S10 E7 | Vite’s Next Chapter: Alex Lichter on Rolldown, Oxc, and Beyond - The Angular Plus Show

Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:08):
Welcome to the Angular plus Show. We're app developers of
all kinds share their insights and experiences. Let's get started.

Speaker 2 (00:21):
Welcome to another episode off the Ango plus Show. I
am super excited, even though I feel like I'm going
to see our next our guest tomorrow again, which is
of course I'm blessing, I guess super.

Speaker 3 (00:34):
Excited to be here.

Speaker 2 (00:35):
But before we get to that, I don't want to
like spill the bean, so let me introduce my co host, Lara.

Speaker 4 (00:40):
How's it going.

Speaker 5 (00:41):
I am great, It is lovely to see you.

Speaker 4 (00:44):
It's been a while.

Speaker 5 (00:45):
It's been a while, yeah, a couple of weeks off.
It was good, like I had. Work's been real busy,
so it's been nice to because on my calendar, I
just have this podcast moment blocked out, so I just
don't ever tell anyone I'm not actually doing the podcast,
so I get so much stuff done on Wednesdays if
I'm not doing the podcast. So but I'm happy to

(01:07):
be here for this one.

Speaker 2 (01:08):
The nice tip for that before we get to the guests,
just building it up a little bit more working for
most of European company but living in the US. Except
that I would not recommend living in the US in general.
But if you happen to be here, because basically, like
after ten am, there's like just quite yeah.

Speaker 5 (01:25):
Yeah, that's why I used to go to the office
really really early in the morning. I would go at
like six am because I had three good hours before
anyone else showed up.

Speaker 2 (01:32):
So all right, I'm genuinely excited to introduce O guest, Alex.

Speaker 6 (01:40):
How's it going, good? Good curse and the blessing. You
have to talk about that?

Speaker 5 (01:45):
No, it's for you, like you have to deal with
yon two days.

Speaker 3 (01:49):
Okay, okay, if you are the blessing, I'm the curse.
You get it now?

Speaker 4 (01:55):
God, Okay, that needs an explanation.

Speaker 6 (01:57):
Okay, here here we are also happy to be on
again a second time and hopefully not the last time.

Speaker 4 (02:04):
So yeah, this.

Speaker 2 (02:05):
One plays out just to make sure you're still like
on our good list, you know.

Speaker 4 (02:09):
Ye, not that I become the curse in you the
blessing right up here?

Speaker 2 (02:16):
So you recently, somewhat recently recently ish.

Speaker 3 (02:20):
Became DeFore for void zero a year ago.

Speaker 2 (02:23):
Yeah, signed, there's this one.

Speaker 4 (02:28):
That's this one. That one.

Speaker 2 (02:30):
Let's just completely level the playing field and assume that
I have no idea what Void zero is just like
no idea, So explain it to me like I'm five.

Speaker 6 (02:40):
Yeah, that's actually also a big part of my job
because literally a lot of the operas have no idea
of the company, because why would they. They even have
no idea less often but still sometimes of the tools
we're building.

Speaker 4 (02:52):
So well, let's start with that.

Speaker 6 (02:54):
If you developer adaptive oper to be precise and do
something with front end, then you slightly use v jas
underhood somewhere. So no matter if it's like I don't know,
React project, the view project, even angular right has VJA
we built in even as death sorr. Yeah, TIY tell you,
but what are getting? We're getting there step by step.
So unless you use next JS, something is usually running

(03:17):
the feed or you have like some other set up.
But that's that's the way it goes, and fat itself
is in a way, it's always been a multi stakeholder project. Right,
There were people and still are people around employed by
different companies. But when Avenue founded Void zero roughly two
years ago, announced it a year ago, I think around
October something like that. So then yeah, we can't exactly exactly.

(03:41):
That'll be a recurrent topic I suppose I hope you
all enjoyed this. This is an episode or like this
not this episode as well, but this year is we
canmed their first in person one also with live stream.

Speaker 4 (03:51):
Back to wid zero.

Speaker 6 (03:52):
So when you found that there were like he was,
he created weed right, So he started like thinking of
the shortcomings of feed there and there were a few
things that could be improved.

Speaker 4 (04:01):
Plus of course.

Speaker 6 (04:02):
It's also about well sustainability of open source. That's another
big part there. So we found and avoid zero to
work on tools like VAT, but also like like roll
down and OEC so roll down as a rustified new bunder,
it's not just a rusty fied roll up. It's very
important and to in the end work on something called
the unified JavaScript tool chain. So if you think about that,

(04:23):
like when we start a new greenfield project, we think of, okay,
what do we use? Okay, we use vidus ability maybe
in our framework template there, and then we use es
lint or maybe should we use bay them or auxland,
And which formator should we use we use like Prettier
or do we use deep print or do we use
like esl stallistic rules And we have all these choices
which are well painful in a certain way. Of course,

(04:45):
you all want to use typescript, so that's definitely said.
That's also part we don't really want to touch there.
But then it continues with like, okay, maybe you want
to use a certain library or tool, and then that
uses other.

Speaker 4 (04:54):
Things under the hood.

Speaker 6 (04:55):
If we even like look a level deeper, like level lower,
so like different parser or a different I don't know,
a bit different resolver, and then you have a lot
of effort because then okay, use feed. We uses things
under the hood, and then there's a plug and I
use for example, Babel, and things get like parsed and
transformed like gazillion times for no real reason, and people

(05:18):
have hard time picking their tools and are quite well
we all own the term JavaScript fatigue, right, so they're
quite tired of all of that, And the idea of
foid zero is to make that obsolete, to make like
one unified two chain that just works TM right with
all the major frameworks, with feed as the core, like
in Centerpiece, with a regod foundation that is fast and well.

(05:40):
That solves all the needs. So that's our mission to
make JavaScript developers as productive as they have never been before.

Speaker 3 (05:50):
It sounds like a great mission statement for a startup.

Speaker 6 (05:53):
Yeah, absolutely, And I mean we got the venture capital right,
so like I haven't raised some seed money, and people
are always skeptical with like manure capital open source, which
I understand, but like I funnily I had an interview
with him after henounced the company when I wasn't part
of the company yet. It was like I was just
generally interested. A lot of people were like wondering, Okay,
how should it work out and buy venture capital, et cetera.

(06:15):
But to be fair, like to build such a tool
chain out like from from zero to something like from
okay we have veed to everything around, then you somehow
need to keep up velocity and that doesn't work on
like the nation based framework like for like veed for example,
which is still in dependent not related to avoid zero.
That's running the nation based and that's great and it works.

(06:37):
But for for such a big project, this is really
not feasible, especially because you need people working that full time.

Speaker 4 (06:43):
And yeah, then you need money from somewhere.

Speaker 2 (06:46):
Yeah.

Speaker 3 (06:46):
I absolutely don't think that this is unreasonable.

Speaker 2 (06:50):
I rather think that the underlying expectation that is often
there of like oh, just it's open source, so it's
free and people are just doing that in their free.

Speaker 3 (06:57):
Time as a form of modern exploitation. But let's not
go there.

Speaker 4 (07:02):
I'm to that.

Speaker 2 (07:04):
That, Yeah, what I'm curious about because quite honestly, from
my expectation being in the Anger ecosystem for the last
ten years, I absolutely had that JavaScript fatigue moment in
like the Anger JS and beginning of Anger two phase
where it was like, well, are we going Webpack? Are
we going system Jazz? And like I completely felt that.

