Indexing the Ruby World with RubyEvents.org - RUBY 677

Hey everyone, Charles Max Wood here! It’s great to be back behind the mic with Ayush Nawatia for another episode of Ruby Rogues. This time, we’re diving into the fascinating world of Ruby community resources with two amazing guests: Adrien Poly and Marco Roth. They’re the creators behind RubyEvents.org, an ambitious project that’s aiming to centralize and modernize access to Ruby-related videos, conferences, meetups, and more.

In this episode, we explore the vision and evolution of RubyEvents.org—originally RubyVideo.dev—and how it’s grown into a hub for the Ruby community. We discuss the technical stack (think SQLite, Tailwind, Vite, and Hotwire), how they’re leveraging LLMs to auto-tag thousands of conference talks, and what’s on the roadmap next. Whether you're a conference organizer, speaker, or just a Ruby fan, this one’s packed with insight and inspiration.

 Links & Resources

• RubyEvents.org – Discover and explore Ruby conferences, talks, and speakers
• Ruby Video GitHub Repo – Open source code for web and mobile apps
• StimulusUse – StimulusJS utility library created by the guests
• Hotwire DevTools – Browser extension for debugging Hotwire apps
• HYROX Fitness Race – Hybrid CrossFit endurance competition mentioned by Adrian
• Camel Up Board Game – Charles’ board game pick
• The Overview by Steven Wilson – Ayush’s music pick
• Sword of Truth Series by Terry Goodkind – Fantasy book series
• Catalyst (Web Components framework) – GitHub’s lightweight JS framework

Become a supporter of this podcast: https://www.spreaker.com/podcast/ruby-rogues--6102073/support.

Episode Transcript

Speaker 1 (00:04):
Hey, welcome back to another episode of the Ruby Rogues podcast.
This week, on our panel, we have a Yushwatya.

Speaker 2 (00:11):
Hello. Hello.

Speaker 1 (00:13):
I'm Charles Maxwood from Top End Devs. Man, it feels
like it's been a while since we've recorded one of
these dozen.

Speaker 2 (00:18):
Yeah, it's been a few months. Yeah, at times, just
been flying for me. So it's all gone back quite quickly.
But it's nice to be back.

Speaker 1 (00:29):
Yeah, definitely. For those of you who are listening and going,
didn't you release one like last week or the week before, Well,
there were a few things that happened. I'm just going
to fill you all in. I don't know if you
need all the insight baseball, but it's interesting, and then
I'll introduce our guests. I wound up getting a job
at Price Picks at the end of March. I also
hired my daughter to be putting the episodes out and

(00:52):
she didn't and so I fired her. I had to
fire my own kid, and so Jib who's our editor,
took over posting the shows, and so he went and
found all the ones that hadn't been posted and started
posting him again. So the gap that you all, as
listeners saw was probably back in like April. Right, there

(01:13):
was like a month in March or April that you
didn't see episodes. So you're getting episodes that we recorded
a month or two or three ago, and now we're
kind of back at it and recording episodes again. So anyway,
your gap is different from ours. So but it feels
so good to just get back and talk about the
stuff we love. And we've got two terrific guys who

(01:36):
were here who are working on a very interesting project.
We have Adrianne Poli and Mark and Marco Roth.

Speaker 3 (01:47):
Hey, thanks for having us.

Speaker 1 (01:50):
Yeah, absolutely, and yeah, I think I'm the only one
of you guys that's not in Europe, so thanks for
coming in the evening or afternoon whenever it is there.
Uh yeah, So I reached out because I ran across
a really interesting project. It's it's Ruby video or Ruby
videos dot dev, which one is it.

Speaker 4 (02:12):
It started as a Ruby video and uh, like two
months ago we decided to rebrand it for a bigger
picture on the under the Ruby events dot org. The
project has more ambitions, but we can dive a bit

(02:34):
more into the details and that's later on.

Speaker 1 (02:39):
Yeah, makes sense. So when you started, I'm kind of
curious what what was the vision I was?

Speaker 4 (02:50):
It was several things occurred at the same time. First,
I visited a similar project in the Python world that
is called the t Video or something like this, which
is a website that indexes something like fifteen thousand Ruby

(03:10):
Python videos of conference and so on. And I was
looking at something similar in the Ruby industry and I
couldn't find anything. Of course, you can always go to
YouTube or Google search and search for whatever things you need,
but I couldn't find a single place where you could

(03:31):
search for a specific topics and get all the conference
on that particular topic in the Ruby industry. So at
that time, I also I was interested to test several
things that were coming out that there was come out.

(03:54):
There was this Esclie trends, and I wanted to build something.
So I started to build a small website that started
with I think eight hundred videos, mostly rails cones, Ruby
coones and so on. That was the proof of concept,
the first proof of concept so largely inspired by a

(04:17):
Byton video, and it had a bit of traction at
the beginning. When it was announced. Of course, it was
open source, and very quickly Marco joined the venture and
came Marco was also maintaining Ruby Conference. If you want

(04:46):
to add more details on this, maybe Marco.

Speaker 3 (04:49):
Yeah, so I was maintaining Ruby Conferences dot org, which
used to be this simple website where you can add
o where you can see all the confidences are happening
in the Ruby world.

Speaker 1 (04:59):
I use all the time. By the way, it's like
what's coming up and where can I submit calls for
proposals and stuff? So yeah, thanks for that.

Speaker 3 (05:08):
Yeah, that's that's been like going for a long time.
It's I have been just trying to bring in the
up to date and kind of kept adding these new
events that I saw coming on. But then I realized
that they also had like video links to YouTube playlists
and all these other like topics they had on the website.

(05:28):
And then we decided to add meetups, and then we
were like, yeah, let's maybe redesign this webpage because it
does look dated in a way, so you wanted to
kind of add more features to it. But it kept
kind of like his ideas kept growing and they couldn't
really find a good way to make it happen on
this checker website. And at the same time, then Ruby

(05:50):
video came around and it was like, Oh, actually, this
is very much like what I kind of hadn't visioned
for Ruby Conviences to be. That's why I decided to
kind of join forces with oddly On to kind of
make this one website or kind of contribute to Ruby
Video to make this the thing that I wanted Ruby
Confidence to be. And that's kind of what ties into

(06:11):
this rebranded bed a few months ago, where you want
to bring these two websites together as one. This is
going to be Ruby events dot org and that's in
the future where you will find all the upcoming events,
previous events, and even if you're at the event at
the moment, you can see the schedule and all of that.
So it's kind of just a deep place to go

(06:32):
for anything Ruby events related.

