رونوشت 🔗
1
00:00:00,005 --> 00:00:05,025
So I was woken up this morning at 5 a.m. by birds screaming over what was inside.
2
00:00:05,025 --> 00:00:07,505
A barbecue, which, by the way, once I got in there, they weren't going to be
3
00:00:07,505 --> 00:00:08,805
all that – they weren't going to be that happy.
4
00:00:08,845 --> 00:00:12,785
But we just moved a bunch of gear into this space, and the birds came down and
5
00:00:12,785 --> 00:00:13,925
were fighting over it this morning.
6
00:00:15,005 --> 00:00:18,865
And I could go out there and lift one of them and put them under my arm and
7
00:00:18,865 --> 00:00:22,285
carry them around like a small dog. They're just these huge creatures.
8
00:00:23,725 --> 00:00:28,025
And, man, oh, man, I want one as a pet so bad. Train it. Have it go collect things for me.
9
00:00:28,145 --> 00:00:30,865
Levi's going to get a little jealous, though. You know, watch out.
10
00:00:43,125 --> 00:00:47,865
Hello, friends, and welcome back to your weekly Linux talk show. My name is Chris.
11
00:00:48,005 --> 00:00:48,645
My name is Wes.
12
00:00:48,885 --> 00:00:49,725
And my name is Brent.
13
00:00:50,225 --> 00:00:55,285
Hello, gentlemen. Well, coming up on the show, I tried to build a Linux desktop
14
00:00:55,285 --> 00:00:58,225
using nothing but AI text editors.
15
00:00:58,865 --> 00:01:02,345
I never touched a single config file. I just vibed it all the way.
16
00:01:02,585 --> 00:01:06,185
But did it work or did the bots break it? I'll tell you about that.
17
00:01:06,385 --> 00:01:09,905
And then Wes is sharing his tales from DEF CON and Nix Vegas.
18
00:01:10,165 --> 00:01:12,245
Then we're going to round the show out with some great emails,
19
00:01:12,445 --> 00:01:15,165
some boosts, some tremendous picks, and a lot more.
20
00:01:15,285 --> 00:01:18,505
So before I get any further, let's say time-appropriate greetings to that virtual
21
00:01:18,505 --> 00:01:19,565
lug. Hello, Mumble room.
22
00:01:20,145 --> 00:01:22,145
Hey, Chris. Hey, Wes. And hello, Brent.
23
00:01:22,885 --> 00:01:27,985
Hello, hello, hello. Hello. We have a small team in Mumble today.
24
00:01:28,225 --> 00:01:32,505
I think it's maybe this is the live stream summer doldrums, I suppose.
25
00:01:32,765 --> 00:01:35,485
We're grateful for everybody that can make it. We're also grateful for the chat room.
26
00:01:36,145 --> 00:01:39,125
And we're grateful for the friends over at Define Networking.
27
00:01:39,225 --> 00:01:43,005
Go check out Manage Nebula, define.net slash unplugged.
28
00:01:43,265 --> 00:01:49,425
It's a decentralized VPN built on an open source platform that we love. It's called Nebula.
29
00:01:49,625 --> 00:01:53,365
And the entire thing, top to bottom, is open source.
30
00:01:53,525 --> 00:01:58,125
So you can self-host all of it, or you can have the blissful piece of Manage Nebula.
31
00:01:58,225 --> 00:02:02,945
And it is optimized for speed, simplicity, and they're using industry-leading security.
32
00:02:03,385 --> 00:02:06,525
Nebula has a decentralized design that keeps your network resilient.
33
00:02:06,685 --> 00:02:10,865
If you want to manage it for your home lab or for an enterprise of tens of thousands
34
00:02:10,865 --> 00:02:13,325
of systems, Nebula is already doing it.
35
00:02:13,605 --> 00:02:17,885
In fact, Nebula was doing some heavy lifting when you're down there at DEF CON
36
00:02:17,885 --> 00:02:19,185
slash Nix Vegas, Mr. Payne.
37
00:02:19,405 --> 00:02:23,085
Yeah, that's right. I mean, what happens when you're trying to host,
38
00:02:23,365 --> 00:02:27,645
you know, you need to go get Nix packages even when you're doing a proxy or a build cache.
39
00:02:27,645 --> 00:02:31,105
And you're trying to fight, I mean, not only just conference Wi-Fi in general,
40
00:02:31,325 --> 00:02:36,125
but a whole bunch of hackers on said Wi-Fi, well, you bond as much as you can,
41
00:02:36,245 --> 00:02:39,865
and then you use Nebula to tunnel all of that,
42
00:02:41,073 --> 00:02:46,453
dedicated VPS upstream to just protect things and to make sure you have the
43
00:02:46,453 --> 00:02:50,553
best possible routing out of there so you don't get stuck in a big Vegas mess.
44
00:02:52,033 --> 00:02:55,133
That's really cool. And then they could just stand it up like that too.
45
00:02:55,573 --> 00:03:00,213
Nothing else offers Nebula's resilience, its speed, or scalability.
46
00:03:00,313 --> 00:03:04,213
And that stuff, it comes in just the network traffic usage. It comes in the
47
00:03:04,213 --> 00:03:08,093
uptime and the battery usage for mobile devices. It's really fantastic.
48
00:03:08,113 --> 00:03:11,953
It's why we're switching our infrastructure over to it and also the team behind
49
00:03:11,953 --> 00:03:15,973
it are friends of the show so go get started with up to 100 hosts absolutely
50
00:03:15,973 --> 00:03:22,233
free no credit card required just go to defined.net slash unplugged support
51
00:03:22,233 --> 00:03:26,333
the show and visit them at defined.net slash unplugged,
52
00:03:28,533 --> 00:03:34,353
all right so just a few housekeeping items before we dig into this uh the boys are remote this week.
53
00:03:35,093 --> 00:03:42,113
Brent's off in Eastern Canada and taking over a parent's cottage and has a little spot over there.
54
00:03:42,253 --> 00:03:44,993
Are the kitties with you right now, Brent, or did you leave them somewhere else
55
00:03:44,993 --> 00:03:45,573
while you're recording?
56
00:03:45,813 --> 00:03:49,273
I left them somewhere after last week when they had to just sit around while
57
00:03:49,273 --> 00:03:52,273
I was doing the show. They asked to not be a part of it this week for some reason.
58
00:03:52,493 --> 00:03:55,913
I understand. And Wes, you're in your hotel in Vegas right now.
59
00:03:56,033 --> 00:03:58,173
Can you, do you have any kind of view? Do you see anything?
60
00:03:59,093 --> 00:04:03,233
Well, you know, I do, I do have, I can see the strip. I can
Load full transcript
see the monorail,
61
00:04:03,413 --> 00:04:06,893
except I turned off the AC to do this here show.
62
00:04:07,213 --> 00:04:13,153
So I also closed the blinds to, you know, fight, fight that horrible sky death globe.
63
00:04:13,573 --> 00:04:18,293
Yeah. So I'm just chilling here. No, I'm not going to complain at all then about the weather.
64
00:04:18,333 --> 00:04:20,753
It's a little warm today, but I'm not going to complain at all.
65
00:04:20,833 --> 00:04:21,593
It's nothing like Vegas.
66
00:04:22,273 --> 00:04:25,953
I also wanted to give a shout out to a project that we love here at the show.
67
00:04:26,233 --> 00:04:28,593
It's the Podverse Podcast Catcher.
68
00:04:28,973 --> 00:04:34,453
It's a GPL open source podcasting 2.0 player. You see we also integrated into our website.
69
00:04:35,733 --> 00:04:39,793
And they're undergoing a big rework. And it's going really well.
70
00:04:39,893 --> 00:04:43,433
And they're looking for some help with a web app, perhaps their mobile app as well.
71
00:04:43,733 --> 00:04:48,353
And they let me know that now is the time to engage. Mitch wrote,
72
00:04:48,433 --> 00:04:50,953
I believe the new infrastructure is powerful and will hopefully be the best
73
00:04:50,953 --> 00:04:53,153
podcasting 2.0 compatible software available.
74
00:04:53,233 --> 00:04:57,453
And the backend components are mostly finished and the front-end website is in progress.
75
00:04:57,653 --> 00:05:00,373
So if you know anybody that could help us, please send them our way.
76
00:05:00,793 --> 00:05:05,133
So go find Podverse on GitHub. If you would like to help out what I think is
77
00:05:05,133 --> 00:05:09,173
an absolutely critical project, we need more open-source podcasting apps.
78
00:05:09,173 --> 00:05:12,633
It's not much of a thing. It really should be.
79
00:05:14,166 --> 00:05:18,526
Also, call out for a sponsor that wants to work with us to get us to Texas Linux Fest.
80
00:05:18,746 --> 00:05:23,706
The Linux Unplugged show is looking for someone to help us attend and cover
81
00:05:23,706 --> 00:05:27,346
Texas Linux Festival, which is October 3rd through the 1st.
82
00:05:27,726 --> 00:05:31,146
If you would like to help us get there and do some co-coverage with us or something
83
00:05:31,146 --> 00:05:34,626
like that, email me, chris at jupiterbroadcasting.com.
84
00:05:34,966 --> 00:05:37,666
And if you want to get there and do a talk or something like that,
85
00:05:37,746 --> 00:05:42,566
sometimes that helps the old boss, grease the wheels, get a talk in.
86
00:05:42,566 --> 00:05:44,726
Their call for papers has been extended.
87
00:05:45,026 --> 00:05:48,446
So you do have a little bit of time still. And it's October,
88
00:05:48,906 --> 00:05:50,486
Austin, Texas Linux Festival.
89
00:05:50,826 --> 00:05:57,906
It's one of the high signal, smaller events in terms of like scale or Linux Fest Northwest.
90
00:05:58,546 --> 00:06:02,486
It's still a good amount. And the signal to noise ratio
is really good.
91
00:06:02,986 --> 00:06:05,146
We've always really enjoyed Texas Linux Festival.
92
00:06:05,326 --> 00:06:09,326
Oh yeah, we had a great time last time we went. I'm getting excited already.
93
00:06:09,326 --> 00:06:15,506
And, yeah, you know, I think there's a certain smoky flavor to those Texas-style
94
00:06:15,506 --> 00:06:17,526
Linux hackers that you just don't get anywhere else.
95
00:06:21,346 --> 00:06:28,746
Okay, Wes, so you are there in Vegas right now, and I'm just curious about how
96
00:06:28,746 --> 00:06:32,026
the whole thing went, really. I know you had a talk, I think it was on Saturday,
97
00:06:32,026 --> 00:06:36,326
and I think you had a chance to meet some folks. Tell us, give us a little flavor.
98
00:06:36,846 --> 00:06:41,726
Oh yeah okay so like first time at defcon first time in vegas actually somehow
99
00:06:41,726 --> 00:06:43,906
i've yet to yet to be here um you.
100
00:06:43,906 --> 00:06:44,906
Are kidding me.
101
00:06:44,906 --> 00:06:45,566
No wow.
102
00:06:45,566 --> 00:06:46,746
Did not know that.
103
00:06:46,746 --> 00:06:49,986
Um but you know overall i gotta say very impressed excellent
104
00:06:49,986 --> 00:06:52,966
conference and just like super super friendly
105
00:06:52,966 --> 00:06:57,486
folks i mean i ended up having a neighbor at the hotel in the room with me you
106
00:06:57,486 --> 00:07:00,346
know we were able to just like walk over together we became friends for the
107
00:07:00,346 --> 00:07:05,686
the whole conference and of let me uh tell him way too much info about nix so
108
00:07:05,686 --> 00:07:10,226
first of all just like you know i'm no i'm no pen tester i'm not a reverse engineer
109
00:07:10,646 --> 00:07:14,366
so i wasn't sure i was coming for the nix subconference we'll get more into
110
00:07:14,366 --> 00:07:17,386
that but just overall i would say like if you if you are interested if you have
111
00:07:17,386 --> 00:07:21,566
a budget to come to this kind of thing um much like the vibe at like a at a
112
00:07:21,566 --> 00:07:23,786
linux fest but at a whole different scale like,
113
00:07:24,466 --> 00:07:27,906
it's just super friendly and welcoming and so many people are here either their
114
00:07:27,906 --> 00:07:32,246
first time or their 17th time but uh it doesn't really matter people are looking
115
00:07:32,246 --> 00:07:35,266
to be friendly and welcoming and make new friends.
116
00:07:36,503 --> 00:07:39,803
I was really here, of course, for Nix Vegas, which is, I guess,
117
00:07:39,863 --> 00:07:41,323
the first time they're doing this.
118
00:07:41,463 --> 00:07:45,643
And it's a community sort of sub-conference inside the larger DEF CON.
119
00:07:46,143 --> 00:07:48,443
It's a common thing, right? They've got whole different villages,
120
00:07:48,623 --> 00:07:51,823
they've got sub-treks, and they've got these communities.
121
00:07:52,243 --> 00:07:55,683
And man, I just, like, huge shout-out to the folks behind all this.
122
00:07:56,423 --> 00:08:01,803
Many, many of them, but in particular, Morgan Jones, our buddy Tristan Ross,
123
00:08:01,903 --> 00:08:05,883
who was on the show and was a release manager recently for the most recent NixOS release.
124
00:08:05,883 --> 00:08:08,783
Also Tom Barrick from Phlox and a
125
00:08:08,783 --> 00:08:12,323
shout out to Dan Baker who also really helped organize this
126
00:08:12,323 --> 00:08:15,083
whole thing who sadly got the flu and was not
127
00:08:15,083 --> 00:08:17,983
able to make it to the conference but still so much
128
00:08:17,983 --> 00:08:20,723
hard work you know going there and of course Phlox and
129
00:08:20,723 --> 00:08:24,683
Determinant Systems were sponsoring to help make that possible and because
130
00:08:24,683 --> 00:08:27,563
these Nix folks are just crazy they also
131
00:08:27,563 --> 00:08:30,803
got some hardware support from System76 and Ampere
132
00:08:30,803 --> 00:08:34,263
and Protectly and Next Computing because the amount
133
00:08:34,263 --> 00:08:37,563
of infrastructure they stood up they weren't just like gonna be at
134
00:08:37,563 --> 00:08:40,463
defcon they stood up like whole domains they had
135
00:08:40,463 --> 00:08:43,383
this this impressive nebula infrastructure they
136
00:08:43,383 --> 00:08:46,723
had powerline networking going they ran a pixie server
137
00:08:46,723 --> 00:08:51,343
they basically rebuilt all of nix packages and pulled down all of nix packages
138
00:08:51,343 --> 00:08:55,083
to build a binary cache running on their network that you could get to over
139
00:08:55,083 --> 00:08:59,483
the powerline or the wi-fi that they were broadcasting and all of that is powered
140
00:08:59,483 --> 00:09:02,863
by nix and reproducible And they're pushing as much as they can upstream to
141
00:09:02,863 --> 00:09:04,923
a GitHub repo called NixVegas,
142
00:09:05,103 --> 00:09:08,423
or the GitHub org called NixVegas that we'll definitely have a link to.
143
00:09:09,083 --> 00:09:11,843
So they're doing Nebula over Powerline networking?
144
00:09:13,103 --> 00:09:18,103
Well, I don't know that Powerline was in the path out, but Powerline was a feature that was attacked.
145
00:09:18,183 --> 00:09:18,563
Love it.
146
00:09:18,763 --> 00:09:21,723
Yeah, so they could go elsewhere in the convention center, plug in,
147
00:09:21,783 --> 00:09:24,163
and then Pixie boot, because they were also running a Pixie server, of course.
148
00:09:24,942 --> 00:09:29,662
Oh, my gosh, that's so cool. So I'm guessing pretty good vibes.
149
00:09:30,042 --> 00:09:33,622
When we go to these kind of events, especially these more sub-conferences,
150
00:09:33,622 --> 00:09:37,482
it's a pretty tight crew. Like you say, everybody wants to chat and learn.
151
00:09:38,162 --> 00:09:42,582
Yeah, absolutely. And also just shout out to them because I got a little sneak
152
00:09:42,582 --> 00:09:46,742
peek behind the scenes at the first night as they were getting everything set up.
153
00:09:46,742 --> 00:09:53,002
And I can tell you that, like, they did not have a projector or a screen until,
154
00:09:53,202 --> 00:09:57,602
like, 1030 or 11 the night before things were supposed to get going.
155
00:09:57,802 --> 00:09:59,442
So, they were, like, improvising a backup.
156
00:10:00,062 --> 00:10:04,722
Shout out to Andrew and Anna, who came up with, like, a sheet-based backup and
157
00:10:04,722 --> 00:10:08,042
went and got a backup projector to try to make that work.