(07:25):
But since with the Anger CELI, it's basically somewhat gone
from my expectation. Pretty much all projects I've seen in
the last ten years are using the Anger cel i
X as an extension of the Anger CLI and therefore
relying on the provided tools like first web Pack, ear
is built e Islin, Prettier, those are already configured, and

(07:48):
I have not yet seen a single one that was like,
you know what, let me figure out the system jazz
thing and let's go back to system jazz instead of Webpack.
The APIs are somewhat public, so you could do this now,
but still I have not seen anyone do this, So
I'm kind of curious of and I completely agree with you.

(08:09):
I see the JavaScript fatigue in my position and like
literally every other framework literally except Angular except so I'm
kind of curious if Angler is even like a relevant
market for you in that sense.

Speaker 6 (08:26):
Yes, So it's pretty easy to say yes because the
JavaScript fatigue is one part I think, especially in Angular,
like if you have a bigger application and like I
mean like enterprise level and you run esland, then how
long does it take? Like this is the other part, right,
or two chains because they're not very well optimized. And
also a lot of things are written in JavaScript. That
doesn't mean the necessarily slow, but very commonly things are

(08:48):
slow and written JavaScript. That's that's another part there, where
like sure you can run your prettier, you can run estland,
you can run your webpac based bills for example, like
if you use es bill, that's a bit faster because
it's written in goal. But then you have other issues
like chunking, where Angler now introduced roll down for like
their experimental chunking. So that's also great to see some

(09:09):
adoption there. So that's definitely another point to make a
two chain. That's not only like okay, we solve this
problem of choice, but it's also the best and fast,
like fastest choice, so to say, and especially for Yeslan,
it's like we worked Onlinter is still working on called Auxland.

Speaker 4 (09:25):
The idea is so, first of all, we.

Speaker 6 (09:27):
Ported a lot of rules over more than five hundred
and forty rules are ported from JavaScript to rust so
that they run fast, so you can say, okay, even
just for very costly rules like I don't know cyclical
imports or no barrel files because you need to well
scan everything, you can run Auxland. But then you can
just run straight away everything in there, and at the
time the episode comes out, there will also be the

(09:49):
support for custom JavaScript plug. And so the idea was
always like, okay, if you write russ toooling for JavaScript,
how to like add extensions, et cetera, And what if
this this weird one little ears on plug I want
to use or like this custom rule from my company. Yeah,
you can do that with custom jazz plug and stand
technical preview.

Speaker 4 (10:06):
But they work very well.

Speaker 6 (10:07):
They're es when compatible, even though you have the option
to write it in a slightly different alternative API that
is then even faster than Auxceland and even will give
out first any as incompatible like rule like or plugin
and also an auxlent one, so you can kind of
kill two birds with one stone, and we're curious if
some people will adopt that.

Speaker 2 (10:28):
So I should have probably prefaced that because I'm I'm
a huge fan of everything you're doing. I'm just trying
to play Devil's advocate to get absolutely to share more
like perception and get more information that might be relevant
from the Anger community, right I quite honestly, if I
would have VIC money, I would bet money on everything
Evan does or touches. So yeah, obviously.

Speaker 6 (10:52):
I'm biased there, right, But I think it's good to
point it out, Like I think it's good to point
out these things that especially from the Angular community that
I like, I know a little bit. I know quite
a few people from the Angler community, but I haven't
built an Angler application in a while, right, Yeah, I
might if I would have time, like I have to, like,
you know, explain it in your spare time.

Speaker 5 (11:13):
That's when I do all of my.

Speaker 4 (11:15):
Other what is this spare time cons.

Speaker 5 (11:19):
Like never let anyone know you have any because it
will become theirs.

Speaker 2 (11:22):
Yeah, I like I send them calendar and writes just
to block time off their calendar.

Speaker 4 (11:30):
Nice, that's a good idea.

Speaker 5 (11:32):
Yeah, yeah, definitely, yeah. And I think what I actually
what I liked about Angular when I started was I
didn't have to care a lot about what happens under
the hood. But I know that there's been a lot
of you know, it was not easy to implement beat
in Angular for the dev like the local builds and stuff,

(11:53):
and so having I love this direction where instead of
having all these weird special tools that we have to
use that we're all kind of moving towards, like let's
streamline it so that we can kind of start to
fill instead of people having this decision fatigue, they can say, Okay,
this is trusted, and then there's room for people to say, yeah,

(12:14):
it's trusted, but what if I do it like this,
because we still need that innovation. We always are going
to need that innovation. And so it's nice to not
have to fiddle around with everything.

Speaker 6 (12:25):
Fully agreed and the best the best part is if
you say, like, Okay, I don't like one part of
the too chain, I just want to try something else,
like nobody will stop you, right, and all the foundation
like as I said, like we'd roll down or ex seed,
they're they're all open source, like their m I T license.
So if you also like, hey, I want to build
it together on my own, that that's also where to go.
But of course, like also with the announcement of feed plast,

(12:47):
they can recently at feed conf So for us in
the future and for all of you listening, uh, like
a few weeks ago, that also changed as like having
that like unified toochin like the cargo for JavaScript.

Speaker 4 (12:59):
If some RUST fans are are in the.

Speaker 6 (13:01):
Room to have that, that's that's definitely also like there
everything is put together, everything is ready for you. And
if you say, hey, I don't want to use the
I don't know oxen, I want to use different lintern,
go ahead. That's that's fully up to you. But you
at least don't have to like do all the choices.

Speaker 2 (13:16):
Okay, as as VT plus is for us in the future,
and except the slide that Avon shared literally a year
ago at the last week on where it was a
fancy phrasing of it's going to be the cargo for
JavaScript some light Yeah I.

Speaker 4 (13:31):
Can, I can.

Speaker 6 (13:32):
So, so the idea is so okay, let's start, let's
prefest this with that VAT plus is is not open source,
Yes it is it is source available though, so first
of all, at that moment it's in a big weight list,
so it is not available for public right now, but
if be interested after what I'm saying, there is like
a way to register on the website, which is also

(13:54):
then in the show notes. So the idea is there
will be general launch launch like early next year. That's
our goal there, but for now it's like trying out.
We tried out with like a couple of companies already.
That's also having seen the website for example of frame
Er Linear work together with them on certain parts of
that as well, because they have quite a big code

(14:15):
bases and also quite some goals there to achieve, and
our need for some very custom features like custom chunking
for example, that rolling can do very well. But back
to the plus, so it's not open source. It's source available,
but it will be free for most people using it
because we said, okay, you know.

Speaker 4 (14:32):
What, we don't want to gate keep anyone.

Speaker 6 (14:34):
The idea is also like okay, it works like a
lot like that in different scenarios. Right the big enterprises
have the money, they can pay for it and make
the whole development possible and make the tool chains somewhat
self sustainable whilst there for everyone and everybody can.

Speaker 2 (14:51):
Use it well.

Speaker 5 (14:52):
And it's smart too because like I can't just go
to Cisco and say we're going to use this. Now,
they're going to want to see pocs. They're going to
want to know how it's different. And if I haven't
been able to try it out, then how am I
supposed to tell them that that's a good choice.

Speaker 6 (15:06):
So exactly, and as I said before as well, like
the idea is to have that unified toochin available for everyone, right.
We don't want to say like, hey, oh you're locked
out because you can't pay, and plus I don't know.
It's still has its source available, so people can take
a look the source cade, they can try all all things.
So even though it's not open source by definition, it's
still something people.

Speaker 4 (15:26):
Can take a peek in.

Speaker 6 (15:27):
People can like they can just use it and they
don't have to worry. Like in the end the productivity
boost through it that will be paid.

Speaker 4 (15:35):
That will be paid either.

Speaker 6 (15:37):
When I don't know, I have a great app and
that works so super well and you raise venture capital area,
do you boostterp yourself to whatever money? And then then
it will be fine, will be worth the hat start.