Speaker 1 (06:36):
Nice. I usually you have questions because I've got a
ton and I don't want to.

Speaker 2 (06:41):
I have a ton of why didn't you continue, I'll
jump in as.

Speaker 1 (06:44):
And when okay, sounds good. So I'm kind of curious.
So you decide you want this website, and I think
the closest thing we've had in the past was Confree's
used to host all the videos from all the conferences
that they recorded. But I went and looked, and I
and find them anymore. And I don't know, I don't
know what the story is there, if I'm just not

(07:05):
doing the right Google search or whatever. But anyway, so
you start putting the site together and I mean, how
are you finding the videos? We're just embedding them from YouTube?
And were there any challenges that you're into kind of
pulling this together?

Speaker 4 (07:23):
Yeah? For now, we don't host any videos, so we
just index them. So we mostly find them on YouTube
because that's where mostly events are hosting their videos. A
few do host them by themselves, like Brighton. I think
Brighton is an example for something quite recent, but mostly

(07:51):
it's YouTube videos, so we do get the meta data
from YouTube. There is there has been a very big
work and Marco contributed a lot on this to build

(08:11):
all of those yanal file that will describe all the
meta data for all the talks, all the events with
both their links and so on, with the correct speakers,
the correct description. Lots of cleaning of the meta data
has been done.

Speaker 1 (08:28):
So do you import it and then clean it up
by hand or do you go find the videos that
you want and then just capture all that information somehow?
Or how does that work.

Speaker 2 (08:39):
Yeah.

Speaker 3 (08:39):
So we have a few what we call metadata parsers,
and some events have the same structure in the title
of how they arrange their titles, so we can kind
of do some smart logic by splitting on the word
by or by dashes or commas, and then you get
there like ninety percent of the way and then there's

(09:01):
like some left over to do where it just go
in and cleaned up by hand. But I guess this
could also be improved in the future now that we
have LMS and all of that to be more automated.
But for now this has been proven to work out
really well.

Speaker 1 (09:18):
Right, And you said that you have it all in
YAMO files. So when there's a new conference or I
was kind of assuming it was built on like rails
or Hanami and had a database behind it. But is
this more along the lines of something like a jackal
or Bridgetown or something where you know, maybe you say, hey,
my data is in these files, and so I'm just

(09:40):
you know, I just build every time I get a
new conference.

Speaker 3 (09:44):
Yeah, so for now, it's it is a rails up
and it has a database. Okay, but what we do
is that we go through all the AMMA files and
import them into the database, just so we have a
nicer interface to queer the data and kind of half
speak profiles and event pages that's just easy to quer

(10:05):
you and kind of build the websites around them, so
they are all genamically generated on the service side and
then served using the traditional rails requests response cycle.

Speaker 1 (10:18):
Gotcha, And I guess the other question I have is,
you know, are you mostly focused on conferences or are
you looking to do things with like meetups or if
somebody puts up a good tutorial or something like that,
you know, those are all videos that you may or
may not want to pull in. I guess tutorials aren't events,
but meetups are.

Speaker 2 (10:38):
Yeah.

Speaker 3 (10:39):
That was one of the reasons why I personally also
wanted to move away from Ruby video, because this was
meant to be focused on events, which also includes meetups,
and we have a few meetup groups on there too,
and not all the meetups are recording videos for the
events and talks, which is also the other thing that

(11:00):
I want to get away from Ruby video, because we
want to still index all the talks that are happening at
these meetups, even if they have no recording to it,
So that you can at them as the titles, at descriptions,
at the media data, at some of the slides if
they have them available, or some of the resources like reposts,
that you can still see that this talk and event

(11:22):
took place at this time by the speaker, and this
is all the information we know about it as well.
But that's the thing. Like with tutorials, I think this
is something that's harder to index in that sense that
it keeps like it stays relevant, which I think this
is still really good on YouTube or anywhere else where

(11:44):
you host these tutorials. So I personally don't see at
least in the near term that we kind of host
these tours on on Ruby events as well.

Speaker 2 (11:56):
Okay, So I wonder if we could dig into these
technicals a little bit, because you guys have built an
actual production website using sequel Light and there some new
feed reels features like solid que and also I think
it's the first website that I can definitely remember has
used page view transitions. So in terms of all these

(12:20):
technologies in an actual app, what was your experience.

Speaker 4 (12:24):
Like, Yeah, as I said, when I started to build,
I love to build things, but I am an engineer,
so I am an engineer builder, so whenever I tried
to build something new, it also has to be a
way for me to discover tests some new technologies. So

(12:48):
real builders will use proven tech and they just focused
on the on the product, the feature, the and so on.
So for me, as I said, when I had this
initial idea of Ruby video, there were several technology that
all came out and I wanted to try, and page

(13:09):
view transition was one of the technology that I really
wanted to try on. They had this demonstration where there
is the same effect when you click on the small
thumbnail of a video, then it's zoom up and you
get this nice transition from the index to the show view,

(13:33):
and I think that's pretty cool. At the same time,
I had this feeling that in the rails industry, you know,
always with the same topics, rail is dying and so on.
A lot of the the impression that rail was old
comes from some website built on rails that are that

(13:57):
have some old look that they are not modern. Uh,
that the back end is great, but the front end
is built by back end developers, and I want it also, uh,
this website to be some kind of flagship and that

(14:18):
demonstrates that we we can build. We can get a
nice uicks uh and UI with rails. Uh. You don't
need to react, you don't need whatever front end framework
to do this. And there are lots of technology availables
nowadays with base you transition with Delawyn and some some

(14:41):
component libraries that you can use to get something that
is nice.

Speaker 2 (14:45):
Yeah, it looks great as as quite impressed. We just
are clean and simple designers, which is awesome. Does the
page transition to API hooks into tablet drive? Doesn't it?

Speaker 4 (14:59):
When Hm, I started the project, it wasn't available with
Turbo drive, so I use a package I don't recall
the name, yes exactly, that provide the the bridge between

(15:19):
Turbo and Pagroot transition. Since then, uh, Patriot transition has
been built in into Turbo, so now it's uh, it's
a native feature.

Speaker 2 (15:32):
Cool. And how was this the first time you'd use
sequel light on a production app or how do you
used it before?

Speaker 4 (15:41):
For me, it was the first time. Yes, how did you.

Speaker 2 (15:44):
Find that change from I'm guessing postgrass that you used
in the past. Was it was it a big change
in terms of thinking and designing or was it pretty
much seamless.

Speaker 4 (15:56):
Mm hmmm. I would say it was seamless in terms
of a developer experience. What is nice is that doing
a backup or if we need to dump some database,
it's just a single fight that you need to to move.

