JS Party ā€“ Episode #306

New Year's Party šŸŽŠ

with Jerod & the gang

All Episodes

Itā€™s our 5th annual New Yearā€™s party! Jerod & the gang review our predictions from last year, discuss whatā€™s trending in the web world, make a few predictions for 2024 & even set some new resolutions for this year.

Featuring

Sponsors

Socket ā€“ Secure your supply chain and ship with confidence. Install the GitHub app, book a demo or learn more

Fly.io ā€“ The home of Changelog.com ā€” Deploy your apps and databases close to your users. In minutes you can run your Ruby, Go, Node, Deno, Python, or Elixir app (and databases!) all over the world. No ops required. Learn more at fly.io/changelog and check out the speedrun in their docs.

Notes & Links

šŸ“ Edit Notes

Chapters

1 00:00 It's new year's party time, y'all 01:04
2 01:04 Welcome to 2024 02:03
3 03:07 KBall's bad prediction 05:57
4 09:04 Nick's good prediction 07:01
5 16:05 Jerod's good (?) prediction 07:02
6 23:07 Junior engineer concerns 13:40
7 36:47 KBall makes a prediction 02:14
8 39:01 Chris makes a morbid prediction 01:12
9 40:13 Nick makes a prediction 03:39
10 43:51 Amal makes a prediction 04:13
11 48:04 How many smoots? 00:53
12 48:57 Jerod rains on Amal's prediction 01:44
13 50:42 KBall defends Amal's prediction 02:58
14 53:39 Sponsor: Socket 03:24
15 57:03 Resolutions time 00:32
16 57:35 Chris' resolution 03:26
17 1:01:01 Amal's resolution 01:37
18 1:02:38 Nick's resolution 01:46
19 1:04:24 KBall's resolution 01:20
20 1:05:43 Jerod's resolution 00:58
21 1:06:41 A big announcement 02:46
22 1:09:28 Wrapping up 00:43
23 1:10:10 Next up on the pod 01:00

Transcript

šŸ“ Edit Transcript

Changelog

Play the audio to listen along while you enjoy the transcript. šŸŽ§

Threeā€¦ Twoā€¦ Oneā€¦ HAPPY NEW YEAR!! [laughter] Spectacular welcome to 2024 from your friends at JS Party. Iā€™m Jerod, and I am here with Amal. Happy New Year, Amal.

Happy New Year, Happy New Year. Weā€™re having a little too much fun already. Letā€™s just put it this way - this show was supposed to start 30 minutes ago, our recording time anywayā€¦ So weā€™re like ā€“ weā€™ve just been *bleep* about, soā€¦ Happy New Year.

We are going to have some fun today. Kball is also here. Whatā€™s up, Kball?

ā€™m celebrating, and I just learned I have a built-in kazoo.

Maybe even better than b0neskullā€™s, who had a kazoo that he brought with himā€¦ Because thatā€™s Chris. Hey, Chris.

Thanks for bringing that kazoo.

Youā€™re welcome.

Happy New Year, man.

Yeah, itā€™s really exciting to be in 2024 right now.

A-ha. So exciting.

The excitement is palpable. [laughter] Nick is here as wellā€¦ Whatā€™s up, Nick?

Ahoy-hoy. New year, new me. I wear sunglasses now.

He wears his sunglasses at nightā€¦ So he can, so he canā€¦ Throw TypeScript shadeā€¦ [laughter]

Shameless for not using Neovimā€¦

Yeah. Right, right.

Well, you have arrived at our fifth annual JS Party New Yearā€™s celebration. This is a fun way that we like to kick off what is this new year, without us knowing what is this new year. Weā€™re going to guess; we are going to reflect a little, we are going to review past predictions, we are going to perhaps cast forward some predictions, and maybe talk a few of us into being so foolish to make resolutions right here for everyone to hold us to account. So that is the plan. Where should we start? Should we start with a quick lookback at last yearā€™s predictions and see how they held up?

Letā€™s do it!

I was just looking backā€¦ I definitely missed it. But maybe I should re-up it.

Yes. Kballā€™s big prediction - in fact, this one was so bold that it turned into a bet, I think, amongst participantsā€¦ Twitter is gonna go into bankruptcy in ā€™23. [laughter]

Wait, wait, had Elon taken over at that point? I canā€™t remember.

He had. He took over in, I think, November of ā€˜22, and so we were right during the big hubbub, and multiple organizations were suing Twitter, and he had laid off two thirds of the staff, or whatever that number isā€¦ And Kball thought this couldnā€™t possibly continue forward, and yet here we are, a year later, and not bankrupt.

Well, but Twitter is dead.

Not his prediction. His prediction was Twitter will go bankrupt. Actually, Twitter will file for bankruptcy. He should have been more vague; he should have been much more vague.

Twitter will be deadā€¦ Yeah. I mean, I think itā€™s a reminder of a variation on an old thingā€¦ Thereā€™s this saying that says the market can stay irrational far longer than you can stay solvent. In this case, itā€™s Twitter is staying solvent far longer than seems rational.

There you go. Would you like to ā€“ well, I guess Feross would be here to be on the winning side of the bet, but heā€™s not here, so I guess heā€™s still is a loserā€¦ Sorry, Feross. [laughter] Would you like to re-up it? Would you like to double down, maybe go all-in twice and say by 2025, by the end of 2024ā€¦? Or do you think at this point youā€™re wrong?

What was our wager? Did we have a wager?

It was like internet points, or somethingā€¦

Okay. Well, Iā€™ll go double or nothing on my internet points, for sure.

[laughs] Okay. Alright. So weā€™ll see you here next year for Kball to eat crow another time. I think itā€™s going to just continue along. You know, things tend to go slower and steadier. Maybe Elon will land this plane in a controlled fashion. Who knows? Call back to the pre-show. Amal, what do you think? Do you think Twitterā€™s going anywhere? X, f.k.a. Twitterā€¦

Honestly - so I have very close friends of mine who judge me on the daily, whenever I send them links from Twitterā€¦ And I can basically hear their eyeroll. Iā€™ll text them a link and I could hear the eyeroll on the other side, like ā€œOh, are you still on there? When are you gonna get off?ā€ And hereā€™s the thing, I was off. I was off for almost two years. I just came back a few months ago, andā€¦ I think Iā€™ve been having fun. Itā€™s definitely a lot different. It feels like an echo of itself. Itā€™s not the same platform from 2018, where I think for me tech Twitter kind of peakedā€¦ But, you know, itā€™s still the fastest way to get news and reach out to people and share information.

[00:06:06.05] I mean, real talk, I just ā€“ I just needed help from someone the other day on something; this person is like a really big deal on the internet, and I was like ā€œThis person is uniquely suited to help me with my problemā€, and I literally reached out to them, and in two minutes they responded back to me, in 30 minutes we were having a conversation on our cell phones. I mean, it was that quick. And so I feel like weā€™re still ā€“ I donā€™t know whatā€™s going to be that when it comes to kind of news connection and information.

Yeah. Thereā€™s still value there.

Yeah. But ā€“ yeah, do I need to feel like I need to take a shower after using Twitter? Sometimes yes. But am I going to stop using it right away? I think Iā€™m gonna maybe go down with the platform; that might happen for me. But weā€™ll see. Weā€™ll see.

Yeah. I always thought Twitter was a cesspool from the beginning. So for me - you know, does it feel different? I thought for like a while it did; it was like it was too much about the main character being Elon Musk every single day, and I feel like that started to fade slightly, perhaps. And there was ā€“ too much Twitter was talking about Twitter for a long time, and so that was annoyingā€¦ But I feel like as that kind of settled out, and it seems like ā€œOkay, itā€™s going to kind of continue onā€¦ā€ Itā€™s still the place, I think, where the big stories are broken and talked about first. For example, the whole OpenAI Sam Altman debacle, that five-day drama, it pretty much unfolded on Twitter. Now, was it valuable for all of us to be part of that? I donā€™t think so. Did we gain anything there? No, but thatā€™s really where the story unfolded, so it still is a bit of a gravitational pull for stuff like thatā€¦ But it was always for me trashy, and so now it just feels like more trashy, but I was already ready for it, so I find value where I can find valueā€¦ And I roll my eyes, maybe more than I used to.

