October 2, 2024 • 39 mins
The Elixir Mix Panel discusses the history of Elixir and the high points and big changes in the language and ecosystem. They go into the big changes that brought about growth in the ecosystem, ease of use in the language, better features, and much more.
Links
Picks
Become a supporter of this podcast: https://www.spreaker.com/podcast/elixir-mix--6102049/support.
Links
Picks
- Adi- Simplebet - Software Engineer
- Adi- Career Opportunities | Annkissam
- Allen- Just-in-Time Mode - Tailwind CSS
- Charles- Premium Podcast Feeds | Devchat.tv
- Charles- Who Not How
- Charles- The Miracle Morning
- Charles- Psycho-Cybernetics
- Charles- As a man Thinketh
- Charles- Dev Influencers | Devchat.tv
- Sascha- Bypass - bypass v2.1.0
Become a supporter of this podcast: https://www.spreaker.com/podcast/elixir-mix--6102049/support.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:06):
Hey everybody, and welcome back to another episode of the
Elixer Mix. Today on our panel we have Audi, Hello, Alan.
Speaker 2 (00:14):
Hello, Hello, Sasha, Hello everybody.
Speaker 1 (00:18):
Is this your first week back?
Speaker 3 (00:19):
Yes? Regular it is.
Speaker 1 (00:23):
How how cool you are why we invited you back?
Speaker 3 (00:28):
So yeah, a while ago, I as the guests joining
the podcast and talking about the library of Mind, Nigger
and like behaviors and looking. I've been using alex for
like a four years now and focus the libraries published.
So it's great to know be here and talk to you.
(00:48):
Exchange oudiens, happy to be here.
Speaker 2 (00:50):
Nice.
Speaker 1 (00:51):
I'm Charles Maxwood from dev chat dot TV, And yeah,
so we decided to talk about kind of the history
and evolution of Alixir. Yeah, I think I kind of
want to start this one off and kick it off.
So the first podcast episode that I ever did related
to Alixir was on Ruby roachs and he has no
surprise to anybody, I think if you've been in the
Elixir community for very long. We interviewed this guy who'd
(01:14):
been in the Ruby community for a long time, written
a whole bunch of Ruby libraries, ruby gems, started a
Ruby consultancy, and then he went off and was fiddling
with this airline stuff and decided to write Elixir. So
it was jose Vaalin and so yeah, we were like, well,
come come tell us about this this thing, and so
he came on and he talked about Elixir and it
(01:35):
was kind of like this funky thing that he did
on the side while he was still writing all the
ruby stuff that we all used. So anyway, it's come
a long way since then, obviously, But yeah, I like
that flavor, right, because you never know if the wacky
project is going to turn into something that's it this
ecosystem that all these people earn a living at, right.
Speaker 4 (01:58):
Yeah, It's it's kind of crazy how fast things have
kind of been rolled along, right that this language just
came out what came out quite some time ago, but
it still feels like quite new to me, although it's
been changing a lot, especially with this recent one at
twelve release.
Speaker 1 (02:11):
Yeah, I'm kind of wondering what are kind of the
big milestones or sign posts I guess in Elixir's history
that you guys see, right, I mean some of them
are going to come in I guess after you discovered Elixir,
But what what were the turning points I guess that
got you in in the first place.
Speaker 3 (02:31):
Right, It's actually something which I'd say that got me in.
When I started with Elixir, it was just because my
employer used the Lixir, so I basically had to. But
when I started, I think that was one point four
or one point three, we didn't have a formata back then,
so it was always like we always had these discussions about, Okay,
how does the code look like superformata. It is like
(02:52):
the superf like that. So it was pretty cool when
I'm not sure which version accepted, but when Lyxi finally
got the opinionfomata and a lot of these discussions just
died down because the formata did this for you.
Speaker 1 (03:06):
Right. I have a question for you, Sasha, regarding that.
So you said your employer was using it. Were they
using it with Phoenix or were they using it for
something else like Nerves or some other project.
Speaker 3 (03:18):
Yeah, at that point it was Phoenix. I think Nerves
wasn't even in its existence back then, so I mean
that was probably about four years ago. Maybe it was
already there, but that definitely not a speaks today. So yeah,
it was. The company I used to work back then
was doing a lot of I WoT stuff. So the
real time capabilities of Felix. They were pretty nice in
(03:39):
that area, which is why are you using it?