Speaker 2 (15:46):
So to say, I love how you started with a
licensing model, how it's going to work. I still kind
of cannot grasp what value proposition doesn't offer? Why should
I pay money for it?

Speaker 4 (16:00):
Yeah?

Speaker 6 (16:01):
First of all, if you want to try it, as
I said, you don't have to. So there there are
a couple There are a couple of things there. So
the first thing it is if you if you think
on front for a company perspective, like, first of all,
you want to make sure, okay, we have certain let's
say security guarantees. This is one big dependency. Uh, that's
in the end in lines things like v test ox,

(16:22):
lend ox, format, et cetera.

Speaker 4 (16:23):
So the things we're building underhood already.

Speaker 6 (16:25):
It is basically a super set of feed You can
use it as drop and replacement for VAT itself. Uh,
and it comes with a lot of things. It comes
of course that like with Flinting, we're formatting out of
the box, you have a death server. Just with VEAT
we'll have a task runner, mono repo support out of
the box, so things will just work. Also with caching,
there's also a thing that like why making mono repo support,

(16:47):
or like why making mono repo tooling complex? If well,
everything we have there, there's a weed contig everywhere. It's
easy to detect and task runner. So we basically have
this like all in one solution. Then whatever frame we
use on with a wet plug and like I know,
React view as well, et cetera.

Speaker 4 (17:03):
You name it and from there you get to.

Speaker 3 (17:06):
Go by no means. I mean that's in a condescending way.

Speaker 2 (17:10):
But so it's an opinionated configuration for bead with some
add ons.

Speaker 6 (17:15):
So if you think about that, like, we have a
lot of parts that are open source right now, So
like Rodin is open source ox on, open source ox format,
it's ceteraly whole os projects.

Speaker 4 (17:24):
Open source VAT.

Speaker 6 (17:25):
So we can say, okay, I build my own tool
chain together and just take off of that, and then
I have to explain it to my company and to
my team and say, hey, this is how it works,
and I have to go through all document documentations of
all these different projects, cetera, et cetera. Or you go
to the VAT plus website and there everything is nicely explained.
Plus we bump the pencies for you. We address security
issues also then in the like open source packages, but

(17:48):
still is also like there if you're interested as a company,
say hey, I want to have some slats also something
you can talk to us.

Speaker 4 (17:53):
So there's this like one single point.

Speaker 5 (17:57):
That's that's actually like a big cell point I think
for enterprise apps there because one of the hardest parts
about bringing in a new dependency is getting through security
with it and making sure that it's a secure package,
that it doesn't introduce vulnerabilities, and so yeah, having being
able to outsource any of that is nice.

Speaker 2 (18:18):
So absolutely, what is feedback and that you received from
the clients that you worked with on on those things,
Like I mean, I think I would assume that the
most thing you're interested in is like developer productivity in
a sense, so like the bill times go down, this
does this improve developer productivity?

Speaker 3 (18:35):
Like what are the things you looked at.

Speaker 2 (18:36):
And what is the feedback you see so far? And
you don't need to name names just like the general.

Speaker 4 (18:42):
No, no, of course, of course, I mean I will
e names.

Speaker 6 (18:46):
The name is because we also had a few speakers
that be kind of that gave talks about exactly that
mainly linear and frame them mentioned before, and and like
there it's for example, with linear, we we worked lot
getting everything ready in their repositories with like rolldin with
auxcellend et cetera, and we've seen like big gains there.
It was very helpful also to get these this feedback
also early on from like real world mono repos big projects,

(19:09):
because I remember from my position as a next team
member from like also an open source maintainer is like, Okay,
how do people use certain parts?

Speaker 4 (19:17):
What are they actually doing?

Speaker 6 (19:19):
And especially if these enterprises contact so it's it's sometimes
really difficult to talk like you can talk about it,
but you have to see it.

Speaker 4 (19:26):
And there it was. It was really.

Speaker 6 (19:27):
Helpful that like parts of the core team had access
to the code and take can take a look at
and also solve issues, also deal with progressions. So we're
really really thankful for that. But besides that, there's also
a repository that tracks all the performance wins for roll
down vats specifically, so there we see like sixteen eighteen
acts built time reductions so it's not percent so like

(19:50):
actually times which which is pretty big. And with rolldown
becoming a default in VD eight at some point that
will be available for everyone also no matter if you
use plus or not.

Speaker 5 (20:03):
So so just for the listener. In my daily experience
at work, our build step on our whenever I try
to merge a pr our build step takes like fifteen minutes.
Sometimes that means I have to wait that entire time
for that step to pass. And if I have to

(20:24):
rebase against Maine, if I made a mistake and I
need to push a change, it starts that whole thing
over again. And the longer that pipeline takes to run,
the longer you're sitting idle waiting to get your branch merged.
And usually if anyone listening is like me, I like
to do pull requests and steps, and it is painful

(20:45):
if you get one step and you have to wait
a whole day to merge it, and then oh, I
tried to build on top of it. Now I'm going
to rebase, and now I'm going to pull my hair
out because oh my gosh, that rebase was terrible. And
so there's a lot of like it. I think it's
easy for people to say, oh, well, you know, fifteen
minutes isn't that big of a deal, But over the
course of the day it adds up and adds up
and adds up and it's a big deal.

Speaker 6 (21:07):
So, as you said, it's just this is the bill
step right. Then we don't even talk about linking or formatting,
which is yeah, this is like yes. So if if
you like have your your web Storm editor open for example,
or Vius code uh because yeah, exactly reb store there
you go, like if you if you say okay on safe,

(21:28):
I want to like run my linter and your linter
takes I know, even if it just takes three seconds,
it's like, okay, what do you do these three seconds?
It's it's idle time and over like over the day,
we all know we hit like control save for commands
save quite often. This is like a lot of time
wasted in the end. So if your Linda would just
run like subsecond for example, perfect, why why not happen

(21:50):
that straight away?

Speaker 4 (21:51):
Or like a pre committal where I was like.

Speaker 6 (21:52):
Okay, then we just you know, we say there's no, no,
very fine, and we like just give over and.

Speaker 4 (21:59):
That's also not really the way to go.

Speaker 5 (22:00):
But no verify. Yeah.

Speaker 6 (22:03):
So these are like all real problems that a lot
of a lot of developers had. I've went to a
conference recently and I asked like, okay, who is using
island and most people raise their hand, And asked, okay,
who's happy with how fast there are island runs? And
nobody a lot of single person raised their hand. Was like, yeah,
there is a problem here that needs to be solved.

Speaker 2 (22:21):
I think this is a little bit of a black
and white question because if you were asked a question
the other way around, who is not happy with the
Islam performance? I would also be sure not everyone raises
the hand.

Speaker 6 (22:31):
Okay, maybe this was not the most useful question, but honestly,
like to be fair, to be fair, I've talked like
one on one with a few of these people in
the break and people said like, yeah, you know lint
or like formating or linting for the whole code base,
like it takes forever and it's a performance.

Speaker 4 (22:49):
Cool. I mean, I don't.

Speaker 6 (22:49):
I think especially in bigger projects, like if you have
I don't know, just a one hundred components, it's pretty okay, right,
it's pretty fine.

Speaker 4 (22:56):
But like in.

Speaker 6 (22:57):
Bigger projects and then in a mono reppo contact et cetera,
et cetera, it's not that trific.

Speaker 4 (23:02):
I think we can all agree on that.

Speaker 5 (23:03):
Yes, oh yeah, it's it's yeah, like you're totally screwed
if you accidentally touch like the main pack, you know,
like anything that is depending on it, like yeah, the
affected breath. Your affected files are like more.