I was wearing high heels over the trash. Like, I felt like that bougie woman that was wearing a fur coat and some high heels, and like stepping over trash, like ā€œEwā€¦ Get out of my way.ā€ Like, thatā€™s how I treat Twitter. I definitely am selective what I engage with and who I followā€¦ And you have to really curate it heavily. Otherwise, you will go nuts. Thereā€™s a lot of triggering stuff there. Itā€™s meant to be triggering; the whole point of the platform was to invoke rage.

Yeah, exactly.

But yeah, you were saying, Nickā€¦?

I said you definitely grok it as a callback to the ā€“

Ohā€¦

[unintelligible 00:08:31.26]

Good one.

Oh, my Godā€¦ [laughs]

You know itā€™s a good callback when you have to explain it.

For the record, for everyone listening, Nick is still wearing his sunglassesā€¦

I donā€™t think he has any plans for taking them off.

Well, he was rocking out like ā€“

They look like EDM DJ Ultra Music Festival sunglassesā€¦

The best part is I can see myself in his glasses now, so now Iā€™m staring deep into his eyesā€¦

I can see the whole monitor reflectionā€¦ [laughs]

Iā€™m very aware of whatā€™s on my screen nowā€¦

Yeahā€¦ [laughs] We can tell when youā€™re not paying attention.

Well, Nickā€™s jokes may crash and burn, but his prediction from last year wasnā€™t as bad as Kballā€™s. Nick predicted that we have Bun, we have Deno, we have Nodeā€¦ His prediction would be that Deno and Bun end up making Node better, rather than one of them supplanting Node. What do you guys think of that prediction and how thatā€™s played out this last year?

I have so many thoughts on thisā€¦ Yeah. Like, I just want to say, Nick, that was a really good prediction. Congratulations.

Thank you.

ā€œCongratulationsā€¦ā€ [laughs]

That is definitely happening. Iā€™ll actually link to some discussions on Twitter, the forbidden platform. Weā€™ll put some links in the show notes to some conversations I was kind of a sidecar on, where this exact thing happened, where Jared was like ā€“

Jared Sumner.

Yeah, Jared Sumner, to be clear; not Jerod Santo.

Just for clarity purposes.

[00:09:55.13] Yeah, Jared found this area of Node that was slow, and then the maintainers were like ā€“ or I take that back. It wasnā€™t that he found it, itā€™s that somebody when they were doing benchmarking found this new areaā€¦ And then the maintainers looked into it, and it was like ā€œOh, it has to do with some old dependency that hadnā€™t been updatedā€, and yada-yada-yada. Itā€™s just fascinating how these discussions are happening in the community now, and Node is getting better as a resultā€¦ Iā€™m very, very happy about that.

Chris, your take on this sentiment. Do you think itā€™s true? Are you happy about it? Youā€™re a ā€“ I think of you as a Node guy. Iā€™m not sure if thatā€™s the right way to think of you, but I very much associate you with Node.js. What are your thoughts on this?

Yeah, I mean, I think the same thing is gonna happen; you know, after Yarn came out, it pushed Npm to add some features and stuff that it didnā€™t haveā€¦ And it made Npm better, and I think the same thing is gonna happen with Node.

I agree with all of that.

Of course you do, itā€™s your prediction.

And this is why itā€™s good. Itā€™s healthy. Itā€™s so healthy. This is what you want. Otherwise, you have stagnation. I think competition is really healthy. So itā€™s important for us to kind of have thatā€¦

Absolutely. And thereā€™s a lot of new cool things coming to Node proper, that have been in Bun or Deno before. One that I can think of off the top of my head is specific permissions. You can disable networking, or disable file system access, probablyā€¦ And thatā€™s something that Deno - like, that was there from the outset; you can only run this in a secure way, and get access to the things that you need access to, and nothing else. I think thatā€™s really cool. And thereā€™s a huge push towards more secure Node modules, too. Or maybe thatā€™s just the Feross propaganda in the back of my headā€¦

[laughs] Very true. I agree with all that. I think itā€™s kind of clichĆ© at this point, but competition breeds innovation, and lack of competition just creates stagnationā€¦ And itā€™s awesome to see new entrants come in, and projects get a new reason to do awesome things, and new ideas, like ā€œOh, we didnā€™t think about doing it that wayā€, or ā€œWe have these constraints, you donā€™t. Maybe we could change one of those and modify that we work to get 80% of that, versus the 100% that you have.ā€ But itā€™s all good to see. Whether or not Deno and/or Bun will actually gain a significant foothold on like user base, I donā€™t really have much prediction there. I think probably Bun has a better shot, because Denoā€™s more differentā€¦ But gosh, who knows? Time will tell. And Node has such a huge user base right now that it would be an upstart browser trying to take browser share from Chromeā€¦ Which could happen, butā€¦

The Bing versus Google Search analogy I think is [unintelligible 00:12:43.01] in terms of scale. Most projects in the wild use Node, period, for lots of different things.

On some of the shows this year, when we talked about these different things, one of the things we talked about is that you donā€™t have to have Node versus Bun in your decision-making. You can use Bun in particular parts of your deployment, where performance is particularly importantā€¦ And other deployments that are more focused on backwards-compatibility, or need certain features that it doesnā€™t implement, just donā€™t use it. It can become a tool in the toolchest without having to supplant Node.

Yeah. I do have to say though, for me as a tech lead - or former tech lead at this point - I would say the fear I have when I hear about that is it just kind ofā€¦ Like, I think really hard about introducing new things into my stack, and whatā€™s the benefit that you get versus the consā€¦ And for me, the cons usually outweigh. And so if I can get 80 with the existing tools that we have, then itā€™s worth the compromise, because thereā€™s lots of other better ways to spend that engineering time.

Not to cross-promote hereā€¦ I donā€™t want to cross-promote things that Jerod does, that arenā€™t TypeScript-relatedā€¦ But that episode of Changelog & Friends with Searles about - I think it was It Dependenciesā€¦ Itā€™s gotten me constantly thinking about ā€œDo I really need to add this new dependency?ā€

[00:14:05.13] Oh, I love that.

Thatā€™s good.

Yeah, that was a great show.

I have to listen to that.

I love that we have you thinking that way. I think itā€™s a healthy way of thinking. It doesnā€™t mean you donā€™t end up adding it, it just means that you stopped to consider it, and you weigh the pros and cons, and you think ā€œHmā€¦ This has a total cost that I need to consider before I add it.ā€ And maybe you end up doing it anyways, but at least now youā€™re taking intentional steps with your codebase.

Yeah. And I think for me, just again, to give some more context to where my kind of extreme trepidation or what seems like a very conservative answer - Iā€™ll give you some contextā€¦ Itā€™s not just me and my team and my engineers that have that cost. Itā€™s like our ops team, our security teamā€¦ Thereā€™s a huge ripple of people in different parts of the company that are impacted when you introduce large new technology into the stack. So I just want to put that out there. Like, itā€™s not just devs. Itā€™s SRE, itā€™s ops, itā€™s securityā€¦ And for teams itā€™s platform teams, itā€™s all kinds of stuff. So just think about it that way.

Yeah, I would also strongly consider the licensing and ownership of the projectā€¦ Because stuff like Bun and Deno, they could pull those permissive licenses if they want, right?

And there might be an incentive or pressure to do soā€¦ Which if youā€™re a small company, maybe youā€™re not going to be able to afford it, or something. And so youā€™ve really got to be careful about the ā€œWhere does your open source come from when youā€™re picking open source?ā€

Yeah, well said. Weā€™ve definitely seen prior art now, of that not just being a hypothetical situation that may happen to you. Itā€™s actually happened on more than one occasion of a relicensing of a majorly dependent upon open source software, until weā€™re suddenly - everybody needs to sit down and decide how are they going to forge a path forward given these licensing changes. So itā€™s no small thing, even if itā€™s a small package that you rely upon.

