WEBVTT 00:00:03.100 --> 00:00:09.300 Hello and welcome to another Retro and React. We've got big news in the tech space industry. 00:00:10.580 --> 00:00:19.060 Yeah, this episode we want to talk about both some anthropic drama that happened about a week ago 00:00:19.500 --> 00:00:24.640 and also some tailwind drama. So we're covering both our bases. We're covering AI and we're 00:00:24.560 --> 00:00:30.280 covering front-end development, which is kind of where Bikeshed lives most of the time. 00:00:32.340 --> 00:00:33.480 Scott, where should we start? 00:00:34.300 --> 00:00:36.880 I think we're going to start with the Anthropic drama. 00:00:38.900 --> 00:00:39.040 Yeah. 00:00:40.180 --> 00:00:46.880 Yeah, for those that aren't terminally online, like Scott and I are, about a week ago or 00:00:47.020 --> 00:00:54.080 so, Anthropic basically put the kibosh on open source or third-party agent harnesses. 00:00:54.480 --> 00:00:59.400 So those are tools like Claude Code, like OpenCode, Clawdbot, like Pi. 00:01:00.580 --> 00:01:01.880 The list goes on and on and on. 00:01:02.460 --> 00:01:05.120 There's probably 20 more being made as we're recording this episode. 00:01:06.980 --> 00:01:10.460 But yeah, Anthropic basically said you should stop, 00:01:12.320 --> 00:01:17.780 or you can't use those tools if you're using one of the Claude subscriptions. 00:01:18.280 --> 00:01:20.340 So the Pro or the Max subscriptions. 00:01:22.200 --> 00:01:30.400 Previously, these tools were kind of doing some shady things to let users leverage those subscriptions and use them. 00:01:31.760 --> 00:01:42.580 We were talking just before the episode, but one example is I think I've seen a few snippets in some code bases that has a prefix prompt that these harnesses would send to Anthropic models and say, 00:01:42.640 --> 00:01:48.860 like you are a Claude Code like you're a great coding tool um and uh and that was like basically 00:01:49.340 --> 00:01:55.320 kind of the check that uh maybe the back end was doing for anthropic to determine if if this is 00:01:55.420 --> 00:01:59.260 actually coming from cloud code or not i think they would they were also like spoofing you know 00:01:59.260 --> 00:02:04.180 you can think of it kind of like you know maybe user agent or some headers on the requests uh to 00:02:04.190 --> 00:02:08.519 try to make it look like the user was using Claude Code but they were actually using something like 00:02:08.539 --> 00:02:09.000 open code. 00:02:11.280 --> 00:02:12.280 Yeah, so anyway, 00:02:12.640 --> 00:02:14.440 Anthropic basically said, nope, that's 00:02:14.440 --> 00:02:16.240 no longer happening. If you want to use those open 00:02:16.600 --> 00:02:18.440 harnesses, then you need to pay for 00:02:19.220 --> 00:02:20.380 the sort of the API billing, 00:02:20.630 --> 00:02:22.040 which is like kind of per token 00:02:22.460 --> 00:02:24.000 billing instead of using the monthly 00:02:25.060 --> 00:02:26.280 subscriptions that Anthropic 00:02:26.440 --> 00:02:26.600 offers. 00:02:28.230 --> 00:02:30.300 That totally covers it. I just want 00:02:30.300 --> 00:02:31.900 to break it down even further. 00:02:32.330 --> 00:02:33.280 So it sounds like 00:02:34.040 --> 00:02:36.180 for the most part, Anthropic has 00:02:36.180 --> 00:02:37.940 a suite of tools. They want you to, 00:02:38.260 --> 00:02:41.160 when you pay for pro or max accounts, 00:02:41.630 --> 00:02:42.900 you use those suite of tools. 00:02:43.130 --> 00:02:45.740 And it sounds like open code in this example 00:02:47.299 --> 00:02:51.000 was essentially, I guess, reverse engineering, 00:02:51.170 --> 00:02:54.220 but hitting the API and kind of saying that it was, 00:02:55.380 --> 00:02:57.240 I'm Claude Code with headers 00:02:57.680 --> 00:03:00.140 and also prompting it with that it's Claude Code 00:03:01.000 --> 00:03:03.760 and basically just using the API directly. 00:03:04.120 --> 00:03:05.660 And it seems like Anthropoc is trying 00:03:05.740 --> 00:03:07.100 to basically put their foot down and say, 00:03:07.160 --> 00:03:13.260 know you use our suite of tools. Something we had talked about, Matt and I, is when we worked 00:03:13.900 --> 00:03:21.360 together at an AI startup, it was really important that you get users sticky into your suite of tools. 