's when I sat down and I'm like, okay, I want 00:57:30.940 --> 00:57:32.061 to do this. 00:57:32.081 --> 00:57:34.044 I'm going to write this article and I'm not just 00:57:34.084 --> 00:57:35.265 going to write it in one language. 00:57:35.285 --> 00:57:37.087 I'm going to write it in all three languages that 00:57:37.147 --> 00:57:38.148 I speak. 00:57:38.168 --> 00:57:40.631 Hugo has these facilities for taking the same 00:57:40.731 --> 00:57:43.233 article and publishing them on all three sites 00:57:43.293 --> 00:57:44.775 and linking them across each other. 00:57:45.430 --> 00:57:48.974 And ox-hugo, that passes through. 00:57:48.994 --> 00:57:51.856 ox-hugo uses Hugo as a compilation target. 00:57:51.996 --> 00:57:56.061 ox-hugo, I was able to set the metadata on there. 00:57:56.081 --> 00:57:58.462 If you've noticed my file name in this editor... 00:57:58.463 --> 00:58:04.169 I name everything after Simpsons puns: cromulent.en.org. 00:58:04.189 --> 00:58:09.073 And then I also have... I do have them separated by language. 00:58:10.415 --> 00:58:12.737 Here's my Spanish document and then, 00:58:13.983 --> 00:58:16.507 very predictably, here's my German document. 00:58:17.007 --> 00:58:19.850 By doing this in Org Mode with ox-hugo, 00:58:19.851 --> 00:58:22.035 I'm able to link together the posts that are 00:58:22.075 --> 00:58:24.378 translated into multiple languages. 00:58:24.398 --> 00:58:26.040 And then those all appear, if you go out to the 00:58:26.101 --> 00:58:29.085 Hugo site, like the rossabaker.com, you can go 00:58:29.165 --> 00:58:30.787 there and anything that's available in multiple 00:58:30.827 --> 00:58:32.450 languages, that'll show up. 00:58:32.470 --> 00:58:34.052 You can click through the Spanish or German 00:58:34.092 --> 00:58:35.434 translation of those. 00:58:35.617 --> 00:58:37.019 [Sacha]: I didn't notice any shortcuts in 00:58:37.039 --> 00:58:39.363 your config for quickly switching to another, 00:58:39.503 --> 00:58:42.407 like the Org mode source in a different language 00:58:42.468 --> 00:58:43.309 for the same post. 00:58:43.389 --> 00:58:46.373 Would you happen to have any conveniences like 00:58:46.414 --> 00:58:46.614 that, 00:58:47.095 --> 00:58:49.078 or do you just manually switch to the other file 00:58:49.138 --> 00:58:50.520 and then find it? 00:58:50.540 --> 00:58:52.763 [Ross]: I just manually switch to that, but 00:58:52.843 --> 00:58:53.965 the data is all there. 00:58:54.025 --> 00:58:58.933 The file names have to line up, so that's 00:58:58.973 --> 00:59:00.195 definitely a helper I could write. 00:59:00.515 --> 00:59:02.518 If I can find time to get back into this, that 00:59:02.558 --> 00:59:04.341 would be an excellent. I love your idea. 00:59:04.827 --> 00:59:05.989 [Sacha]: Yeah, because I figured, 00:59:06.029 --> 00:59:07.832 if you're going to write something, you might 00:59:07.872 --> 00:59:09.775 want to switch between languages very quickly. 00:59:09.795 --> 00:59:12.860 If you make an edit, okay, "I'm going to add this thought to it. 00:59:13.140 --> 00:59:14.162 I want to switch to the other ones." 00:59:14.422 --> 00:59:15.944 [Ross]: Yeah, typically I'll just take care 00:59:15.984 --> 00:59:17.927 of that with a split window and I'll have the two 00:59:17.988 --> 00:59:18.849 up there that way. 00:59:19.610 --> 00:59:21.513 That way, I'm comparing them side by side. 00:59:21.573 --> 00:59:23.777 That's probably why I haven't had that yet, 00:59:23.817 --> 00:59:26.200 but it would still be nice to see, okay, "I do 00:59:26.260 --> 00:59:27.583 want to make a quick edit in there." 00:59:27.603 --> 00:59:29.365 It would at least scroll me to that point in the 00:59:29.405 --> 00:59:30.447 other window quicker. 00:59:31.523 --> 00:59:33.466 [Sacha]: If you add more language-related 00:59:33.486 --> 00:59:36.291 stuff to your extended config, I'm very 00:59:36.351 --> 00:59:39.516 interested in what people use to look up 00:59:39.577 --> 00:59:41.440 dictionaries, look up example sentences, and 00:59:41.460 --> 00:59:42.722 things like that. 00:59:42.802 --> 00:59:45.527 @blaiseutube says, "Recovering linguist here. 00:59:45.667 --> 00:59:47.690 English, Spanish, and French for work. 00:59:47.730 --> 00:59:50.595 Japanese, Portuguese, Sanskrit, and Swedish for fun." 00:59:50.575 --> 00:59:54.720 So I'm guessing there are a lot of linguists or 00:59:55.320 --> 00:59:58.363 people who do language learning as a hobby who 00:59:58.384 --> 00:59:59.345 use Emacs. 