158
00:10:08,042 --> 00:10:12,062
People like you know chris and rj there's just this whole crew of folks behind
159
00:10:12,062 --> 00:10:18,662
the scenes making whatever needed to happen happen up to and including at the last minute and then,
160
00:10:19,442 --> 00:10:23,582
morgan sitting there running everything not only sort of like looking at bgp
161
00:10:23,582 --> 00:10:26,682
looking glass to make sure that the peering they have on the vps that's doing
162
00:10:26,682 --> 00:10:30,702
all the nebula streaming has is like optimized but also running an own cast
163
00:10:30,702 --> 00:10:34,122
server and obs and like doing all the stuff we do all at the same Yeah.
164
00:10:34,602 --> 00:10:37,222
Right. Because as we know, although at a smaller scale, it's there,
165
00:10:37,262 --> 00:10:39,302
you have the, you have the networking you have to solve.
166
00:10:39,482 --> 00:10:42,442
You have the production technology you have to solve them. The audio,
167
00:10:42,582 --> 00:10:47,162
I mean, there's so much, and I can so, so picture one of us doing like a backup
168
00:10:47,162 --> 00:10:50,582
run for another solution while we're trying to fix the primary solution. Like,
169
00:10:51,257 --> 00:10:53,377
You can really appreciate the hustle.
170
00:10:53,757 --> 00:10:57,757
Exactly. It definitely had our kind of last minute, just figure it out energy
171
00:10:57,757 --> 00:10:59,777
going on. So I naturally like that.
172
00:11:00,057 --> 00:11:04,097
Also very cool is Morgan and Tristan worked on an incredible badge.
173
00:11:04,257 --> 00:11:08,397
You know, like cool, hacky, embedded, styled badges are a big thing at DEFCON.
174
00:11:08,577 --> 00:11:12,897
And the Knicks team showed up for Knicks Vegas and DEFCON.
175
00:11:13,157 --> 00:11:17,157
Not only were they able to get the official SVG of the Knicks and Knicks OS logo,
176
00:11:17,157 --> 00:11:20,277
though um but like you know they got this sent over
177
00:11:20,277 --> 00:11:23,377
to a lab in um a fab in china
178
00:11:23,377 --> 00:11:26,397
you know they got it all ordered uh it's powered
179
00:11:26,397 --> 00:11:29,097
and built by nix i mean there's like key morgan gave a
180
00:11:29,097 --> 00:11:31,757
great talk uh but like keycads in there they had
181
00:11:31,757 --> 00:11:35,637
to use something called sv svg to shenzhen all
182
00:11:35,637 --> 00:11:38,377
kinds of like hacky scripts and adapters to make this thing as
183
00:11:38,377 -->
00:11:41,097
reproducible as possible so you can actually just go up
184
00:11:41,097 --> 00:11:44,177
to github and like go get this awesome badge it's
185
00:11:44,177 --> 00:11:47,037
running um some expressive like risk 5
186
00:11:47,037 --> 00:11:50,257
chip on there the badges also have
187
00:11:50,257 --> 00:11:54,337
wi-fi they join a wi-fi mesh and sadly
188
00:11:54,337 --> 00:11:57,997
there was like a little bug in some of the routing so the sd card doesn't work
189
00:11:57,997 --> 00:12:03,657
but it was going to be a full nix cache but it can also just be a nix cache
190
00:12:03,657 --> 00:12:07,897
proxy so you can you can be cloning down nix packages and you can point to your
191
00:12:07,897 --> 00:12:12,677
badge as a substituter and have it go over the conference like their,
192
00:12:13,277 --> 00:12:17,517
sub-conference Wi-Fi to get your Nix packages, which is just incredible.
193
00:12:18,957 --> 00:12:22,357
Holy. You had me at reproducible badge, but that's next level.
194
00:12:23,617 --> 00:12:26,577
Yeah. And there's a bunch of Easter eggs on there. I haven't even gotten,
195
00:12:26,857 --> 00:12:29,637
I'm excited to go, I'll bring it up and so you guys can check it out.
196
00:12:29,717 --> 00:12:32,017
We'll have to plug it in and see what we can do with it.
197
00:12:32,137 --> 00:12:33,937
I imagine this thing draws attention.
198
00:12:34,357 --> 00:12:37,757
Yes. So that was what worked out. I was super glad. Next time,
199
00:12:38,525 --> 00:12:43,885
A, I need more JB swag I can be given out, but then also Nick stuff,
200
00:12:43,885 --> 00:12:48,685
because once I had that badge, suddenly I found myself being an impromptu Nick's
201
00:12:48,685 --> 00:12:51,725
ambassador, which it turns out is a role I'm very happy to play.
202
00:12:51,865 --> 00:12:55,205
So whether it was just hanging out while talks were going on at the edge of
203
00:12:55,205 --> 00:12:58,725
the subconf area and talking to people wandering by, asking about it,
204
00:12:58,785 --> 00:13:02,085
or as I'm wandering around at night, at parties, at meetups,
205
00:13:02,185 --> 00:13:05,545
at events, folks are all like, oh, cool badge, so that's one line.
206
00:13:05,545 --> 00:13:08,905
But a lot of folks would just be like, oh, NixOS, or I've been curious about Nix.
207
00:13:09,025 --> 00:13:12,705
Oh, I've really been meaning to learn Nix, especially met a ton of Arch users
208
00:13:12,705 --> 00:13:16,945
in that camp, where it's like, they just needed a little bit more of a push,
209
00:13:17,085 --> 00:13:19,105
a little bit more of like, is it okay?
210
00:13:19,265 --> 00:13:21,445
Is it going to be horrible to try to onboard?
211
00:13:21,805 --> 00:13:24,705
So the badge was a huge win for me, because I just got to have a lot of great
212
00:13:24,705 --> 00:13:27,505
conversations, chatting with random people, trying to spread the good word.
213
00:13:27,765 --> 00:13:31,605
So is this how you ended up getting pulled into a 2 a.m. rescue session?
214
00:13:31,965 --> 00:13:37,265
Yes, absolutely. So I was, you know, trying to pace myself, trying not to stay
215
00:13:37,265 --> 00:13:41,225
up too late, but winding down, working on heading back from the convention center
216
00:13:41,225 --> 00:13:43,345
to the hotel I was staying at.
217
00:13:43,545 --> 00:13:47,865
And I ran to the gentleman I met that first day, Paul, and he was with some
218
00:13:47,865 --> 00:13:50,245
of his friends and they were working on a CTF.
219
00:13:50,605 --> 00:13:54,465
And when I walked over, they'd said something, what really made me stick around
220
00:13:54,465 --> 00:13:59,385
rather than just saying hi was the combination of WebAssembly and DOS.
221
00:13:59,525 --> 00:14:04,045
I was like, what are you doing with WebAssembly and DOS? And it turns out that
222
00:14:04,045 --> 00:14:05,505
this CTF they were doing,
223
00:14:05,785 --> 00:14:11,265
what they gave you to start was just a Z5 file, which is a game data file used
224
00:14:11,265 --> 00:14:14,685
by the Z machine, which is a virtual machine for running interactive fiction
225
00:14:14,685 --> 00:14:17,725
games, like from the DOS era.
226
00:14:18,405 --> 00:14:18,845
Amazing.
227
00:14:19,165 --> 00:14:22,225
So you could just play the game, right? So that was one way to try to go through
228
00:14:22,225 --> 00:14:23,305
the capture the flag event.
229
00:14:23,725 --> 00:14:27,085
But, and of course, probably I'm sure the folks behind this knew,
230
00:14:27,365 --> 00:14:32,545
there's a debugger for these game files called Z tools. And so,
231
00:14:32,865 --> 00:14:35,785
unfortunately, these folks had only brought a Mac with them.
232
00:14:35,885 --> 00:14:38,145
I think their other laptops were back at the hotel at this point because it
233
00:14:38,145 --> 00:14:40,665
was, you know, 2 a.m. And they'd been partying.
234
00:14:41,645 --> 00:14:44,965
So they had a Mac. This thing would not run on a Mac. It would run on Windows.
235
00:14:45,205 --> 00:14:47,645
You know, you could probably get it running. You could compile it for Linux.
236
00:14:48,005 --> 00:14:51,985
So they were going down the path of trying to get WebAssembly and DOS so that
237
00:14:51,985 --> 00:14:54,965
they could upload this debugger and the game file to that, which,
238
00:14:55,025 --> 00:14:56,725
of course, was just going terribly.
239
00:14:57,045 --> 00:15:00,505
So I started, like, I don't know, just trying to help. I was like,
240
00:15:00,565 --> 00:15:03,705
okay, what about FreeDOS? Maybe like we spin up, you know, ultimately UTM,
241
00:15:03,925 --> 00:15:06,705
which is QEMU, trying to run free DOS.
242
00:15:06,865 --> 00:15:11,005
And then, of course, then we had to figure out how to actually get the files in there.
243
00:15:11,165 --> 00:15:16,525
And are you using UTM x86 emulation for this, which is slow as hell on a Mac?
244
00:15:16,785 --> 00:15:20,205
Yeah, I think so. I don't know which
era of Mac it was, but yes,
245
00:15:20,365 --> 00:15:21,485
presumably that must be it.
246
00:15:21,665 --> 00:15:22,665
Oh, my God. All right.
247
00:15:22,865 --> 00:15:25,745
Yeah. So they were finally able to get free DOS to download.
248
00:15:25,825 --> 00:15:28,725
You know, you still have to somehow they have a free DOS live CD,
249
00:15:28,865 --> 00:15:32,845
which is almost a thing that doesn't quite compute for me. But it is a thing. It does work.
250
00:15:33,005 --> 00:15:36,805
And then it turns out there is a QEMU flag that you can just pass through like a fat...
251
00:15:38,839 --> 00:15:41,679
Format it'll treat it like a fat device to attach but it has
252
00:15:41,679 --> 00:15:44,519
some weird limitations around like being read only maybe and like
253
00:15:44,519 --> 00:15:48,099
only 512 bytes and how do you then you have to make sure you can pass extra
254
00:15:48,099 --> 00:15:51,959
arbitrary flags through utm so while that was like we were debating that i was
255
00:15:51,959 --> 00:15:55,559
like okay well let me pull out my laptop uh i know i could at least like make
256
00:15:55,559 --> 00:16:00,919
a like a loopback raw device that we could format like fat 16 and mbr and like
257
00:16:00,919 --> 00:16:03,779
try to stuff the files and then maybe mount that as a second drive.
258
00:16:04,519 --> 00:16:07,199
But that's when I finally, because I was kind of getting, you know,
259
00:16:07,319 --> 00:16:10,259
trying to figure out the whole path they had gone down to get where they were
260
00:16:10,259 --> 00:16:13,419
in terms of like actually what the problem up the tree was they were trying to solve.
261
00:16:14,039 --> 00:16:17,139
And then I realized that the whole problem was really just trying to get those,
262
00:16:17,999 --> 00:16:19,139
ZTools debugger working.
263
00:16:20,779 --> 00:16:26,059
And yeah, that was just packaged in Nix packages. So I could just literally clone that down.
264
00:16:26,239 --> 00:16:29,259
It took, you know, two seconds. And then I was able to immediately dump all
265
00:16:29,259 --> 00:16:33,299
of the game data and share that with the team. and then they could start like
266
00:16:33,299 --> 00:16:34,959
grepping through it, trying to find clues.
267
00:16:35,239 --> 00:16:37,859
Of course, if you just grep for the word flag, there's like a,
268
00:16:38,039 --> 00:16:40,679
there's an F you in there from the people behind it. Like, of course we wouldn't
269
00:16:40,679 --> 00:16:42,579
just put this in here like that. What are you doing?
270
00:16:43,079 --> 00:16:46,679
So I don't know how far they eventually got, but I was able to like give them
271
00:16:46,679 --> 00:16:51,159
those text file dumps and at least felt like I was participating and helping
272
00:16:51,159 --> 00:16:53,299
out. And hey, the power of Nix.
273
00:16:53,679 --> 00:16:56,959
I have one question though. Were you out and about at 2 a.m.
274
00:16:57,219 --> 00:17:00,739
With your laptop or did you discover them, come across them,
275
00:17:00,739 --> 00:17:05,679
realize their plight go back to your hotel room get your laptop come back and do this.
276
00:17:06,499 --> 00:17:09,259
No you know so that day not only did I not manage to
277
00:17:09,259 --> 00:17:12,819
actually have dinner but I did not manage to escape the conference at all so
278
00:17:12,819 --> 00:17:15,759
I had my laptop from being there during the day and I just never had a chance
279
00:17:15,759 --> 00:17:19,779
to go back to the hotel so I was bopping around having having a couple of drinks
280
00:17:19,779 --> 00:17:26,739
dancing all with my laptop so oh my gosh that's amazing think that it is uh it is a trooper.
281
00:17:27,645 --> 00:17:31,985
Yeah, really? Wow. Well, also, before the show started, you kind of teased us
282
00:17:31,985 --> 00:17:36,265
that you may have heard about a new feature coming to Nix that you are particularly nerded out about.
283
00:17:36,545 --> 00:17:41,165
Yeah, right. I'm a functional programming guy. I like Clojure and other functional
284
00:17:41,165 --> 00:17:43,065
programming. And Nix is a functional programming language.
285
00:17:43,365 --> 00:17:47,205
And one of the things with functional programming languages is you're doing function calls.
286
00:17:47,345 --> 00:17:50,625
And often you might need to chain some data through a couple of functions,
287
00:17:50,745 --> 00:17:52,325
right? You call one thing to get your initial data.
288
00:17:52,405 --> 00:17:55,585
You call another function on that to maybe modify it a little bit.
289
00:17:55,585 --> 00:17:58,705
And then you finally pass it off to like the final thing that needs that data.
290
00:17:59,125 --> 00:18:03,805
And with Nix, right now, and in a lot of functional programming languages without
291
00:18:03,805 --> 00:18:07,185
these tricks, you kind of have to read it backwards, right?
292
00:18:07,245 --> 00:18:11,105
So it's like it's nested with the last thing, you read it first,
293
00:18:11,145 --> 00:18:13,785
and the first thing is all the way inside the nesting.
294
00:18:14,365 --> 00:18:17,485
That can be confusing. If you do a lot of functional programming language,
295
00:18:17,665 --> 00:18:19,925
you kind of get used to it. But no one really likes it.
296
00:18:20,165 --> 00:18:23,385
Often you then have to kind of leave more comments in the code to try to explain
297
00:18:23,385 --> 00:18:27,165
exactly what's going on with the flow. So I had no idea that this was already
298
00:18:27,165 --> 00:18:32,085
a shipping experimental feature in Nix, and there's an open RFC going for it.
299
00:18:32,765 --> 00:18:36,705
They're introducing a pipe operator, which just lets you do exactly that,
300
00:18:36,765 --> 00:18:37,805
except it flips the order.
301
00:18:38,025 --> 00:18:42,745
So you run, you produce your first data, and then you can just pipe it just
302
00:18:42,745 --> 00:18:47,705
like you're used to in a good old
Unix shell to build a pipeline that looks
303
00:18:47,705 --> 00:18:50,945
in the proper readable order right out of the gate.
304
00:18:50,945 --> 00:18:54,165
So it's like a small change on the language, but because this pattern is so
305
00:18:54,165 --> 00:18:57,645
prevalent basically everywhere, it could be one that would be a huge win for
306
00:18:57,645 --> 00:19:01,525
readability and hopefully something that makes it easier for folks not only
307
00:19:01,525 --> 00:19:03,085
coming from other functional programming languages,
308
00:19:03,305 --> 00:19:08,565
but actually the person who's the steward of the RFC was not a functional programmer before Nix at all.
309
00:19:08,565 --> 00:19:11,985
And so they were actually doing this just because it made it so much easier for them.
310
00:19:12,105 --> 00:19:16,425
So I had no idea. Some of the other folks I was chatting with who were definitely
311
00:19:16,425 --> 00:19:18,425
Nix nerds in the audience also had no idea.
312
00:19:18,625 --> 00:19:22,345
So yeah, they were also asking for explicitly support.
313
00:19:22,525 --> 00:19:26,625
Well, I'll link to the RFC in the show notes. And go chime in.
314
00:19:26,725 --> 00:19:30,385
Go try it out. You can enable it now if you have a new enough Nix with an experimental feature.
315
00:19:30,605 --> 00:19:33,385
Then they need users to go test it and break it and figure out if it's going
316
00:19:33,385 --> 00:19:38,385
to work so they can get it across the finish line and shipping and actually ready to go.