Well, letā€™s now turn the focus to me. I had a prediction last year; I donā€™t know, you guys be the judge. I feel like itā€™s kind of happened, but not 100%ā€¦ I predicted - and remember, this was the end of 2002. Weā€™d just had the Twitter takeover, weā€™d just had Chat GPT end of November. The AI Revolution had begun. The buzz was peaking. Maybe not peaking - itā€™s gone higher since then - but what I said was that I felt we would see AI plateau in the second half of the year. I thought it would continue to improve, but not at the same pace that it had improved up until then. I specifically said weā€™ll still be writing our own code this time this year. What do you think? How did I fare?

Well, according to the advertisement that I see constantly on GitHub, theyā€™re not writing 55% of my code.

[laughs]

I thought it was 40-something last time I checkedā€¦

Iā€™ve seen that one, too. Theyā€™re really shoving thatā€¦

Maybe theyā€™re A/B testing that numberā€¦

[laughs]

Well, no, no, I heard that 40% of code on GitHub now is generated by AI, but I donā€™t know if thatā€™s actually true. That sounds ridiculous if thatā€™s the case.

Oh, youā€™re right. It says ā€œMake me 55% faster.ā€

Thatā€™s cool.

I think thereā€™s probably a research study that went behind that number, but I think other numbers are most likely misinformation flying around at a paceā€¦ Because itā€™s very popular right now to tell people how good you are at coding, now that Copilotā€™s writing your codeā€¦ But I donā€™t know, most of the demos to me donā€™t look like they could stand [unintelligible 00:17:25.11] for more than the 30 seconds they were recorded onā€¦

70% of the time it makes me faster 100% of the time.

I was gonna say, I definitely experimented with some of that, and took it out of my config, because I did not find it overall sped me up.

Yeah, same here. I think for me itā€™s only useful for boilerplate code. Other than that, Iā€™m always refactoring around it. But itā€™s really great for learning a new language, or a new tool, which is how Iā€™ve been using itā€¦ Because Iā€™m learning C++. I learned it a long, long time ago, and I just blocked everything I knew out of my brainā€¦ So Iā€™m relearning that; so thatā€™s been helpful.

[00:18:07.25] I will put an asterisk on that; it is really good for learning an old tool that is new to you. If you try to use it to learn an actually newly developed tool, it will mislead you all over the place.

Yes. Lack of data.

Yeah, lack of data. Right. But no, I think for me itā€™s been a good learning aid, but when using it in a context where Iā€™m an expert, itā€™s only useful for boilerplate code, or helping me come up with regexā€™es, which is how Iā€™ve also been using it quite a bit.

Now, are yā€™all talking about specifically Copilot, or are you also lumping in like talking to ChatGPT, or Bard orā€¦?

I find ChatGPT to be more useful as a ā€œpersonā€ that I go ask questions to, versus anything that generates code in-line inside my editor. When I started type, even - Iā€™ve never been a huge fan of IntelliSense, which I know makes me like a very small minority of developersā€¦ Itā€™s great if youā€™re trying to know what functions are available in this scope. I like it for that. But anytime Iā€™m like halfway typing something and stop, itā€™s not usually because Iā€™m trying to figure out, itā€™s just because Iā€™m pausing my brainā€¦ And then what Iā€™ve found specifically with Copilot, but Iā€™ve also tried Cody, and Tabnine - this was probably two years ago, so Iā€™m not sure the current state of Tabnine. If I do it that way, it just puts out like ā€œYou should autocomplete thisā€, and itā€™s almost never what I want, at leastā€¦ Even if it is, Iā€™m stopping and then Iā€™ll think about code that I wasnā€™t caring about before, and now Iā€™m like thinking about something else. And so it takes me out of my flow. And maybe that just takes time to get used toā€¦ But I like to go somewhere; instead of going to Google, almost 9 times out of 10 I will go to ChatGPT and say ā€œWrite this regex for me. Write this [unintelligible 00:19:46.25] And for that class of problem, where I already know how to solve it, I just donā€™t want to, I find it to be a big productivity boost. But Iā€™d rather go to it than have it integrated. You guys?

Thatā€™s 100% where I find the benefit of Copilot. Like you said, it will generate crap. And then I feel like it pulls me outā€¦ Because I already know what I want to write, and itā€™s trying to help me by autocompleting thatā€¦ But then more times than not, it just ā€“ I have to pause my train of thought, read over what [unintelligible 00:20:14.24] and then either decide yes or no. And if itā€™s not right, then I have to get back to my train of thought, which is not always foolproofā€¦

Rightā€¦? Failsafe?

Yeah. But the only place where Iā€™m totally like ā€œYes, just generate, and I will accept yours, and then try and tweak that from thereā€ is unit testing. Itā€™s actually pretty darn good at generating unit tests.

Absolutely. Agreed on that.

Makes sense. If youā€™re into writing tests. Chris, what do you think?

Yeah. I mean, Iā€™ve I think Iā€™ve had good success with Copilot. I havenā€™t tried to use ChatGPT directly for coding, but itā€™s cool that they have ā€“ I donā€™t know what they do now, but it has some sort of understanding of the project youā€™re working in. And so I would be like - Iā€™d select this class, and be like ā€œOkay, refactor this class in this particular wayā€, that would be tedious to do manually, and it does it. And it uses my types, and everything; it tries to keep the same style. And with unit tests and testing in particular, it knows Iā€™m using a weird, a non-popular, just kind of an obscure assertion library. But it knows all the assertions Iā€™ve written before, and itā€™s able to figure out what to use. So itā€™s not like Iā€™m asking it to write unit tests and itā€™s going and using try syntax, or [unintelligible 00:21:45.06] syntax, or somethingā€¦ Because it doesnā€™t; it knows what Iā€™m using in it, and it does it correctly four out of five times.

[00:21:55.02] I think this is kind of the natural process of maturation of a new set of capabilities that was developed, and weā€™re trying to learn the best ways to use them. And I think, for example, writing unit tests, doing other things - like, that is a productivity boost. I do think thereā€™s a lot of hype cycles around the productivity boost that you get from writing new code, that I think are baloneyā€¦ I donā€™t think itā€™s actually ā€“ I mean, to Amalā€™s point, itā€™s really useful when writing boilerplate. And if youā€™re writing a set of boilerplate scripts, or youā€™re trying to do little things to push data from one service to another, I think it is a dramatic performance boost there.

I have definitely seen, particularly like a junior using code completion, and then getting things where it has to go through many cycles of review and feedback, because itā€™s not matching styles, or other things from the rest of the projectā€¦ And I think itā€™s a performance cost. I mean, I think we need to figure out where itā€™s useful and where itā€™s not, but some of that is getting past the hype cycle and saying ā€œOkay, itā€™s a useful tool in these scenarios, and it is not a useful tool in these other scenarios, and we should stop trying.ā€

Yeah, yeah, absolutely. I couldnā€™t agree with you more, Kball. And thank you so much for bringing up the juniors thing, because I have thoughts on this, Jerod, which I really want to share on airā€¦ So Iā€™m really, really concerned about the arc of our junior engineers, actuallyā€¦ Because I think remote work has been terrible for them. Iā€™ve managed and led junior engineers in this remote environment, and trust me, itā€™s not the same as being in an office with them, where they can just walk over to you, and they get that ambient learning from their team.

So I think remote work isnā€™t the greatest. It does provide access to work, but itā€™s not the greatest when you are very ā€“ you know, in the first one to three years of your career. The other thing is, AI - I look at tools like [unintelligible 00:23:50.24] VZero from Vercel, and all these other kind of code generation AI tools that are specific to kind of developing UI components, etc. - this is fine. I mean, this is the work that - listen, as somebody who has been in this industry for, I donā€™t know, 10 plus years, I donā€™t want toā€¦ This is not work that I want to do. This is actually work that I have not been doing for years. I canā€™t remember the last time a company paid me money to make a component. You know what I mean? Like, Iā€™m managing infra, and architecture, and all these other thingsā€¦ And whoā€™s writing components? Our junior engineers. Theyā€™re learning the ropes, theyā€™re doing the gruntwork, so that they can kind of get their reps in, the repetition, so that they understand how these systems are connected together. Theyā€™re the little building blocks, right?

And so Iā€™m just really worried about what does the future look like for them? How are we going to have ā€“ are we going to even have junior engineers? Because AI is doing the work of what I would consider junior people, and I think starting at a higher level of just ā€“ I donā€™t know; I donā€™t know if that is really doing them justice. I donā€™t know. Maybe this is an irrational fear, I have no ideaā€¦ But Iā€™m just saying Iā€™m worried.