00:59:59.365 --> 01:00:01.267 It's great because you can switch to all the 01:00:01.427 --> 01:00:03.309 different input methods very easily too. 01:00:03.489 --> 01:00:05.311 It's all text. 01:00:08.214 --> 01:00:09.035 Oh, this is fantastic. 01:00:09.115 --> 01:00:10.617 [Ross]: For all the diacritics that we 01:00:10.657 --> 01:00:13.839 don't have on the US layout keyboard, 01:00:13.840 --> 01:00:17.224 the `C-x 8` key binding. 01:00:17.592 --> 01:00:21.817 Autocomplete if you need an obscure Unicode 01:00:21.857 --> 01:00:24.240 character, and it'll autocomplete on all those. 01:00:25.141 --> 01:00:29.186 So, okay, I want nice little built-in there. 01:00:30.488 --> 01:00:32.090 [Sacha]: Yeah, on Linux, I've just been 01:00:32.170 --> 01:00:34.673 using `setxkbmap` to switch my layout 01:00:34.693 --> 01:00:37.697 temporarily, but I also like just being able to 01:00:37.777 --> 01:00:41.361 set the input method in Emacs in case I just want 01:00:41.401 --> 01:00:43.704 to write something quickly, then I can just 01:00:43.744 --> 01:00:46.268 `C-\`, I think. 01:00:46.548 --> 01:00:51.816 Okay, I've got about one minute 01:00:51.856 --> 01:00:53.599 before the kid does lunch break. 01:00:53.619 --> 01:00:54.761 Thank you so much for doing this. 01:00:55.442 --> 01:00:58.667 Was there anything that you wanted to pass on as 01:00:58.728 --> 01:01:03.195 a key tip that you'd like people to know from 01:01:03.235 --> 01:01:07.362 your experience with using Emacs? 01:01:07.763 --> 01:01:11.309 What's something we haven't mentioned that you'd 01:01:11.329 --> 01:01:11.950 like to share? 01:01:13.246 --> 01:01:17.159 [Ross]: I guess I didn't mention the things that I think 01:01:17.279 --> 01:01:18.483 are exciting outside. 01:01:18.503 --> 01:01:20.289 We talked about the built-in configs, so the 01:01:20.349 --> 01:01:24.522 things that I missed the most, I would say... 01:01:25.480 --> 01:01:27.602 Well, I guess look at my crab juice config. 01:01:27.743 --> 01:01:29.685 Those are the packages that I found that I 01:01:29.765 --> 01:01:31.307 absolutely cannot live without. 01:01:31.867 --> 01:01:33.589 Magit, a few language modes. 01:01:34.651 --> 01:01:36.673 You can get by on the built-in thing a lot better 01:01:36.713 --> 01:01:38.315 than you think, but there are a few things that 01:01:38.335 --> 01:01:40.417 are out there that are the really special ones, 01:01:40.437 --> 01:01:42.339 and I think that's the distilled essence of that. 01:01:42.580 --> 01:01:44.482 Just take a look at that package. 01:01:44.502 --> 01:01:46.504 Those are the ones that, yeah, I tried really 01:01:46.544 --> 01:01:48.427 hard on the built-ins and I failed. 01:01:48.467 --> 01:01:50.269 These are the ones I really need. 01:01:50.289 --> 01:01:51.951 I'd encourage people to take a look at that as well. 01:01:52.772 --> 01:01:53.192 [Sacha]: All right. 01:01:53.493 --> 01:01:56.745 For folks who are watching, you can find all 01:01:56.765 --> 01:01:58.893 of that stuff at rossabaker.com. 01:01:59.615 --> 01:02:00.378 Thank you so much. 01:02:00.659 --> 01:02:04.212 You've got links to your GitHub and Berg. 01:02:04.883 --> 01:02:10.188 Thanks to everyone on the stream for hanging out 01:02:10.268 --> 01:02:11.890 and for sharing your questions and comments. 01:02:12.471 --> 01:02:17.716 I almost forgot to mention [@ispringle] had a tip 01:02:17.776 --> 01:02:23.242 about putting emacs-lock-mode on a keyboard 01:02:23.262 --> 01:02:27.606 shortcut so you can toggle it if you wanted to. 01:02:28.307 --> 01:02:29.147 Nice idea. 01:02:29.428 --> 01:02:30.329 Again, thank you everyone. 01:02:30.349 --> 01:02:31.650 We're going to end the stream here. 01:02:31.630 --> 01:02:33.494 I will work on the transcripts and all that stuff. 01:02:33.574 --> 01:02:36.500 Thanks again Ross for doing this, and I'll see 01:02:36.540 --> 01:02:38.905 everyone around probably in September or October 01:02:38.945 --> 01:02:44.998 because I have to be on summer vacation.