00:03:21.580 --> 00:03:31.980 So you start creating multiple tools for users to get into your ecosystem and try to get them 00:03:31.920 --> 00:03:38.440 stuck in your ecosystem this is um a classic way to get people using your products and and get them 00:03:38.700 --> 00:03:38.700 ingrained with your products so you have these customers for a long period of time 00:03:38.701 --> 00:03:38.701 yeah and i 00:03:43.500 --> 00:03:51.020 think i mean i i think honestly with with um opus 4.5 it's maybe not the case but models prior to 00:03:51.140 --> 00:03:54.840 that the there wasn't that much stickiness so i think this is kind of like what you're getting 00:03:54.890 --> 00:04:01.880 at scott where it's like just using for example sonnet or haiku um from anthropic it's like okay 00:04:01.900 --> 00:04:07.520 okay, they're pretty good, but someone could easily switch to maybe GPT 5.2 or something 00:04:08.040 --> 00:04:10.800 and get maybe equal or better results. 00:04:10.910 --> 00:04:17.100 And so from that sort of perspective, Anthropic, I think, wants to or maybe sees and realizes 00:04:17.420 --> 00:04:20.820 that they can't compete just on the model. 00:04:21.109 --> 00:04:25.200 The model itself is not the thing that their customers care about. 00:04:26.540 --> 00:04:31.200 It's the sort of the integrations and the tooling and the capabilities that Anthropic can offer 00:04:31.200 --> 00:04:35.800 on top of it things like deep integration in terms of like the model being the best at calling tools 00:04:35.860 --> 00:04:41.340 for example like that's not just the model it's like also all the other stuff they do on top of it 00:04:41.680 --> 00:04:46.840 um and so i think that's the like the thing that they're trying to like accentuate is that 00:04:47.660 --> 00:04:54.919 their suite of software on top of their offering is what they see the value prop as and so that's 00:04:54.940 --> 00:05:01.540 what they're trying to like um you know maybe like get people more in line with is is like yes if 00:05:01.540 --> 00:05:05.100 you use if you're paying these subscriptions you should be using our software to get the best value 00:05:05.320 --> 00:05:11.960 out of it um i think also like i i don't i don't know if this is actually hard data or not like or 00:05:12.060 --> 00:05:16.980 like if i don't know where i've seen this but i have a sense that anthropic is losing money on 00:05:16.980 --> 00:05:22.780 these subscriptions um and maybe making their money more so on things like enterprise contracts 00:05:22.800 --> 00:05:26.020 which use API billing, I believe, under the hood instead of these subscriptions. 00:05:27.880 --> 00:05:31.980 I think it's just like the subscriptions allow you to get far more tokens out for what you pay 00:05:32.160 --> 00:05:35.260 than if you were to doing the equivalent on their API billing setup. 00:05:36.560 --> 00:05:41.500 And so from that lens, of course, they might be losing money on some people 00:05:41.640 --> 00:05:43.520 that are really pushing their limits to the max. 00:05:44.980 --> 00:05:48.760 Basically, the big thing is they're recognizing that the product is not the model itself. 00:05:49.220 --> 00:06:00.260 And both OpenAI and Anthropic have been kind of really pushing down this path now where, you know, I think it's OpenAI is working on a browser and Anthropic now has a Chrome extension. 00:06:00.590 --> 00:06:04.620 I haven't integrated that, but they have more products like the Claude app. 00:06:05.490 --> 00:06:08.140 So they're really trying to build out a suite of tools, like you're saying. 00:06:09.190 --> 00:06:16.900 But this also, it kind of reminds me a little bit of when we would talk about, like, what is the future of models? 00:06:17.040 --> 00:06:23.300 Is it like there's a lot of little models that do one thing really well or they orchestrate it into one big model? 00:06:23.430 --> 00:06:26.400 I don't understand how Opus is created. 00:06:26.430 --> 00:06:29.520 I haven't looked into that or any of the GPT models. 