(16:23):
Combined with Camal and a small VPS hosting made the
things amazingly cheap to host. That's that was I mean,
compared to Heroku or whatever past where you always end

(16:43):
up with a minimum of fifty year old per month,
we could divide this almost by ten and get better
performance out of it.

Speaker 1 (16:54):
So the performance language, I love that. I love that
because yeah, I've had the same experience with Kamal and
some of those tooling, so.

Speaker 4 (17:05):
So it's it's yeah. I mean, I wouldn't say that
it was always perfect and that I sometimes spent more
time that than I should have learning all of those
Camal settings and so on, but that's that's something that
I wanted to do also as part of growing experience,

(17:27):
so that that it was okay for me to spend
some time on this. But I'm not sure how you
feel Marco about it. But for me, the the experience
with as with Circulite, has been amazing. I mean that
this the speed, the the API. There is nothing that

(17:48):
really that I couldn't do up to now that I'm
able to do in post grace or whatever database.

Speaker 3 (17:57):
Yeah, it's been similar experience for me. I didn't really
have any problems with adapting from coming from prostgress. Mostly
pretty much everything worked as I expected to work, and
also from the query speed and all of that, it
didn't really make any difference for me on how I
built apps. But I think this is mostly also thanks

(18:18):
to active record that does the heavy lifting for us
in the bag we.

Speaker 4 (18:22):
Had for the full tech search to do a bit
a bit more integration. I guess that this will with
Rail's eight point one should be mostly covered with something
official from what I understood, but it's not yet. Really

(18:44):
we don't know exactly the boundaries. But now we have
a full tech search that is performed by esk light
and that works quite nicely.

Speaker 2 (18:59):
Did yeah have to change your database schema to be
a sequel file instead of an RB file or is
that something that they've fixed Because I knew until like
a couple of years ago, if you wanted to do
full tech search your sequel light, you couldn't have a
schema dot orb it had to be a structured dot sequel.
Is that something that they've changed now?

Speaker 4 (19:19):
I think so, because we are still using the RB schema.

Speaker 2 (19:24):
Okay, great, that's good to know.

Speaker 1 (19:26):
So have you thought about doing like a semantic search
of some kind on there?

Speaker 4 (19:31):
Yes, the the very first version of Ruby video I
used malisearch combined.

Speaker 1 (19:43):
That was nice to esq light because.

Speaker 4 (19:48):
I think at that time uh full tech search. It
was probably possible with esclight, but either I couldn't find
a up to date tutorial or things that that were
built in Ruby. At that time, I feel like it

(20:09):
was still a bit something new in the Reils world
for building a production website. So I started to use
a mail search and we we did some proof of
concept for doing semantic search and like similar topics, similar

(20:33):
finding similar dogs to one tales by. We created some
embeddings and so on. That's something that we didn't push
to production yet, but I'm really interested to work on
this feature uh in in the upcoming months or so.

(20:56):
That's that's really something where I think we can improve
the overall discoverability of the content.

Speaker 1 (21:03):
Yeah. I've played with it some, mostly using vector vector
stuff with either elastic search or postgress and yeah, I
mean it's cool stuff and it works out because then
you can if I don't put in exactly the term
that a full tech search would, you know, miss on,

(21:24):
it can figure out what I'm looking for and kind
of give it to me anyway.

Speaker 4 (21:27):
So it's definitely something that we'll want to improve to
get some semantic search.

Speaker 1 (21:33):
So what other things are you looking to add to it?
It sounds like you're going to merge the Ruby Conferences
dot org and Ruby events dot org. Do you do
you have other ideas of things that you want to
add to this or bring into it, or do you
have kind of a kind of a limited scope that

(21:54):
keeps it simple.

Speaker 3 (21:55):
Yeah, I mean there's a bunch you can do and
a bunch of stuff you want to do. But then again,
it's kind of tricky to build out all these features
and make it work, especially because Ruby conscious that didn't
really have a lot of meta data around any of
the events. It was pretty much the name, the dates
in the location right and now on Ruby Events we

(22:17):
have the speakers, the talks, all the descriptions, so it's
not just a one to one copy that we can
take over and then it's good. So yeah, are looking
on ways to bring over all the events. We have
the events index now, but I have no speakers and
no talks touched to them. So that's what be one
thing to just go back to all these events and

(22:38):
to try to add all these old talks so you
have the data for it. But also then looking at
the future. Having attended a bunch of conscious in the
last few years, I noticed that the people that organized
these events struggle with just the whole workflow as well.
And one of the aspects of that is also the

(22:58):
call for papers and finding the right platforms to host it.
It seems to be quite an issue that's quite expensive
for conference organizers to just be able to host the CFP.
So we were also looking in two ways to make
it easier to have like a full process for conference
organizers to go from announcing the events up until the

(23:19):
point to leasing the videos. To have this full workflow on.
Let's announce the CFP, let's collect the proposals, let's finish
it up. Then let's publish a schedule, let's publish the
actual talk titles, Let's publish the event as I think
that's happening. Then during the event, maybe there's some updates.

(23:41):
Maybe there's just the currently the talk is currently running,
and then after the talk, after the event, you don't
get the recordings attached to the same the same talks.
So just to kind of have this full experience on
for conference organized, but also for attendees, to have this

(24:03):
one interface platform where you can just go to and
find the information you need. Another aspect also was to
add photos official photos from the event, but also from
social media. So if you use the conference hashtag, that
we show these photos right next to the event itself
and give some more impressions of what the event was

(24:25):
like from people that's attend the event over talking about
the event in some way or a number just to
kind of bring the whole community aspect together, especially now
when we have the social media breakup where everybody is
just on some social platform but not everybody is on
the same platform. At least with these rupy events, that

(24:46):
we have one place where we see the stuff that
we want to learn from, at least for the engineering
part of our careers.

Speaker 2 (24:58):
Nice with this, like just like running this project, do
you do you guys always see it always as a
passion project? Do you reckon? You might like the sponsorship
or some other kind of revenue model in the future,
because if you're going to do something like come up
with the whole workflow for conferences to run, see I

(25:18):
FeAs and so that seems like quite a lot to
do just for fun.

Speaker 3 (25:22):
I mean, I haven't really thought about how you would
monetize it. I'm sure there are ways to do it,
but then again, in the terms of community, it doesn't
really seem to make sense to just monetize the whole
platform for that. So at least for me, it's I
plan it's to be for it to be a passion
project that continues to be the case. And but maybe

(25:47):
has another view about this, No.