317
00:19:38,565 --> 00:19:42,565
By default well Wes I got to know how was the Vegas experience given it was
318
00:19:42,565 --> 00:19:46,825
your first time there but also we've heard some things that Vegas has changed
319
00:19:46,825 --> 00:19:48,985
in a little bit so how was Vegas.
320
00:19:50,390 --> 00:19:55,470
You know, I have not yet had to get to see as much of it as I would like. It's definitely hot.
321
00:19:56,150 --> 00:20:02,710
It definitely doesn't sleep. But I like it, you know? There's entertainment wherever you'd like.
322
00:20:02,890 --> 00:20:04,950
Always something going on. I know that's your style.
323
00:20:05,550 --> 00:20:09,490
Yes, absolutely, right? There's no problem. I was able to go pick up some good
324
00:20:09,490 --> 00:20:13,070
show supplies at like 3 in the morning last night when I was getting back to the hotel.
325
00:20:13,630 --> 00:20:17,490
It is giant, though. So thankfully, the hotel I'm at is attached to their monorail
326
00:20:17,490 --> 00:20:21,210
system on the Strip. So that helps out a lot because otherwise you really don't
327
00:20:21,210 --> 00:20:24,250
want to walk, you know, 30 minutes and 106 degrees.
328
00:20:26,650 --> 00:20:29,270
But I'm excited after the show today. I'm hoping I can, well,
329
00:20:29,350 --> 00:20:33,630
maybe take a nap, but then go get off the strip, go explore some more of,
330
00:20:33,710 --> 00:20:36,790
you know, the rest of Vegas. So hopefully I'll have something fun.
331
00:20:36,910 --> 00:20:39,510
You know, maybe that vitamin D will keep you from getting sick.
332
00:20:39,810 --> 00:20:45,610
Did you get to bump into any listeners or anybody that maybe you've known from other conferences?
333
00:20:46,050 --> 00:20:49,690
Yes, I definitely saw our buddy Rob, who we had on the show,
334
00:20:49,790 --> 00:20:52,990
who helps organize scale. He was here. That was great to see him.
335
00:20:53,290 --> 00:20:57,070
And I ran into a few folks who've seen other conferences kind of all around.
336
00:20:57,190 --> 00:21:00,730
But in particular, shout out to Britton, who gave a great talk about Klan,
337
00:21:01,010 --> 00:21:03,610
which is a cool system we might have to talk more about.
338
00:21:03,730 --> 00:21:07,910
He's also working on a pretty great looking, like based on his talk and based
339
00:21:07,910 --> 00:21:11,130
on looking at his Hyperland setup, a very nice Hyperland setup,
340
00:21:11,250 --> 00:21:15,610
sort of Omarchi style, but on top of Nix, powered by Klan. And I don't know
341
00:21:15,610 --> 00:21:18,270
that any of that's ready or anything, but something I'm looking forward to.
342
00:21:18,550 --> 00:21:24,010
And then also a big shout out to listener Lucas, who heard about Nick's Vegas
343
00:21:24,010 --> 00:21:28,870
on the show and then flew out from Missouri just to come and check it out.
344
00:21:29,070 --> 00:21:32,410
So super great. Had a lovely time talking with Lucas.
345
00:21:32,630 --> 00:21:36,790
And he's actually in between jobs right now. So if you're looking for a sharp
346
00:21:36,790 --> 00:21:42,210
person who can run your infra and knows Nick's, we'll have a link to lucasr.com,
347
00:21:42,290 --> 00:21:43,190
which is his LinkedIn profile.
348
00:21:46,650 --> 00:21:51,230
1Password.com slash unplugged. Take the first step to better security for your
349
00:21:51,230 --> 00:21:54,270
team by securing credentials and protecting every application,
350
00:21:54,570 --> 00:21:56,830
even unmanaged shadow IT.
351
00:21:57,110 --> 00:22:03,630
That's 1Password, the number 1Password.com slash unplugged. And it's all lowercase.
352
00:22:03,870 --> 00:22:07,930
Go learn more at 1Password.com slash unplugged and support the show.
353
00:22:08,190 --> 00:22:10,050
Here's something I can really connect with.
354
00:22:10,550 --> 00:22:14,350
And if you're in IT or if you're in security specifically, you probably can too.
355
00:22:14,950 --> 00:22:18,650
The reality is there's a mountain of devices out there and more and more SaaS
356
00:22:18,650 --> 00:22:20,170
applications all the time.
357
00:22:20,290 --> 00:22:23,030
And that creates a mountain of security risk.
358
00:22:23,290 --> 00:22:26,410
The reality is that's getting harder and harder to deal with.
359
00:22:27,450 --> 00:22:32,190
Unless you check out 1Password Extended Access Management. That can help you
360
00:22:32,190 --> 00:22:35,190
conquer
the mountain of security risks that are coming at us all the time.
361
00:22:35,970 --> 00:22:39,990
When surveyed, over a half of IT pros say that securing their SaaS apps is the
362
00:22:39,990 --> 00:22:42,750
biggest challenge, specifically for the reasons I just outlined.
363
00:22:43,430 --> 00:22:49,930
Well, Trellica by 1Password can discover and secure access to your apps, managed or not.
364
00:22:50,090 --> 00:22:55,150
It inventories every app in use at your company. It has pre-populated app profiles
365
00:22:55,150 --> 00:22:57,790
that know where to assess and what to look at for different SaaS risks.
366
00:22:57,930 --> 00:23:03,350
And it lets you manage access, optimize spend, and even enforce best practices
367
00:23:03,350 --> 00:23:05,790
across every app your employees use.
368
00:23:05,890 --> 00:23:10,050
So you can manage the shadow IT stuff, which is something I used to struggle with so much.
369
00:23:10,150 --> 00:23:15,570
You can securely onboard and off-board employees, have a real process for that,
370
00:23:15,670 --> 00:23:17,090
and meet compliance goals.
371
00:23:17,410 --> 00:23:22,530
That's what Trellica by OnePass provides, a complete solution for SaaS access governance.
372
00:23:23,030 --> 00:23:26,510
It's just one of the ways that extended access management helps teams strengthen
373
00:23:26,510 --> 00:23:28,330
compliance and security.
374
00:23:28,870 --> 00:23:34,050
I had so many clients over the years that I'd show up and they'd have passwords under their keyboard.
375
00:23:34,650 --> 00:23:39,730
And this was a big battle for many years, was helping users understand the risks
376
00:23:39,730 --> 00:23:42,770
of that and then getting them tooling to solve that problem.
377
00:23:42,990 --> 00:23:45,970
And of course, one password came in and beautifully provided that tooling.
378
00:23:46,250 --> 00:23:50,470
Well, we're in a new era now, and that's where Trelica and extended access management
379
00:23:50,470 --> 00:23:54,530
come in. So go take the first step to better security for your team by securing
380
00:23:54,530 --> 00:23:59,050
credentials and protecting every application, even the unmanaged shadow IT.
381
00:23:59,550 --> 00:24:05,010
Learn more at 1password.com slash unplugged. That's the number 1password.com
382
00:24:05,010 --> 00:24:06,930
slash unplugged, all lowercase.
383
00:24:07,170 --> 00:24:09,550
You check it out, they have more information, and it's a great way to support
384
00:24:09,550 --> 00:24:12,930
the show. Just go to 1password.com slash unplugged.
385
00:24:16,330 --> 00:24:22,450
Well after the show last week kind of just on a whim um i decided to show you,
386
00:24:23,010 --> 00:24:25,930
a little tool i'd been trying out for vibe
387
00:24:25,930 --> 00:24:29,690
coding i i thought you know mostly it was just look at how nice the interface
388
00:24:29,690 --> 00:24:33,530
is it's really pretty i think chris will like it and hey it's packaged in nix
389
00:24:33,530 --> 00:24:36,430
so we could just give it a try but like we were sitting around you know finishing
390
00:24:36,430 --> 00:24:40,770
out the show notes i didn't think too much of it but it turns out that little
391
00:24:40,770 --> 00:24:43,870
event seems to have totally reshaped your week.
392
00:24:44,710 --> 00:24:48,470
Yeah, I kept playing with it. You know, the first thing when you showed it to
393
00:24:48,470 --> 00:24:52,070
me is like, oh, this would have been cool while we had the TUI challenge. This is a fun toy.
394
00:24:53,470 --> 00:24:57,090
And then I just decided to see how far I could push it. And I kind of created
395
00:24:57,090 --> 00:24:59,970
myself a little mini challenge over the last few days.
396
00:25:00,150 --> 00:25:08,670
Could I could I vibe code a riced up Hyperland desktop from essentially a base headless Nick system?
397
00:25:09,290 --> 00:25:12,410
But before I get into all that, I do want to set the tone for a moment.
398
00:25:13,290 --> 00:25:19,230
Because I realized, you know, we kind of touch on these types of topics once
399
00:25:19,230 --> 00:25:25,630
or twice a year, but we first started talking about this with no hype in like around 2022. And,
400
00:25:26,899 --> 00:25:30,259
I couldn't believe it when I went back and looked at our back catalog when we
401
00:25:30,259 --> 00:25:32,399
started talking about LLMs and this type of stuff.
402
00:25:32,959 --> 00:25:36,919
And I think we've consistently been really good about separating the practical from the hype.
403
00:25:39,019 --> 00:25:44,919
And I think it's one of these things where we have very high expectations and
404
00:25:44,919 --> 00:25:48,659
we kind of miss the trees in the forest, as they say, or whatever.
405
00:25:49,739 --> 00:25:53,779
And I think because part of it is some of the things we ask these LLMs to do,
406
00:25:53,999 --> 00:25:55,719
humans are pretty good at already.
407
00:25:56,579 --> 00:26:00,139
And so we kind of are judging it with a high bar. And so one of the things that's
408
00:26:00,139 --> 00:26:06,199
in the discussion right now around these LLM tools is can they code? Can you vibe code?
409
00:26:06,399 --> 00:26:09,219
Can you build application using these tools? And then, of course,
410
00:26:09,299 --> 00:26:12,259
people build these applications and they're pretty much universally mocked.
411
00:26:13,239 --> 00:26:16,079
But that's a pretty high bar, you know, because even humans aren't very good
412
00:26:16,079 --> 00:26:16,879
at developing software.
413
00:26:17,839 --> 00:26:22,139
And there aren't a ton of great examples online. So if you index the Internet,
414
00:26:22,479 --> 00:26:23,699
the average is pretty low.
415
00:26:25,279 --> 00:26:30,399
But I had this theory, and I thought, well, what if I lowered the bar a little bit?
416
00:26:30,779 --> 00:26:35,719
And maybe instead of trying to build a complex application, what if I just try
417
00:26:35,719 -->
00:26:37,179
to build some configuration files?
418
00:26:37,799 --> 00:26:42,559
You know, what if I try to just generate some YAML, or a Docker Compose file, or Nix, and so on.
419
00:26:42,839 --> 00:26:46,319
And so that's where I started with Crush. It's like, well, I don't need to create
420
00:26:46,319 --> 00:26:49,259
an app. I just, I want like a config template to start with.
421
00:26:50,178 --> 00:26:53,878
And I started building on that idea. And what I realized, and I'm curious what
422
00:26:53,878 --> 00:26:56,458
your thoughts are, Wes, is I have a theory.
423
00:26:57,058 --> 00:27:00,238
And that theory goes that this stuff is a little bit easier.
424
00:27:00,238 --> 00:27:01,938
It's a little bit simpler of a language.
425
00:27:02,218 --> 00:27:06,958
And there's probably more examples of it online that are better on average than,
426
00:27:07,058 --> 00:27:11,958
say, maybe code that these models have been trained on. So it's like it's indexing text.
427
00:27:12,438 --> 00:27:16,018
And this stuff is just a little more straightforward. So maybe these LLMs are
428
00:27:16,018 --> 00:27:20,538
just perhaps a little bit better at generating config files than application code?
429
00:27:20,918 --> 00:27:24,158
Yeah, I mean, I think the limited scope definitely helps, right?
430
00:27:24,298 --> 00:27:29,738
You're not trying to create a sketchy dating app so people will upload all their
431
00:27:29,738 --> 00:27:33,938
ID pictures that you leak out on the internet or whatever. You kind of have some nice bounds.
432
00:27:34,378 --> 00:27:39,458
There is a schema. When you go to try to rebuild or build whatever, it can shout at you.
433
00:27:39,578 --> 00:27:42,498
It can tell you what's wrong. not always with the best error messages,
434
00:27:42,518 --> 00:27:45,338
but often with a good error message or at least good enough,
435
00:27:45,438 --> 00:27:49,038
you can pipe that back in and get some suggestions on what to do next.
436
00:27:49,158 --> 00:27:52,618
Because, yeah, you're not trying to architect a giant distributed system with,
437
00:27:52,618 --> 00:27:56,598
you know, a Rails application and JavaScript and how do you figure out the right
438
00:27:56,598 --> 00:27:58,918
security and protocols and database and caching.
439
00:27:59,138 --> 00:28:01,978
Like, it's a better scoped problem.
440
00:28:02,458 --> 00:28:06,958
Yeah. And so, gentlemen, this was my hypothesize, is that perhaps this would
441
00:28:06,958 --> 00:28:10,218
be something these things would be better at. And I sort of got to this after
442
00:28:10,218 --> 00:28:11,238
playing with this crush tool.
443
00:28:11,498 --> 00:28:14,018
But I wanted to know how good.
444
00:28:14,618 --> 00:28:18,198
And I thought, well, how can I figure out how far somebody could take this?
445
00:28:18,558 --> 00:28:22,158
Somebody who's looking at this as a system administrator or somebody building
446
00:28:22,158 --> 00:28:25,618
a Linux box, not somebody trying to develop an application, not somebody that's
447
00:28:25,618 --> 00:28:28,998
trying to create, you know, some sort of wild out there AI app.
448
00:28:29,238 --> 00:28:32,098
I'm just trying to manage a dozen config files on my box.
449
00:28:32,418 --> 00:28:35,998
And so I took my existing Nix OS system that I have at home.
450
00:28:36,995 --> 00:28:42,115
And I essentially commented out everything but just the really base boot to
451
00:28:42,115 --> 00:28:45,275
a terminal, hardware setup, you know, left the file systems in place,
452
00:28:45,315 --> 00:28:48,595
but essentially just stripped it all the way down.
453
00:28:48,935 --> 00:28:53,035
Which, by the way, is so great to build a load into a previous version when
454
00:28:53,035 --> 00:28:56,875
you screwed something up, fix it, and then try it again and try it again.
455
00:28:57,015 --> 00:28:58,455
Like, I cannot tell you how great that is.
456
00:28:59,035 --> 00:29:01,415
Yeah, it's like, oh, I deleted too much in that build.
457
00:29:01,595 --> 00:29:08,075
Nope, nope, nope. So I stripped it down, got it essentially headless,
458
00:29:08,255 --> 00:29:13,775
and then proceeded to vibe code, at least attempt,
459
00:29:14,055 --> 00:29:19,095
a complete Hyperland desktop with riced out way bar and settings,
460
00:29:19,295 --> 00:29:21,895
custom key binds, theming, and more.
461
00:29:22,055 --> 00:29:26,235
And I quickly discovered I needed to SSH into this thing because I have this
462
00:29:26,235 --> 00:29:31,155
wide, ultra wide screen and Crush was horrible on there because it's on the terminal.
463
00:29:32,155 --> 00:29:35,195
But using it i was able to build a real basic config
464
00:29:35,195 --> 00:29:37,975
through through an ssh session running crush on that
465
00:29:37,975 --> 00:29:40,935
local box i was able to build out a basic
466
00:29:40,935 --> 00:29:44,595
config that i never wrote myself and what it really did is it i told it what
467
00:29:44,595 --> 00:29:48,115
i wanted and it went through and it walked my config and said well okay i'll
468
00:29:48,115 --> 00:29:51,775
re-enable this i'll re i'll re-enable this and i'll add this package and i'll
469
00:29:51,775 --> 00:29:54,815
re-enable this and so it was really intelligent about well if i already had
470
00:29:54,815 --> 00:29:58,235
it in there it would just uncomment it for me oh you already need you you already
471
00:29:58,235 --> 00:30:00,235
have this and it's fun, right?
472
00:30:00,335 --> 00:30:04,295
Because you pointed a directory, it scans that directory, it indexes everything
473
00:30:04,295 --> 00:30:06,235
and it has kind of a cool look to it.
474
00:30:06,555 --> 00:30:09,635
I mean, I thought it looked really neat. Brent, maybe you thought it looked,
475
00:30:10,715 --> 00:30:14,215
I don't know, like it was trying too hard, like an 80s retro computer interface.
476
00:30:14,375 --> 00:30:15,655
I know you had a chance to play with it.