00:06:29.650 --> 00:06:35.740 I would suspect that maybe they take some of the best parts of smaller models and somehow route them into a larger one. 00:06:36.240 --> 00:06:37.500 But it is just interesting. 00:06:37.740 --> 00:06:50.400 I think we're seeing how these AI companies are trying to build out their product in real time and react to how they can best position themselves to be on top. 00:06:52.240 --> 00:06:59.980 Yeah, in terms of like the model architecture, I don't think Anthropic is doing sort of mixed mixture of agents or mixture of models or anything like that yet. 00:07:01.060 --> 00:07:07.360 it is interesting i feel like maybe every couple of days someone's like dude like like claude just 00:07:07.460 --> 00:07:12.040 dropped off like the quality that i'm getting back from my claude sessions is like terrible now 00:07:12.680 --> 00:07:16.580 they must have pushed some sort of hot fix in the background or whatever that changed it 00:07:17.040 --> 00:07:20.980 and every time i see someone tweet this like the you know the t i don't know i think it's like the 00:07:21.040 --> 00:07:24.540 team lead of Claude Code or at anthropic or something responds saying actually we haven't 00:07:24.620 --> 00:07:28.680 changed anything but share your feedback like we want to figure out like what's causing it um 00:07:29.580 --> 00:07:34.640 i so it's like i don't know it's interesting maybe they're pushing things in the background and like 00:07:35.760 --> 00:07:40.280 and maybe they're not trying to say that or maybe they're not actually and it's just like people 00:07:40.410 --> 00:07:46.220 being sort of delusioned by you know maybe opus had like a really good result on one problem and 00:07:46.220 --> 00:07:49.680 then they changed their prompt slightly and has like a really poorly poor result and it's like 00:07:49.820 --> 00:07:55.280 sort of you know problem exists between the keyboard and the chair kind of issue um i don't 00:07:55.260 --> 00:08:02.580 know but uh yeah i think that sort of covers it the the anthropic you know juicy news the drama 00:08:02.770 --> 00:08:07.480 i think what an interesting aspect of this is like open code i think even prior to this so open code 00:08:07.580 --> 00:08:12.860 for those don't know is like a very popular open source harness a lot like Claude Code but uh you 00:08:12.860 --> 00:08:17.440 know it's fully open source it's built on bun and um and they support a lot of different model 00:08:17.640 --> 00:08:23.319 providers they also offer some free model usage here and there i'm not sure how they do that or 00:08:23.340 --> 00:08:28.760 who they partner with to do that. But anyway, it's been sort of teased for a while now that 00:08:29.780 --> 00:08:33.960 they were going to offer this thing called OpenCodeBlack, which is like their own subscription 00:08:34.550 --> 00:08:39.539 service. I don't know who's backing their inference. As far as I know, they don't run 00:08:40.469 --> 00:08:45.340 their own inference. So I'd be curious to see more details there. But I think they started to open up 00:08:45.460 --> 00:08:51.680 the wait list for that. So if people are looking to move off of Claude Code or move off their Claude 00:08:51.700 --> 00:08:58.600 subscription, that might be the place to go to. I think they have a lot of good positive energy 00:08:58.740 --> 00:09:05.240 and ecosystem at the moment. Let's dig into Tailwind a little bit. I think, Scott, you probably have 00:09:05.240 --> 00:09:08.360 the best perspective of this. I think maybe I stumbled across the thread and shared it your 00:09:08.420 --> 00:09:14.240 way, but then you were doing maybe more digging into it. I'm curious to, for those that aren't 00:09:14.300 --> 00:09:20.420 aware, give us a brief background on what this is. What is this drama? What's happening now? 00:09:20.920 --> 00:09:21.560 You're way too kind. 00:09:23.040 --> 00:09:25.680 On probably seven or nine days ago now, 00:09:26.340 --> 00:09:31.680 Matt sent me a GitHub thread that was kind of going viral on Twitter. 00:09:32.650 --> 00:09:37.540 But essentially someone wanted to add, and I didn't rehash this, 00:09:37.540 --> 00:09:39.840 but basically someone wanted to add a feature that was AI related. 