Speaker 5 (25:49):
No, I mean for me it's those always I've always
wanted to be and to remain a passion project. So
that's also one of the reasons where I know that.

Speaker 4 (26:05):
We have lots of ideas of what we can do,
but I mean, we also have day jobs, and so
it has to remain a passion project and so things
might not arrive tomorrow, it might take a bit more time,
but we are doing it this with our little hands

(26:27):
and some help of the community. Also, there's been lots
of contributors lots of very interesting contribution from a various persons.
Maybe some person that you met Marco, but I think
also some person that came naturally to the ripple and

(26:50):
suggested some nice improvements.

Speaker 2 (26:54):
Yeah.

Speaker 3 (26:54):
The only other aspect that I could see this making
some money at least two account for the cost of
running it would be so if we want to add
sponsors anyway, we want to kind of index all the
sponsors that have sponsored an event and give them a
profile page two to say this company sponsored all these
events and that's what they kind of gave back to

(27:15):
our community. Also that it's more interesting for companies to
sponsor events not just for the duration of the events
or leading up to it that they can see their names,
but also in the archives say these are the companies
that made this event possible. And maybe there are some
companies that would love to still help out or kind

(27:35):
of have the name on the platform to say we
support the rugby community in that way. And I could
see that kind of workout, But I don't see it's
being monetized in a way for you to access any
information or running a call for papers or anything like that.

Speaker 1 (27:51):
Yeah, I think that's a natural thing. So I've been
working on a side project that it does index the conferences,
but it also indexes like gems and tutorials and you know,
lets people know when the meetups are coming up, and

(28:12):
you know, basically it's like an all in one directory
for anything having to do with Ruby. It turns out
it's probably a lot more work than what you guys
are doing. But that was what I was looking at too,
was just you know, it's like, hey, you know, this
naturally fits in here where you know, you're looking at
error handling tools right for dev tools under dev tools,

(28:35):
and so then you'd have that banner there because somebody
paid to kind of have their banner or be at
the top of the list or something like that. And
so I think something like this makes a lot of
sense in your case too. Yeah, where somebody is you're
not gatekeeping any content. You're just making people aware, Hey,
they sponsored this video and you know you can find

(28:56):
them here, and it just kind of extends whatever it
was that they got from the conference exactly. Yeah, So
I'm kind of curious, and I guess I could probably
go look this up, But how many conferences of you
indexed and put onto the website?

Speaker 3 (29:16):
There are five hundred forty conferences on the website right now.
Oh wow, and I think there are around two hundred
and fifty to half actual talks touched to them. Okay,
so there's still like half of them don't have any
like talks touch to them, but there's still a lot
of data already on the platformance off.

Speaker 4 (29:38):
And that's so then seven hundred and ten.

Speaker 1 (29:43):
Dogs, That's what I was going to ask, how many talks?
Seven thousand plus talks.

Speaker 4 (29:50):
That's amazing, almost three five hundred speakers. Wow, yes, and
that's that's also a nice feature of Ruby Events is
that speakers have their own page where the they we

(30:10):
get some information from their guitar profile and also will
index will list all of their talks. They can if
they connect with Guitar, then they can get a little
mark that will certify their their profile and they can

(30:31):
do some small edits to the content on their their profile.
But we start to have yeah, there's lots of lots
of speakers. Some it's it's interesting when you discover a
speaker and that you can also see what other talks

(30:51):
he gave. So that's uh, brings a lot of serendipity
into the the search experience with the the whole content.

Speaker 1 (31:01):
Nice. Yeah, I found's page and yeah.

Speaker 2 (31:05):
Yeah, I went looking for my own page in a
very narcissistic quay as well.

Speaker 1 (31:11):
Did you I'm spoken? Oh, it says right there, claim
your profile.

Speaker 2 (31:18):
Yeah, I'll have found this.

Speaker 1 (31:24):
Oh you've got my old stuff in here.

Speaker 3 (31:28):
There you go.

Speaker 2 (31:29):
Wow, I'm looking for your page, not chuck.

Speaker 1 (31:36):
Yeah, and the most recent talk in there for me
is from twenty thirteen. It was a Ruby Rogues panel
I did. Yeah, I did some other talks.

Speaker 3 (31:47):
But I'll rest on twenty fifteen ls though, But I
think it's if your middle name there, so it's just
Charles Wood, which has another talk in twenty fifteen.

Speaker 1 (31:56):
Oh okay, yeah, we do plant it on speaking more
so so you may see other stuff pop up if
you're wondering that.

Speaker 2 (32:06):
Sweet I as I found the topic section where you
can jump into topic and see all the talks that
fit under that, which which I think is super cool.
But just curious about how you tagged each talk with
a number of topics or is it a manual process
or g use like l l m's or something of

(32:28):
that for it.

Speaker 4 (32:29):
Yeah, it's it's l M based. We it's we try
to to to Once we get the meta data from
the for for for the does those you file, we
we ingest them into the ESCALI database and then we

(32:50):
try to enhance those meta data. First, if we can
get the transcript of the talk, we'll add the transcript
to our database. We'll do a first cleaning of the
transcript because most of the time the transcript that we
get are not so accurate, and we run them through

(33:14):
a first ai agen that will produce something that is
much cleaner, not perfect, but cleaner. Out of it this,
we build a small summary of the talk where you
can get i don't know, like the main key points

(33:36):
that are covered during this talk. And then out of
this the transcript and the summary, we feed this to
an l LAM with the existing topics that we have
because we try to avoid doing too much. Duplication doesn't
work perfectly, but does a bit and then uh, this

(34:01):
will create the taggings. So in the perfect world, well,
of course we'd like to get some more manual tagging,
but it wasn't given the amount of talks we had
and all the the history that we needed to cover.
The backlogs, the DLM really help us improve do this very.

Speaker 1 (34:27):
Quickly, are you Yeah, you were talking about the cost
of running this. Are you are you paying for the
LLLM that you're accessing? Are you running something like olama
on your own thing? On your own.

Speaker 4 (34:42):
It's mostly open ai tokens, so okay, and frankly speaking, yes,
I'm paying for this, but I think I probably put
like twenty US dollars or an ace.

Speaker 1 (34:58):
Okay, so you have it, spend time. I've seen those
costs go up when people use some of those models, So.

Speaker 4 (35:06):
Yeah, no, for now, the models that we are using
for topics and summarize those things are pretty stand out.
Those are the cheap.

Speaker 1 (35:20):
Models of I was going to say, you're not paying
for the most recent expensive models.