Speaker 4 (23:18):
Than any one thing.

Speaker 2 (23:22):
Though that those tools, particular Islam and Prettier, are extremely
good at is having the ecosystem support, having holistic support
across entire angulo projects, having support for literally every file
type that is out there, even though you have never
heard about it. I would be curious, for instance, who
knows about things like ejs that used to be like

(23:42):
super popular ten years ago.

Speaker 3 (23:46):
Doesn't matter.

Speaker 2 (23:46):
Those two supports that, and they grew into that ecosystem.
Whereas the last time we talked in April, oh excellent
not or excellent oh excellent? Um and I also you're
far didn't support template parts basically basic basically focused on
the script parts of like types strip and JavaScript, whether

(24:08):
it's the status and or a vision there.

Speaker 4 (24:11):
Yeah, updates.

Speaker 6 (24:12):
So update for the linter is so we focus in
JavaScript and typescript first and foremost. That means if you
use some kind of like custom component templates, I think
of swell components, few components, Axtra components, et cetera. Angler
like there if you put if you use just template
and then the back takes, for example, then it's just a.

Speaker 3 (24:31):
Property, right right, But you're not linking.

Speaker 6 (24:35):
So that's and that's exactly that's exactly the point where
we're not linking these parts. We don't lend any like
htmlish language CSS each language at the moment, so that's
that's sort of But the good thing is we do
lind the rest. So if you would even use like
let's say, using a file components right, and angler also
has like standalone component sports, then you can at least

(24:57):
have the script part lintered fully. If you say, okay,
these like these template rules are super important to me,
I don't miss them out. You might have to use
es lind in tandem with axceland, so you can say
I run oxland for a first and then I run
island with we have a plug in also migration script.
So there's something called es limp plug in auxland. It's
a bit confusing, but the idea is you put it

(25:18):
into your eslin configured will disable all the rules you
have enabled an auxland, and then you only run the
missing rules that are.

Speaker 4 (25:25):
Not support it for whatever reason.

Speaker 6 (25:27):
And that should give you, first of all, really fast feedback,
because if there's something in the script part that you
accidentally did wrong or a rule you you didn't here
to you straight away like okay, fine, it will will
stop there and will be good. And in the future
we also want to support these template parts. But the
problem with that is for JavaScript and also for Typescript.
Of course we have a parser, we have a resolver,

(25:49):
et cetera, like under the hood, but doing that customly
for like I don't know, I said, like angler templates,
for for view templates, for fell components.

Speaker 4 (25:58):
As for components, it's we'd.

Speaker 6 (25:59):
Have to a like implement parsers for these traits.

Speaker 3 (26:03):
So tell me more about it.

Speaker 4 (26:04):
Yeah, exactly. Someone working with webtooling someone Yeah.

Speaker 3 (26:08):
Yeah.

Speaker 6 (26:10):
In that sense, JSX is a blessing from the tool perspective.
I just want to quote you exactly the words of
your Nicholas.

Speaker 5 (26:18):
Worth Man quit stro Man.

Speaker 3 (26:21):
No, No, he's great.

Speaker 4 (26:25):
Looks good for a third episode, I see, I see
checking all the box.

Speaker 3 (26:28):
Yeah, you're still on the good list for next one.

Speaker 4 (26:31):
Lovely.

Speaker 3 (26:32):
Yeah.

Speaker 4 (26:32):
For the formator. Just to to finish that up.

Speaker 6 (26:35):
So the formata works with JavaScript fully with typescript like
almost hundred percent. We still have a few test cases
that might might be at the point of the episode
coming out might also have like ninety nine something percent.
There's also an RFC open in general, because for example,
what we want to do there is we want to
do sorting. We want to use Prettier frombattered languages, so

(26:56):
we don't have to deal with like, oh, can you
form at CSS for me, It's like, okay, then we
have to run Prettier. But we also want to make
sure that all the concerns with the tool like line
with right or like this, that the printing line wrapping
and print with that. That's the way around that all
of this. This is also something that can be solved.
So it's not just like hey, another tool in the mix,
but it's actually something that will let us, say, make

(27:19):
most of the people happy out there.

Speaker 2 (27:22):
Quick question on that, because Island and Prettier kind of
grew historically, so like with Island nine, we had that
hard cut where the team said, okay, we separate the
stylistic rules with like line with and those things. Is
that something that you kind of like lesson learned. Okay,
let's focus on semantics versus stylistic things with those things

(27:43):
and oh excellent, excellent or.

Speaker 4 (27:45):
Excellent yeah both right axcelland yeah, I like.

Speaker 2 (27:48):
That doesn't even bother about those things, and we just
recommend you to use the formata if you have strong
feelings about space versus tap or how do you look
at those things?

Speaker 6 (28:02):
Yeah, so at the moment, these are exactly like, we
don't implement any stylistic rules in Auxland, except they are
in a core and careful good reasons like I don't know,
curly brackets or something like that.

Speaker 4 (28:12):
Because yes and core.

Speaker 6 (28:15):
But other than that, yes, this is part of the
formula regarding the stallistic rules. And I know a lot
of people out there are big fans of Anthony Foo's
Eslin confixts.

Speaker 4 (28:23):
I was like, hey, can we port it over? What
can we do?

Speaker 6 (28:25):
And this is also definitely something we want to take
a look at. Like with the custom javas could plug
and support technically you could run most if not all
of the rules already in there. We're not talking about
performat running, but at least there is a way for that.
But it's definitely also a target we want to look
into and see at the beginning, not every eslent API
will be available in auxcelland so we make sure, okay,

(28:46):
what is there so people can use that if we
want to, but for us want a separate that the
formator is doing formating and the lanitur is doing think
correctness not style.

Speaker 4 (28:56):
But that's also like getting.

Speaker 6 (28:57):
The tools out there and then evolving and also listening
to community feedback. So if there's anything in terms of
like I don't know feedback for Deformata RFC is similar,
that's always there's always great to let us know. So
even if it's just the thumbsup on like yeah, I
want I don't know build and tail on CSS support,
which we will definitely want to do, go ahead and.

Speaker 4 (29:18):
Let us know.

Speaker 5 (29:20):
Nice, So you just added for the show notes the
RFC for the format ort. Yes, this is like sometimes
people are very opinionated and they love to share their
opinions and this is a good place to do that.

Speaker 4 (29:31):
Exactly.

Speaker 6 (29:31):
Always always please in a constructive way, right Like I
I like, I don't let me be blunt here. I
don't care if you like tailment or hate it, like
it's fine, but we will support tailwands build Informator like
we will do that. And if you don't use tailwent,
you don't have to use that. All good, No harm
done here, right, So this this is actually discussion that

(29:55):
came up where it's like, hey, this is not the
place to discuss what do you like or hate tailent,
what do you think is just the most horrible thing
ever created or it's a god blessing. This is about
what you want to support and there is a big
need for that, so.

Speaker 4 (30:06):
We want to bring that in.

Speaker 5 (30:08):
Yeah, that's also it's sound that we have to point
that out to people, but thank you for bringing that
up in a way.

Speaker 6 (30:15):
In a way, it's always good to make that clear
upfront because it's like, it's not about what I personally
use or not. It's also like we take a lot
what people use. There's for example, so if you take
a look at the tea as conflict and the compiler options,
you have like a path option there, and everybody says like, hey,
this is not where you should actually write your resolution path.
This is more like to inform typescript that you have

(30:36):
set them somewhere else. And that's like just hey, please
respect them.

Speaker 4 (30:41):
But people do that. People do it all the time.