Speaker 1 (03:42):
Right? The other question I have related to that is,
so you came to it through a job, And it
seems like a lot of people when they go look
for their next job, it's like, Okay, I've got fifty
zillion years doing JavaScript or Ruby or GHP or Python
or whatever, right, And so they go and they just
(04:03):
kind of stick with that technology because they've got so
much seniority and they can kind of talk their way
into whatever kind of job they want. So what made
you take a job where you were going to be
working in a different technology.
Speaker 3 (04:17):
The main reason was actually because I was switching over
from a mobile developer job with like occasional backend what development,
to like a full time back and development job, like
junior side. Because I'm a previous employer, I did mostly
iOS development, but I did the occasional PHP back end
and that was always something which I enjoyed more So
(04:38):
I was looking for an opportunity to work more in
that area. And then I came across this other company
and they used the licks here and the functional language,
and I was like, oh, it sounds interesting. Enough and
now I'm here today, so things, I guess things work out.
Speaker 1 (04:52):
That makes sense.
Speaker 5 (04:53):
Yeah, so me.
Speaker 6 (04:54):
It's one of the things I remember is so something
very simple and trivial like the with statement that we
now take for granted ALEXA. When it came out, it
did not have that, and I remember even using I don't.
Speaker 5 (05:05):
Think that was even unless in all those things.
Speaker 6 (05:08):
And I know even using IF was like, don't use
it unless they really want to use it, because it
IF kind of uses case within like cases like the
main statement. And it was actually my first job too,
doing elexir in twenty fifteen, and so it looks it
was like just one pointer at that time. And when
they introduced with statement, it really kind of cleaned up
(05:29):
our code quite a bit, because you know, now you're
thinking in terms of the railway pattern, right, like returning
okay result or error error, and that just cleans up
your entire controller.
Speaker 5 (05:40):
And then Phoenix kind of.
Speaker 6 (05:41):
Adopted that pattern with the fallback controller, right, like cleaning
up that error as well, and like standardizing what error
is your APIs or return and stuff. So I think
that with kind of catalyzed that entire list of tools.
Speaker 5 (05:55):
I was a really big fan.
Speaker 3 (05:56):
Of with Ford was basically nested case statements all the
way down, right, like a lot of case statements, and
there wasn't much anything else you could do. Like that
was the way things were written it which also made
sense when you think where Alexi came from from what's
of Berlin, because that's how you do handling and arline.
You do the case statements, or you pet on if
you match explicitly on the okay with radio when you
(06:19):
say okay, I want to raise you, but yeah, right.
Speaker 6 (06:22):
Or use the pipe operator, but the functions kind of
can take okay or error energy like your entire pipeline.
Speaker 3 (06:31):
Definitely you have written code like that. I've just looked
at the release notes from the LICKS and like the
format actually came in one point six, but there's a
lot of good stuff in one point six. Also, a
dynamic supervisor came in there, so that was something we
didn't have. I mean, I remember that that the supervisor
still has the simple one for one strategy, right, which
is like kind of akin to the dynemic supervisor. But
(06:53):
there's one thing about the simple one for one strategy
that it's like a lie. It's not simple. It's it's
definitely not simple understand that. I mean, we had to
use it in one project, and like, I mean, it
was pretty new to addix. Yeah, so I was like
trying to read the documentation and I grabbed my head
around and I just I just failed. Just went to
(07:14):
a colleague at some point and please explain to me.
Speaker 1 (07:17):
Yeah, I'm kind of curious because I can't like pick
out the features or anything because I've just kind of
been watching the community more than participating in it. And
over the years, like it was kind of this thing
that was just kind of out there, you know, and
and the Airline community was also another community. It was
just kind of I just kind of watched it and
(07:37):
it's kind of steadily grown. And then I don't know
what it was, but at some point it seems like
the Elixir community just really took off. And I'm not
sure where, I'm not sure why, I'm not sure what,
but it really took off. And about I want to say,
about four or five years ago was when it really
kind of kind of caught up with me that a
(08:00):
lot of the friends of mine that I had connected
with over the years in the Ruby community, it all
moved to it a lot of the people that I
knew from some of the other communities that moved to it.
I was actually getting asked on a regular basis when
I was going to start an Elixer show, so we
started this one. I mean, was there something around then
that really just gelled? Was it that they finally there
(08:25):
was Alixer comp and there were some of these other things.
Those tend to be more the symptoms of the community
growing rather than the cause. I'm trying to put my
finger on, you know, Was there any one or several
things that happened around then that that kind of came
together for it?