Amal, you have nothing to worry about, because the AI companies have this all figured out; if you ask them, there wonā€™t be any engineers in five years. Soā€¦ Problem solved.

Oh, goshā€¦ Iā€™m just saying itā€™s a concern, you guys. And this is an us problem, it affects our industry; we should be figuring this out, and making sure we have a path for people at all levels, senior to junior. And I just want to make sure that people can progress those paths in a way that is conducive, and healthy, and sustainable. And Iā€™m not saying that even being a junior engineer ā€“ some people get through junior in six months, and theyā€™re done. And theyā€™ve moved on to mid. Some people are there for a few years. So itā€™s not even something thatā€™s like a universal standard, or timeline, or whateverā€¦ But Iā€™m just saying, I just want to make sure that we have a path for people to learn, effectively, so that they can manage these complex systems. And when things go sideways on the AI written code, they know how to debug at 3am, if they need toā€¦ Because thatā€™s what happens. Incidents happen.

[00:26:01.16] Well, I think Iā€™m with you. I think that thinking back to when I was a junior, I didnā€™t know how to debug at 3am either, until I had to debug at 3am, and then I had to figure it out. So I think that still all those same processes apply, just perhaps maybe in scarier waysā€¦ I donā€™t know; maybe it was pretty scary back then for me as well.

I think itā€™s more that youā€™re debugging at 3am as a senior, or mid-level person, or above. Itā€™s just that like ā€“

No, no, I mean, when I was a junior, I said; back when I was a junior. How I learned to do it was by doing it.

Oh, I see.

Thereā€™s certain things that you learn to do by doing. And you canā€™t actually impart that to somebody until they go through the fire, hopefully with somebody alongside them whoā€™s been there before, because itā€™s a lot better that wayā€¦ But I donā€™t know, The Kids Are Alright.

I hope so. I hope so. Iā€™m so worried about this younger generation. I feel so privileged and blessed to be in my mid to late 30sā€¦ [laughter] Inside joke thereā€¦

It is concerningā€¦ Yeah, I do have concerns, because I use these tools every day; I go to ChatGPT before I go to Google now, almost 100% of the time.

[laughs] We should make that a T-shirt.

But I also have ā€“ almost every single time, Iā€™m asking it to help me write this thing, and then I have to literallyā€¦ I either literally copy and paste the error that I got, or Iā€™m like ā€œThat method doesnā€™t exist. What does it always do? Youā€™re correct, that doesnā€™t exist! Let me correct that, and go for [unintelligible 00:27:30.00] And having to just constantly do that as a junior, where you either have ā€“ you have to do something, and itā€™s constantly giving you this BS all the timeā€¦ It would feel overwhelming.

Yeah, but think of the alternative. Go back, Nick, when you were writing C for the first time, in like Intro to Computer Science class, or whatever language youā€™re writing. Java. And you would type out the thing and youā€™d say ā€œOkay, GCC, whatever, compile this.ā€ And it would be like ā€œCompilation error.ā€ And it would just stare at you. And youā€™d be like ā€œUhhhā€¦ Okayā€¦ā€ It wouldnā€™t give you anything besides maybe chunking out a whole bunch of binary-looking crapā€¦ And youā€™d be like ā€œWhat do I ā€“ā€ Think what the kids can do right now. They take the output, they put it to ChatGPT, they say ā€œThis didnā€™t workā€, and it says ā€œYou are correct. Try this.ā€ And then they try that.

And so like, we were just out there ā€“ I mean, Iā€™ll speak for myself, because Iā€™m older than you, Nick. I was on an island when I was learning. I didnā€™t have any help at all. Pretty much. I had a book that didnā€™t change the words when things changed. And I had a compiler that didnā€™t really say much when it wouldnā€™t compile. And we all survived. So I donā€™t knowā€¦ The Kids Are Alright, guys. Theyā€™ll be alright.

So I think there is a key thing hereā€¦ Amal, you brought this up and you said ā€œI donā€™t know if this is just me.ā€ Itā€™s definitely not just you. I think this is a real concern. Itā€™s a concern that happens anytime technology changes.

Weā€™re not talking about it, though, Kball.

Well, I mean, you say weā€™re not talking about itā€¦ It was the subject of one of the big talks at a conference I went to.

Okay, thatā€™s good.

There is some amount of this in the discourse. It was a management-focused conference, so it may not be ā€“ I donā€™t know. But yeah, there was definitely a question. Now, the speaker in that case was sort of pushing his book; not his physical book, but like his stance, because he runs a training companyā€¦ And he was saying, ā€œOh, if youā€™re not going to learn these things through the course of doing anymore because of AI, we need to do more formal training.ā€ That is one angle to this. Like, if you think about calculators, we force kids to practice doing calculations without calculators, because it helps them start to learn the underlying pieces, even though at some point they pretty much start using calculators for everything.

There may be an analog to what we need to do in software. Maybe itā€™s more formal training, maybe itā€™s something else, but there may be something where we say ā€œIf you are learning, you donā€™t use these tools, even though they would help you, because you need to learn the underlying fundamentals before you start applying these tools.ā€

[00:30:08.01] And we separate out the doing for learning from the doing for doingā€¦ Which is not something that happens right now a lot in entry-level positions. Thereā€™s a lot of ā€œYouā€™re doing this thing, and youā€™re expected to learn on the job, and kind of go.ā€ Maybe we need to formalize the separation a little bit more, of like ā€œHereā€™s a set of things you do to learn principles, and hereā€™s a set of things you do to get the job done.ā€

In college, all I learned was principles, and I didnā€™t know anything ā€“ I didnā€™t learn anything about doing a job. And you read articles where employers say thatā€™s a huge problem about college grads, where itā€™s just theory and first principles, and they donā€™t know how to ship software.

Yeah. But yeah, I think the applied part of engineering is youā€™ve gotta go to a tech school for that. [laughs]

Well, yeah, but I mean, if you go through college and you havenā€™t used AI ā€“ Iā€™m sure you do now, butā€¦ If you donā€™t know how to do it, and theyā€™re gonna expect you to know how to write a prompt, I meanā€¦ I donā€™t know.

Yeah, no, I hear you. Iā€™m sorry, and just ā€“ Iā€™m very biased towards tech schools, because I went to one. I didnā€™t graduate. Thatā€™s like its own show, fun backstory thereā€¦ But I did go to a tech school, and I felt like they are the only ones that I saw doing a ton of like hands-on learning projects. They make sure you can do the thing that you learned. Iā€™d love to see more universities doā€¦

Well, the problem with college isnā€™t merely that you donā€™t get applied education, itā€™s that you end up with a giant heaping debt that you never pay off your entire life. And itā€™s the only thing you cannot file bankruptcy against. Thatā€™s the problem with college. But weā€™re on a whole other podcast talking about that.

I do wonder, coming back to the generative AI thing, if part of the challenge right now - and this is coming in from the chat, but the challenge right now is sort of this assumption pushed by people who have a lot of investment in AI, that AI can actually replace people coding. And theyā€™re pushing ā€“ you have all these studies, ā€œOh, itā€™s improving your productivity, itā€™s doing all thisā€, and so everybody thinks ā€œOh, Iā€™ve got to do this or Iā€™m falling behind.ā€ And as we were highlighting, for some problems itā€™s a big help, and for other problems, itā€™s crap. And itā€™s not helping you get more done, and at the same time, itā€™s inhibiting your ability to learn.

So I think some of this is pushed by people like the Sam Altmans of the world, who - they have an agenda. And itā€™s very clear that they have an agenda, but theyā€™re kind of taking thisā€¦ Itā€™s like the NFT boom. ā€œThis is gonna replace all these things. This is gonna be doing these things.ā€ Well, thereā€™s more value in AI than I think there was in crypto, but thereā€™s also a whole lot of junk.