477
00:30:16,555
--> 00:30:19,735
I actually appreciated how it was kind of active.
478
00:30:20,055 --> 00:30:22,255
You know, you were doing something, if you were waiting for it,
479
00:30:22,315 --> 00:30:27,535
it had this sort of, I don't know, animation, but it wasn't over the top or it was super annoying.
480
00:30:27,695 --> 00:30:31,815
It was just like, things are happening so surprisingly i liked it.
481
00:30:31,815 --> 00:30:37,415
And it lets you switch between multiple models it supports open router as well
482
00:30:37,415 --> 00:30:42,875
which is what i was using um so that's kind of nice did you accomplish anything when you tried crush.
483
00:30:42,875 --> 00:30:45,835
I attempted something a different
484
00:30:45,835 --> 00:30:52,735
direction than you i attempted to um pull the irc feature out of our jb website
485
00:30:52,735 --> 00:30:56,075
because it's a thing that i've wanted to do for a while but I haven't sat down
486
00:30:56,075 --> 00:31:00,155
and actually did it it didn't go so well for me but this is the first time I
487
00:31:00,155 --> 00:31:04,055
try to you know vibe vibe so maybe that's on me sometimes.
488
00:31:04,055 --> 00:31:04,715
The vibe isn't.
489
00:31:06,104 --> 00:31:11,664
I do want to call out that the folks behind this, the Charm slash Charm Bracelet
490
00:31:11,664 --> 00:31:13,884
team project, I don't know the full details.
491
00:31:14,324 --> 00:31:18,184
But I mean, not only did they make it super easy to try with all kinds of Nick's
492
00:31:18,184 --> 00:31:23,484
instructions and other distributions on Crush here, but it's all powered by
493
00:31:23,484 --> 00:31:27,404
Bubble T, which is a powerful TUI framework written in Go that they make.
494
00:31:27,504 --> 00:31:31,464
And then it powers other stuff like Gum, which is a tool to,
495
00:31:31,604 --> 00:31:35,824
you can call out to Gum from shell scripts to like make fancy looking interfaces.
496
00:31:36,344 --> 00:31:41,064
They've also got VHS, which is a CLI home video recorder for your terminal.
497
00:31:41,284 --> 00:31:44,204
And they've got Glow, which is a tool I've used all the time because it renders
498
00:31:44,204 --> 00:31:46,304
markdown on the command line that looks really nice.
499
00:31:46,464 --> 00:31:50,184
So like, it's very cool to see them continue to advance and Crush is a very
500
00:31:50,184 --> 00:31:53,304
impressive tool powered by some really good primitives.
501
00:31:53,604 --> 00:31:57,564
And it is slick looking. It really is. And what's neat is, of course,
502
00:31:57,624 --> 00:32:00,424
this is designed to look at a project directory for software,
503
00:32:00,544 --> 00:32:05,684
but it treats the configuration directory like a project. And And then it understands my system.
504
00:32:05,724 --> 00:32:08,404
It understands my sub configuration files, my includes it.
505
00:32:08,544 --> 00:32:13,144
It knows I'm on a flake based system. Like it gets my box. And then so everything
506
00:32:13,144 --> 00:32:15,624
it's doing is with that context and that understanding.
507
00:32:16,903 --> 00:32:20,143
That's powerful. That's really, really powerful. And I'll come back to that later.
508
00:32:20,343 --> 00:32:24,123
I do wonder, just because you mentioned, which you said you were using Open
509
00:32:24,123 --> 00:32:28,163
Router. Were you using a variety of models or did you just stick with one model for this?
510
00:32:28,563 --> 00:32:32,183
I did end up changing it. So I'll mention the model I use later,
511
00:32:32,403 --> 00:32:34,843
the second model. But I started with Claude, which seemed pretty good.
512
00:32:34,983 --> 00:32:39,163
And it's nice because with Open Router, as you probably could explain better
513
00:32:39,163 --> 00:32:42,943
than me, Wes, is I've just put the credits in one place and then it kind of
514
00:32:42,943 --> 00:32:46,063
divvies it out depending on which model I select on the tooling I'm using.
515
00:32:46,063 --> 00:32:49,643
And you've been telling me to do this for months so I finally went and funded
516
00:32:49,643 --> 00:32:54,983
an Open Router account and then yeah I selected Claude through that and I think
517
00:32:54,983 --> 00:32:57,383
it's one of their later ones that's available.
518
00:32:57,743 --> 00:33:00,523
I have a hard time differentiating all the model names to be honest with you
519
00:33:00,523 --> 00:33:01,703
it's kind of ridiculous the branding.
520
00:33:01,983 --> 00:33:05,543
Yeah it's getting confusing but you're totally right Open Router is just a nice
521
00:33:05,543 --> 00:33:08,823
little service that sits in the middle so you pay them instead of having to
522
00:33:08,823 --> 00:33:14,303
pay like six different folks and then they basically proxy to all the various
523
00:33:14,303 --> 00:33:15,563
different LLM providers.
524
00:33:15,803 --> 00:33:19,243
If you spend enough on credits, you get some free stuff. You can choose Claude
525
00:33:19,243 --> 00:33:23,323
or GPT or Gemini or DeepSeq or open source ones.
526
00:33:23,883 --> 00:33:29,523
And then it implements like an open AI style API. So it's super compatible with a lot of stuff.
527
00:33:29,783 --> 00:33:35,503
So I was surprised how simple it was to get a basic desktop without ever writing
528
00:33:35,503 --> 00:33:36,843
or touching a config file.
529
00:33:37,223 --> 00:33:41,323
But it was really basic. It's like, you know, all the little,
530
00:33:41,483 --> 00:33:45,923
because Hyperland is what you make of it. It's very, very simple by default.
531
00:33:46,163 --> 00:33:48,523
And so, you know, you got to rice it up, as they say in that community.
532
00:33:48,803 --> 00:33:51,723
And there's lots of little iterative changes that you want to make,
533
00:33:51,803 --> 00:33:54,123
and you want to make them quickly. Because one of Hyperland's cool features
534
00:33:54,123 --> 00:33:56,623
is that it auto-reloads when you change the config.
535
00:33:56,883 -->
00:33:59,923
And that's, you know, when you're making little tweaks to your bar or this or
536
00:33:59,923 --> 00:34:01,403
that, super great feature.
537
00:34:01,823 --> 00:34:05,603
And you can move quickly if you have the right tooling.
538
00:34:05,903 --> 00:34:09,863
And so this is where I decided to step it up. For the first time ever,
539
00:34:10,123 --> 00:34:14,403
I tried out the Cursor AI Code Editor, which is a fork of VS Code.
540
00:34:14,763 --> 00:34:17,943
I know there's others out there like Void, although I think development is paused for the moment.
541
00:34:18,857 --> 00:34:22,877
Similar to Crush, you know, in VS Code, you point it at a directory.
542
00:34:23,677 --> 00:34:27,217
So I could point it at my Hyperland, my Waybar, my style sheets,
543
00:34:27,377 --> 00:34:30,197
everything that makes up a Hyperland desktop configuration.
544
00:34:31,157 --> 00:34:34,917
And then I could ask questions against that configuration. So I could say,
545
00:34:34,997 --> 00:34:37,197
for example, what key got, did I bind to Firefox?
546
00:34:37,417 --> 00:34:40,237
And it would, it would tell me, but I, I could also ask it, what can I do?
547
00:34:40,637 --> 00:34:43,097
What, what, what, what software is missing in order to accomplish this?
548
00:34:43,757 --> 00:34:46,397
And as I started to use it, I started to
549
00:34:46,397 --> 00:34:50,017
realize that you can kind of build a plan with this thing and then they'll go
550
00:34:50,017 --> 00:34:54,377
and it'll execute through all of them and you can even let them run on their
551
00:34:54,377 --> 00:34:58,397
own on the command line with your user privileges including running the build
552
00:34:58,397 --> 00:35:04,057
tests checking the output when the when the build fails resolving those problems
553
00:35:04,057 --> 00:35:06,057
and then building again until it works.
554
00:35:06,057 --> 00:35:09,897
It's kind of wild right like i mean don't maybe don't give it access to your
555
00:35:09,897 --> 00:35:13,877
production database just yet but like yeah i mean that's where i think the kind
556
00:35:13,877 --> 00:35:17,557
of the agent side of stuff right like you not only does the lm generate the
557
00:35:17,557 --> 00:35:21,357
plan but then like it will actually step through and execute all that and then
558
00:35:21,357 --> 00:35:23,997
collect more data to keep this the feedback cycle going.
559
00:35:23,997 --> 00:35:29,397
And then gives me a summary of what it did formatted a markdown which is great.
560
00:35:29,397 --> 00:35:33,157
Right yes of course right turns out these lms they are good at summarizing so.
561
00:35:33,157 --> 00:35:38,117
And and because all of this is just text files it's really pretty simple but
562
00:35:38,117 --> 00:35:43,177
also the way cursor works is as it's going it's showing me a diff of what it's
563
00:35:43,177 --> 00:35:47,737
adding and removing and I can approve each step or I can deny a step or I can stop it.
564
00:35:47,917 --> 00:35:51,137
Right. That kind of critical human in the loop step, you're still in control,
565
00:35:51,377 --> 00:35:55,197
especially because, you know, despite how much Brent and I try to lecture you,
566
00:35:55,357 --> 00:35:57,257
you know, you don't always put all of your stuff in Git.
567
00:35:57,417 --> 00:36:00,117
So it's kind of nice that you get that extra diffing that you,
568
00:36:00,337 --> 00:36:01,937
you know, so you don't have to go full tool.
569
00:36:02,197 --> 00:36:05,957
You guys were ringing in the back of my head, though. So what I did is I took
570
00:36:05,957 --> 00:36:11,657
all my config from like my Etsy Nix OS and I staged it in a config folder in
571
00:36:11,657 --> 00:36:16,297
my home directory so i had this working against that and then when i was done
572
00:36:16,297 --> 00:36:21,197
and i was satisfied i would just like a monkey on the back end cp all the files over and then build,
573
00:36:21,997 --> 00:36:26,997
so you know it was so i wasn't letting it just run wild on the bear system i
574
00:36:26,997 --> 00:36:29,657
did move them over and then i would kind of like check it and then deploy it
575
00:36:29,657 --> 00:36:32,437
and cursor because the time i was using it which was uh.
576
00:36:33,737 --> 00:36:36,537
August 7th 9th 8th whatever they had
577
00:36:36,537 --> 00:36:39,197
free gpt5 integration and it's funny because i've seen
578
00:36:39,197 --> 00:36:43,837
a lot of people crapping on gpt5 this weekend but i definitely noticed a step
579
00:36:43,837 --> 00:36:51,277
up in the thoroughness because they can research so if you have a model that
580
00:36:51,277 --> 00:36:55,277
can can do a lot of things and that's one of the things gpt5 can do is it can
581
00:36:55,277 --> 00:36:58,437
do research and the other type of stuff whatever so i could tell it,
582
00:36:58,637 --> 00:37:03,777
I want a super riced out Hyperland theme that has my current IP address up in
583
00:37:03,777 --> 00:37:05,417
the Waybar, my CPU, my GPU.
584
00:37:05,557 --> 00:37:08,417
I want my network stats up there. I want the current weather.
585
00:37:08,517 --> 00:37:09,937
I want a clock with a calendar.
586
00:37:10,357 --> 00:37:13,437
Also, I want you to make sure that all of this works with a dark background.
587
00:37:13,617 --> 00:37:17,157
And by the way, I don't have an NVIDIA GPU. I have an AMD GPU.
588
00:37:17,557 --> 00:37:21,457
And keep in mind, I'm also on NixOS. So we have to work within these limitations.
589
00:37:22,357 --> 00:37:26,217
It then kicks off a research job. It cites 16 different sources.
590
00:37:26,677 --> 00:37:31,057
It reviews them against my current configuration. It then proceeds to modify
591
00:37:31,057 --> 00:37:33,017
those configs to enable them.
592
00:37:33,137 --> 00:37:37,197
And then if I need to install a package, it updates my configuration.nix.
593
00:37:37,837 --> 00:37:43,397
It runs a build test for me, make sure that works, and
then also gives me an
594
00:37:43,397 --> 00:37:45,277
output of any additional software I need to install.
595
00:37:45,337 --> 00:37:47,837
And I can either have it do it or I can do that. I just had it do it.
596
00:37:48,477 --> 00:37:53,117
I mean, you can really get moving here because I don't know anything about Hyperland.
597
00:37:53,697 --> 00:37:56,737
I don't know the names of these tools, these modules that you can put up,
598
00:37:56,837 --> 00:37:59,117
and there's scripts that you can write. I don't know any of this stuff.
599
00:37:59,217 --> 00:38:01,997
I don't know what launcher people like because there's multiple launchers.
600
00:38:02,237 --> 00:38:05,597
Right. That's really sticking out to me. I mean, just like, right,
601
00:38:05,637 --> 00:38:07,977
you were able to say, like, I want a RICED configuration.
602
00:38:08,057 --> 00:38:11,817
I want something that's going to look nice like the folks over on Unix Porn Share, right?
603
00:38:11,937 --> 00:38:14,697
Like, without having to have a super concrete list.
604
00:38:14,717 --> 00:38:18,017
I linked Unix Porn. I linked some GitHub's in there for it to go check.
605
00:38:18,617 --> 00:38:21,957
But that's, you know, it's one thing to know exactly what you want and then
606
00:38:21,957 --> 00:38:23,277
try to get these tools to do it.
607
00:38:23,377 --> 00:38:27,097
And it's kind of the next level up when you can just be like, I have a high level goal.
608
00:38:27,257 --> 00:38:31,037
I'll know it when I see it. But otherwise, like I need you to suss out what
609
00:38:31,037 --> 00:38:33,617
the even the possible steps are to try and get there.
610
00:38:34,393 --> 00:38:38,533
Well, here's, and tell me if you think this is true. Here's why I think I was
611
00:38:38,533 --> 00:38:42,193
having kind of a success after success for a while. It was going really well
612
00:38:42,193 --> 00:38:46,113
because everything Nix configuration is text-based.
613
00:38:46,373 --> 00:38:50,193
Everything for configuring Hyperland is text-based. And most of the stuff that
614
00:38:50,193 --> 00:38:53,033
makes like the Waybar modules work is just bash scripting.
615
00:38:53,373 --> 00:38:53,573
Yeah.
616
00:38:53,853 --> 00:38:55,853
So this is like, this is its power zone.
617
00:38:55,993 --> 00:38:59,653
It totally is. Yeah, you're able to turn so much of this stuff into,
618
00:39:00,033 --> 00:39:02,893
you know, it's captured, it's declarative, it's right there.
619
00:39:02,893 --> 00:39:05,793
And lots of people are pushing and sharing it on the internet.
620
00:39:05,813 --> 00:39:10,173
So you have a good training base and you have the native medium for these tools.
621
00:39:10,353 --> 00:39:12,473
So they don't have to go that far to be able to affect these changes.
622
00:39:13,133 --> 00:39:17,273
I mean, I'm not kidding. I'm saying like this first initial success I was having,
623
00:39:18,133 --> 00:39:20,433
happened in like the span of maybe two, two
and a half hours.
624
00:39:20,593 --> 00:39:22,613
I probably started around 8.30 a.m.
625
00:39:22,793 --> 00:39:26,913
And by 10 a.m., I'm like, I can't believe how well this is going.
626
00:39:27,693 --> 00:39:31,953
And I started, I was actually, I got back to work. I had things humming by 1.30
627
00:39:31,953 --> 00:39:35,933
p.m. I had things humming to the point of where I was getting like that,
628
00:39:36,093 --> 00:39:38,253
oh, no, I think I'm falling in love with this.
629
00:39:38,373 --> 00:39:41,713
I think I'm going to have to change all my systems feeling. You know that how I get, right?
630
00:39:41,873 --> 00:39:47,213
Like, we got to change the studio out. We got to change my workstations out. Like, I got to switch.
631
00:39:47,433 --> 00:39:52,113
I mean, I was really quickly beginning to even develop muscle memory for my hotkeys that I set up.
632
00:39:53,233 --> 00:39:58,353
So it took an LLM to get you finally, after years and years of the audience,
633
00:39:58,553 --> 00:40:00,693
trying to get you to a toiling window manager.
634
00:40:01,413 --> 00:40:01,673
Yeah.
635
00:40:01,893 --> 00:40:02,073
Huh.
636
00:40:02,653 --> 00:40:09,973
That is that is true that was 1 30 by about 2 p.m. it was literally all gone,
637
00:41:46,311 --> 00:41:50,391
This is when it really hit me, like, I didn't know how I had built this, right?