00:09:40.320 --> 00:09:44.460 I think it would allow them to see all of their tokens 00:09:44.720 --> 00:09:48.620 or let the AI know exactly like what they wanted the tailwind output 00:09:48.730 --> 00:09:49.340 to be for something. 00:09:50.400 --> 00:09:52.600 And Adam, is it Waltham? 00:09:54.320 --> 00:09:57.780 Well, Adam, I apologize if I mispronounce his last name, 00:09:57.780 --> 00:09:58.780 but Adam of Tailwind, 00:09:59.720 --> 00:10:02.360 I think he's a co-creator of Tailwind, works there. 00:10:03.040 --> 00:10:08.940 He had a pretty strong response 00:10:09.360 --> 00:10:10.660 about not adding it in at the time, 00:10:11.300 --> 00:10:15.660 which that PR creator was like a little bit taken back, 00:10:15.980 --> 00:10:18.460 but he eventually went on to explain that, 00:10:18.620 --> 00:10:23.420 Hey, we, Tailwind laid off 75% of the company about seven to nine days ago. 00:10:24.300 --> 00:10:31.079 And the way they saw it is because of AI, their traffic website was down about 40%. 00:10:32.180 --> 00:10:35.140 And their overall revenue they saw was down about 80%. 00:10:36.160 --> 00:10:42.480 And their model as an open source software is to drive traffic to their website, their doc site, 00:10:43.140 --> 00:10:45.180 to get people interested in their product 00:10:45.740 --> 00:10:47.620 and then to reach out about their product 00:10:47.860 --> 00:10:48.640 to make their money. 00:10:48.900 --> 00:10:50.420 That's how their open source software 00:10:52.380 --> 00:10:53.160 gets funded essentially, 00:10:54.900 --> 00:10:57.420 which I'm sorry, it's just crazy to me 00:10:57.560 --> 00:10:59.780 that maybe I'm a Tailwind Stan. 00:11:01.200 --> 00:11:02.520 I'm a shill for Tailwind. 00:11:03.140 --> 00:11:05.760 I think it's probably the best CSS solution out there. 00:11:06.480 --> 00:11:10.240 It's hands down and clearly, I got a blog post about it. 00:11:10.860 --> 00:11:12.700 The approach you should take 00:11:13.720 --> 00:11:15.360 in a modern architecture, 00:11:16.280 --> 00:11:16.520 there's almost, 00:11:17.720 --> 00:11:18.560 there's few reasons. 00:11:18.800 --> 00:11:19.680 If you're building something small, 00:11:19.900 --> 00:11:21.140 you don't need the overhead, I guess. 00:11:21.940 --> 00:11:22.660 There's almost no reason 00:11:22.720 --> 00:11:23.860 if you're using Next.js 00:11:24.120 --> 00:11:24.820 or one of these things 00:11:24.820 --> 00:11:25.800 that bundles with it, Waku. 00:11:26.400 --> 00:11:27.540 But essentially, 00:11:28.900 --> 00:11:30.240 it's crazy for me to see 00:11:30.940 --> 00:11:31.560 such a popular, 00:11:31.760 --> 00:11:34.120 it's the most popular CSS solution 00:11:34.420 --> 00:11:35.180 on the web right now, 00:11:36.220 --> 00:11:37.020 have this problem. 00:11:37.819 --> 00:11:38.940 Basically, can't get funding. 00:11:39.400 --> 00:11:41.360 I mean, there's maybe a better way 00:11:41.620 --> 00:11:42.160 for it to get funded. 00:11:42.240 --> 00:11:48.960 people were already talking about how it could essentially get bought and they should be looking 00:11:48.960 --> 00:11:54.380 to get acquired i'm sure the folks there have been looking to get acquired and that's not something 00:11:54.960 --> 00:11:58.800 that they haven't thought about but i was also reading down some reddit threads and they were 00:11:58.900 --> 00:12:06.859 saying uh it's unclear but there there's definitely less than 50 people at at um working on tailwind 00:12:07.260 --> 00:12:12.420 their LinkedIn, I apparently said two to 10 people, but some folks were saying there were literally 00:12:12.680 --> 00:12:17.480 four engineers on it. So it might've just been three engineers are now gone. And there's, 00:12:17.620 --> 00:12:22.580 there's just a single engineer working on it, which is really sad to the best open sourced 00:12:23.480 --> 00:12:27.740 CSS front end solution on the, on the web might have one engineer working on it. 00:12:29.380 --> 00:12:35.140 Got to show Tailwind some love. Got to say, it's like crazy how many things have migrated to it. 00:12:35.180 --> 00:12:36.860 And like, this is the situation. 