And the question is, that none of us know the answer to, is like ā€œWhere will the pace of progress take us over n years?ā€ Which goes back to my prediction; I predicted weā€™d have another plateau. I havenā€™t seen a step change like we saw with transformers since thenā€¦ And now weā€™re just kind of like at this new plane, where we can do generative AI, now we have these tools, and now weā€™re gonna integrate them, weā€™re gonna test them, weā€™re gonna figure them out, weā€™re gonna have these minor improvements to our work and life, that donā€™t replace humans at all. The question is, will the next step get us to a place there, or not? Or is it or is there going to be another one? And when will those step changes happen? This last one took about five years. Is there going to be another one at the pace of ā€“ that was what the buzz was about back in the spring, was like ā€œLook how fast weā€™re moving now. Itā€™s going to just hit the stratosphere from here.ā€ We have seen that not take place. We are learning how these have improved our lives, but now weā€™re seeing where they stop, and the question is [unintelligible 00:33:54.09] in five years there wonā€™t be any engineers; like, that could be true. I donā€™t think it is, I donā€™t think the history has played out that way. Past results do not predict future performance. But maybe these people are right that you wonā€™t need humans in the loop in 5 to 10 years, and the AI will do it all. I think all five of us are probably on the betting against that, but weā€™ve been wrong before, havenā€™t we, Kball? [laughter]

[00:34:20.13] I think that thereā€™s less of a concern of it fully replacing us, but I do think that there is some truth to the idea that humans who use the AI well will outpace the humans that donā€™t.

And that might lead to ā€“ I mean, Iā€™ve heard the argument both ways. That might lead to less overall jobs, because you can really have a ā€œ10x developerā€, whoā€™s 1x plus the 9x AIā€¦ Or is it gonna be like the ā€“ another argument I saw was the push from assembly languages to higher level languages didnā€™t make it so that there were fewer programmers because it was easier to program, it made it so there were way more programmers.

Yeah, because thereā€™s more opportunity, and cheaper ā€“ more people can play the game. Yeah.

Yeah. I mean, I think thereā€™s a huge swath of businesses whose economics donā€™t make sense at the current cost of software development, that might make sense if you reduce the cost of software development by 10x, or 5x, or something in those domains.

Thereā€™s a huge swath of businesses that exist today, whose economics donā€™t make sense based on software development, and they only are alive because capital was so cheap, and weā€™re learning that.

But maybe those sure businesses do make sense in a world where AI makes software developers 10 times more productive.

I just wish those businesses would die. Iā€™m sorry to say that. And it sounds so harsh, but really, itā€™s really bad for all of us. Because what it does is it creates this churn, yā€™all; all these incredible, smart, kind human beings go work at these companies with a hope and a dream, and then they get laid off, and it just creates churn. And all they want for Christmas, New Yearā€™s, whatever, is stability tech. Stop creating churn in our community; stop it. It sucks.

Itā€™s a natural process though, of all creation, is churn, and life and death, and things survive and things donā€™t, and thatā€™s the way it plays out. You canā€™t just know whatā€™s going to work; you have to try things.

Too much churn. We can do better. We can do better, you know? And I would rather we be growing at a slower rate and have it be more sustainable and less churn-heavy than what we have today.

The last year and a half have been brutal in this industry. Theyā€™ve been absolutely brutal. And I think a lot of that has been a hangover of a couple of years that were manicā€¦ Itā€™s almost like the industry itself is doing a manic depressive thing.

Coming back to predictions, I would actually predict that towards the second half of next year, we get back to something that feels a little bit more stable in the tech industry. I think weā€™ve still have some overhang to work through. I think thereā€™s going to be more startups going bust. But I think both because weā€™ve worked through a fair amount of it, and because I expect financial conditions to ease up, the second half of this year I think is going to look more stable in the tech industry.

Yeah, I think so, too. I think overall 2024 Iā€™m hearing is looking up, even just from recruiters and people that are waiting on hiring.

Define stability. I want Kball to define that, so I can hold his feet to the fire later. This is way too vague. I feel like youā€™re gonna be right, even if itā€™s not specific. What do you mean more stable?

Iā€™ve learned my lesson from Twitter. [unintelligible 00:37:32.02]

Well played, well played.

Itā€™s a good question.

Less layoffsā€¦?

Maybe itā€™s less layoffsā€¦ Which weā€™ve had reduced layoffs over the course of the year; we will probably see ā€“ I think we saw a few coming through in December again, and we may see a spike up at the beginning of the year again, just because everybodyā€™s doing their annual ā€œOh my god, what are we trying to do next year?ā€ So I think that. I think a part of it is big tech will be expanding and hiring again.

[00:38:03.12] Which weā€™ve seen of the big tech. I think Facebook/Meta is the only one so far that has shown that they are, which may be just be the first of all, or maybeā€¦

Oh yeah, no, Iā€™m being poached for roles at Meta and Microsoft and Google right now, soā€¦

Youā€™re being poached? I thought you were funemployed.

Oh, I guess Iā€™m funemployed, but I mean, theyā€™re poaching me from my first choice. [laughter]

Youā€™re being recruited?

Theyā€™re poaching you from your life of leisureā€¦

Yes. Yes. Yes.

Well, thatā€™s good to know.

Yeah, Iā€™m being recruited. Iā€™m being recruited. Yeah.

What about us plebs? Can us plebs get a job offer, Amal?

Yeah, I guess. Sure. And technically, another letter in the FAANG, but Iā€™m not gonna say which one. And no offense to anyone who works at any of these companies, Iā€™m not trying to work at any of these companies. Iā€™m trying to work somewhere else. But Iā€™m just saying that the recruiters are back.

Recruiters are people too, you know? Thatā€™s what I always have to remind myself. Cool. Chris, any predictions for this year, 2024? Youā€™d like to put your name on any future goings on?

Well, weā€™ve already heard that Twitterā€™s gonna go bankruptā€¦

[laughs] You can put your money behind that one.

Yeahā€¦ Can we do like a death pool? Or no?

[laughs]

Thatā€™s where Iā€™m at.

Itā€™s gonna get pretty morbid around hereā€¦

Okay, letā€™s not do that.

Who would you like to see die this year? No, uh ā€“

I mean, I donā€™t have a preference, butā€¦

Okay, but who do you think might die?

Well, Jimmy Carter doesnā€™t have longā€¦

[laughs]

Thatā€™s trueā€¦ But then again, heā€™s been in hospice for like the last five years.

Thatā€™s true. I thought he was gone already. He hasnā€™t died? Wowā€¦

No, heā€™s hanging on.

Thatā€™s amazing. Thatā€™s a sound choice.

So that must be it. Thatā€™s my prediction.

I hope Jimmy Carter is alive. I hope he stays alive. Heā€™s one of my favorite living presidents.

I hope he survives, yes, just so that we can call Chris wrong last yearā€¦ But thatā€™s an epic call. Alright, Nick, can you top that? 2024 prediction, better than what Chris just didā€¦ Because oneā€™s amazing.

I donā€™t knowā€¦ I like too the idea of a Deadpoolā€¦ Not that I want any one, or any ā€“

Like a sequel?

Not that I want anything to die, but I just feel like thereā€™s going to be churn enough that the kings are going to at least be challenged more than they have been in the past.

Which is good.

3; And the things that are making me think that ā€“

Get specific.

I donā€™t know if I want to call it out specifically, butā€¦ [laughs] I feel like the close-knit [00:40:47.11] ā€“ I donā€™t know my words todayā€¦ My CO2 is off the charts because of podcasting.

[laughs] Oh, goshā€¦ The quantified self.

Yeah. I think that the really close marriage between React and Next is going to turn out to be a mistake that hurts React more.

Ohā€¦ Do you think thereā€™s gonna be a divorce?

No, no, theyā€™re not gonna get divorcedā€¦ But something else is gonna supplant it, or at least supplant it in the mindset, right? Because nothing is supplanting jQuery yet.

Okay. Mindshare.

Iā€™m worried that ā€“ Iā€™m not worried, Iā€™m predicting potentially that Nextā€¦ Everybodyā€™s gonna start seeing through the Next facade to the Vercel advertisement that it isā€¦

Okay, I think thatā€™s already happened, to a certain extent. You think itā€™s gonna go viral?

Itā€™s gonna really affect Reactā€™s market share.

People are actually going to move away, versus just talking about it.

To Astro.

How do we measure that? Oh, to Astro. You have a specific one.