Speaker 6 (08:40):
I guess must be like the bigger companies kind of
adopting it, like Pepsi e commerce and I know, cars
and you know, like the like who were in like
more legacy kind of a programming language, just them adopting it.
I actually remember like a more quantifiable like difference, the
difference in the number of responses, the number of people.
In alexir conf twenty seventeen and twenty eighteen, they announced
(09:03):
it at the keynote that the number of sponsors they
have more than quadrupled in one year.
Speaker 5 (09:10):
And it's just that what happened during that time.
Speaker 6 (09:13):
I don't know.
Speaker 1 (09:14):
Yeah, I guess that makes sense. I mean, I think
another thing that I saw at the time was and
I guess this isn't a technical milestone, but so I
got invited to CODEBAM in twenty twenty eighteen or twenty nineteen.
But there were a whole bunch of talks for Yeah,
these bigger companies that were talking about and they weren't
like talking. Some of them were talking about how they
(09:35):
had built their infrastructure on the Beam and how they
had built a lot of it on a elixer. But
more than that, they were talking about how they had scaled,
how they had lowered the response times, and a lot
of the concurrency setups that they were doing and things
like that. And so they had actually solved a lot
(09:57):
of the issues that existed in some of the systems
that they it had before by migrating to Elixir. And
some of them were like these huge setups like WhatsApp
I think was one of them. I think Discord does
a whole bunch of stuff in the back end with Elixir.
There are a whole bunch of other ones and they
were like these massive massive back end systems that either
(10:20):
run in like dedicated data centers or running the cloud,
and I mean they just like crush the amount of
bandwidth and data that just just course through them. And yeah,
I think people really saw what elixor systems were capable
of that you couldn't do with say a no JS
or a Ruby on rails, and people started saying, Okay, well,
(10:45):
if I need say data overforms, then dot net Ruby
on rails, that makes sense, right, But if I need
if I'm sending raw data, audio data, if I'm sending
gobs and gobs and gobs of data, it makes a
lot of sense to have a system that can handle
the volume of data, but not just the volume, but
(11:08):
the sheer number of discrete messages at a time. And
that's that's where I think people figured out, oh, yeah,
I need something that can do this. And Elixir was
easier for the developers to digest than earline.
Speaker 3 (11:24):
I think a huge part of that is also how
easy it is to pick up Phoenix, because I mean
you just also, yeah, to like submit the form right
safer form. It's super easy in Rails, but it's also
super easy in Phoenix, so you can get started pretty
easily there too. But then the skating story is a
completely different one. So while earline, yeah, it is certainly
(11:48):
something which could also super easily handle these these these
skating scenarios, it's not as easy to pick up as
serious So I think a lot of success came from
that too, that that people could see, Okay, yeah, this
can scale two different in different ways and maybe better
than than other technologies out there, but it's also not
as super hard to pick up as some of these
(12:09):
more let's say these things totally.
Speaker 6 (12:12):
Yeah, that's definitely I think a huge point. And as
you guys are talking about it, I just like it
kind of reminded me of my older daid Alixa, and
like why did I like Elixer, And I just like
I also realized, like everyone around me seemed to just
leve Alixer and the involvement of you know, Jose and
Christma code in the community and like listening to feedback
(12:32):
and advertising, and Jose was being involved in like Phoenix
and writing marks and like kind of driving the community's
vision forward while keeping it fun for developers, right, Like
I mean, like everyone who's worked with Alixa wants to
work in the Elixer. So like companies who've adopted a
lix a little bit have no choice but to adopt
it more because people who work there want to keep
(12:54):
working in Alixer. So yeah, like in twenty fifteen, my
first job project we built with the company's first Elixa app,
And in twenty twenty we're probably building it's thirty thirty
of the Elixa app.
Speaker 5 (13:08):
So just like because I didn't want to build Ruby
on rails app.
Speaker 1 (13:13):
So yeah, it's interesting too. A friend of mine, he's
actually the one that recruited me at my current employer.
He just let everybody know he was leaving, and he's
going to a company that was founded by the folks
that founded the company that I guess I'm it's kind
of the company I'm at that was acquired by the
company that I'm at, So they they're his former bosses,
(13:34):
not mine, but because he'd been there for several years.
But yeah, they're doing rails and Elixer, and I'm wondering
if it was that same kind of thing right where
they started with rails, because this company uses rails pretty
much exclusively, and then that somebody brought in the Elixer
and it's just kind of taken over most of the
new stuff it sent me.