Speaker 4 (35:28):
Well, they are recent, but they are not the largest one, right. Okay,
we have seven thousand videos. It's a lot, but it's
not it's not that much neither, so we don't We
don't ingest thousands of video every day, so the cost remains.

Speaker 1 (35:53):
Oh yeah, that makes sense.

Speaker 4 (35:55):
Yeah, you know, I haven't counted exactly, but it's it's
something that as I can't afford, I don't need to
find a business model yet for this spot.

Speaker 1 (36:15):
Sounds good.

Speaker 2 (36:16):
I wonder if if you guys want to switch gears
a little bit and just talk talk a little bit
about stimulus, because you guys are both core maintained as
of that, and is that something you want to discuss.

Speaker 3 (36:28):
Do you mean, like SIMS in general, on the context
of in.

Speaker 2 (36:31):
General in general and in the context of the website.
I guess.

Speaker 6 (36:37):
Sure, I mean yeah, so it's it's just the setup
of hot Wire in terms of governance is a bit strange,
because it's like Reels has a core team and an
issue to even a committed and it's all fairly clear.

Speaker 2 (36:54):
With hot Wire and the constituent libraries, things seem a
bit unclear. What is your impression and experience about being
maintained as a stimulus where in slightly less of a
democratic governance structure. I guess from my point of view,
thirty seven signal still holds the keys essentially.

Speaker 3 (37:17):
I mean yeah. From my perspective, it always was that
I helped out with triaging issues, trying to respond to issues,
trying to see if you can bring in some of
the small changes at our like no brains, to merge
for bigger features. It was always just me proposing some
features and then somebody from the actual company wouldn't merge

(37:42):
those requests, So there wasn't that fever kind of steering
the whole direction of the framework. It was just us
trying to at least me trying to help out in
some ways. That is straightforward to just do when you
can kind of sit down for a few minutes and
look at these issues that we're coming in.

Speaker 4 (38:03):
I think it's reated the same thing on my side,
mostly trying to help. I would say that my involvement
into this stimulus ripple in the past months years has
been way less than what it could have been a

(38:24):
few years ago. But I think it's also in general,
I'm writing way less stimulus controllers since hot wire in general.
I don't know if this is something it might be

(38:46):
feeling too, but we with hot wires and the terminal
streams and morph and so on, Lots of the things
that in the would require some stimulus controlers are now
somehow already included into Turbo and hot wire, So I'm

(39:11):
using way less stimulus in general.

Speaker 1 (39:14):
I'll chime in and I'll say that that's my experience
as well. Initially, when they came out with Stimulus, it
was like, this is a breath of fresh air and
I could just put this where I want, And yeah,
then I have to give credit to I used because
I was reading his book and I figured out that
Turbo did a whole bunch of other things for me
that I liked. And so yeah, I've been writing a

(39:38):
lot less Stimulus and just using Turbo streams mostly.

Speaker 3 (39:43):
Now.

Speaker 4 (39:43):
But the governance, yeah, we know it's a thirty seven
signal repo. They have the final decision on this. I mean,
besides Stimulus, there is with Marco, we built stimulus hues,

(40:05):
which at that time, uh, the idea was okay stimulus
because at that time stimulus was even more uh frozen.
This was really a frozen project. The idea was to
get something done and keep it like this. But we

(40:27):
felt in the community that we needed some more tools
around stimulus, so there were no way to get those
tools into the core product. There were no real plugging capabilities.
And that's where Stimulus Shoes brought this pattern and this

(40:48):
API where you could add to your stamus controllers some
new features and some new capabilities with an API that
was almost feeling like you have you added some plugins
to your stimulus controllers. And that was a way to

(41:09):
somehow okay, keep the stimulus and all their decision and
their like this and build something that could fit into
this model.

Speaker 2 (41:22):
Nice. Yeah, I think it's just one of the things
you got to live with with stimulus. I guess it
is what it is. I'm glad that you guys put
in the effort to try and keep things going as
much as possible. But yeah, with with about taking over
so much of the heavy lifting, you only left to
really use stimulus for like the tiny bits where you

(41:44):
need clients out interactivity. I guess so you're connecting that
back to Ruby Events. Was there anything big or heavy
that you felt you absolutely need needed stimulus for on
that website? And if you don't have stimulus, what do
you think you would have used.

Speaker 3 (42:06):
I think the biggest piece that we have on Stimilus
use use on Ruby Events is the video player for
the playing the actual videos. I think there's some logic
around that where we have to embed some plugin or
some player in case it's to be light, and that's

(42:26):
what we have to just wrap up in a stimulus
control to make it work. But I think that's pretty
much like the biggest thing we have. There are like
around fifteen or twenty controls. It's not that many. Most
of them are super small and tiny and super specific
for one use case in the app. So it's not

(42:47):
that it actually makes heavy use of stimulus. I think
it's that's the other thing. I think that's just a
testament turboalytis. You can do so much with Turbo we've
had having to go back and write Similus controls yourself,
and if you do, they are super small and straightforward
to write. So I think that's really the key here.

(43:12):
The other thing about Similus as well is I think
that it is also considered done right, that doesn't I
don't really think there's too much we have to add
to the core itself that you can achieve otherwise. And
with that and having Similous use next to it, I think,
at least for my use cases, it has always been

(43:34):
more than enough. And I guess for the other question
for what I would use if I wouldn't kind of
write stimulus, it would probably be web components. Honestly, I
think they have been very underrated the last few years.
You can achieve very similar outcomes with some basic classes

(43:55):
that you write that are web components. And there's this
other project from Githup, which is inspired by Stimulus, which
I also get the name of that framework, but is
almost like stimulus, but you're for web components, so you
write web components that have targets that have components that

(44:16):
are controls in the same the Yeah, and if I
wouldn't write Seamus, I would probably use that framework or
just pure web components.

Speaker 2 (44:30):
Yeah, I got to look up that framework because I
love web components. Like, if I don't have Stimulus available,
which is usually if I'm doing like a static side,
you're probably using bridge down, then I don't bother pulling
in stimulus because a bit too much. So then I
usually use web components And yeah, I love them. They're
web stand that they just seem to work out of
the box quite nicely. So yeah, A big, big fan

(44:52):
of those.

Speaker 3 (44:53):
Yeah.

Speaker 1 (44:54):
One other thing that I would think about reaching for
is if I didn't have Turbo, i'd be looking and
at HTMX.

Speaker 2 (45:02):
Yes.

Speaker 3 (45:05):
Yeah. HMX is also interesting because it also feels it's
way more lower level than Turpo, so you have to
kind of write a lot more or more explicitly what
you want to do, which is personally what I dislike
about it. But I guess if you couldn't use Turbo,
it will still be better than the alternatives. Yeah.