Speaker 6 (30:44):
And that's also we have a support like there was
a wet plug and now we have it in roll down.
We have that also support it just because this is
such a big use case, not because say, hey, this
is a great idea to do. So there are a
lot of things where in a way, you just can't
ignore that this is something is happening if you if
you don't like it, or if it's not.

Speaker 5 (31:02):
Recommended, yeah, well, and you know, I think it's also
like the RFC is a good place to point out
if something that you do use frequently is not being supported,
bring it up because you are you're probably not the
only person doing that every now and then you come
across a very special situation. But for the most part,
this like we follow the same patterns, and you know,

(31:26):
groups of patterns I guess, you know, like not everybody
wants the name code, but like if you're doing something,
chances are somebody else is too, And if it's something
you need supported, probably somebody else does too. And it's
a good place to bring it up because it's not
going to get supported if you don't bring it up.

Speaker 4 (31:40):
So agreed, agreed.

Speaker 6 (31:42):
For one good example for that is, for example, prettier
is always like doing auto close on the image tech
and other attacks that are not auto like closable and
with probably HMIL you didn't have that, but in JSX
that's the way. So there's also something people say like, hey,
it would be great to have PROPHTML formatting, and they
this is something that was brought up in the RFC
people up or death, so we will consider that and

(32:02):
this is great, so please keep keep the comments there coming,
and if there's anything as well, if you want to
have a chat, we have a lot of discourses for
like roll on for auxen, et cetera, for VAT as well,
so if there's any feedback, quite active in all of
them a certain way. So yeah, make sure if there
stuff of the chat to reach out.

Speaker 2 (32:22):
Good morning, you know that moment when your coffee hasn't
kicked in yet but your leg is already blowing up
with Hey, did you hear about that new framework that
just dropped?

Speaker 4 (32:31):
Yeah, me too.

Speaker 2 (32:33):
That's why I created the Weekly Death Spur, the newsletter
that catches you up on all the web deth chaos
while you're.

Speaker 3 (32:39):
Still on your first cup.

Speaker 2 (32:41):
Oh look, another anger feature was just released, and what's
this typescripts doing something again? Look also through the poor
request and change slot GRAMA so you don't have to
five minutes what's my newsletter on Wednesday morning? And you'll
be the most informed person.

Speaker 4 (33:00):
In your standard.

Speaker 2 (33:02):
That's better the Weekly Desperate because your brain deserves a
gentle onboarding to the week's tech matness. Sign up at
Weekly Brew dot death and get your dose of deaf
news with your morning caffeine. No hype, no clickback, just
the updates that actually matter. Your Wednesday morning self will
thank you.

Speaker 6 (33:21):
Maybe it would be good to like just give a
brief overview because we talked about a lot of tools
here and like just like the levels of abstraction, how
this all belongs together because in the way, okay, we said,
like we have the REAT plus which is a super
set of BEAT, and then we have veed as the
build tool, and then VET uses roll down as it's bundered.
You can also use roll on standalone, but the most
common use case we'll be just using veat. There are

(33:42):
of course companies using this stand alone. Just great for
the built scripts to replace roll up or yes built
or whatever or in case of Angler, as I said before,
the experimental like better chunking because roll on is really
really good at defining certain advanced chunking patterns, and so
we have that the idea, as S. Nicholas also mentioned before,
to unify the like say bundler level and wed with

(34:05):
like es built, roll Up cetera. Just make one thing
and roll up is more or less the best of
all the world, so to say. So we take some
es built behavior, but also of course it's speed in
a certain way.

Speaker 4 (34:16):
Just make it tiny bit faster.

Speaker 6 (34:17):
Even we take roll ups API and compatibility so it's
compatible to all the roll up.

Speaker 4 (34:21):
Plugins out there.

Speaker 6 (34:22):
And then we even borrow An got inspired by by
that pack in the end for all the advanced chunking part,
because that back was really really good at that and
neither es built nor nor roll up can can do that.
They're very limited. So ES built is very very limited
in terms of chunking. Roll up has like manual chunks,
but it's also it could be better, and that's what
we did with rolldown and from there when building a bundler,

(34:47):
we also need a strong foundation. So even before Boshan
joined Avoid Zero and bosh is the project lead of OX,
the choice was to say, hey, let's use Oecus the
foundation because it's had the fast module resolver, the fastest paser,
the fastest transformer, very very good, minifire at amon one
of the best like speed compression ratio because if you're

(35:09):
the fastest, then you don't have the smallest files, and
if you have the smallest files, you're not the fastest,
et cetera. So that that was the idea there. And
oxc also live, ox format the formuda and oxland the
linter nice. So that's like the level of hierarchy. And
on top of like I know, plus your framework in
your code sits more or less.

Speaker 7 (35:28):
Yeah.

Speaker 5 (35:28):
So like for the the everyday dev, especially for an
angular dev, most of this stuff is going to be
just pretty much obstracted away from you and you're not
really going to be worrying about it. But as you
start moving more into the architecture of your app and
like pipelines and all of that kind of stuff, this
is where you really start to care about how these
things work. I mean, you will care as an individual
contributor like oh gosh, this is slow, but you know

(35:51):
it's like it's not like a junior dev is going
to pick up a ticket to go make the linter
faster and actually fine, I mean, I have had some
weird tickets my way my career.

Speaker 4 (36:02):
This is a fun story.

Speaker 6 (36:03):
Someone someone I wouldn't say Junion developer, but someone is like, hey,
I I want to learn Russ start a contributing to
or eccenes. It's not part of the core team. So
sometimes like just curiosity can bring it quite far and good.

Speaker 5 (36:16):
Feedback absolutely, and I you know, I'm like, I also
don't want to bag on juniors because I also usually
prefer to say early career devs because a lot of
times they don't carry the same baggage that I do,
and so they might have an idea that I wouldn't
necessarily have.

Speaker 4 (36:32):
As well as well.

Speaker 2 (36:33):
So marketing taught me this fancy expression of junior professionals.

Speaker 5 (36:40):
They sound small like they sound.

Speaker 2 (36:46):
Yeah, one thing I'm wondering in all this because I mean,
you said it's yourself universal JAVASCPT tool chain.

Speaker 6 (36:57):
But yeah, also universally we're not a run time very important,
We're not a runtime we run on all the run times.

Speaker 2 (37:06):
Sorry, how much are you collaborating with like maintainance off
those things, because like I know for a fact that
Angela is a little bit more doing their own stuff.

Speaker 3 (37:15):
With the Anglo Celi versus.

Speaker 2 (37:17):
Yeah, you, I would assume you have, like White, a
close contact to the view folks for given aspects.

Speaker 4 (37:24):
I think Evan knows the guy there, yeah, like one
or two maybe yeah, possibly possibly, uh No.

Speaker 6 (37:29):
I think especially for the ved ecosystem, we're really close
in touch with most of the maintainers because that's the
big benefit of feed, kind of like bringing the whole
ecosystem together and making that possible. So I would say
very especially for like rolldown feed, when people like okay
want to prepare for rolldown, we've seen like h of course,
like view nuxt people put put things that work in

(37:52):
But before that, like quick migrated to a rolldown even
for their like browser Playground and got for example for
the browser playerund like four times the speed the smallness
work of people who are really like into that. I've
seen regular people from tanstac like, hey, how's it going there?

Speaker 4 (38:07):
Or is that maybe an issue?

Speaker 3 (38:08):
Is that?

Speaker 6 (38:09):
So I think they're all quite involved. Of course, some
more and some some less. But in the end, also
given that Evan is the creator of Weed knows those
a few folks here and there, that's that works pretty well.

Speaker 3 (38:24):
How involved are you still these days with the because
I was talking to.