638
00:41:50,571 --> 00:41:54,871
And not only that, but my documentation at best was a couple of the prompts that I had saved.
639
00:41:56,291 --> 00:41:59,351
Right, you had some intent, you had whatever the diffs were,
640
00:41:59,351 --> 00:42:02,871
and how much you could remember of your vibe high.
641
00:42:03,251 --> 00:42:06,731
Yeah, and this is right where it's like, suddenly, do you have the tools to keep going?
642
00:42:06,851 --> 00:42:10,151
Can you get yourself to a state where you're like, well, I've made myself a
643
00:42:10,151 --> 00:42:14,771
mess here, and it's almost there, but I don't know the right levers to push to keep going.
644
00:42:15,131 --> 00:42:15,451
Yeah.
645
00:42:15,731 --> 00:42:18,511
You had this all committed in your Git repo though, right?
646
00:42:18,771 --> 00:42:21,911
Okay, I admit, I should have. That would have helped the situation.
647
00:42:21,911 --> 00:42:25,151
I at least would have, you know, had a sense of like, okay, I could go check
648
00:42:25,151 --> 00:42:26,591
what happened over there and roll back.
649
00:42:26,711 --> 00:42:30,671
But no, and I started thinking like, am I going to, am I going to like try to
650
00:42:30,671 --> 00:42:31,551
figure out the differences?
651
00:42:31,771 --> 00:42:34,511
Am I going to swap the files? Is there some, but then I started thinking,
652
00:42:34,611 --> 00:42:35,531
well, maybe there's something broken.
653
00:42:35,911 --> 00:42:40,611
Do I try to rebuild this prompt by prompt? And then like, will I get a different
654
00:42:40,611 --> 00:42:42,111
result? which, by the way, I did
test.
655
00:42:42,271 --> 00:42:46,251
And yes, I do get different results sometimes, like when I'm ricing out the
656
00:42:46,251 --> 00:42:48,491
way bar. It depends on what I emphasize.
657
00:42:49,331 --> 00:42:54,251
If I emphasize large text or certain style things, it's interesting what kind
658
00:42:54,251 --> 00:42:55,091
of differences I can get.
659
00:42:55,471 --> 00:42:59,131
So I wasn't sure what direction I was going to take. So I fired up cursor.
660
00:42:59,531 --> 00:43:02,471
I explained what had happened, that I'd done a NixOS rebuild,
661
00:43:02,491 --> 00:43:05,931
and I came back and I was essentially at a stock Hyperland desktop with like
662
00:43:05,931 --> 00:43:07,751
the Omnicube background.
663
00:43:08,151 --> 00:43:12,591
And I'm sitting there looking at my screen, I've been realizing I have been
664
00:43:12,591 --> 00:43:15,791
working at this since about 8 a.m. It's now two something.
665
00:43:17,151 --> 00:43:19,331
I need to just get up and get some air.
666
00:43:56,123 --> 00:44:03,163
Yeah. So one of the things that we did, the LM and I, is we converted this into a multi-host setup.
667
00:44:03,383 --> 00:44:09,283
And so each individual host of mine has a subfolder with its own config.
668
00:44:11,083 --> 00:44:13,643
I forgot. I forgot we did that.
669
00:44:14,303 --> 00:44:15,383
Oh my gosh.
670
00:44:15,383 --> 00:44:17,103
Yeah so i.
671
00:44:17,103 --> 00:44:18,163
Love this journey.
672
00:44:18,163 --> 00:44:21,403
I know so but what's great is the
673
00:44:21,403 --> 00:44:25,523
machine didn't and it explained to me what's going on and it's like would you
674
00:44:25,523 --> 00:44:29,263
like me to just take care of the deploying and staging and then building and
675
00:44:29,263 --> 00:44:33,463
i could just do that if you would like i'm not kidding you and i'm like yeah
676
00:44:33,463 --> 00:44:39,383
why don't you go ahead and and just handle this for me uh because i'm clearly going to screw.
677
00:44:39,443 --> 00:44:42,423
Up and from that point forward it did all
678
00:44:42,423 --> 00:44:45,303
of it it i let it i let it do
679
00:44:45,303 --> 00:44:48,323
the config changes i let it copy over.
680
00:44:48,323 --> 00:44:50,963
My configs once i you know did a quick review and then
681
00:44:50,963 --> 00:44:54,143
i let it do the the rebuild and i.
682
00:44:54,143 --> 00:44:58,063
Let it determine when it should just do a nix os rebuild switch or when it should
683
00:44:58,063 --> 00:45:01,563
do a nix os rebuild boot depending on the significance of the change and how
684
00:45:01,563 --> 00:45:07,523
fast it wanted to test its considerations yeah it's crazy so this was i i recorded
685
00:45:07,523 --> 00:45:11,703
one more clip sort of later on in the evening just sort of capturing my reflections on the moment.
686
00:46:22,421 --> 00:46:26,701
It was sort of blowing my mind. I even had a hard time sleeping that night because
687
00:46:26,701 --> 00:46:33,041
I was
just running through how I just built a full desktop system.
688
00:46:33,261 --> 00:46:35,081
And yeah, it took me a while, but I was playing around, right?
689
00:46:35,241 --> 00:46:39,341
I mean, it actually worked. There was a moment where everything went sideways,
690
00:46:39,341 --> 00:46:41,841
but that was kind of because I was trying new stuff too.
691
00:46:42,901 --> 00:46:46,121
I was shocked that I got this far. Now, there is stuff that's not great.
692
00:46:46,941 --> 00:46:49,421
I went through this morning. I was looking at the config files.
693
00:46:49,541 --> 00:46:53,161
There's some areas where I was trying different things, and it's kind of sloppy
694
00:46:53,161 --> 00:46:56,281
about how it commented a few things out. So it's messy.
695
00:46:56,961 --> 00:46:57,241
Yeah.
696
00:46:57,341 --> 00:47:00,321
Yeah. I mean, it's not bad. I think you could clean it up pretty quickly.
697
00:47:00,521 --> 00:47:02,981
But I looked at it and I was like, oh, that's not how I would do it.
698
00:47:03,101 --> 00:47:07,181
I am kind of curious in that regard. Like, did you hit, because what I'm picking
699
00:47:07,181 --> 00:47:10,841
up, and I found this when I started kind of trying vibe shelling and stuff,
700
00:47:10,981 --> 00:47:15,901
like, I mean, the vibes are good, the vibes are bad, but it is really like the vibe.
701
00:47:16,361 --> 00:47:19,181
It's not just that you're riding the vibes. Like, it gives you vibes, right? you were
702
00:47:19,181 --> 00:47:22,661
feeling super empowered and a little down at one point i'm
703
00:47:22,661 --> 00:47:25,401
curious though like did you run because we're not just trying to sell
704
00:47:25,401 --> 00:47:28,741
cursor subscriptions of course no i don't uh you know unless they want to reach
705
00:47:28,741 --> 00:47:32,361
out um uh but like did you run into any points where like you just were hitting
706
00:47:32,361 --> 00:47:34,841
your head against the wall or it was hitting its head against the wall because
707
00:47:34,841 --> 00:47:38,361
sometimes you can get down loops where like the ai is confusing it and if you
708
00:47:38,361 --> 00:47:41,581
don't have enough context or your own like escape hatches to tell it like no
709
00:47:41,581 --> 00:47:44,521
you've gone down the wrong path here that isn't going to work.
710
00:47:44,521 --> 00:47:49,041
Ah yeah i'm glad you asked this yeah one time only one time but one time.
711
00:47:50,286 --> 00:47:53,186
It just got in this crazy loop of like, okay, I have to figure this out,
712
00:47:53,246 --> 00:47:54,626
but then I'll figure this out, but then I'll figure this out,
713
00:47:54,666 --> 00:47:56,206
but then I have to figure this out, and then I'll figure this out.
714
00:47:56,366 --> 00:47:59,306
And it was like these three things, and it went, and it went,
715
00:47:59,346 --> 00:48:01,486
and it went, and it went, and it went, and I just eventually had to abort,
716
00:48:01,546 --> 00:48:02,626
and I
just kind of restarted.
717
00:48:02,786 --> 00:48:03,526
Yeah, that makes sense.
718
00:48:04,046 --> 00:48:06,986
Again, I'm not trying to sell any particular product, and I would love to know
719
00:48:06,986 --> 00:48:08,066
of free software alternatives.
720
00:48:08,586 --> 00:48:11,746
But one of the things that's handy about Cursor is you can give it particular
721
00:48:11,746 --> 00:48:14,726
config files as context, GitHub repo links as context.
722
00:48:15,246 --> 00:48:19,206
So you can get back up and running pretty quickly after something like that happens.
723
00:48:19,206 --> 00:48:21,926
I'm also curious because i was
724
00:48:21,926 --> 00:48:25,526
running into this where it's like especially with those like when that's happening uh
725
00:48:25,526 --> 00:48:28,266
suddenly i was like becoming more keenly aware that
726
00:48:28,266 --> 00:48:31,786
like right i'm every time it's doing this i'm paying
727
00:48:31,786 --> 00:48:36,026
for that unless i mean either in power and you know the gpus you've bought if
728
00:48:36,026 --> 00:48:39,746
you're doing it self-hosted style or in credits or you know whatever so i'm
729
00:48:39,746 --> 00:48:43,806
curious did you get any kind of rough spend because it's you know if it if it's
730
00:48:43,806 --> 00:48:48,326
three or five or ten dollars that's different than if it takes you know 30 to get.
731
00:48:48,326 --> 00:48:51,126
Yourself a desktop no no it was like eight bucks uh when
732
00:48:51,126 --> 00:48:54,306
i used open router is all i spent when using crush plus i
733
00:48:54,306 --> 00:48:57,686
did a couple on yours so i think i spent like 10 cents on your account right
734
00:48:57,686 --> 00:49:00,346
but uh when i switched over to cursor and this was
735
00:49:00,346 --> 00:49:03,866
kind of a cheat and i'm really glad you brought this up cursor was
736
00:49:03,866 --> 00:49:08,466
giving away gpt5 access for the weekend they might still be doing it oh yeah
737
00:49:08,466 --> 00:49:14,366
so i essentially just went for it i just went for it i mean hours i was like
738
00:49:14,366 --> 00:49:19,446
i'm not going to stop because it's basically unlimited and so yeah that was
739
00:49:19,446 --> 00:49:21,626
i love that you're essentially you're.
740
00:49:21,626 --> 00:49:26,166
You're sucking up that vc money to build uh free and open source desktops keep doing it.
741
00:49:26,166 --> 00:49:29,346
It does make me think how great would it be if you could accomplish this with
742
00:49:29,346 --> 00:49:32,026
a local llm like that imagine if you could have a local llm that could help
743
00:49:32,026 --> 00:49:35,906
you build these setups right because we're we're dealing with you know nix and
744
00:49:35,906 --> 00:49:40,566
yaml and hyperland configurations and css files and bash script that's the entirety
745
00:49:40,566 --> 00:49:41,806
of it it's really approachable i.
746
00:49:41,806 --> 00:49:44,426
Mean and they're getting better and better. I
think that's the future.
747
00:49:44,686 --> 00:49:49,406
And I wanted to try this because Oma Archie impressed me a lot,
748
00:49:49,426 --> 00:49:50,806
but I'd like to have my own take on it.
749
00:49:52,469 --> 00:49:56,049
Here's what I ended up with that I like. You know, the cons were I think some
750
00:49:56,049 --> 00:49:57,489
of the configs could be cleaned up a little bit.
751
00:49:58,129 --> 00:50:02,769
But what I liked is, one, I could generate reports against my configs.
752
00:50:03,329 --> 00:50:07,449
So I had to generate me markdown tables of all my custom keybinds so I can do
753
00:50:07,449 --> 00:50:08,729
a little printout of that.
754
00:50:09,329 --> 00:50:12,509
And you can just ask it things about that, which I found to be really fascinating.
755
00:50:12,749 --> 00:50:18,229
And at the end result, I got a rolling flake-enabled NixOS system using the
756
00:50:18,229 --> 00:50:24,049
Zen kernel with all its optimizations, ButterFS-based file systems with auto-scrub enabled.
757
00:50:24,649 --> 00:50:28,789
Nick Store is auto-optimized, and it also does auto-garbage collection.
758
00:50:29,469 --> 00:50:34,349
ZRAM is set up. I have auto-BPF tune enabled, auto-tuning the system on demand.
759
00:50:34,829 --> 00:50:38,269
And it's a collection of what I think are some of the best-in-class Linux desktop
760
00:50:38,269 --> 00:50:46,189
applications, riding on a totally rad Hyperland setup that I am very, very into right now.
761
00:50:46,289 --> 00:50:49,729
I mean, really love this setup so much. I think this is my new setup.
762
00:50:50,569 --> 00:50:55,829
And in hindsight, I wouldn't start with a naked MVP system, especially if you're
763
00:50:55,829 --> 00:50:59,189
doing it on a declarative system and you have a high-res screen.
764
00:51:00,849 --> 00:51:04,949
Just, you don't need to. I was doing it as an experiment and looking back at
765
00:51:04,949 --> 00:51:09,129
it, I should have just started with the desktop I had on there and then just done this.
766
00:51:09,329 --> 00:51:13,249
Ah, you like overly stripped it back thinking like that would make a cleaner,
767
00:51:13,489 --> 00:51:14,589
easier experience probably.
768
00:51:14,729 --> 00:51:17,489
But it turned out to be a little counterproductive. I suppose especially,
769
00:51:17,669 --> 00:51:21,829
maybe that is true in other systems, but with Nix, I guess it's like it matters less.
770
00:51:21,989 --> 00:51:24,409
So if anybody out there is running Hyperland or has done this type of stuff,
771
00:51:24,749 --> 00:51:27,929
I'd love some advice from you. What to do, what not to do, what I should look
772
00:51:27,929 --> 00:51:29,349
into, because I'm a total noob.
773
00:51:30,369 --> 00:51:33,789
But Rofi is such a cool launcher because it's also a file browser,
774
00:51:33,789 --> 00:51:38,209
and it can also give you a dictionary of all your keybinds, but it's also a
775
00:51:38,209 --> 00:51:39,789
very fast application launcher.
776
00:51:40,209 --> 00:51:43,149
And because Hyperland doesn't really choose anything for you,
777
00:51:43,289 --> 00:51:48,289
I'm kind of picking and choosing. So I have Kate and Dolphin from Plasma,
778
00:51:48,289 --> 00:51:51,909
and then some of the apps I'm using are GTK apps from Gnome,
779
00:51:51,989 --> 00:51:55,149
but it's all running so smoothly on Hyperland.
780
00:51:55,389 --> 00:51:59,669
It is what I've always wanted in terms of performance from the Linux desktop.
781
00:51:59,669 --> 00:52:05,029
That smoothness and responsiveness with a hint of modern animation that isn't gaudy,
782
00:52:05,129 --> 00:52:10,649
but a level of customization where I can do things that take extensions in Gnome
783
00:52:10,649 --> 00:52:15,569
that are really straightforward like Waybar modules in Hyperland that aren't
784
00:52:15,569 --> 00:52:16,949
going to break every time I update.
785
00:52:16,949 --> 00:52:20,569
And so it's this it's this combination of
786
00:52:20,569 --> 00:52:26,129
really practical key binds and navigation with really good tiling that's working
787
00:52:26,129 --> 00:52:31,369
for me way better than I ever expected with this curated setup kind of going
788
00:52:31,369 --> 00:52:34,669
out there and asking the internet like what are some of the most popular things
789
00:52:34,669 --> 00:52:39,529
what are in vogue right now in hyperland and bring it into my setup for me and.
790
00:52:40,849 --> 00:52:46,089
I think my biggest takeaway from this is, say what you will about these LLM tools.
791
00:52:46,809 --> 00:52:51,609
I never wrote a single line of any config file of any of this.
792
00:52:51,749 --> 00:52:55,109
I never touched a CSS file. I never touched my Waybar config.
793
00:52:55,329 --> 00:52:56,869
I never touched my Nix config.
794
00:52:57,149 --> 00:53:01,129
I didn't make the changes to make each machine have their own subconfiguration.
795
00:53:01,509 --> 00:53:04,289
I didn't make the changes to have a declarative Hyperlin setup.
796
00:53:04,509 --> 00:53:07,689
At one point during it, I realized I could do that, and it did all the refactoring
797
00:53:07,689 --> 00:53:10,669
for me to make Hyperlin declarative on my Nix system.
798
00:53:10,849 --> 00:53:11,549
That's awesome.
799
00:53:12,009 --> 00:53:14,709
Yeah. And I didn't, I just told the machine to do it.