Speaker 3 (13:57):
Something I observed with my current employee because also use Elixir,
and we use Rails, and we have some very very
long term Rails developers. We're like really attached to Rails
and Ruby. But we are also some colleagues who like
don't didn't have a lot of Rails or Elixir background
before they joined. And what ends up happening is that
especially the people who don't have like a strong attachment
(14:20):
one way or another after a while, prefer to write
things in elix heir. I can't like point to specific
things and say, okay, this is one reason why. But
it's an interesting trend. I've observed that, like if people
basically don't don't come with pre attached notions, they tend
to like cover Okay, like maybe I write this additional
(14:41):
service in elix Hea just because I like it to
do that that way. That's certainly something I've observed for myself.
I mean, I'm just add the unicorn who did Ruby
after I did elix heir, so I never wrote a
line of Ruby before I've already written a bunch of
lines in elix heir, So my my justment's probably bit
(15:02):
clouded there. Yeah, it's interesting to see how people act
to these technologies, even my dad experience before.
Speaker 2 (15:11):
Yeah.
Speaker 1 (15:11):
I try not to be too precious about it, but
at the end of the day, I get in and
I eventually just want to get crap done. And Rails
is just the thing that I'm most familiar with, right,
and so it's like, oh, I just banged out four features,
right instead of I had to go figure out how
to do this in Phoenix too, because I didn't know
how to do it before. But I can definitely see,
(15:34):
you know it, it has some characteristics and I see
the performance metrics on Alixer and I'm just going, holy crap,
that is just fast. So another milestone that I am
fully aware of and was around when we interviewed Chris
record for or when it came out, was live view
and that was I mean that that was a big deal.
(15:59):
And I have some friends that have kind of duplicated
that for Rails with stimulus right, stimulus reflex, But that
was amazing. That was freaking amazing.
Speaker 2 (16:09):
Yeah.
Speaker 3 (16:10):
I was actually on very elixy, kind of like a ELIXACONPUA.
When when Chris mccle also did a demo and that
was like super impressive because it did like it did
this demo with like the way animated things. By the way,
you also said like, okay, it's a super bad idea
to do the animation with live View, but just to
show you like you can, like it works. It's fast
enough to do like an animation.
Speaker 2 (16:31):
It's smooth.
Speaker 3 (16:32):
So yeah, that was insane. Just did to see like
this this thing on stage live by Chris and then
it really took up steam. I mean they you just
mentioned stimulus Reflex, but there's also like a PHP kind
of version of it, and I'm not the one ninety
percent sure how it's called that.
Speaker 1 (16:47):
No, it's it's rails. It runs over action cable on
the rails.
Speaker 3 (16:54):
Yeah, I know, like as soon as we fixed it's
all raised. But that's also a p spe clone of
Life View. Like I'm not more on hundred percent.
Speaker 1 (17:01):
Yeah, there are other plans of it.
Speaker 4 (17:03):
Yeah, that one's called live Wire And actually, if you
know Alpine Alpine actually came out as from the same
guy and now we're using Alpine within Live View, right,
So there's a lot of things that actually come outside
the community that also inspire or help out us, to
which I think is something that we should talk about
too because or something that needs to be mentioned at
(17:24):
least because it seems like we always inspire people, right,
Live View came out and it seems like an explosion
of people doing similar things like live Wire, and I
think the wine and Rails, if you guys, is that
culled stimulus or is that steamus is something else similar?
And then there's yeah, now we got Alpine, right, and
that came from the PHP community and that's part of
the table, that's part of the pedal stack, right.
Speaker 2 (17:44):
Which is pretty huge.
Speaker 1 (17:46):
Yeah. I keep wanting to do Alpine on jobs for Jabber.
We just haven't managed to get them around yet. But yeah,
it's brought about kind of these mini frameworks bringing them back.
Speaker 3 (17:55):
And I think Alpine was also like a big prompt
for Christmas call too at these script hooks into by
a few. So it was interesting to see like all
these ideas bounced back and forth between like these like
different communities to inspire of it to build something that
bigger than the parts of this the governments, some of
the products. I was curious to see, like, well, what's
(18:16):
going to be the next big thing? Right? Like I mean,
we just talk Life View. It's like thing X yeah yeah,
yeah probably yeah.
Speaker 1 (18:28):
Well that's interesting too because NX is the mathematics library, right, yeah.