Speaker 2 (45:25):
Yeah, I have the same complaint. I just I like
that you can just pull in Turbo and suddenly all
your links are magically using table drive. I don't want
to be sitting and writing data attributes for every single
one to make it work. So yeah, similar complaint on HDMX,

(45:46):
And my experience at Stimulus has also been like what
you mentioned, the very small controller is doing very specific things.
So in your opinions, you think Stimulus should have a
standard library that provides a lot of these small things
that are common.

Speaker 3 (46:00):
On Yeah, for some time I was feeling that there
is a need for that, but I feel like the
more I've written Stimulus, the less I think that's actually
the case. And I also quite dislike the idea of
importing hope control from some packages because usually it's not

(46:23):
the the child skip part that is annoying to write.
It's usually the markup, and we've had the markup next
to the controller it is pretty much useless. So I
don't think it really makes sense to have pre defined
controllers that you're just important use because yeah, it depends
on the markup if you had some way to kind

(46:44):
of share markup and controls, then maybe, But I think
then again that's kind of where we are with web components.
That's pretty much what that is. You import one element,
it has to behavior attached to it and it's like isolated,
and that's kind of what you use. But if you
don't use that, and I think that's what simlous use

(47:04):
is actually really good that too, because you have these
common behaviors that you can use with these hooks. It
just important sooks that you kind of want and then
you get that behavior, but it don't get ready made
controls that are just important in the use.

Speaker 4 (47:19):
Yeah, at some point I thought that for revid and
some standard libraries, it would have a bit more interesting
to get some places where you can basically copy paste
some standard stimulus controllers, because also my experience has been

(47:42):
that I'd never really like to import a stimulus component
from a library. I prefer to have the code because
we are usually talking about what twenty fifty one hundred
lines of code maximum in a stimulus control. What's the
point to import this from a back cage rather than
having this into your into your project directly. So I

(48:07):
prefer to copy and paste directly to the code and
get my own control.

Speaker 1 (48:15):
Yeah, I'm gonna agree with you.

Speaker 4 (48:17):
There.

Speaker 1 (48:18):
There's a stimulus dash components dot com that has a
bunch of ease that pull in other libraries and things,
and every time I've tried to use them, eventually I'm
just like I needed to do something slightly off from
what this does out of the box, and even doing
the kind of inheritance thing that they tell you you

(48:38):
can do with it, it still doesn't quite jive. And
so yeah, I wind up going and finding their source code,
copying it into my thing, and then going all right,
now I can make it do what I want. And
it's usually it's almost always simple enough to where yeah,
I can just reach in and go that changes, and
that changes, and that changes, and I'm good.

Speaker 3 (48:57):
Yeah. The framework I was mentioning earlier is called from GitHub.
That's as I mentioned meant.

Speaker 2 (49:04):
It's called Catalyst.

Speaker 4 (49:06):
Yah.

Speaker 2 (49:06):
Interesting, need to check that out for sure.

Speaker 1 (49:08):
All right, Well, anything else that we want to delve
into here, I mean we've been going for fifty one minutes.

Speaker 3 (49:16):
I guess one thing that I want to just mention
as well is that we have built some mobile apps too.
Now with Hotwaire Native that we just want to kind
of get out there so if you are at these events,
that you can pull it up and see the schedule
and maybe bookmark some of these talks that you want
to check out later if you miss them in person
or just want to revisit them. And that's kind of

(49:38):
the other thing that I think is really nice about
this whole ecosystem that we have now. But everything is
in one place and you just need this one app
for these events. And previously I have seen some point
of this have their own app or their own PWA's
that you have to install, and at least with this,
it will be this one place where the community can
focus on and contry data through and then you have

(50:01):
it on the go as well.

Speaker 4 (50:02):
And what is great is that this is under the
Ruby events all on GitHub, and you have the web
pads that is open source, but also the two mobile
app iOS and Android are both open source projects, so

(50:23):
that's that's also a very nice, uh place to learn
those technology. I wouldn't say that we are providing perfect code,
but at least it's a real projects that are that
are live in production and that you can look at
the code awesome.

Speaker 2 (50:44):
I don't know as a native answer is that is
that downloadable from the app stores?

Speaker 3 (50:48):
Yes?

Speaker 2 (50:49):
Awesome?

Speaker 1 (50:50):
So you just look for Ruby events that's it.

Speaker 2 (50:53):
Yes awesome, guys, just says the last technical question for
me about the webs I just around you it is.
I've been a fairly vocal critic of thellerency. I'm curious
to understand why you guys are using it. Just is
this something you like? And also how come you using
write wheat and hybernand it instead of like, instead of

(51:17):
es bild or something like that. Just curious about those
two choices.

Speaker 4 (51:22):
Can talk about Telwyn and maybe Marco you'll talk about it.
Mm hmm, Well, Telwyn, I think, and we it's not
only using tellwin, but also it's using a set of

(51:43):
components from daisy Ui. Then that is that provides some
some kind of it's a bit the bootstrap uh kind
of components that we used to have in the bootstrap days,
but built for uh delawing. So we use that. I

(52:08):
mean it's it's good in the sense that we didn't
have to make too much decision and and things were
already prepackaged. I know that we had some cases where
uh Tweaking some components was harder than expected. Marco had

(52:32):
some quite a bit of struggle with some component. I
remember we haven't migrated yet. We have a pr open
to migrate to tell in four and day z Ui
five h, which bring more flexibility because we don't we
don't have to import uh the entire day z Ui components.

(52:57):
We can just pick the one that will want so
that that would be a nice improvement. Now, why tellwind
versus other things is probably I'm more familiar to use
telling nowadays than using stand out CSS, but that's I

(53:19):
don't have strong opinion on this.

Speaker 1 (53:22):
Yeah, I like tailwind, so I live in your camp.

Speaker 3 (53:27):
Yeah, it feels like it's just the de fact of
standards at this point, at least in the wider app community,
that this is what you use. But I can also
see how you dislike it, especially if these classes get
super long and you have to think about how you
break it up. But I think it's still better to
have these complexity at least in in one place, in

(53:47):
the components themselves, not like little around the whole app.

Speaker 4 (53:51):
And yeah, we do use a view components a bit.
We should use it more. We have some places where
we know that we would need to refactor a bit
like the the user, the speaker of the nails and
so on. So we know that we have some lots

(54:11):
of duplicates that we could refactor. But we we do
use a view components, and when you combine view components
with Telwin's that makes it quite quite nice to get
your design system the basics, basic building blocks.