800
00:53:15,049 --> 00:53:17,569
I mean, there was literally a point where I was just like sitting there sipping,
801
00:53:17,709 --> 00:53:20,649
sipping on a drink and just watching the machine go and doing little checks
802
00:53:20,649 --> 00:53:23,989
on it and approving things and okay, go for it or asking other questions.
803
00:53:24,789 --> 00:53:29,369
That part was really powerful. As far as like these companies go and the resources
804
00:53:29,369 --> 00:53:32,129
they use, those are all things we're going to have to sort out.
805
00:53:32,229 --> 00:53:36,069
But I think we have to not discount their
capabilities when you lower the bar
806
00:53:36,069 --> 00:53:39,349
of the task a little bit to something that's a little bit easier for it to handle.
807
00:53:39,349 --> 00:53:43,469
So if you wanted to create the next mobile app, like Wes is saying, that's a big ask.
808
00:53:43,549 --> 00:53:46,149
But if you wanted to generate a riced out Hyperland desktop,
809
00:53:46,509 --> 00:53:48,269
that's actually within its capabilities today.
810
00:53:48,889 --> 00:53:52,989
And because it's all just written in text, I can just go through it and learn what it did.
811
00:53:53,169 --> 00:53:56,329
I read my next config, I read the Hyperland config, the Waybar config,
812
00:53:56,389 --> 00:53:58,489
and I'm done. And I understand what it did.
813
00:53:58,729 --> 00:54:01,429
I think this is kind of the similar scale, like when we were at the Red Hat
814
00:54:01,429 --> 00:54:04,669
Summit this year, right? We saw Lightspeed really being talked about as like
815
00:54:04,669 --> 00:54:06,509
a command line tool you can use.
816
00:54:06,649 --> 00:54:10,709
I think it's that same kind of idea, right? Like maybe it's not architecting
817
00:54:10,709 --> 00:54:14,369
your entire OpenShift cluster, but it can definitely help you troubleshoot,
818
00:54:14,729 --> 00:54:16,669
work with text files, modify them.
819
00:54:16,789 --> 00:54:20,169
And then, you know, you layer on more declarative stuff and it just gets better.
820
00:54:20,469 --> 00:54:24,109
But I think you knew I was going to ask this, but you didn't know the full reason.
821
00:54:24,209 --> 00:54:28,309
While I was playing Nix Ambassador this week, I was actually,
822
00:54:28,409 --> 00:54:31,329
because you had already like, you'd been teasing Brett and I in our private
823
00:54:31,329 --> 00:54:34,329
chat, just sort of about what you were working on, how excited you were.
824
00:54:34,529 --> 00:54:38,449
So I was actually using that as a selling point to Nix. I was like, yeah, well, I'm here.
825
00:54:38,649 --> 00:54:43,049
My buddy back home, he's got himself a vibe-coded, super shiny desktop.
826
00:54:43,329 --> 00:54:46,969
And I know at the end of the week, if he just pushes that up to GitHub,
827
00:54:47,269 --> 00:54:50,549
I'm going to be able to just totally clone and copy it. And that's one of the
828
00:54:50,549 --> 00:54:52,109
reasons I was telling people they should try Nix.
829
00:54:52,549 --> 00:54:53,749
Oh, the pressure's on.
830
00:54:53,969 --> 00:54:55,189
Are you willing to share?
831
00:54:56,129 --> 00:55:01,929
Totally. Yeah, let's do that after the show because I would also like a set of eyes on these.
832
00:55:02,169 --> 00:55:05,029
I mean, who knows? Maybe there's something in here that's awful.
833
00:55:05,029 --> 00:55:08,469
I would like people to review it and boost in some suggestions for improvements
834
00:55:08,469 --> 00:55:12,589
or fixes or additions or things to make Hyperland even more enjoyable.
835
00:55:13,629 --> 00:55:16,069
So, yeah, I'd love to get it up there. I might put it on my GitHub,
836
00:55:16,129 --> 00:55:17,609
I guess, and we'll put a link in the show notes.
837
00:55:17,949 --> 00:55:18,269
Perfect.
838
00:55:18,669 --> 00:55:20,269
All right. You talked me into it, Wes Payne.
839
00:55:24,404 --> 00:55:29,004
Unraid.net slash Unplugged. Go unleash your hardware, check out Unraid,
840
00:55:29,124 --> 00:55:34,104
support the show, and celebrate Unraid OS turning 20 years old.
841
00:55:34,344 --> 00:55:39,724
Can you believe it? What a testament to them. And it's also the 20 days of the Unraid summer sale.
842
00:55:39,824 --> 00:55:44,584
It's really just begun, so it runs now through August 26th, and you'll get 20%
843
00:55:44,584 --> 00:55:47,404
off the starter, Unleash licenses, and upgrades.
844
00:55:47,804 --> 00:55:51,644
Also, every item in the brand new merch store, also 20% off.
845
00:55:51,644 --> 00:55:55,444
And they're also doing daily giveaways, free licenses, merch bundles,
846
00:55:55,464 --> 00:55:57,984
and even a link station NAS is up for grabs.
847
00:55:58,384 --> 00:56:03,204
So if you have a unique Unraid setup, you can enter their Show Us Your System
848
00:56:03,204 --> 00:56:05,904
competition. That's Show Us Your System competition.
849
00:56:06,284 --> 00:56:09,044
And you get a chance to win prizes or at least bragging rights.
850
00:56:09,064 --> 00:56:10,024
And I want to know about it, too.
851
00:56:10,224 --> 00:56:13,184
So on August 30th, they're going to have a big bash. Their founder is going
852
00:56:13,184 --> 00:56:17,264
to get together. Some friends are going to get together for Unraid's 20th birthday bash livestream.
853
00:56:17,544 --> 00:56:20,684
And they're going to reveal the Show Us Your System winners at that time.
854
00:56:21,644 --> 00:56:24,804
I think it's going to be pretty neat. They're also going to premiere their first
855
00:56:24,804 --> 00:56:27,784
ever film and much more. Can you believe it?
856
00:56:28,564 --> 00:56:33,384
Unraid.net slash unplugged where you get started. So it is happening right now.
857
00:56:33,504 --> 00:56:37,544
The 20 days of Unraid summer sale where you can get 20% off the starter,
858
00:56:37,924 --> 00:56:40,064
the Unleashed licenses, upgrades, merch.
859
00:56:40,244 --> 00:56:44,144
And of course, they have the show us your unique competition going on right
860
00:56:44,144 --> 00:56:48,104
now where you can send them deets about your setup and maybe get some swag.
861
00:56:48,304 --> 00:56:51,244
I don't know. Maybe you get some bragging rights. I don't know.
862
00:56:51,644 --> 00:56:54,284
You know, maybe there's even a little bit of LinkStation hardware in there up
863
00:56:54,284 --> 00:56:59,344
for grabs. You just never know. You get started at unraid.net slash unplugged.
864
00:57:02,124 --> 00:57:06,044
Well, again, this week we have a few new member shoutouts.
865
00:57:06,344 --> 00:57:11,784
Dan is now a new core
contributor and Jacob as well, which means we ran out
866
00:57:11,784 --> 00:57:13,544
of slots for that bootleg promo.
867
00:57:16,144 --> 00:57:20,524
Amazing. Thank you, everybody. We did, and then I got a couple of requests to
868
00:57:20,524 --> 00:57:23,464
add some more slots, And I thought on it and I thought, why not?
869
00:57:23,524 --> 00:57:26,364
If we get a few more people checking out the bootleg, I think they'd stay members forever.
870
00:57:26,564 --> 00:57:29,404
So the 15% off deal, it's a great deal.
871
00:57:29,924 --> 00:57:34,124
People love the deal. Best deal. It's back. And I added a few more slots and
872
00:57:34,124 --> 00:57:35,104
then people jumped on it already.
873
00:57:35,244 --> 00:57:38,144
So now as of this episode, there are seven redemption slots left.
874
00:57:38,564 --> 00:57:43,144
If you use the promo code bootleg, either as a core contributor or for the Jupiter.party membership.
875
00:57:43,804 --> 00:57:47,004
The Jupiter.party gives you all the shows and extra bonus content.
876
00:57:47,944 --> 00:57:51,444
And including what you get with the core contributor, which would be the bootleg
877
00:57:51,444 --> 00:57:55,524
version, the ad-free version, of course, and the direct support of the show.
878
00:57:55,724 --> 00:58:01,404
So all of that is available with a big old 15% discount, which is a number that
879
00:58:01,404 --> 00:58:03,504
hurts a little bit, but I love you guys.
880
00:58:03,864 --> 00:58:06,564
You can get signed up, use the promo code bootleg, and thank you to Dan and
881
00:58:06,564 --> 00:58:08,204
Jacob for signing up. Appreciate you.
882
00:58:09,755 --> 00:58:15,275
Well, Brent, we had a listener, Aaron, write in, and he says,
883
00:58:15,475 --> 00:58:19,515
Brent and I were trying to get a flake going where one config file has multiple
884
00:58:19,515 --> 00:58:23,775
host names in it so we could have one file for all the store computers.
885
00:58:23,935 --> 00:58:26,235
If any of you have any examples of this, we'd appreciate it.
886
00:58:26,275 --> 00:58:31,355
The goal is to have a different hardware config automatically detected to automatically
887
00:58:31,355 --> 00:58:34,895
format the disk, also have different host names for each computer they want set.
888
00:58:35,055 --> 00:58:38,595
We're trying to manage these machines from a distance using SSH over tail scale
889
00:58:38,595 --> 00:58:42,515
for updates, installing printers, and simple troubleshooting when employees have issues.
890
00:58:42,775 --> 00:58:46,575
So you want to essentially have a system that figures out its host name and
891
00:58:46,575 --> 00:58:47,915
all of that and configs itself, right?
892
00:58:48,135 --> 00:58:51,015
Yeah, I thought, what the heck, anything is possible with NixOS.
893
00:58:51,275 --> 00:58:56,795
So this is actually a little ask for my brother here, who has a few stores with
894
00:58:56,795 --> 00:59:01,955
a few simple computers. They're just being
hosts for employees who are just trying to use a POS.
895
00:59:02,975 --> 00:59:08,295
And I thought maybe, well, it was over the Christmas holidays when I was feeling
896
00:59:08,295 --> 00:59:12,595
all jazzy, about Nix OS that all these computers can just kind of be the same
897
00:59:12,595 --> 00:59:14,035
and have really the same config.
898
00:59:14,035 --> 00:59:20,195
That's a huge benefit and just be deployed from the comfort of the couch at home.
899
00:59:20,475 --> 00:59:23,975
Yeah, I need to go into the stores to, you know, update these things.
900
00:59:24,215 --> 00:59:30,315
So I was trying to help him get a nice little Nix OS config ready for these.
901
00:59:30,495 --> 00:59:33,735
But I have to say, I hit the edge of my
902
00:59:33,735 --> 00:59:38,515
knowledge in nixos and i'm learning a ton so we use disco to try to get all
903
00:59:38,515 --> 00:59:43,715
the disks identical which worked amazingly we use nixos anywhere with nixos
904
00:59:43,715 --> 00:59:48,155
factor to get things automatically detected and deployed and so we learned a
905
00:59:48,155 --> 00:59:51,615
ton over that christmas break when we deployed all this stuff why.
906
00:59:51,615 --> 00:59:54,475
Are we just hearing about this now what is going anyway go on.
907
00:59:54,475 --> 01:00:01,175
I know right but we hit a bit of a brick wall in trying to make all of these
908
01:00:01,175 --> 01:00:05,095
configs into one that we could just deploy by calling a host name.
909
01:00:05,515 --> 01:00:10,615
So that's where both my brother Aaron and I are saying, hey, dear community,
910
01:00:11,697 --> 01:00:16,417
Where do we go now? How do we do this? How do we deploy this in a way that is,
911
01:00:16,417 --> 01:00:19,557
you know, future intelligent and also best practice?
912
01:00:19,777 --> 01:00:22,157
Yeah, I'd be curious, too, because I've got plenty of machines myself.
913
01:00:22,377 --> 01:00:26,197
I'm just wondering if there's a, do you have any code anywhere folks could look at?
914
01:00:26,297 --> 01:00:29,317
Because sometimes that can be helpful to get an idea of exactly like the state
915
01:00:29,317 --> 01:00:31,637
you're at now and then like where you're trying to go.
916
01:00:31,637 --> 01:00:34,517
Boy, this guy, this guy with the put it up on GitHub.
917
01:00:34,897 --> 01:00:38,977
Please put all of your things on. That's actually a great idea,
918
01:00:39,177 --> 01:00:44,177
Wes. I have been scared up to this point to deploy any of my configs to GitHub
919
01:00:44,177 --> 01:00:46,637
because I don't trust myself to keep myself private.
920
01:00:46,797 --> 01:00:50,137
But maybe you can have a look at it and we can deploy something up there just
921
01:00:50,137 --> 01:00:55,277
so that folks can have a look, see what kind of mess we've created and try to help us clean it up.
922
01:00:55,677 --> 01:00:57,957
Yeah, exactly. You know, we could do some filtering or whatever.
923
01:00:58,037 --> 01:01:00,297
You don't have to share all the private details, but just to give it a rough
924
01:01:00,297 --> 01:01:01,557
idea of like where you're at
925
01:01:01,557 --> 01:01:05,197
so that we don't waste time suggesting stuff that might not be relevant.
926
01:01:05,397 --> 01:01:09,557
Though I am curious, maybe you'll have to try after the show uh i wonder how
927
01:01:09,557 --> 01:01:12,997
our friends the uh so-called thinking machines might do with this one.
928
01:01:12,997 --> 01:01:19,377
That is a great question i had not considered trying that but it's worth a try,
929
01:01:20,597 --> 01:01:24,777
so if anybody has any suggestions we'll have a link to that and uh please help
930
01:01:24,777 --> 01:01:27,317
us we're deep deep in it now.
931
01:01:27,317 --> 01:01:33,477
Well it looks like we got another email this week this is cool because apparently
932
01:01:33,477 --> 01:01:38,157
there's someone taking our podcast around the world as they adventure.
933
01:01:38,157 --> 01:01:41,037
There's a little note here hey there just wanted to say thanks
934
01:01:41,037 --> 01:01:44,097
for making such an awesome linux podcast i'm
935
01:01:44,097 --> 01:01:47,017
currently cycling around the world right now in china headed
936
01:01:47,017 --> 01:01:52,197
towards mongolia and your show makes the long hours on the road so much more
937
01:01:52,197 --> 01:01:57,437
enjoyable you probably already know this but your podcast is truly global i've
938
01:01:57,437 --> 01:02:01,617
been listening to it across many countries in central asia already keep up the
939
01:02:01,617 --> 01:02:06,237
great work and thanks again cheers from javier that.
940
01:02:06,237 --> 01:02:07,737
Is so cool thank.
941
01:02:07,737 --> 01:02:10,757
You so much for writing in um amazing.
942
01:02:11,557 --> 01:02:16,077
Yeah keep us posted and keep spreading the word too that's great thank you everybody
943
01:02:16,077 --> 01:02:18,157
who wrote in linuxunplugged.com contact,
944
01:02:23,158 --> 01:02:27,278
And, of course, you can send a message into the show and support this here show.
945
01:02:27,558 --> 01:02:29,938
And we don't have a baller this week.
946
01:02:30,238 --> 01:02:34,718
You know it's going to be a rough week when our top boost is WH-2250,
947
01:02:34,958 --> 01:02:38,678
easy for me to say, with 5,000 sats.
948
01:02:40,378 --> 01:02:43,418
He says, Chris, now that you have Music Assistant connected to Home Assistant,
949
01:02:43,558 --> 01:02:48,018
the next step is to trigger albums, playlists, and books with NFC tags.
950
01:02:48,258 --> 01:02:50,138
Oh, I'm going to pause right here, WH.
951
01:02:50,918 --> 01:02:53,758
After the episode, I was thinking the same thing. I've got a bag of them.
952
01:02:53,918 --> 01:02:55,958
And I do that now for certain lights.
953
01:02:56,478 --> 01:03:02,698
So at night I can just like swipe an NFC tag. But this is a great idea for albums specifically.
954
01:03:02,958 --> 01:03:05,798
And I think it's
actually pretty spouse approval factor high.
955
01:03:06,038 --> 01:03:09,618
I wouldn't say extreme, wouldn't say best, but it's high to medium.
956
01:03:10,018 --> 01:03:13,958
He goes on to say, I printed a one-fourth scale milk crate to hold my LPs and
957
01:03:13,958 --> 01:03:17,758
scan what I want to listen to. Oh, that's so cool.
958
01:03:18,158 --> 01:03:18,598
Brilliant.