Speaker 2 (38:33):
Weeks ago, yes, and he basically said that at some
point when once we took off Evan more or less
handed it to the community to steer.

Speaker 6 (38:43):
So well, I think this is in a way yes,
in a way no, Like as I said before, it's
some kind of a multi stakeholder project. But I think
right now it also transforms a bit more into like
under the vision of Void zero, with a lot of
input from the ecosystem. So the ecosystem is still an
important part in the community, like we couldn't do it
without them, and we're really grateful to see like so

(39:04):
many people not starting framework wards but all collaborating together,
and also like saying, hey, there are things going wrong
in different frameworks, let's find a good solution FEAD so
nobody has to write their own workarounds for that, sometimes
in a more constructive fashion, and then sometimes it's a
it's a quick post on Twitter, which is also a
way to criticize people.

Speaker 4 (39:23):
I get that nowadays.

Speaker 3 (39:26):
Yeah, well, some people on Twitter also deserve criticism.

Speaker 2 (39:30):
I had a good amount of it actual last couple
of days, so exactly what I'm talking.

Speaker 5 (39:35):
About iMore I just like, no, you did not.

Speaker 3 (39:38):
Miss out it's.

Speaker 6 (39:41):
Or a blue sky even even there sometimes but no,
but but in the end, like it's so we have
so just for example, from our perspective, we have two
people working like full time on we have Safi and
Biroschi on.

Speaker 4 (39:58):
Yeah.

Speaker 6 (39:59):
So, but of course the rest of the v team,
like you mentioned Patsak already he is part of stack
blitz right so bold on new We have other like
a lot of other contributors that are not affiliated with
Void zero directly. So in a way, uh, this is
still quite community driven. But of course given that we
ste the direction of OX and roll down, we also

(40:20):
have quite some input there. But it's not like we say, oh,
this is the way and we'll we'll do it that
way and that's it. So did the whole depth community
is important to us? Yeahs sense And actually on the
note and Nichols, you said that earlier like Angler projects a.

Speaker 3 (40:35):
Lot of things.

Speaker 6 (40:36):
That's true, that's true, but that specific part is like, Okay,
they all have very similar use Angler, Steel I and
an Axe, et cetera. I mean, I know they're like
there were projects saying oh, we want to bring other
bundlers like into Angler instead of yes, built and like
hack things around, and I think the same could happen
with rowdown or maybe I don't know. Evan also mentioned
at some point on Twitter, like, hey, instead of using

(40:58):
roller only for experimental chunking, why don't you we all pulled.

Speaker 4 (41:00):
A roll down because it could solve a lot of problems.

Speaker 6 (41:03):
But for that, it needs to be mature and stable
first of all.

Speaker 4 (41:06):
So it's not even one seri.

Speaker 3 (41:08):
Yet that is part of it.

Speaker 2 (41:10):
And the Anger APIs are so encapsulated that makes innovation
from the community very difficult.

Speaker 3 (41:16):
So I know for the fact that analogue who provides.

Speaker 2 (41:19):
A VAT integration uses private APIs, and that is just
possible because Brandon is basically.

Speaker 5 (41:27):
Brandon is the one that's going to do that.

Speaker 2 (41:31):
And like it creates an environment where innovation is somewhat blocked,
and therefore we're completely right lying on the Anger team
Like Wheat was already public and like publicly and popularly
used in every other framework, and Anger thats were like
beat never heard of that.

Speaker 5 (41:50):
Yeah, we literally didn't. We did an episode about it
where we're like, what is Everyone else is like, oh
my gosh, you guys, that's so bat.

Speaker 2 (42:01):
B test is from my perspective, literally the best thing
that ever happened to testing and JavaScript in the JavaScript ecosystem,
hands down.

Speaker 3 (42:10):
I would not use anything else.

Speaker 2 (42:12):
Yeah, and still it's just experimental and angular.

Speaker 6 (42:17):
Yeah, I know, I know, getting there, getting there it's
sometimes it takes a while, especially in the Angler universe.
I've learned, right, that's fine. And also I mean if
the chances, if the stars align, you know, and roll
down as a stable and then be like, hey, maybe
we can do something there. If you using it already
in experimental for a small part, maybe that's yeah.

Speaker 4 (42:38):
Well let's see, let's see.

Speaker 6 (42:39):
But in the end I don't think it's it's super
unrealistic there. But what I actually wanted to say in
extent to that is also like with NX, then like okay,
you can run es and you can earn prettier as
you said, and there are also any innovation might be
but a bit difficult because that's like how do I
run something like oxland or ox foment in there, right,
or even I don't know biome or so I've seen
like someone wrote like an auxland into gration for an X,

(43:00):
but you also need some work to do that get
it done, which is always a bit tricky.

Speaker 2 (43:06):
Yeah, I'm not as concerned about an X as I
am about Anger in general, because they have a very
good pluck in ecosystem and good APIs to foster that
are also more open.

Speaker 3 (43:18):
And supporting other technologies.

Speaker 2 (43:20):
Whereas for the Anger team, it often feels like this
is the tool set that we can reliably support, that
we know that works, and you need to kind of
aggressively slap it in the Anger face. But this way
it's way better and eventually they recognize that, and so
it just takes long. I don't I don't think this
is out of and I absolutely don't assume any malicious

(43:41):
intent with that. It's more like the Anger team tries
to provide the experience that they can guarantee works reliably
for most I mean.

Speaker 5 (43:52):
Like Google, Like Google itself supports a ton of Angular apps.

Speaker 3 (43:56):
And so like they're using that basis.

Speaker 6 (44:00):
So that I just wanted to say as well, like okay,
like Anger needs to work for the biggest client, which
is Google. Right, this this is something that's always have
to be kept in mind. Otherwise migration well we all
know has to be done. And that's also not noting
nothing people.

Speaker 5 (44:16):
Yeah, literally, anyone I talked to about Angler who hasn't
used Angular in a while. The reason they don't is
because they're still so mad about one to two.

Speaker 4 (44:25):
So yeah, back in the day.

Speaker 2 (44:31):
So glad though that the view community can kind of
feel that pain.

Speaker 6 (44:34):
Now, yeah, I think this is a little bit less
drastic than Angler. Jest to Angler, because it's a new
framework right for views. I mean, I'm curious how like
the remix two to three, I'm.

Speaker 3 (44:47):
So excited for so that that's happening, is that Friday.

Speaker 4 (44:52):
It's happening the same time as a week.

Speaker 6 (44:53):
Confident it happens already for people listening, and we have
no idea what because we're still in the past.

Speaker 5 (45:00):
This is the weirdest time machine to be in.

Speaker 4 (45:02):
It's it's wild.

Speaker 6 (45:04):
But one thing to actually like point out there we
talked about like experimental things. One thing the Angler team
also wants to look into is using Nitro, for example,
for the CLI. And this is also interesting because Nitro
became assuming that this is like it's it's there, like
it's a VAT plug in now right, you can use
Nitro as a VT plug And so it will also
be interesting how the full stack story of VAT will

(45:26):
look like because I.

Speaker 4 (45:27):
Think one of the big shortcomings.

Speaker 2 (45:29):
Yeah, so is that basically the Nitro answer to Vincy
is that.

Speaker 4 (45:34):
Well, oh that's difficult.

Speaker 6 (45:39):
I think Vincy was like the idea of saying, we'll
take nitro and then we make some kind of feed
environment API layer on top of that.

Speaker 4 (45:47):
While the main.

Speaker 6 (45:49):
Point now was also like okay, as for example, tan
stacks start and also solid start, like okay, nitro should
be optional. You can use something else and they're just
a VAT plug in, but with nitro being a v
plug and will very easy to integrate because like, okay,
then the plug in here.