And what's interesting to me is that, so I host
a whole bunch of the shows on the chatout TV,
including adventures and machine learning. And what's interesting is is
that we keep having conversations about some of the limitations
(18:49):
of Python and JavaScript, because because those are the two
primary languages, mostly Python and and specifically in their ability
to parallelize, and Alix are laying beam languages kind of
have I figured out yeah, just saying right, And so
I keep thinking that, yeah, if you can train your
(19:11):
your models and you can just here's the data, go go,
go right, your limitation is your number of fours.
Speaker 5 (19:18):
Yeah.
Speaker 7 (19:19):
Hey, folks, if you love this podcast and would like
to support the show, or if you wish you could
listen without the sponsorship messages, then you're in luck. We're
setting up new premium podcast feeds where you can get
all of the episodes released after Christmas twenty twenty without
the ads. Signing up will help us pay for editing
and production, and you can go sign up at devchat
dot tv slash Premium.
Speaker 5 (19:38):
It sounds like a great fit for Aleksir.
Speaker 6 (19:40):
Yeah, but when when I when I learned about that,
I was like this, I thought that that would like
break a lot of things, like bigger than live view
at it. Yeah, it's just it's almost revolution and the
amount that Jose is invested to see us through like
they I know that within seven days of announcing n
(20:00):
they had started and like kind of gotten like a
neural network's library action on like workable.
Speaker 5 (20:07):
Like that's the amount of hours they're putting into it.
It's crazy.
Speaker 3 (20:12):
It's super cool to see how Jose is also contributing
to like stuff beyond the core language, right, Like I mean,
what's Broadway and stuff like a lot of a lot
of really cool things an ex accent also have like
just has Today's do they has his hands in there?
Like it's pretty cool to see that. Yeah, I was
super impressed by an accent, but that they did most
(20:33):
like pretty much all of that was done without getting
new stuff to the language, like having these what's it
called death and when the computation happens on the GPU
if I remember correctly. To be honest, that's not my area.
Speaker 5 (20:46):
Of expert I think so, I think so yeah, I
think you're right.
Speaker 3 (20:49):
Yeah, and just like build all of this with like
the tools which were already there, like super cool, super impressive.
Speaker 6 (20:56):
And like speaking of NX, they are, they're also building
they built. I don't know if you guys have tried
the live book stuff for like like NX demoing, and
I think I know they did like a lot of
like scripting updates and the new ALIXA release for live view,
like the mix install you know, abiliity to like without
having to do a mixed project, you can just to
mix install in a script and like pull a new
(21:18):
package and try it out. I think they did that
for live book, which I think is part of the
NX ecosystem at the very least. And that's also very exciting.
Like I actually have considered try to convince my publisher
should we should we write a live book version of
my book as well.
Speaker 3 (21:37):
Yeah, I certainly play a wrong with that. So at
least one person's interested.
Speaker 5 (21:43):
Awesome, that's all I.
Speaker 1 (21:43):
Needed there, you go. Yeah. I mean the other thing
is is that just the nature of the Beam and
Elixir to me, there's so many directions that could really go.
I mean that's what's hard on how to pin down
what the next big thing is going to be. Right,
the communities that got brought into it were kind of
traditional communities, but.
Speaker 3 (22:06):
Yeah, especially with an accent now, I mean there's there's
an influx of people from different communities now, right, So
these people are going to bring in their ideas and like,
who knows what that's gonna end up. I certainly don't
because like machine learning is something for me, Like, yeah,
that's like there he'll be dragons, magic captains here. That's
not my area of expertise. So I wouldn't have expected
(22:29):
this to work out like it did.
Speaker 2 (22:31):
I was.
Speaker 3 (22:31):
I was a bit when I had first heard about
the announcement of an accent that people thinking of this
might be a machine or anything. It wasn't bit skeptical,
but yeah, that blew me out of the water. So yeah,
it's really really hard to think what what's the next
thing which is going to pop up? Especially with people
coming in and they're new ideas.
Speaker 5 (22:51):
It sounds like whatever chose it gets into.
Speaker 1 (22:56):
That may not be so far off, but it is
it because he got into machine learning, or because he
kind of saw that that was kind of a right
field to pluck.
Speaker 5 (23:09):
I think both, he said.
Speaker 6 (23:12):
He says very casually in his demos that I was
just dabbling with machine learning, and you know, right now,
obviously he thought about its futures, and Stephanie, he won't
be like investing so much time without knowing that the
right fit, like you said, shuck.