Speaker 1 (54:29):
Yep, yeah, I've done that too. I will admit that
I have moved back away from view components and I
just put all my stuff in partials. It's the same
idea right where it's okay, I know I'm going to
have to have a text field or you know, yeah,

(54:49):
an avatar or something like that, and so I just
have a standard partial for it. Just pull it in
and give it what it needs.

Speaker 2 (54:56):
Yeah, view components, can we have you handed? I just
like them because and you need a little bit more logic.
They're instead of putting it in help as I think
the view component helps. But yeah, I think for a
majority of use cases, partials or do just fine. It's
just that a little bit of extraving that view component
gives you is useful, but definitely don't need it for

(55:18):
every project.

Speaker 1 (55:19):
I think that's a good way to put it, right,
because that's the trade off is that, yeah, you can
just hand in the objects that it needs in order
to figure out what to do, whereas with the partials
you kind of have to you have to finesse it
a little bit and put some of that more of
that into the helpers, which aren't always as targeted as
the view component class, which is essentially like a presenter.

Speaker 4 (55:43):
And so.

Speaker 1 (55:45):
Yeah, I see the trade offs. I really do like
the view components. I just kept running into funny things
with some of it, or it felt like it was
overkill for some of the simple stuff. And so yeah,
but that idea I've used extensively where yeah, it's okay,
I've got all my tailwind stuff in the view, view

(56:07):
component or the partial, and so it always looks the
same because I always just reach for it when I'm
putting that in.

Speaker 3 (56:13):
Yeah, and I guess ask for the beat question, I
think it's it's really from coming from the web Pack
days and where you have the web Pack deft server
running like as a separate process. You really got that
nice developer experience where you update the files and you
don't have to reload the page every time. So just
like this whole experience, from hot reloading to adding dependencies

(56:38):
to integration of the acid pipeline, it just feels like
a well like grounded a package. It just included in
the works. There's no configuration, there's nothing you have to
kind of make work. And if E is built, it
doesn't really handle images. It doesn't really handle style sheets.

(56:58):
I think maybe it doesn't now, but not quite sure.
So it's really just this package of you just included
no configuration, and you've got all the nice benefits that
you don't get with import maps but also don't really
get with VS built. You can get some of that
eas built too, but then you have to write your
custom like watching mechanism to make sure that you reload

(57:19):
the pages. But it feels always so hacky compared to
just using VAT. And yeah, WEAT has focused on developer
experience and just using that tool as it is without
some custom rails, Rappo logic magic, it really feels nice
as it should be.

Speaker 1 (57:39):
Yeah, I've used both. The thing that I miss when
I use import maps as the heart reloading of all
the things, you don't have to do hardly any work
to make it work with VAT. The flip side is
is that I like having the I feel like import
maps is more flexible and I can see where everything's

(58:00):
coming from. VAT tends to, at least in the way
I had it set up. It was building out one
file and production for the JavaScript and compressing everything together,
and that made it a little hard to follow sometimes, you.

Speaker 3 (58:14):
Mean, like for debucking like in the browser when yeah, that.

Speaker 1 (58:20):
Those are the trade offs that I saw. I have
to say that between you know, given given the two options,
I've kind of leaned more toward import maps. But I mean,
I I don't know that I have that much more advantage.
And the hot reloading, like you said, is so nice,

(58:40):
So I don't know. I don't I don't have a
strong feel on this one, but there are things to
like about both.

Speaker 3 (58:47):
Yeah. Other thing I like about viata as well is
if you have a syntax error or some like missing
dependencies or some missing or wrong imports, it will tell
you right away with a noble A and say here's
what's wrong on this line. And with import maps you
pretty much get no feedback at all. Yeah, yeah, exactly,

(59:07):
And that's just this a little bit more that you
get with it. But obviously there's more complexity behind the scenes,
which is also what makes it less attractive.

Speaker 1 (59:18):
I guess.

Speaker 4 (59:19):
For in pot maps you can get hot reloading. Now
with hot wire Spark.

Speaker 1 (59:27):
You just made my day. Ye, here's a virtual hu Adrian.
My life is better now.

Speaker 4 (59:37):
I've not used it yet really because I'm not really
using in pot maps. But from what of the project
that is made, it's a whole ray who build it
from a thirty seven signals? I think they are using
it mostly internally thirty seven signals because they are mostly

(01:00:05):
in both maps shop. It does work well with in
both maps and stimulus controller. Then when you have something
a bit more exotic, then you might not have the
full refresh and so on. You can give this try.

Speaker 1 (01:00:23):
Well, if it'll do that, then why the heck are
you using Vietnam Skip?

Speaker 4 (01:00:29):
I don't know. For me, both maps has always been
more trouble than solutions. But it's probably that I didn't
spend enough time getting through the running curve. I don't know. Also,
in my day job, we have more I mean in

(01:00:52):
both maps. I think what's great when you have a
very limited amount of external dependencies that you will get
So if you mostly have your own stimulus controllers with
very few external dependencies. It's probably quite easy to manage.
But as soon as you want to pull in us
some libraries with their own cess and so on, then

(01:01:17):
the whole thing becomes way less easy. Ye sometimes yeah, yeah,
for me, I never really click up to now, but
who knows.

Speaker 1 (01:01:29):
All right, well, I'm going to go ahead and push
us into our picks. Our picks is our final segment
of the show. If you've listened before, you know, we
just shout out stuff that we like. If you haven't
listened before, we just shout out stuff that we like.
And so I'll have our usually go first, and then
I will do picks and then you guys do picks.

Speaker 4 (01:01:50):
Right.

Speaker 2 (01:01:51):
So, well, I spent quite a lot of me, you think,
my favorite musician around Europe because it was like twenty
plus day tour. I went for nine shows because I'm
a bit mad, so as around Europe and then the UK. Wow,
so I have to pick him. So Steven Wilson check

(01:02:12):
out his new album, The Overview. It's ridiculous. It's two
songs of twenty three minutes and eighteen minutes, so it's
completely nuts. It's like something out of the seventies, but
it is very very good space theme progressive rock. If
into that kind of thing, so yeah, check that out.