959
01:03:19,038 --> 01:03:22,458
You could create a miniature bookshelf of your audio books and set up a physical
960
01:03:22,458 --> 01:03:24,758
digital library. Oh, my gosh.
961
01:03:25,538 --> 01:03:29,858
If I had the space. That's a really neat idea. I have been very successfully
962
01:03:29,858 --> 01:03:31,998
enjoying Music Assistant.
963
01:03:32,398 --> 01:03:35,258
Last night, thankfully, there were no kids in the rig.
964
01:03:35,398 --> 01:03:38,918
I hit the book button, and for the first time, and I had been playing music
965
01:03:38,918 --> 01:03:41,278
a couple hours earlier on all the speakers, for the first time,
966
01:03:41,518 --> 01:03:44,498
even though my automation system played specifically on one speaker,
967
01:03:44,718 --> 01:03:46,858
it played on all the speakers in the house.
968
01:03:47,178 --> 01:03:47,618
Ooh.
969
01:03:47,618 --> 01:03:50,718
Which if you know the kids are trying to sleep at that particular time would
970
01:03:50,718 --> 01:03:55,698
have been that would have been an automation faux pas it was no big deal for
971
01:03:55,698 --> 01:03:58,798
us because we were the only ones in there but oh man i was embarrassed for a
972
01:03:58,798 --> 01:04:00,618
second so i gotta track that down.
973
01:04:00,618 --> 01:04:13,498
Mjvc boosts in with 4021 sats and um yeah uh across uh four boosts here regard
974
01:04:13,498 --> 01:04:17,158
first one regarding manuals and generally tracking of assets,
975
01:04:17,418 --> 01:04:20,118
I highly recommend Homebox.
976
01:04:20,398 --> 01:04:24,098
I use it to track where things are, where they should be. You can also associate
977
01:04:24,098 --> 01:04:28,898
manuals, price, warranty, info, maintenance, all that stuff. Wow.
978
01:04:29,538 --> 01:04:32,398
Yeah, it looks good. All right. Taking a look at that. Thank you.
979
01:04:32,898 --> 01:04:40,358
Re-Homebox continuing here. Also, a system I've developed is to use a random base 36 ID, e.g.
980
01:04:41,218 --> 01:04:44,958
X2GS as an example here, to refer to boxes, bins, etc.
981
01:04:45,538 --> 01:04:48,438
This frees me from trying to pre-sort things e.g.
982
01:04:48,498 --> 01:04:52,878
All cables should go into this box when i want to find the cable or random accessory
983
01:04:52,878 --> 01:04:57,798
i just look up which bin it's in so you can stop caring about having perfectly
984
01:04:57,798 --> 01:05:02,398
sorted labels for stuff and just let the computer do it wow that is clever that.
985
01:05:02,398 --> 01:05:06,878
Is and also smart because i use my label maker so infrequently that between
986
01:05:06,878 --> 01:05:12,098
times like the batteries are almost dead so it's always like just on the edge yes i go to.
987
01:05:12,098 --> 01:05:12,878
Use it this.
988
01:05:12,878 --> 01:05:13,498
To be a lot better.
989
01:05:14,358 --> 01:05:21,798
They've also sent in for our binary boost, and they did send a binary sub boost here, a 1-0-1-0-sats,
990
01:05:23,520 --> 01:05:27,660
I wanted to browse using parts of the binary solo from Flight of the Concords.
991
01:05:27,840 --> 01:05:30,480
So we've got a link there. Maybe we can go check that out after the show.
992
01:05:30,620 --> 01:05:35,440
Thank you. We always love links and ideas for that. And Flight of the Concords are just great.
993
01:05:36,440 --> 01:05:40,520
And then one last boost here. Finally, regarding note-taking.
994
01:05:40,780 --> 01:05:48,540
A few months back, I switched to ForresterNotes.org. It generates static websites that is, to me,
995
01:05:48,860 --> 01:05:52,980
kind of like Obsidian if it were to embrace hypermedia and designed by formal
996
01:05:52,980 --> 01:05:55,800
methods and programming language people.
997
01:05:56,380 --> 01:05:59,620
And then, yeah, we've got a link here we'll include in the show notes. Fascinating.
998
01:06:00,040 --> 01:06:04,620
That is interesting because often I would like to just share a really basic
999
01:06:04,620 --> 01:06:08,740
website with the wife or a note if she just needs to review something for working
1000
01:06:08,740 --> 01:06:09,600
on something with the RV.
1001
01:06:09,940 --> 01:06:12,740
That's an interesting idea. Thank you. So that's fosternotes.org.
1002
01:06:12,840 --> 01:06:14,520
We'll put a link to that in the show notes.
1003
01:06:15,320 --> 01:06:18,840
Well, we have a boost here from Wine Eagle, 6,000 sets.
1004
01:06:22,000 --> 01:06:26,400
It's across a couple of boosts first starting i understand that zfs being out
1005
01:06:26,400 --> 01:06:31,100
of the kernel is more risky to my boot process but not to my data integrity
1006
01:06:31,100 --> 01:06:34,400
and that's why zfs replication exists,
1007
01:06:35,120 --> 01:06:39,940
that guarantees zfs provides our light years ahead of butterfs butterfest send
1008
01:06:39,940 --> 01:06:46,960
is a week compared to the sanoid stack unlike my arch dkms days nixos also provides
1009
01:06:46,960 --> 01:06:51,580
some nice guarantees I like Butterfess running on a fresh kernel and getting
1010
01:06:51,580 --> 01:06:54,860
bit was probably not a great idea.
1011
01:06:55,100 --> 01:07:00,080
Combus got me to triple boot my MacBook back in 2007 in seventh grade,
1012
01:07:00,260 --> 01:07:02,500
but ZFS got me into self-hosting.
1013
01:07:03,100 --> 01:07:08,380
It's home for me now, and I suggest Nix OS users make three separate datasets,
1014
01:07:08,840 --> 01:07:13,100
one for slash Nix, slash home, and of course, the root.
1015
01:07:13,260 --> 01:07:18,220
So backups avoid the Nix store. also shout out to breeze for
not even needing
1016
01:07:18,220 --> 01:07:23,000
an email super duper no kyc and go podcasting.
1017
01:07:23,740 --> 01:07:25,960
Oh i don't have the podfather clip but uh,
1018
01:07:29,752 --> 01:07:32,552
I think that recommendation of three separate datasets for Nix,
1019
01:07:32,672 --> 01:07:34,352
Home, and Root is clever.
1020
01:07:35,012 --> 01:07:38,772
So all my home systems are ButterFS for the data and the root.
1021
01:07:39,172 --> 01:07:45,212
But here at the studio, we're running Arch with the LTS kernel and ButterFS
1022
01:07:45,212 --> 01:07:47,692
on the root and then ZFS for the dataset.
1023
01:07:48,092 --> 01:07:51,632
And that's a really great combination too. And then you don't really have to
1024
01:07:51,632 --> 01:07:55,712
worry about, you know, some sort of regression hitting the most recent kernels
1025
01:07:55,712 --> 01:07:57,052
because you're on an LTS kernel.
1026
01:07:57,612 --> 01:08:01,952
So you can have both. It can be peanut butter and jelly, and that's just delicious.
1027
01:08:02,472 --> 01:08:06,592
Yeah, also, big shout-out to Breeze, B-R-E-E-Z, a great way to boost the show
1028
01:08:06,592 --> 01:08:09,032
without having to switch podcast apps and no KYC.
1029
01:08:10,032 --> 01:08:13,432
I'd also just point out that for folks, Wine Eagle here mentioned Sanoid,
1030
01:08:13,552 --> 01:08:16,112
but on the Butterf side, there's also projects like Butterback,
1031
01:08:16,152 --> 01:08:17,692
which add some of the same functionality.
1032
01:08:17,812 --> 01:08:20,812
I'm not going to say they're equivalent or anything, but there are some options
1033
01:08:20,812 --> 01:08:24,672
there, some improvements around the sound receive stuff, and just more automation
1034
01:08:24,672 --> 01:08:27,372
on top of the tooling that exists for Butterfess natively.
1035
01:08:27,912 --> 01:08:32,952
Mentad's back with a row of decks. I think I was also hit by that Butterfest bug.
1036
01:08:33,092 --> 01:08:38,312
I had hard powered off my Steam Deck running Bazite, then it wouldn't boot or even mount afterwards.
1037
01:08:38,652 --> 01:08:41,152
Fortunately, standard Butterfest file system fixes brought it back.
1038
01:08:41,372 --> 01:08:45,052
I don't think this will change my use of Butterfest, but it was definitely a shock.
1039
01:08:45,652 --> 01:08:50,192
That's a great in-the-field report. Thank you so much for sending that in.
1040
01:08:50,592 --> 01:08:51,772
Sorry you had to go through that.
1041
01:08:53,152 --> 01:08:57,692
If you feel like sharing, I would be curious to know how you rescued a Steam
1042
01:08:57,692 --> 01:09:00,492
Deck. Did you plug it into like an external monitoring keyboard?
1043
01:09:00,932 --> 01:09:05,612
Did you boot it somehow? Did you have like a dock to share that USB-C port?
1044
01:09:06,272 --> 01:09:09,592
That just seems like that's sort of a worst case scenario. Like I was picturing
1045
01:09:09,592 --> 01:09:14,472
my desktop or laptop where it's pretty easy to pop in a thumb drive and,
1046
01:09:14,532 --> 01:09:15,632
you know, I have a keyboard hooked up.
1047
01:09:15,792 --> 01:09:19,192
So Steam Deck is a little more challenging. Good job. Well done on getting that
1048
01:09:19,192 --> 01:09:20,292
fixed. Thank you for the boost.
1049
01:09:20,492 --> 01:09:24,492
Yeah, how much of the Steam Deck, you know, controls interface work when you're
1050
01:09:24,492 --> 01:09:30,292
in an emergency rescue in it ramfs i don't know fuzzy mistborn comes in with a row of ducks,
1051
01:09:31,839 --> 01:09:37,179
Regarding those Ikea speakers, unfortunately, Ikea and Sonos have now ended
1052
01:09:37,179 --> 01:09:39,879
their partnership and the speakers are discontinued.
1053
01:09:40,139 --> 01:09:43,019
Given Sonos' good history of device support, I'm sure the speakers will continue
1054
01:09:43,019 --> 01:09:47,619
to work for a while, but you'll probably need to buy used ones going forward, just an FYI.
1055
01:09:47,919 --> 01:09:52,659
Also, for podcasts and music assistant, check out G Potter Sync.
1056
01:09:52,759 --> 01:09:54,199
It can even run in Nextcloud.
1057
01:09:54,619 --> 01:09:57,959
Yeah, how about that stack? So you get Nextcloud going, then you get G Potter
1058
01:09:57,959 --> 01:10:00,219
server going, and you connect that to music assistant.
1059
01:10:01,919 --> 01:10:06,159
So regarding those Ikea speakers, which we linked last week, you can still get them.
1060
01:10:06,439 --> 01:10:11,579
So yes, Ikea low-key announced in May that they are discontinuing that partnership.
1061
01:10:12,059 --> 01:10:16,139
And it's sort of a bummer because these speakers are essentially,
1062
01:10:16,399 --> 01:10:17,959
the Sonos equivalent is $300.
1063
01:10:18,419 --> 01:10:22,559
And these speakers are under $200, but they have the Sonos guts in them.
1064
01:10:22,659 --> 01:10:26,539
And the nice thing about that is, is it's all local APIs. So Home Assistant
1065
01:10:26,539 --> 01:10:31,959
should still be able to continue to control it unless there's some massive iteration that I don't foresee.
1066
01:10:32,279 --> 01:10:36,499
But because those speakers don't require the cloud service, they don't even
1067
01:10:36,499 --> 01:10:39,219
require the app, probably beyond initial setup. I'm not sure.
1068
01:10:39,859 --> 01:10:42,999
So I had my wife buy two of them. So she bought two for her clinic.
1069
01:10:43,499 --> 01:10:48,039
And she's liked it so much at home that she's asked for a music assistant setup for her clinic.
1070
01:10:48,219 --> 01:10:53,419
And we're going to put a couple of these speakers in. And this is how long she's had a clinic.
1071
01:10:54,359 --> 01:10:59,199
She's going to finally retire her CD player It's like one of those little portable CD player boomboxes,
1072
01:11:00,442 --> 01:11:04,422
And that's how she's done like the lobby kind of like, you know,
1073
01:11:04,522 --> 01:11:07,162
the music for a decade plus.
1074
01:11:07,442 --> 01:11:09,342
And so we're transitioning her to a music assistant.
1075
01:11:10,182 --> 01:11:13,042
I'm going to give her a couple of buttons to press to, you know,
1076
01:11:13,102 --> 01:11:14,462
play different types of music.
1077
01:11:14,882 --> 01:11:19,862
And then I'm also going to experiment with a Zigbee based. I couldn't find a
1078
01:11:19,862 --> 01:11:22,002
Z-Wave version. A Zigbee based volume knob.
1079
01:11:22,402 --> 01:11:25,982
So a physical knob that she can turn to increase and decrease the volume.
1080
01:11:26,202 --> 01:11:27,182
That seems killer.
1081
01:11:27,662 --> 01:11:32,782
Yeah. Yeah. So, yeah. My advice, Fuzzy, is if you can find those IKEA speakers,
1082
01:11:32,882 --> 01:11:36,742
which we have linked in episode 626, get them while you can.
1083
01:11:36,962 --> 01:11:40,502
You can find them used on eBay, but they're not much cheaper.
1084
01:11:40,842 --> 01:11:43,242
And who knows what people played through them in the past?
1085
01:11:46,302 --> 01:11:52,122
Well, Gene Bean boosted in a row of ducks. He's replying to a question we asked
1086
01:11:52,122 --> 01:11:57,242
in Linux Unplugged 624, the one we talked about tiny PCs. He says,
1087
01:11:57,322 --> 01:12:01,062
hey, I don't use the Etsy's directory for my Nix configs at all.
1088
01:12:01,242 --> 01:12:05,682
I have my Flate cloned under my home directory and include a shell alias for
1089
01:12:05,682 --> 01:12:09,182
updating in my config that references where I checked it out.
1090
01:12:09,802 --> 01:12:14,342
Here's the alias I use on NixOS and a little variant for macOS.
1091
01:12:14,342 --> 01:12:17,782
Any links to a GitHub blob that you can have a look at?
1092
01:12:17,942 --> 01:12:21,682
I would like Gene's eyes on mine because I'm doing that now,
1093
01:12:21,682 --> 01:12:27,422
too, as of, you know, this weekend. but maybe i'm not doing it right so if you
1094
01:12:27,422 --> 01:12:29,922
have a chance gene to look at and let me know i'd appreciate it.
1095
01:12:29,922 --> 01:12:34,382
Yeah and gene's using here um you can specify a dash dash flake argument when
1096
01:12:34,382 --> 01:12:38,802
you do your rebuild and then you can just tell it you know whatever directory you store your flake in.
1097
01:12:38,802 --> 01:12:43,902
Thank you gene nice to hear from you brewer seth is here with 4 242 syats,
1098
01:12:45,962 --> 01:12:50,342
ice is longtime linux and homelab tinker but a new listener hey welcome aboard
1099
01:12:50,342 --> 01:12:54,202
and thank you for boosting you got me into bitcoin and lightning through your
1100
01:12:54,202 --> 01:12:59,742
value for value agus advocacy never had a reason to explore crypto bitcoin before
1101
01:12:59,742 --> 01:13:04,182
this week's project is a self-hosted albihub deployment yes yes,
1102
01:13:07,619 --> 01:13:11,359
Uh, next zap will come from my own lightning node. Thanks for inspiring a new
1103
01:13:11,359 --> 01:13:13,939
project. Isn't it fun? It is a lot of fun.
1104
01:13:14,379 --> 01:13:17,579
That is a great boost to end on. That really cheers me up. Thank you,
1105
01:13:18,199 --> 01:13:19,779
Brewer Seth, for updating us on that.
1106
01:13:19,919 --> 01:13:22,559
It's, it's such an interesting world to explore.
1107
01:13:23,079 --> 01:13:27,479
Um, you know, we, uh, just crossed the three and a half years of boosts on the
1108
01:13:27,479 --> 01:13:28,519
show, if you can believe it.
1109
01:13:30,059 --> 01:13:34,659
And Bitcoin is up 187% since we started doing boosts on the show.
1110
01:13:35,439 --> 01:13:39,679
The BlackRock and some of the other ETFs are now the most successful ETFs that
1111
01:13:39,679 --> 01:13:41,379
are based on Bitcoin in the history.