Astro is awesome, okay?

I think thatā€™s how we put a nice, hard prediction on it. Astro overtakes Next.

Yeah. And we can just use the State of JS survey results.

Not every siteā€™s a static site, though. I mean, I think Astro is great for ā€“

Theyā€™re moving the other way.

How so?

Isnā€™t Astro becoming less of a static site generator and more of theā€¦?

[00:42:10.09] Yeah, yeah, yeah. Yeah, I guess youā€™re right with the newerā€¦ Yeah. I mean, thatā€™s fair. I mean, you could still do very rich interactions with Astro sites, itā€™s just that because itā€™s not the hot path, which I think is the right thing, itā€™s just one more thing to do. And so if thatā€™s all youā€™re doing, itā€™s maybe not the best choice, I would say; maybe use Lit, create a bunch of Lit componentsā€¦ You know, thatā€™s the way I would go.

Yeah. I was gonna say, maybe Angular is cool again.

Angular is cool again.

Oh, weā€™re gonna find out ā€“ weā€™ll find out in January, as Kball and Amal host a few eagerly-anticipated and requested Angular coverageā€¦ Coming to JS Party. Thereā€™s my prediction.

Sorry, we suck at Angular cover coverage, guysā€¦ Just so you know. I think itā€™s just like not a lot of us are using it anymoreā€¦ I was managing a team for a little while, an Angular teamā€¦ But I was a manager, so I was dealing with all the other problems, and didnā€™t get a chance to really dig into Angular too much. But Iā€™m excited to do that this upcoming year.

Yup. January, February, weā€™ve got a couple on the booksā€¦ And if those go well, I think now we have some connections into those teams, so we can get more.

Yeah. And I have also just a really great community show planned with a couple of really nice folks from Poland as wellā€¦ I would love to do that in like March, April. Because the Angular community is amazing. Theyā€™re the best community, I feel like, in the frontend space; theyā€™re just the most lovey-doveyā€¦ I donā€™t know, theyā€™re great.

Have you ever heard of Ember.js?

Oh yeah, I take that back. Ember is great, too. I just know less people. Soā€¦ Ember is amazing, too. But those are definitely the best communities, by far.

Any other predictions for 24, Amal?

My prediction is that ā€“ so first, I want to clarify something. Some of the smartest people I know in our industry are writing components, and they write way more complicated components than I could ever even writeā€¦ Thereā€™s nothing wrong with writing components. I just maybe wanted to clarify that Iā€™m not being paid to create buttons and interactions on a screen anymore. Like, Iā€™m paid to do the other stuff. So I just wanted to correct that, because I feel like [unintelligible 00:44:14.23]

But I thought your point recently was youā€™re not being paid to do those things yet.

Oh yeah, yeah, Iā€™m not being paid to do these things yet. Listen, do you understand the meaning of itā€™s really hard to quit funemployment? Iā€™m having too much fun. I donā€™t ā€“ like, if I could do this for another year, Iā€™d be straight. But alas, I have to get a job. Gotta help my husband pay the bills.

My predictions are we are going to start embracing simplicity. So I think weā€™re kind of at the point of the pendulum swing where our problems have problemsā€¦

Oh, thatā€™s a good episode title.

Yeah, our problems have problems. So itā€™s like, weā€™re just gonna start to realize that we have dug ourselves into way complicated mess, and CSS and JavaScript and web APIs are giving us some really great primitives, that are gonna potentially help us write more idiomatic code all around, including type annotations, which Iā€™m super-excited aboutā€¦ Weā€™re a few years away from that unfortunately hitting a browser, I thinkā€¦ But still, thatā€™s very exciting.

So I think weā€™re going to start embracing simplicity, because I think weā€™re going to realize that less is more, and when you have more time, especially, it means you can do more things for your users. So thatā€™s what Iā€™m hoping for.

I think weā€™re gonna see a lot more projects embrace JSDocs over TypeScript as one small step towards simplicity, I think [unintelligible 00:45:44.28] around like build architectures, and stuff like that. I think weā€™ll start to see a lot more apps shipping unbundled JavaScript, but I think minified unbundled JavaScriptā€¦ So a little bit better than what DHH recently shipped with Once, but very close to itā€¦

Hundreds across the board, wasnā€™t it?

[00:46:08.28] Ah, sure, but I think you have to understand thereā€™s a big difference between Lighthouse and simulated performance, versus like in the wild. Very different numbers. And I think thatā€™s why a lot of performance experts ā€“

Donā€™t they create those, so that we can actually do our tests against our ā€“ I mean, arenā€™t we all striving for that though?

Itā€™s that whole thing about once the metric becomes the goal, it gets gamed. So those metrics were useful when we werenā€™t targeting them as much. And now that we target them, we game them in a way that isnā€™t necessarily representative of real world performance.

But you have information about Once, that shows its real-world performance?

Well, I have a friend who I really wish I could ā€“ they just donā€™t want to start a public beef war with DHHā€¦ I havenā€™t started a beef war, but I definitely disagreed with him publicly, and I dealt with all his reply guys for a few daysā€¦ So that was funā€¦ But basically, his whole thing was like ā€œI want to just take his side, and just run those files through a few build steps, and show him how itā€™s faster.ā€ But minifying and compressing those files is gonna be faster than what he has. It just is. You can do benchmarks with any other website in the wild and see that, soā€¦ So hopefully ā€“ anyways. Either way though, heā€™s taking a step in the right direction, so I donā€™t really care about the specifics. I think, in general, a step towards simplicity. So thatā€™s my prediction.

How do you measure that?

I donā€™t know, Jerod, thatā€™s a you problem.

[laughs]

Any measurement I come up with, youā€™re game, soā€¦

Twitter sentiment is measure is ā€“ that is a scientific unit of measurement, okay?

Oh, Twitter sentiment?

Wait, wait, weā€™re gonna run a survey this year, right? So maybeā€¦

Weā€™re gonna ask people if they value simplicityā€¦ Who says no to that? Nobody says no.

I donā€™t know. But maybe we can come up with some sort of proxy metric that we donā€™t expose to people, but you figure out how you want to measure it, put it behind the survey, run a survey beginning of year and end of year, and there we have a way to start to ā€“

Yeah. You guys, Iā€™ve gotta tell you a story about this thing called a smoot. Thereā€™s this bridge right next to MIT, where they literally took some fraternity guy and they just measured how many times can we lie this person down across the bridge. They basically came up with a unit of measurement which was his height, and it was like ā€œHow many smoots does it take to get across the bridge?ā€ So they came up with this unit, andā€¦ Anyway, so all Iā€™m saying is that we have something called a smootā€¦

Where is this going?

ā€¦that was a made-up unitā€¦ So we need something like that thatā€™s like a Twitter sentiment type thing, you know?

Ohā€¦

So how many squirrels does it take before you see someone complaining, or praising something?

[unintelligible 00:48:48.11] already something. Wasnā€™t it called clout?

Prediction - clout will not be renovated in 2024. Okay, so assuming that we can measure sentiment towards simplicity, itā€™s going to increase in 2024.

I think so, because I think enough is enough. Weā€™re really peaking on complexity, and making our lives so much harder than they have to beā€¦ And I think as a result weā€™re hurting our industry, because people on the outside are just looking at us solve problems in different flavors, andā€¦ You know, JavaScript just kind of gets a bad rap in enterprise as a result. Soā€¦

Could we rank different open source frameworks or other things in terms of complexity, and then measure that, or do something ā€“ like, thereā€™s a couple different angles I could see taking this. So one is like certain frameworks are more complex than others, and maybe we could just have a ranking and then see how many people are starting new projects based on those frameworks. We could do dependency counts. Number of dependencies is like a rough correlation with complexity of simplicity. We could just do how many monoliths are getting started on GitHub, versus non-monolithic applicationsā€¦ I donā€™t know, I feel like this is something that we could put up some proxy metrics for.

Yeah. Well, I can ask Twitter, to be metaā€¦

[laughs]

Letā€™s just say ā€œHey, Twitterā€¦ā€

[00:50:10.15] There you go.

ā€œā€¦if I wanted to measure this metric, what are anchor posts that we can use?ā€ And whatever Twitter suggests, we canā€¦

But Twitter is going to be bankrupt.