Speaker 2 (46:02):
And sorry I'm a little oblivious, but would then night
would just be be used as a death server to
ship the files or what are we talking as a
what is the purpose of the slug?

Speaker 6 (46:14):
So the thing is, let's let's think about it differently.
If you just have a VT application, right, let's let's
think it's just VAT and then whatever framework it's the Yeah,
and it kind of been just like playing JavaScript correct
or like Vanilla, how do you write your back end?
Commonly you say like okay, I spin up a different server.
I don't know, use some Nitro deer or like I

(46:35):
don't know hon no, FASTI FI, express CoA, whatever, right,
any other framework that wasn't named here, But what if
you could have it all one's like an actual full
step application. I think this is one of the features
that a lot of meta frameworks provide, like NA Swell, Kid,
et cetera, based on top of FEAT, but they don't
do it through WEED because VAT is not really having

(46:56):
some kind of serve API. So that's definitely something that
that we're looking into as well to give this like
full stack experience out of the box and also close
the gap to say, like, oh, should I use VT
or like, for example, something very common or VT can't
do se O and SSRs that are which is not
even true. We just don't know how and we won't
make it easier so you can actually write full stack applications.

(47:18):
So I having some kind of server API will be
We'll be very interested there, and Nitro could be potentially
a good partner there or a tree at least right
the let's say lower level bullion look of that.

Speaker 3 (47:31):
Sorry a little bit.

Speaker 6 (47:33):
Okay, a lot of a lot of info at the
same time, So I get that, but it's also the
way the idea, right, we want to have this Nitro.

Speaker 4 (47:43):
Yees, So Nitro is a server tool, right, this is great.

Speaker 5 (47:48):
Okay, I'm going to google to Alex what is nitro? Yeah,
tell me here's.

Speaker 6 (47:53):
My explanation of nitro. So, yeah, so nitro is a
Natro is a server tool. Basically it power is different
frameworks like analog like handstag asauls start right now, but
they want to make an option which also find like
next and you can also use the standalone is like
basically an express replacement with more integrations like a better
key value storage integration web soog is that just work,

(48:15):
et cetera, et cetera. So, and the idea is that
I think of like VT itself is like the shared
infrastructure of like the front and part and like assets handling,
et cetera, et cetera. And this like server infrastructure part
was missing a little bit and this is what Nitro
is doing. So I think like if we can collect
Vita Nitro somehow, that would be that would be pretty

(48:36):
powerful to say, okay, you can deploy two different providers
like cloth there like I don't know, netlifier, your own VPS.
You have different presets that you don't have to select necessarily,
like if it's in a Netlifi build step, it will
detect that and we'll just make sure. Okay, I use
the settings there right the right disc folder. I can
then export a Netlify function for example, for like service functions,

(48:57):
and my VPS is just JavaScript workers. It's a worker
file in the end, so you don't have to deal
with that.

Speaker 4 (49:04):
The only thing.

Speaker 6 (49:05):
Yeah, the death server is the process right now, only
working production, so having deaf server prices would be interesting.
But also there right now we have a lot of
VAT plugins from like the different platforms, like software like Netlify,
so it would be.

Speaker 4 (49:17):
Interesting how that will turn out.

Speaker 6 (49:18):
But at least for the production to say, okay, I
can just use my code in Nitro and I can
deploy anywhere without changing much.

Speaker 4 (49:26):
That's that's the idea right now at the moment without
VAT integrated.

Speaker 2 (49:31):
Okay, I think I can formulate my question now. So
if you have a Nitro VAT plug in, so usually
takes all the files and those set at the right
plug and so like build view file or like compile
view YadA, YadA, YadA, you know the draw and would
then for night to just kind of be like a

(49:51):
pointer to say, hey, those files everything and slash server
is server files or how does this look like?

Speaker 6 (49:58):
I'm yeah, so actually, so you can take a look
at it all right, there's an example for Nitro already.
I'll paste it here for.

Speaker 4 (50:07):
The show notes real quick. And the idea would be
there are.

Speaker 6 (50:10):
A couple couple of examples around and it's also not
like it works with I don't know, for example, reacting
as is already or similar. So you have that v
confiict and you basically say, okay, you add your your
Nitro plugin in there, and from there you can handle
that by just saying okay, I have a source folder
and like I don't know, a server dot ts and
the serve dots you you would handle the rest if

(50:32):
you want to, or like a different, different way of entry.
So it would be it would be able to be
configured there, but it's also at the time of recording
all very much to do. You can always take a
look at Preas talk from Vcon if that's probably still
online because it was live streamed, where he showed a
few things around that, and how that will look like.
So yeah, I guess in a couple of weeks from now,

(50:54):
we'll we'll know more. That's for Nitro itself. It's it's
super exciting to see like these more let's say server
side capabilities to like close kind of bridge the gap there,
because there's definitely something I was missing from just plain
VIAT for for a while.

Speaker 3 (51:11):
It is cool.

Speaker 2 (51:12):
I'm coming from the deaf tool side being very concerned
about like sharing context. I mean, I know you had
the issues a nuxt with before of like making the
server part not detectable in the client part with different teas,
convict parts and those things. Yeah, and I can already

(51:32):
see the complaints and bugs from people that just can't
figure their project wrong.

Speaker 4 (51:38):
Yeah.

Speaker 3 (51:38):
Fair.

Speaker 6 (51:38):
I mean the best part is if if FEAT would
provide some kind of native solution to that, right, I
think this would be like the ideal scenario where it
would just be like, look, here's some kind of FEAT
serve API, and yeah, let's let's see what will what
will happen here?

Speaker 3 (51:54):
That is cool? That is cool. I like that.

Speaker 5 (51:57):
Yeah, it'll be fun to have you on when this
has matured more to talk about it again.

Speaker 6 (52:03):
Yes, right now, it's a lot of speculations and what
if and like experimental parts.

Speaker 4 (52:08):
So there there we have a topic episode.

Speaker 5 (52:10):
I guess exactly, yeah, I'm in another kind of time machine.

Speaker 6 (52:13):
Now then we should react in episode three through episode two,
it's like yeah, back then, okay, yeah, that will be confusing.

Speaker 5 (52:22):
We can play back things you said that we're wrong.
We'll be like remember when you said this, like, oh
my god, I just didn't know.

Speaker 4 (52:29):
Yeah exactly, it's like, yeah, back then that was the plan.
But actually yeah.

Speaker 6 (52:34):
So also there, if you're if you're interested in trying
out like the nitro beat pluggin, there is a big
issue uh for like a big discussion for for nitro
beat feedback, as nitrogen free is just becoming a fat plugin.
So also there, if you maybe have some nitro application
already or just curious to poke around, please do that
and always reach out.

Speaker 3 (52:55):
Okay, we talked about bazillion things.

Speaker 2 (52:58):
Yes, so if you if you would want the listener
to walk away with one key information of this episode.

Speaker 6 (53:06):
Well I would say it's a pretty pretty straightforward thing.
If you have tooling pain, void zero can solve it
either now or in the future. But the actual t
ILR is Look, if you're interested in v plus, feel
free to join the waitlist. But more interesting, take a
look at all the tools that you mentioned, Take a

(53:28):
look at the show notes. There will be something for
you in there where you think, hey, either I've never
heard of that, or it sounds so interesting I have
to take a look. Because we all want faster tuling
that's also ecosystem compatible. So even said like, hey, I'm
running my Angle application, I can't use I know we'd
only as step server. I can't can't rest. Maybe you
heard about the experimental chunking with roller. Now you can

(53:49):
enable it and try it out and get some well
better chunking out there which means a better performing application
eventually at run time.