1112
01:13:42,179 --> 01:13:46,239
And that's open source software. In the three and a half years,
1113
01:13:46,559 --> 01:13:49,779
the shift is unbelievable.
1114
01:13:50,139 --> 01:13:54,759
As of last week, it's now included in 401k plans, if you want.
1115
01:13:55,079 --> 01:13:59,419
The level has really shifted and the Lightning Network has grown significantly.
1116
01:14:00,339 --> 01:14:05,859
Uh so if if you've been you know kind of on the fence i don't think it's going away also.
1117
01:14:05,859 --> 01:14:09,779
Albi hub just keeps getting better and better and better and easier to set up
1118
01:14:09,779 --> 01:14:12,399
and work with too so it's easier to onboard now too i think.
1119
01:14:12,399 --> 01:14:15,499
You know it was really it was it was out there on a ledge three
1120
01:14:15,499 --> 01:14:18,179
and a half years ago uh but it's really it's really gone
1121
01:14:18,179 --> 01:14:20,999
very mainstream speaking of albi hub this week in
1122
01:14:20,999 --> 01:14:24,019
bitcoin we'll have an interview with the co-founder of
1123
01:14:24,019 --> 01:14:26,719
albi coming out on monday and it is
1124
01:14:26,719 --> 01:14:29,699
fascinating what's going on over there there's a lot to talk
1125
01:14:29,699 --> 01:14:32,999
about so go check out thisweekinbitcoin.show on
1126
01:14:32,999 --> 01:14:37,299
monday for a new episode thank you everybody who did boost we didn't have a
1127
01:14:37,299 --> 01:14:43,259
huge episode a huge amount of support for this episode we had 19 of you stream
1128
01:14:43,259 --> 01:14:49,739
and we stacked 21,199 sats and when we bring that together with our boosters who we do appreciate,
1129
01:14:50,739 --> 01:14:56,359
We stacked 50,448 sets Of course that is split between the three of us Editor
1130
01:14:56,359 --> 01:14:58,579
Drew and your podcast app of choice,
1131
01:15:04,928 --> 01:15:13,348
I think maybe, maybe our worst week ever in three and a half years.
1132
01:15:14,008 --> 01:15:17,248
Yeah, the total stack wasn't great. The quality was high. We learned a lot.
1133
01:15:17,328 --> 01:15:19,428
We got some great tips, some things to follow up on.
1134
01:15:19,768 --> 01:15:22,948
And we did get some new members on
board. So maybe some folks who transitioned
1135
01:15:22,948 --> 01:15:25,648
to membership, which we always really appreciate too.
1136
01:15:26,008 --> 01:15:28,568
But the systems are just getting better and better over time.
1137
01:15:28,768 --> 01:15:31,888
Fountain FM is getting easier to use. They self-host it for you.
1138
01:15:32,048 --> 01:15:36,048
The Fiat on-ramping is getting easier. apps like striker and more countries
1139
01:15:36,048 --> 01:15:40,908
and you know so bitcoin like i mentioned earlier is up somewhere between 182
1140
01:15:40,908 --> 01:15:46,048
187 percent since we started three and a half years ago the most generous,
1141
01:15:47,048 --> 01:15:49,948
generous version is the u.s dollar is down 10
1142
01:15:49,948 --> 01:15:52,588
percent in that time but it's likely down a lot more and we
1143
01:15:52,588 --> 01:15:56,128
could not have afforded a 10 percent loss that would have things
1144
01:15:56,128 --> 01:15:59,388
things are so expensive for us these days that would have been right
1145
01:15:59,388 --> 01:16:04,368
on the razor thin margin so the boosters they're still continuing to support
1146
01:16:04,368 --> 01:16:07,748
us so even if even if you're not boosting every episode your boost do continue
1147
01:16:07,748 --> 01:16:10,948
to work for us and if you would like to get into this fountain fm is an easy
1148
01:16:10,948 --> 01:16:14,428
and straightforward way to do it and it continues to go to work for us and we
1149
01:16:14,428 --> 01:16:17,428
really appreciate it i also appreciate those members i.
1150
01:16:17,428 --> 01:16:21,688
Do want to add uh just shout out to hybrid sarcasm and everyone helping out
1151
01:16:21,688 --> 01:16:24,588
in our matrix i mean even just this morning folks were asking about like how
1152
01:16:24,588 --> 01:16:27,528
do i get started with boosting and so if you are.
1153
01:16:27,528 --> 01:16:28,308
Curious about that kind.
1154
01:16:28,308 --> 01:16:31,748
Of thing definitely go hop in the matrix we can link there's an unofficial boost
1155
01:16:31,748 --> 01:16:35,708
support room or just chat in any of the rooms and we'll try to help you out
1156
01:16:35,708 --> 01:16:37,028
and or other folks will help you out.
1157
01:16:40,310 --> 01:16:42,990
But sincerely, too, thank you to everybody who is supporting us through our
1158
01:16:42,990 --> 01:16:46,930
membership. We hope you enjoy the perks and that Autopilot support. That's our foundation.
1159
01:16:49,410 --> 01:16:54,110
Now, we got a couple of picks for you this week. The first one is called FF Share.
1160
01:16:54,390 --> 01:16:58,970
It's an Android app that, as you can probably guess, is using FFmpeg on the
1161
01:16:58,970 --> 01:17:02,910
back end to compress your images, your videos, or your audio files before you
1162
01:17:02,910 --> 01:17:05,270
share them, which is kind of nice.
1163
01:17:05,410 --> 01:17:09,170
But you can also, I thought this was cool, for compatibility reasons,
1164
01:17:09,330 --> 01:17:13,170
just tell it every single video format i throw at you just make it a simple
1165
01:17:13,170 --> 01:17:15,150
mp4 that any stupid player can play.
1166
01:17:15,150 --> 01:17:18,710
Yes right exactly it doesn't have to matter about optimizations you don't have
1167
01:17:18,710 --> 01:17:22,050
to matter how much they know about web m just like just translate it please.
1168
01:17:22,050 --> 01:17:23,870
We have that problem in our group chats.
1169
01:17:23,870 --> 01:17:25,290
Yeah absolutely i mean like.
1170
01:17:25,290 --> 01:17:27,450
Pj drops a web m and nobody can play it.
1171
01:17:27,450 --> 01:17:32,190
I've been the guy who then downloads it on my laptop re-encodes mp4 re-uploads
1172
01:17:32,190 --> 01:17:35,730
it just because it's like ah it's bothering me but i'm just doing it on my phone
1173
01:17:35,730 --> 01:17:40,330
i mean come on yeah right i was just gonna say uh right there at the top of
1174
01:17:40,330 --> 01:17:44,710
their readme get it on fdroid get it on obtainium who needs the play store that's.
1175
01:17:44,710 --> 01:17:45,630
Right i'd love to see that.
1176
01:17:45,630 --> 01:17:48,930
I will say the fdroid version is quite hefty out
1177
01:17:48,930 --> 01:17:51,850
of 54 megs and my very first question was what is
1178
01:17:51,850 --> 01:17:56,470
going on here but thankfully they've written a little explainer says the apk
1179
01:17:56,470 --> 01:18:01,030
downloaded from fdroid is the app full versions with all the features and abi
1180
01:18:01,030 --> 01:18:06,770
compatibility included you can find a choice of smaller apks at the github link
1181
01:18:06,770 --> 01:18:08,970
which you can download using something like Obtainium.
1182
01:18:09,450 --> 01:18:13,910
Ah. Yeah, I would bet, though, they must be bundling FFmpeg, which is no slouch.
1183
01:18:14,210 --> 01:18:14,610
Yeah.
1184
01:18:14,890 --> 01:18:19,590
So they also have a little, just a little checkbox, remove all of your XF metadata
1185
01:18:19,590 --> 01:18:23,210
with, like, you know, your camera and location. You can include that or not.
1186
01:18:23,550 --> 01:18:27,590
I wish every OS had that in the share sheet by default, but this adds that.
1187
01:18:27,770 --> 01:18:31,110
And then for the geekiness factor, you can actually see the CLI output of the
1188
01:18:31,110 --> 01:18:33,590
FFmpeg command it's using right there in the Android app.
1189
01:18:34,410 --> 01:18:37,070
And then when the job's finished, it just brings up a share sheet.
1190
01:18:39,115 --> 01:18:43,615
It's pretty cool. Nice find. FF share link in the show notes for that.
1191
01:18:43,795 --> 01:18:48,935
And then we have, I don't know, maybe the second ever listener submitted pick on the show.
1192
01:18:49,315 --> 01:18:53,215
Alex Gates writes in, the podcasting 2.0 consultant. Hey, fellas,
1193
01:18:53,355 --> 01:18:54,655
I got a pick for you this week.
1194
01:18:55,055 --> 01:19:00,275
I've
been a big fan of envelope style online budgeting made popular by YNAB,
1195
01:19:00,435 --> 01:19:02,055
which stands for You Need a Budget.
1196
01:19:03,095 --> 01:19:07,315
I recently decided to migrate to an open source, self-hosted alternative,
1197
01:19:07,315 --> 01:19:08,815
and I've been really happy.
1198
01:19:09,115 --> 01:19:14,215
It's called Actual Budget. It can be used with or without a server if you just
1199
01:19:14,215 --> 01:19:17,575
want to play around or if you don't need syncing across devices.
1200
01:19:17,815 --> 01:19:19,995
And of course, it has a desktop and web apps.
1201
01:19:20,575 --> 01:19:24,635
Recently, they added integration with a service for bank syncing called Simplefin.
1202
01:19:24,915 --> 01:19:29,075
I think it's worth checking out. And budgeting like this has really changed my life.
1203
01:19:29,135 --> 01:19:33,175
And he links us to the Actual Budget documentation, the Simplefin website,
1204
01:19:33,375 --> 01:19:35,495
and their bridge where you can connect your bank.
1205
01:19:35,615 --> 01:19:40,095
He says, with regards, your podcasting to know consultant alex gates oh.
1206
01:19:40,095 --> 01:19:45,415
This looks like some good stuff i've not heard of simple fin before but uh stands
1207
01:19:45,415 --> 01:19:49,075
for simple financial interchange and makes read-only financial interchange simple
1208
01:19:49,075 --> 01:19:55,255
it's like rss for financial information so hey you can see why yes our pod sage was interested huh.
1209
01:19:55,255 --> 01:20:00,495
I would very much very much like to get something like this running and i would
1210
01:20:00,495 --> 01:20:04,315
for me need to be self-hosted it would need to be something i run on my own
1211
01:20:04,315 --> 01:20:07,335
infrastructure so i i'm going to take a look at this after the show thank you alex.
1212
01:20:07,335 --> 01:20:11,995
I've heard repeatedly good reviews of actual budget this is maybe the fifth
1213
01:20:11,995 --> 01:20:14,955
time i hear this so i think i'm going to dive in as well.
1214
01:20:14,955 --> 01:20:17,875
I always just thought people are saying you need to actually budget man
1215
01:20:17,875 --> 01:20:23,715
you need to actually budget i didn't realize it was a thing um so also i'm going
1216
01:20:23,715 --> 01:20:26,495
to put myself out there boosting tell me how much you hate me now for playing
1217
01:20:26,495 --> 01:20:30,035
around with ai and how i boiled the oceans i'd be willing to hear those arguments
1218
01:20:30,035 --> 01:20:34,375
or if you've had any success like this i'd also like to hear those you know
1219
01:20:34,375 --> 01:20:37,655
boys i think the real like ultimate place for this to go is,
1220
01:20:38,467 --> 01:20:41,747
Is something you would run on your own LAN,
1221
01:20:42,047 --> 01:20:46,227
hopefully that doesn't consume a lot of resources that could help you with this
1222
01:20:46,227 --> 01:20:50,167
kind of stuff that maybe maybe isn't a model that tells me how to deal with
1223
01:20:50,167 --> 01:20:54,627
a foot rash or can tell and can tell me what temperature I need to boil eggs at.
1224
01:20:54,827 --> 01:20:58,447
But maybe it's a model that can tell me how a hyperland config works.
1225
01:20:58,587 --> 01:21:01,447
So it doesn't need to be huge. And it's something maybe I even run on my own
1226
01:21:01,447 --> 01:21:06,607
machine or I run on a local network server or on a Nebula network.
1227
01:21:06,987 --> 01:21:10,387
And I just point these editors at that, right? That to me is where this needs
1228
01:21:10,387 --> 01:21:13,547
to go next for probably folks like us to use this kind of stuff long-term.
1229
01:21:13,987 --> 01:21:16,627
Like, I don't think this is going to be a habit for me. I think this is sort
1230
01:21:16,627 --> 01:21:19,987
of a once and done sort of thing to see if it was possible.
1231
01:21:20,747 --> 01:21:24,067
Yeah, explore the edge of technology, see what you can do with it.
1232
01:21:24,187 --> 01:21:27,147
But you're right, like, you could, you know, you could have all kinds of different variants.
1233
01:21:27,287 --> 01:21:31,627
You could have them, you know, trained per operating system or the tools that you want.
1234
01:21:31,727 --> 01:21:34,487
You could probably also, like, you know, down the future, you could customize
1235
01:21:34,487 --> 01:21:37,447
it. You could retrain it. You could fine-tune it for the stuff that you've actually
1236
01:21:37,447 --> 01:21:39,647
been doing over time. There's so many possibilities.
1237
01:21:40,447 --> 01:21:46,527
And hearing about your adventure, Chris, I caught little glimpses of what customizing
1238
01:21:46,527 --> 01:21:50,887
a desktop might be like in the next, you know, let's call it five years.
1239
01:21:51,307 --> 01:21:54,567
Felt like there were some strengths here that we can't ignore,
1240
01:21:54,567 --> 01:21:58,387
and I want to rice my stuff.
1241
01:21:59,087 --> 01:22:05,067
Yeah, I mean, it was pretty wild. I kind of hope somebody over in the sort of
1242
01:22:05,067 --> 01:22:10,387
bluefin Aurora type distributions or silver blue, somebody does a really,
1243
01:22:10,687 --> 01:22:12,507
really nice opinionated Hyperland,
1244
01:22:13,067 --> 01:22:18,387
like an Oma Archie, or somebody releases a Nix OS based one that is really,
1245
01:22:18,407 --> 01:22:21,267
really opinionated. Maybe people know of some out there also boost that in.
1246
01:22:21,427 --> 01:22:24,647
Did I just hear Brent say he hates rising PCs?
1247
01:22:26,447 --> 01:22:28,487
I hate rising PCs.
1248
01:22:30,247 --> 01:22:32,387
Somebody give me a clip of that I want that,
1249
01:22:33,235 --> 01:22:37,515
All right. So that's it. That's just, you know, as we're kind of wrapping up,
1250
01:22:37,595 --> 01:22:41,695
that's on the back of my mind is this felt more like like an incredible technology
1251
01:22:41,695 --> 01:22:43,375
demonstration, but probably not
1252
01:22:43,375 --> 01:22:47,635
something I'm going to use every day until it's maybe under my control.
1253
01:22:47,635 --> 01:22:50,675
And if it can be limited to my systems or just the things I need to use,
1254
01:22:50,935 --> 01:22:53,455
maybe it's approachable. But let me know what you think out there.
1255
01:22:53,675 --> 01:22:57,995
And of course, we would love it if you joined us live. It just gives it a certain vibe.
1256
01:22:58,155 --> 01:23:01,835
We have that mumble room with those low latency open streams right off the mixer.
1257
01:23:01,835 --> 01:23:04,595
We do it right here every single Sunday.
1258
01:23:08,435 --> 01:23:12,375
Yeah, make it a Tuesday on a Sunday. We start at 10 a.m. Pacific, 1 p.m. Eastern.
1259
01:23:12,635 --> 01:23:16,735
You can get it clack-related in your local time at jupiterbroadcasting.com slash calendar.
1260
01:23:17,035 --> 01:23:21,155
And if you've got an app that supports it, Wes, what do we have in store for them?
1261
01:23:21,795 --> 01:23:26,695
Oh, buddy. Not only do we have chapters, which let you skip right to the content
1262
01:23:26,695 --> 01:23:31,275
that you want, But if you want to dive even deeper, we've got transcripts so
1263
01:23:31,275 --> 01:23:35,695
you can read along, throw it into an LLM, whatever you like.
1264
01:23:35,875 --> 01:23:39,155
So check the feed or a podcasting 2.0 app for that.
1265
01:23:39,915 --> 01:23:44,715
Yep. Have fun. Thanks so much for joining us on this week's episode of Your Unplugged Program.
1266
01:23:44,895 --> 01:23:48,255
And we'll see you right back here next Tuesday, as in Sunday!