00:12:37.780 --> 00:12:38.380 It's unfortunate. 00:12:38.990 --> 00:12:46.160 It shows that the open source software model is a difficult model to be, to make money on. 00:12:46.960 --> 00:12:54.840 But also, this is one that really wanted to get, prove usability and gain traction with folks. 00:12:55.030 --> 00:12:56.400 And they definitely have done that. 00:12:57.020 --> 00:13:11.260 But they don't have any gated features that I know of that will stop you from getting more out of them or having a reason to partner with them other than being like a big fan, really believing in the product and wanting to help it grow. 00:13:13.680 --> 00:13:16.580 As you were kind of saying or sort of alluding to, 00:13:17.850 --> 00:13:19.860 one of the other things that they make money from, 00:13:19.950 --> 00:13:22.980 so I think primarily they have two revenue streams as I understand it. 00:13:23.280 --> 00:13:26.000 One is through sponsorship and that you go to GitHub 00:13:26.300 --> 00:13:28.320 or I don't know, whatever service they're partnering with 00:13:28.320 --> 00:13:30.240 and you can say, oh, my company uses this 00:13:30.270 --> 00:13:32.360 and so I want to give you some money to support it. 00:13:33.620 --> 00:13:35.980 But the other revenue stream I think they still offer 00:13:36.260 --> 00:13:40.660 is sort of a paid bundle of custom components 00:13:40.750 --> 00:13:42.780 or pre-made components that are using Tailwind. 00:13:43.860 --> 00:13:49.860 I don't know was it Tailwind Labs was it I don't know I think they offered like sort of a suite of 00:13:50.060 --> 00:13:54.620 components and like sort of different pre-made templates and whatnot I'm blanking on the name 00:13:55.060 --> 00:13:55.060 Tailwind Plus maybe 00:13:55.061 --> 00:13:55.061 yeah something like that and so I think that that was what they're also 00:14:02.490 --> 00:14:06.740 hoping to drive traffic towards in terms of like you go to the Tailwind dock site and you see this 00:14:06.750 --> 00:14:10.680 big banner saying oh like are you trying to build this component like actually like we have one 00:14:10.700 --> 00:14:16.640 pre-made for you and you can buy it here um i think really this is like again you know not to 00:14:16.640 --> 00:14:23.260 make this whole episode about ai but like the cost of code is extremely low now right like 00:14:23.600 --> 00:14:29.160 and uh someone can like an agent can easily write the component you're talking about and you give it 00:14:29.160 --> 00:14:32.820 a screenshot or you could say like i want to design something that looks like this that that's over 00:14:32.880 --> 00:14:38.660 here on this other website like just do it and it could do that now so like offering a bundle of 00:14:38.680 --> 00:14:45.800 pre-made components is not that um competitive anymore right like 10 years ago yes right 100 00:14:46.180 --> 00:14:51.840 in a sense of like it it took a lot of effort to try to write really good high quality components 00:14:52.260 --> 00:14:57.980 accessible components things like that um but obviously you know ShadCN is because like really 00:14:57.990 --> 00:15:03.660 eaten up the market there um but then just generally like these agents are really good at generating 00:15:03.680 --> 00:15:09.760 those writing components. And so I think that's just like, it's hard for them to, yeah, it's hard 00:15:09.840 --> 00:15:09.840 for them to compete on that front anymore. 00:15:09.841 --> 00:15:09.841 Yeah, really well put. I would like to say, 00:15:16.340 --> 00:15:22.320 and I'm in agreement with you, I don't, there's little, I don't want to say there's little value. 