Right. We canā€™t do that anymore.

This sounds like a huge waste of time.

Yeahā€¦ Hereā€™s the problem - youā€™ll never get past the part where you have to define what simplicity is. Youā€™re just never gonna get past it. Thereā€™s PhDs about this thing. So I think we just keep it simple, and just declare Amal wrong at the end of the year. I mean, thatā€™ll be the simplest, right? [laughter]

Okay, whateverā€¦ It wouldnā€™t be the first time I was wrong, soā€¦

I will say, based on Twitter sentiment, it does seem there is a swing back towards more monolithic approaches, all in one frameworks like Ruby on Rails, and Elixir, or Phoenix, and things like thatā€¦

Htmxā€¦ And the kind of commentary Iā€™ve heard about it is like ā€œOh, because the tech industry is shrinking at the momentā€, or if not shrinking, at least struggling a little bit more, ā€œthereā€™s a lot more emphasis again on productivity over pure play scalability, or technical perfectionā€, or whatever. And arguing that some of those other approaches were a zero interest rate phenomenonā€¦ But I donā€™t know. I mean, I think, to Amalā€™s point, there is a sentiment shift that seems like itā€™s going on, and one could make arguments for whether thatā€™s continuing to be trueā€¦ But yeah, an objective measure seems beyond us.

Yeah. And I think the React Server Components - it was a big catalyst, I think, in kind of making that a higher-level conversationā€¦ Because I can tell you, this simplicity thing has been happening ā€“ these are conversations that have been happening, at least in my circles, for a while now. For a while, like, years. [laughs] I think the React Server Components thing really just pushed it to like this other level, where people were just like ā€œWhoa, whoa, whoa, whoa, whoa, whoaā€¦ What problem are we actually trying to solve here?ā€ So I in that sense, I think itā€™s a good thing. But I think for me as just someone who cares about the health of the web ā€“ I just want us to build cool stuff for people, and makeā€¦ We need to focus on the next generation of problems, like how are we going to get our sites translated in all these different languages, and how are we going to make performant and accessible experiences that are for emerging marketsā€¦? And how are we going to do cooler things on mobile? How are we going to just give people no reason to use a native app? How can we create web experiences that are just good enough, you know? So weā€™re just not there yetā€¦ We have so many other things to be focused on, and weā€™re just like circling around these same problems, and Iā€™m like ā€œLetā€™s move on alreadyā€, you know?

Well, not to be a downer, but Iā€™ve had enough trips around the sun now to know that when it comes to the software development community, shiny always beats simple. And we will be interested in simple until the next shiny. And then we will all quickly abandon our principles and chase shiny, until shiny stops shining, and some new shiny comes out.

But is generative AI stopped shining yet? Or is it still shiny?

Good question. Find out next time, on ā€“ no, I donā€™t know. I ran out of stuff to say.

Letā€™s do resolutions, as we are getting near the end. Does anybody ā€“ we have a few predictions; weā€™ll see what happens next year, of course. All in good fun. Resolutions, how about? I tend to just keep mine private, so I can fail to myselfā€¦ But some of yā€™all like to put them out there to the public, so that we can help you stay accountableā€¦ Does anybody want to be resolved in this new year to maybe change something about their lives, or have a goal for your career, or for your funemployment? ā€¦anything, before we get into the big announcement.

So Iā€™m working on a project, and ā€“ well, Iā€™ve already published it, but right now Iā€™m working on a docs site, and a major rewrite, and all that stuffā€¦

Really?

And I want to get it out there, and try to get it in more peopleā€™s hands, andā€¦

Awesome. Tell us about it. I didnā€™t know you were working on something new.

Yeah, itā€™s like a thing that lints your package. So if youā€™re gonna publish to registry, it lints the tarball, the contents of the tarball, and how [unintelligible 00:58:11.06] you install from that tarball. And how does it work? And is it broken? And these are typically things that tests donā€™t catch. The idea is you run this, it will do stuff like it packs your project, installs it into a [unintelligible 00:58:30.16] and then runs a bunch of checks against it, to make sure that consumers can I actually use your package. So thatā€™s the idea.

Thatā€™s very cool.

Thatā€™s very cool.

Is it gonna be called Shiny?

No. Itā€™s called Midnight Smoker.

Is it really called Midnight Smoker?

Thatā€™s kind of awesome.

Whatā€™s the backstory behind that? That sounds like the name of a blues album more so than a JavaScript packageā€¦

No, itā€™s like smoke tests, right?

Smoke tests, but also the Joker, right?

Right. So one of its features is basically you can run arbitrary scripts against the package as it would be installedā€¦ And so that would be a smoke test. ā€œOkay, does this thing actually ā€“ can I require this thing? Can I run the CLI?ā€ And you can do stuff like that. But thereā€™s also like built-in checks if you donā€™t want to write a custom script. So thatā€™s kind of where the name came from, smoke tests.

So that sounds as if itā€™s from the perspective of a package publisher, but as you were describing it, Iā€™ve found myself wondering if thereā€™s a security version of this that a package installer could run this against a package theyā€™re going to install in a sandbox, and see ā€œDoes this make weird network requests?ā€ or something else.

Noā€¦ I mean, I think Iā€™m probably going to just stay away from ā€“ thereā€™s other tools that can do things like that a lot better, I think. Iā€™m trying to be careful not to do things like ā€“ I donā€™t want to necessarily be a tool that checks and makes sure all the dependencies are there, or theyā€™re all used, or somethingā€¦ Because there are other tools that do that, and I donā€™t want to step on ESLintā€™s toes either. I donā€™t want to do things like ESLint does.

You mean you donā€™t want to solve a solved problem?

Right. Iā€™m trying to limit the scope, and I think security checks are betterā€¦ I mean, the other thing I work on, which is LavaMoat, would maybe help more with that sort of thing.

Weā€™ve gotta do a show on LavaMoat next year.

Yeah, we do.

Thatā€™s super-cool. Yeah.

Anybody else? I think thatā€™s awesome, Chris. I think it sounds like a very achievable goal, since youā€™re almost finishedā€¦ Of course, the 80/20 rule and documentationā€¦ So best of luck to you still. Anybody else want to go on the line, put themselves on the line, I donā€™t know, stay on the line with their resolutions?

For me, Iā€™ve got to stay moving, like as in I have to stay unconstipated, I think, with my creativity.

Oh, goshā€¦ [laugh] Well, that paints a pictureā€¦

Yeah. Whatever. Everybody uses the bathroom, including the former Queen of England, you know? And Beyonce.

Unless youā€™re constipated. Then you donā€™t use it.

Anywaysā€¦ [laugh]

But thatā€™s how you think about your creativity, just gotta flush the system?

Just lush it. Keep it moving.

Yeah, Iā€™ve just gotta keep it flowing and moving. Keep it regular. I think thatā€™s the ā€“

ā€œGot to keep on movinā€™ā€¦ā€

Yeah, yeahā€¦

Amal and her metaphors, you know?

Yeah, I have a birthing ā€“ I donā€™t know, like Iā€™m gonna birth a lot of creativity.

[laughs] Youā€™ve got birthing on the mind.

Yeah, in 2024. No, I have some really fun projects coming up. Iā€™m actually stepping into Node as a formal kind of contributor, and Iā€™m going to be leading up to a very exciting project that Iā€™m like ā€œOh, should I ā€“ā€ I wanna have a few more conversations with people before I publicly announce it, but yesā€¦ This is not my announcement, no, but Iā€™m saying Iā€™m excited about that. Iā€™m gonna be doing some cool stuff with Node, and Iā€™m gonna be writingā€¦ Iā€™m gonna not let perfect get in the way of goodā€¦ Yeah, Iā€™ve got to just ā€“ Iā€™ve gotta writeā€¦ So yeah, Iā€™m excited to have a lot of creative energy, kind of, I think, released into the world a bit more this year.

[01:02:28.01] Got it. More creativityā€¦ More flowā€¦ More birthingā€¦ And flushing.

And flushing. Yes.