Speaker 4 (53:56):
Or you say, oh yeah, linting. Of course this is
a pain point. I try out ox et cetera, et cetera.

Speaker 6 (54:00):
I think this is like a bring take your own
whatever we talked about away from here.

Speaker 4 (54:06):
But if you look for a tooling, you know where
to find it.

Speaker 5 (54:09):
I would. I would also point out to like Yan
you pointed out how sometimes it's hard in the Angular
community to bring these tools in. Yes, but if we
aren't aware that these tools exist, how can we know
to ask for them? Because a lot of times the
Angular team is taking feedback from the community, and if
nobody's asking for oxcellent, they're never going to like it's

(54:30):
not going to be on their radar. Necessarily, it might be,
but like, yeah, I'm going to un video, I'm gonna
hide you with my microphone.

Speaker 2 (54:39):
The thing is like on a fundamental level, I agree
with you, right, But then the fact that the Anger
team is not.

Speaker 3 (54:45):
Aware that there might be no.

Speaker 5 (54:47):
No, no, they are aware, but they might not understand, they
might not see that the community wants it, right, Like
letting them know that the community wants it, like, maybe
that helps.

Speaker 7 (54:57):
I agree with that, but all so like, if you
make the cognizant decision redctate the tooling at your use,
then you cannot put yourself in the position and be like, oh,
actually sorry.

Speaker 3 (55:08):
We did not know.

Speaker 7 (55:08):
That's that's not how this works.

Speaker 5 (55:11):
Well it is that is like.

Speaker 3 (55:13):
Literally a toxic relationship.

Speaker 6 (55:16):
What I would what I would definitely agree is like,
if if there are tools that you haven't heard of,
then there's a big chance that other people haven't heard
of fighters, so try it out and spread the word.

Speaker 4 (55:25):
I think that never hurts.

Speaker 5 (55:27):
Yes, well, and like I mean honestly, like I wouldn't
know what to I wouldn't know what I wanted if
I didn't know what's out there. Right, Like when we
were using webpack, it was fine because I didn't have
time to go deal with, oh is there something better?
Because that wasn't my job, Like my job was write
the code, you know, But that's just my point.

Speaker 3 (55:47):
I mean, you also didn't ask for us built.

Speaker 5 (55:50):
No, I know that. I know that, but at least
when it came when if I knew about Like, if
if I take the time to understand what Nitro can
bring us, then I can be excited when it's coming
out and I'll understand. I can talk to people about
how it might help us.

Speaker 3 (56:05):
Like, yeah, I agree with that.

Speaker 2 (56:07):
I just say if the the Anger team literally put
themselves on the position by saying, no, those APIs are
not public, wait, wait, take care of that for you.
Wait wait, we got you bow, and I'd be like, well,
but but VAT is pretty popular, Well.

Speaker 3 (56:23):
Maybe next year. That's just not how this works.

Speaker 5 (56:26):
I feel like if we pester them even more about VITAS,
they'll just go ahead, DIDs you not.

Speaker 2 (56:33):
I think they're going to kick me from the GD
channel at some point.

Speaker 6 (56:41):
But see this is the thing. You just need more
allies there easy. So this is the way spread the
word and till like, can you also send the message
about want to all.

Speaker 5 (56:50):
Show up in the GD space demanding vitas easy.

Speaker 3 (56:54):
We can make like little signs.

Speaker 8 (56:56):
Like it, get special hats you know, it's for the
Google I shop like angular and vould be hilarious.

Speaker 2 (57:08):
Like a non violent protest.

Speaker 5 (57:12):
At the IOWA event.

Speaker 2 (57:14):
You didn't get the idea for me here and it's
not Oh wait, well it's not recorded, no quoting non violence.

Speaker 5 (57:22):
It's totally fine.

Speaker 6 (57:24):
One good change for the framework ecosystem.

Speaker 2 (57:29):
Having said that, test is still the best thing and
they should absolutely do it.

Speaker 4 (57:32):
Yeah, agreed, Thank you so much, Alex, Yes.

Speaker 5 (57:35):
Thank you so much for joining us and sharing your
knowledge and just always being on our good list of.

Speaker 2 (57:41):
Guests always is a strong statement.

Speaker 4 (57:45):
No, no, I go with always. It was said, it's recorded,
we have all the way.

Speaker 2 (57:49):
I don't even think I'm always going to be a
house always on the good listness.

Speaker 3 (57:57):
That's how No, No, I'm my vote doesn't count anymore,
doesn't count.

Speaker 5 (58:03):
No superseded, Thank you, no problem. If somebody wants to
reach out to you, what's the best way to do that.

Speaker 6 (58:13):
Yeah, best way is either socials on Blue Sky or
on Twitter, which I will always call Twitter. I think
if you shoot, the md ms are open there. Otherwise
one of the project this course. If you have a
general question and yeah, I don't at one of the
countless conferences for this year not that many, but definitely

(58:33):
next year are isn't not around again, but yeah, it
might be the easiest. Feel free to shooting questions. And
if you have tried out it's also very important. If
you tried our tools and you feel some kind of
block or issue, then please let us know. Also there
once again, we contest all the frameworks with everything that
we set up so that we rely on of you
to well. We try the dog food as much as

(58:55):
we can, so please please do.

Speaker 4 (58:59):
Yeah, we need dox exactly.

Speaker 5 (59:02):
Excellent, and I can't confirm that. If you see Alex
in person, go up, introduce yourself, say hi, just friendly,
great person. It's always I'm always happy to talk with you.
And you just bring a positive influence to the communities
that you're involved with.

Speaker 4 (59:18):
So that's the goal. Yep, absolutely, thank you so much
for having me.

Speaker 3 (59:24):
Absolutely apparently any time.

Speaker 5 (59:27):
Anytime you're the best guest. You're the new host. Yon
just got kicked up.

Speaker 6 (59:34):
I have to learn a thing or two more about
Anglo though. Okay, question, how is it T used signals
or r x chs.

Speaker 2 (59:43):
Both?

Speaker 4 (59:46):
At least right now it depends that.

Speaker 5 (59:50):
Same thing is both depends? All right? Well, thank you
so much for joining us to the listener. Don't don't
for get to subscribe if you like this content. Your
subscriptions definitely help us keep making episodes and we will
catch you next time.

Speaker 9 (01:00:11):
Hey, this is Prestolin. I'm one of the NGI Champions writers.
In our daily battle to crush out code, we run
into problems and sometimes those problems aren't easily solved. NGCOMF
broadcasts articles and tutorials from Angie champions like myself that
help make other developers' lives just a little bit easier.
To access these articles, visit medium, dot com, forward slash, ngcomp.

Speaker 1 (01:00:33):
Thank you for listening to the Angular plus show in
CHICOMFF podcast. We'd like to thank our sponsors, the NGCOMF
organizers Joe Eames and Aaron Frost, our producer Gene Bourne,
and our podcast editor and engineer Patrick Kay's. You can
find him at spoonful ofmedia dot com

All Episodes

S10 E7 | Vite’s Next Chapter: Alex Lichter on Rolldown, Oxc, and Beyond

Episode Transcript

Popular Podcasts

Stuff You Should Know

Dateline NBC

The Bobby Bones Show

.css-15opob5{left:0;position:absolute;top:0.8rem;} All Episodes

.css-14f5ked{margin:0;word-break:break-word;display:-webkit-box;-webkit-box-orient:vertical;box-orient:vertical;-webkit-line-clamp:2;overflow:hidden;}S10 E7 | Vite’s Next Chapter: Alex Lichter on Rolldown, Oxc, and Beyond