00:15:23.500 --> 00:15:29.400 The problem, if you're going to make UI nowadays and you want some value from it, the only way you 00:15:29.420 --> 00:15:34.080 can is to use bleeding edge technology that might not be available in all browsers that 00:15:34.400 --> 00:15:41.120 that all the AI models might not know about yet or it is so intricate and maybe use case specific 00:15:41.660 --> 00:15:47.540 like if you if you prompt AI to build you the perfect button or modal or component you know 00:15:47.700 --> 00:15:52.400 first it probably skips accessibility then you have to tell prompt it to do that that's not too 00:15:52.460 --> 00:15:59.500 much more effort, right? But my point is, um, there are, there are maybe very niche use cases 00:15:59.780 --> 00:16:04.340 where you can get the perfect thing. Uh, and, but there are things also like shot CN out there that 00:16:04.460 --> 00:16:08.920 might just give you that, uh, and you have to like configure it a little bit. So there is like, uh, 00:16:09.260 --> 00:16:14.000 uh, there are some pros to still building some things. There are some pieces that you can still 00:16:14.220 --> 00:16:18.920 add to it. I remember there was a big fad about six, eight months ago where it's like, oh, uh, 00:16:19.160 --> 00:16:21.640 AI can't do 3D things yet. 00:16:21.920 --> 00:16:23.420 Like if you write your CSS in 3D, 00:16:23.860 --> 00:16:25.280 like make 3D buttons, 00:16:25.760 --> 00:16:26.620 AI can't handle it. 00:16:26.920 --> 00:16:27.900 Yeah, but most websites 00:16:28.140 --> 00:16:29.040 don't have 3D buttons, right? 00:16:29.120 --> 00:16:29.880 Like it's cool. 00:16:30.200 --> 00:16:30.860 It's a good point. 00:16:31.340 --> 00:16:33.120 But my point being here, 00:16:33.880 --> 00:16:35.140 there are like these niche cases 00:16:35.680 --> 00:16:37.420 where AI might not be able 00:16:37.420 --> 00:16:38.480 to do it right off the bat. 00:16:38.500 --> 00:16:39.820 You could maybe prompt it. 00:16:39.860 --> 00:16:40.760 It might be more effort. 00:16:41.360 --> 00:16:44.700 But for 95% of the majority, 00:16:45.280 --> 00:16:46.220 AI is going to cover it 00:16:46.220 --> 00:16:47.320 or ShadCN's got you covered 00:16:47.400 --> 00:16:48.340 or there's something out there. 00:16:48.560 --> 00:16:53.160 So you really have to find a niche to make what you're building more valuable. 00:16:53.820 --> 00:16:58.960 Yeah, I think, and yeah, like Tailwind certainly, or like their component offering certainly 00:16:59.320 --> 00:17:00.620 wasn't competing in that niche. 00:17:00.620 --> 00:17:03.480 It was just, you know, as I understand it, again, I've never paid for it. 00:17:03.600 --> 00:17:09.160 So I've only like sort of perused their offerings online or what they sort of market as available 00:17:09.209 --> 00:17:09.560 for sale. 00:17:09.760 --> 00:17:14.100 But it's sort of run-of-the-mill setup and components. 00:17:14.410 --> 00:17:17.979 And like obviously also some like pre-made templates in terms of like landing pages and 00:17:18.000 --> 00:17:23.699 dashboards and things like that but um it uh yeah certainly it's stuff that you could sort of 00:17:24.300 --> 00:17:29.220 uh you know ask claude to do in the background for uh 30 minutes and then you have your dashboard 00:17:29.340 --> 00:17:34.900 built out for example i'm not trying to say that what happened to tailwind is is good obviously it 00:17:35.040 --> 00:17:40.860 sucks that people are getting laid off but um i think it's just like pointing that it's a 00:17:41.620 --> 00:17:45.879 i think it's just a really hard thing to do is try to monetize an open source product 00:17:46.600 --> 00:17:54.480 it's very difficult to do that you know either you you start to split and bifurcate like what 00:17:54.480 --> 00:17:59.840 you're offering and and you offer a free open source thing and then you offer a paid thing and 00:18:00.640 --> 00:18:05.220 you know obviously that's what kind of what they were doing already but i think they also are just 00:18:05.280 --> 00:18:09.719 like sort of so they're both like competing in a very difficult market in terms of offering a paid 00:18:09.760 --> 00:18:16.020 like sort of paid but also free and open source thing but then also competing where these these 00:18:16.280 --> 00:18:22.140 models and these agents are are really really good at being able to replicate what they're they're 00:18:22.240 --> 00:18:25.560 offering for sale uh you know it's like someone came into the market and said actually the thing 00:18:25.740 --> 00:18:30.420 that everyone else was selling at that market is now basically uh you know cents on the dollar to 00:18:30.450 --> 00:18:35.219 buy and so obviously everyone's gonna flee all the other stalls and and go to this new new vendor 00:18:35.540 --> 00:18:40.280 that's maybe a terrible metaphor but um yeah i don't know it's just like a they're in a you know 00:18:40.410 --> 00:18:40.410 stuck between a rock and a hard place 00:18:40.411 --> 00:18:40.411 i just want to piggyback on like calling out um the the tailwind 00:18:47.690 --> 00:18:52.860 uh design system maybe maybe it wasn't promoted even enough uh because we weren't 100 sure of the 00:18:52.860 --> 00:18:57.960 name but um it probably had great tailwind integration but unfortunately like that's not 00:18:57.960 --> 00:19:05.200 a differentiator when ShadCN like it also does right and um like you said you could just prompt 00:19:05.260 --> 00:19:10.200 claw to do that work for you. And like, basically what Tailwind gives you out of the box is like, 00:19:10.380 --> 00:19:15.880 all right, you know, build me, like customize me really simply. So they, they almost are like, 00:19:16.440 --> 00:19:20.120 so good. I don't want to say that they shot themselves in the foot, but, but they're so, 00:19:20.360 --> 00:19:25.580 it's so sad to see something that is such a good open source solution struggle to make money when 00:19:25.880 --> 00:19:31.980 it basically is like, it's almost like too good that it's caused, I don't want to say it's caused 00:19:32.000 --> 00:19:37.120 its own collapse. Like it's so good. Everyone uses it. They're just going to have AI solve the 00:19:37.370 --> 00:19:44.040 problem. And then, um, the folks building it, uh, are in a tough place. It's just hard to see that 00:19:44.140 --> 00:19:49.460 it's a victim of its own success. I think is what I'm, um, the words I'm looking for. Awesome. 00:19:52.100 --> 00:19:57.560 Great. Um, yeah, I think that covers the news that we have for today. Curious to hear, you know, 00:19:57.740 --> 00:20:01.960 for other folks thoughts on these topics and, uh, and let us know if there's something that's 00:20:01.980 --> 00:20:03.860 that we maybe haven't caught on yet 00:20:03.950 --> 00:20:05.740 and we should cover in our future 00:20:06.150 --> 00:20:06.900 retro and react episode. 00:20:07.350 --> 00:20:08.540 Any closing thoughts, Scott? 00:20:08.780 --> 00:20:09.660 Otherwise we can wrap it up. 00:20:10.490 --> 00:20:11.020 Keep coding. 00:20:13.240 --> 00:20:13.780 Keep coding. 00:20:14.340 --> 00:20:14.580 All right. 00:20:15.040 --> 00:20:16.240 Thanks everyone for tuning in this week. 00:20:16.600 --> 00:20:17.780 Remember to leave a review for the episode 00:20:18.040 --> 00:20:19.500 and the show and your favorite podcast network. 00:20:19.830 --> 00:20:21.880 We appreciate 12 and a half stars out of 10. 00:20:22.220 --> 00:20:24.860 Those are the ones that really like bring us joy 00:20:25.080 --> 00:20:26.160 when we see those come through. 00:20:26.920 --> 00:20:28.260 And if the system isn't letting you do 00:20:28.350 --> 00:20:29.160 12 and a half out of 10, 00:20:30.020 --> 00:20:31.680 just send bug reports to them. 00:20:32.060 --> 00:20:32.740 and just keep trying. 00:20:33.400 --> 00:20:34.720 Got to be determined to put in that review. 00:20:35.320 --> 00:20:37.740 Feel free to join the Bikeshed community Discord server as well. 00:20:38.400 --> 00:20:39.840 It's getting a little bit more lively. 00:20:39.990 --> 00:20:42.040 We're trying to bring some more activity 00:20:42.760 --> 00:20:44.940 and sort of fun things there. 00:20:45.430 --> 00:20:48.100 So definitely join it and tune in to see more there. 00:20:48.660 --> 00:20:50.760 You can see a link to join that from our website, 00:20:51.160 --> 00:20:52.020 bikeshedpod.com. 00:20:52.900 --> 00:20:54.300 And additionally, if you enjoyed this episode 00:20:54.430 --> 00:20:55.020 or other episodes, 00:20:55.320 --> 00:20:56.720 please share them with your friends or your enemies. 00:20:57.580 --> 00:20:59.400 Really, we just need and want more listeners 00:20:59.700 --> 00:21:00.780 to feed our egos. 00:21:01.660 --> 00:21:03.480 Great. Thanks, everyone, for tuning in. 00:21:04.030 --> 00:21:05.600 See you next week. Peace out.