Nick, you wanted to be on more conferences in ā€˜23ā€¦ I see that youā€™re speaking at that conference in Januaryā€¦ So how do you feel about that resolution, and are you resolved to continue that flow, or how do you feel about that? Maybe you have new ideas for this yearā€¦

Yeah, I am pretty excited about thatā€¦ Kicking it off right away with that conference at the end of January, which will be a lot of funā€¦ And I havenā€™t submitted to any other ones, but I have ideas for talks that I think are really good, and theyā€™re things Iā€™m really passionate about. Iā€™m really getting into code [unintelligible 01:03:10.27] and manipulation, and I think that thereā€™s a lot of really cool stuff to do with thatā€¦ And Iā€™m kind of working on a project around that right now, that I hope to continue on in the new year.

But yeah, I think thereā€™s a lot of changes coming up, and Iā€™m pretty excited about it. I think ā€“ thereā€™s a podcast that I listen to called [unintelligible 01:03:34.10] where they talk about a yearly themeā€¦ And I just verbatim stole the yearly theme from the host for me, and that is the year of small improvements. So I have this idea of - Iā€™m just going to make a tiny improvement at least once a week. So thatā€™s 52 improvements over the next year. And thatā€™s going to be like overall awesome. And thatā€™s things like Iā€™m going to actually install a proper light in my office, orā€¦ You know, just like small bug fix, things that are annoying me Iā€™m going to fix, and not let them just be annoying.

Like your internet provider? [laughter]

Like my internet provider.

Too soon?

I hope notā€¦

Yeah. I love that idea.

And Iā€™m gonna wear sunglasses.

Kball, would you like to go on the line?

Sure. I have made and failed at physical resolutions the last two years weā€™ve done this, so I am not doing anything in that domain. So last year, the first half of the year I was purely running my own business, and in some ways because it takes a long time to ramp up a business, I was undercommitted for much of the year, and then I ended up getting a job, which has been awesome and challengingā€¦ And Iā€™m still running my own business. So I went from undercommitted to overcommitted.

So my resolution for this year is to get to a place where I feel like Iā€™m moderately balanced. I am not so overcommitted that I feel like Iā€™m pulled in a million different waysā€¦ Which - it was trending there towards the year, as I got more onboarded with the job, and it was less overwhelmingā€¦ But also not so undercommitted that Iā€™m looking for something new, like a new job. So I think thatā€™s the big goal.

The other thing is to keep working on my coaching skills. Thatā€™s what the business was around. Thatā€™s also the job that I got, is director of engineering at a coaching company called Mento. Thereā€™s a lot of opportunities for me to think about and learn to be a better coach, and I want to keep doing that. That is something that I enjoy tremendously, and I think is also something where Iā€™ve been able to help people a lot, and itā€™s super-funā€¦ So I will keep working on that.

Love it, love it. Well, Iā€™ll finish up quickly, and weā€™ll go to the big announcement and call it a show. They say the best way to predict the future is to invent it. So I will combine predictions with resolutions, and share a little bit. I think we are resolved here at JS Party to add some new voices to the pod this yearā€¦ And so there we have both a prediction and a resolution.

[01:05:59.18] Weā€™re also resolved to bring back an old podcast here at Changelog, to revitalize and bring back an old show that we used to produce, and produce it once againā€¦ And Iā€™m also resolved to take Changelog++ to the next level in 24.

So those are three things that I think are achievable, two of which are already in motion and happeningā€¦ So itā€™s easy to predict things that are 80% done, but weā€™ll see if the last 20% gets finished in time for it to be successful. Definitely going to add some new voices around here, so stay tuned for that. Hopefully, theyā€™ll be somewhat familiar, yet new, as regulars; itā€™s something weā€™ve been working on behind the scenes for the last couple of months. So thatā€™s exciting.

Okay, Amal - big announcement. So much has been building up to this moment. Kball, b0neskull - set her up. That means Kball does something and Chris just sits there.

Thank you. Yeah, I didnā€™t want to have to prompt Kballā€¦ So yeah, I want my ā€“

Go ahead, Kball. Announce the announcement.

Announcements, announcementsā€¦ Announcements.

[laugh] There you go, Amal.

Chris is shaking his head like nothingā€™s gonna happen. Heā€™s like ā€œNope, nope, nope. Not doing it.ā€

Heā€™s like ā€œThat shouldnā€™t have happened.ā€

Wait, I could do the nose kazoo again if you want.

Okay, do it. It sounds good.

Yeah, please. It sounds like a horn. [nose kazoo 01:07:13.18]

So my announcement is that Iā€™ve pinged these dudes that Iā€™ve worked with on JS Party a few weeks ago to say ā€œHey, weā€™d love to do a survey this yearā€, this year being next year. ā€œHow do you all feel about it?ā€ And they were all ā€œYeah, that sounds good.ā€ Okay, cool. So next steps are for me to create a repo; I want to kind of work on this in the openā€¦ But essentially, what Iā€™d like to do is create a JS Party listener survey, an annual listener surveyā€¦ And so itā€™d be great to work with some of you on the questions, and what you all think are usefulā€¦ Weā€™ll kind of seed some initial things there, but obviously, we can discuss itā€¦ And so what Iā€™d like to do is everything will be done in the open; the survey questionsā€¦ I think weā€™ll probably use the ā€“ I think The State of JavaScript has this platform, the survey platform that they use, so I think we can actually technically host it thereā€¦

But anyways, so weā€™d love to kind of like hear more from you all a bit more directly on what you want more of, less of etc. and just have a way to kind of touch base annually. And so Iā€™d like to give us - because everyone has 17 million other things that theyā€™re working on, and Iā€™d like to give us at least like six to eight months to kind of actually work on this, starting in mid-January. And then Iā€™d love to launch the survey in the last three months of this year.

So yeah, so I think thatā€™s kind of what Iā€™m thinking. So again, weā€™ll put a link in the show notes for the repo. Weā€™ll be having these discussions and working in the open. If youā€™re interested in contributing or chipping in or watching along, please. Yeah, we would love your help, and we welcome you. And I think one thing weā€™ll have to decide is ā€œOh, does it make sense to translate the survey into a different language?ā€ Maybe not, because the show is in Englishā€¦

TypeScript?

ā€¦but I donā€™t know. You all decide.

There you go. There you have it. The link will be in your show notes. You can get involved, help shape this, help make it better than we could surely make it on our own. That would be awesome. And weā€™d love to hear from you all about how to make that great for everybody. Okay, this has been our New Yearā€™s party. Happy ā€™24 to all you all out there listening. Any final words from the panel before I hit that outro song? Except for Nick. Anybody but Nick.

Oh, come onā€¦ I had a good one.

[laughs] Go ahead, Nick.

Did you know that if you long-press the Share button on Threads, you get the share sheet?

Thatā€™s it.

For Nick Nisiā€¦ Amal, Chris, Kballā€¦ Iā€™m Jerod, this is JS Party. Happy 24, yā€™all. Letā€™s make it an awesome one. Weā€™ll see you next week.

Well, no, no, no, I have an announcement to make.

Okay, announcements. Announcements. This is part of our ā€“

So you let me know if you want the announcement to be now, or later.

Letā€™s save it for later. You always save the announcement for the end.

Announcements, announcements, announcementsā€¦

Yeah, Kball, youā€™re always the best at announcing announcements, I have to say. Youā€™re always like [01:11:29.12] That was your job in a past life, you know?

I donā€™t know, Chris has a literal kazoo over there. He could probably do a better [01:11:35.28] than Kballā€¦

Yeahā€¦ You know what - can we do a song between Kball and Chris? Can you do ā€œAnnouncements, announcements, announcementsā€ and Iā€™ll be maestro, and he kazoos?

Can your kazoo do an announcing sound?

No. [laughter]

I saw that coming a mile away, and I still wanted it, too. So I allowed it. Itā€™s amazing.

Did he really mean it, or was that just like ā€“

Oh, he means it.

Oh, noā€¦

Itā€™s both.

Alright. Well, weā€™ll just have to imagine it, Kballā€¦ Okay, do you wanna do the announcement?

Um, nope. Letā€™s save it for the end. Always save the best for the end.

Okay, weā€™ll do that. Weā€™ll do that.

Changelog

Our transcripts are open source on GitHub. Improvements are welcome. šŸ’š

Player art
  0:00 / 0:00