(01:02:34):
He's sturing America and the Autumn, I think, so if
you're in America, maybe if you like what, yeah, you
can see him live now. I watched a couple of
movies last week. Coincidentally, both were Woody Allen movies, so
might as well just pick those because I like them both.
One was match Point, which I really liked, I think

(01:02:56):
not a really really good opening line, which talks about
luck and how people are reluctant to admit just how
much luck plays a role in life, and it compares
it to a ball hitting the top of a tennis
net and for a second it can go either cross
the net or fall right back and depends on your

(01:03:17):
locks is. And it's just very clever writing and imagery
throughout that movie, which is really good. And then the
other movie I watched was almost the opposite, which is
called Hollywood Ending, which is about as stupid as match
Point was intelligent and deep. Hollywood Ending was about as stupid,
but I really enjoyed it because the jokes were The

(01:03:40):
jokes were funny, like no matter how stupid the movie
wasn't the ending was. It was really well written and
really funny. So yeah, I'm gonna go with those two
movies as my picks. Yeah, I don't have a tech pick.
I'm just gonna call it there.

Speaker 1 (01:03:53):
All right, good deal. I'm going to throw in some picks.
I always start with the board game. The board game
I'm gonna pick is Camel Up. We had four couples
we got together for my friend's birthday and this is
one of the games we played. It's a pretty simple
game board game. Geek waits it at one point five,

(01:04:14):
which means that it's you know, your casual gamer can
pick it up pretty easily and run with it. Age
is eight and up can play it. It takes about
a half hour, and it's a camel race around a pyramid,
and so you roll the dice and you move the camels,

(01:04:37):
and yeah, you bet on the camels. You also can
reserve a space and if a camel lands on it,
then you know it'll move it forward or back and
you get money at the end of each round. If
you bet on the right camel, you get money. And
then at the end, whoever picked the right camel to
win the overall race. The first person to do it

(01:04:57):
gets more money, the second person gets less, person gets less.
You can bet on more than one camel. If you
bet wrong at the end, then you lose a coin
or a dollar or whatever it is. I mean, that's
essentially the whole game. You just play until one of
the camels gets across the line, and then you finish
the round and the dice you stick them in the

(01:05:19):
top of the pyramid, and then you flip the pyramid
over and push a little thing on the side and
it drops one of the dice, and whatever color the
dice is, you move the camels, and then the only
other rule is is if you move a camel onto
the same space as another camel, it goes on top
of the stack of camels. And then if a camel
underneath any of it, So if a camel has other

(01:05:41):
camels on top of it, when it moves, it moves,
and it moves with all the other camels on top
of it too, and so anyway, it gets kind of interesting,
you know, betting and doing all the things there, But
super fun game if you're looking for a quick game.
Like I said, it plays up to eight people and
so yeah. That was part of the reason why we

(01:06:03):
were playing it with that group, right, is because it
kind of fit everybody in. So, yeah, I'm gonna pick
camel up as as my pick for a board game.
I've also I'm in the middle of the last book
of the Sort of Truth series by Terry Goodkind and

(01:06:26):
those I've really really enjoyed these books. The first one
is called Wizard's First Rule, and so I'll put a
link to that in the in the show notes as well.
But yeah, it's it's a fantasy, it's high fantasy, and
I've I don't even know how to describe them to you,

(01:06:50):
so just go pick them up. I think there was
a TV series that they made out of it, too,
but I haven't watched it, and I don't know how
close they stuck to the book. So if you've seen
the TV series and you know, I don't know how
how good the context is there, but I'm going to
pick that. I don't know that I have any technical picks,
so I will just let Adrian and Marco do their picks. Adrian,

(01:07:14):
do you want to go first?

Speaker 4 (01:07:16):
Okay, I wasn't really prepared for this. I might go
a bit of topics I don't know. I'm not sure
about the exact rule of the pig, but I'll talk
about a recent passion that I had discovered. I've been
doing quite a lot of sports recently and discovered CrossFit

(01:07:39):
and iras. I don't know if you're familiar with iracks.
It's kind of competition a raise of CrossFit, and just
purchased some tickets for the next event in Bordeaux November
and we are and it has been quite amazing how

(01:08:03):
I could progress in sport doing this and how motivating.
So if you are looking for something quite intense and
to to to to get back to to sport, it's
it's a nice challenge to to to put to the
race isn't easy and one of the hardest parts nowadays

(01:08:24):
is to get a ticket for some of those race
because it's very very become very very popular. But it's
sounds quite challenging and nice. And that's a nice thing
that I discovered recently. So that's my passion pick, I

(01:08:44):
would say. And in terms of technical what.

Speaker 2 (01:08:51):
What did I.

Speaker 4 (01:08:56):
Well, we are doing a lot of things with.

Speaker 3 (01:09:00):
With LLLM.

Speaker 4 (01:09:01):
Of course nowadays it quite impressive how we can We
are doing a lot of adaptive learning, and it's pretty
impressive how we can do this very easily versus what
we could do in the past.

Speaker 1 (01:09:21):
Cool. And just to put it out there, the hierrox
is h y r o X. It took me a
minute to find it on the Google. So awesome, Marco,
What are your picks?

Speaker 3 (01:09:33):
Yeah, I have one for music as well. One of
my favorite DJs and producers they are They have residency
now in a new club that just opening Ibisa called Universe,
and it's this new hyper club they call it. It's
like space themed. It's really big, but it's like really
nice and all is a custom build for that specific club,

(01:09:55):
and they have like a new show called Holo Sphere
and it's like this hollows with like visual LEDs on
the into the and the outside of that sphere, and
the DJ is performing in the center of the sphere
so you get these really nice free D effects and
it looks really awesome. And I want to go see

(01:10:15):
a show of his, like when he's like doing his residency,
it's like every Monday till September, so I really want
to go see that. And if you are not familiar
with his music, then check it out. And I guess
for the technical pick, I have been quite enjoying a
tool called hot Wire left tools by a friend of mine.

(01:10:38):
It's like a brows extension for hot wire, so you
can kind of inspect the stumus controls your turbo frames
and all of that, and there's a lot of stuff
you can just see in debug using that in the browser.
And they have been working on a native what do
you call it, like inspect panel in the browser, so

(01:10:59):
like next to you notes, so inspector in the theft
tools you also get like this new panel now for hatwire.
So this is really cool to see how this came about. Yeah,
that's called what do I F Tools and it's available
in all platforms breaking download extensions.

Speaker 1 (01:11:18):
For grosss awesome. All right, Well, how do people find
you guys online? If they want to connect?

Speaker 3 (01:11:23):
For me, it's my first and last name, dot dev
Macro dot dev now Arian Paulie pretty much everywhere, mostly
on Twitter, a bit on the Blue Sky.

Speaker 4 (01:11:39):
Then what else. Yeah, that's the two main stings and
on getub.

Speaker 1 (01:11:44):
All right, cool, Well, thanks for coming and thanks for
putting this together, because it seems pretty handy and something
that the community I think needed. Awesome. Thank you so
much for having us all right, well, folks, until next time,
Max out