Speaker 1 (23:26):
Yeah, we're going to have quantum elixer. That'll be awesome,
just running on some cue bits IBM, we're coming for you.
But yeah, I think it occupies a lot of space,
and I think some of it is right for disruption.
I honestly think the traditional web development is right for
disruption at this point. I don't know what's going to
(23:48):
come in and hit it, but it is. And so yeah,
you know, kind of seeing the next thing. It's exciting
to see that they're looking at it. People like Jose
You don't know.
Speaker 6 (23:59):
I know they were like box of like web assembly
and Alixir or Beam at least last year. I don't
know if anything happened with that. I don't know if
you guys know.
Speaker 3 (24:08):
Lumen, it's called SOMs like this that assembly thing written
in Rust where they try to basically get at Beam
languages run in the browser and they deliberately don't try
to like build the full beam because madness, but like
a browsery five version of it very cold. But I'm
(24:31):
not sure that's standing.
Speaker 1 (24:32):
I mean, yeah, we had a really interesting discussion on
jobascript jabber about web containers and I don't think I
could even do it justice, But they managed to get node,
like full on node running in the browser. And yeah,
the guys aren't stack blits. They were explaining how they
(24:52):
put it all together and crap like that, and it
was it was pretty wild, right, And it runs on
web assembly and you is web assembly threads and things
like that, and yeah, you're bringing up WASM. I mean,
it does. It crosses my mind. It's like, how much,
even even a slightly hampered version of Elixir running on WASM,
(25:14):
what can we do in this kind of secure, sandboxed
arena that is a web container. What kind of apps,
what kind of functionality, what kinds of things can we
do with that that may consume some of this? I
don't know. I mean, it's it's it's crazy what people
are coming up with.
Speaker 5 (25:34):
Yah.
Speaker 2 (25:34):
Yeah.
Speaker 3 (25:35):
One thing I think is also interesting there is that
when you actually in for example, look at the performance
characteristics of life, you're a lot of time to spend
like encoding and decoding jacent. So for example, if you
would then actually have something running in the browser understanding
native Elixir data types, you lose that overhead. So maybe
that also boosts performances into new areas which we haven't
(25:57):
seen before, right, like who knows? Who knows?
Speaker 1 (26:01):
Yeah, I don't know.
Speaker 6 (26:02):
Ye, speaking of like similar projects, there was an open
gl Elixir port that was happening at one point too.
Someone gave a talk about that in Alexicon twenty seventeen,
or at least like open gel like interface. I don't
want to say open gl itself, but just graphics bringing
graphics to Elixir.
Speaker 5 (26:19):
I don't know about that's it's state. Also, I'm not
sure if it's a thing anymore.
Speaker 2 (26:23):
Yeah, that was void, wasn't it. I believe the guy
came from Xbox.
Speaker 5 (26:27):
I think the name sounds familiar.
Speaker 2 (26:29):
Yeah, yeah, so that's still going on.
Speaker 4 (26:31):
Actually he created a whole company with that, and he's
kind of working a lot of stuff and I think
he presented something a year or two ago where it's
really looking quite nice and you can remotely log into
these guys and play with them like you're extra there
so he's really moving forward with that. I fact, I
think he just got a chair in the Alixir in
(26:53):
the Airline Foundation. Oh wow something if I remember. Yeah,
So it may seem like it's not going for it,
but as far as I know, like if you make
a company that's going to be selling this kind of service,
you're moving forward. You're just you're selling to businesses, right,
because what he's investing heavily into it. Business had the
cash to pay for that, not individual people and you know,
(27:15):
open source kind of communities.
Speaker 6 (27:16):
Oh wait, I think I think we might be mixing up. Yeah,
I just I just googled it. I think what you're
talking about is Scenic, right.
Speaker 2 (27:24):
Yeah, yes, that one I think also uses open gl. Right.
Speaker 3 (27:28):
I just want to just mention scene.
Speaker 6 (27:31):
Yeah, I don't think Scenic uses open gl at least
I don't think it did. I thought it was just
a very simple way to build. I thought he's used
web gl for the desktop environments. But my understanding might
be off.
Speaker 3 (27:45):
From what I thirsd Scenic is mostly used like where
like more embedded areas, Like it's not really like BVV
reactive your eye scenario or what monogopa. I know where
I have a screen of size and x son of
some kind of the face of that. But it might
be off here, but that wasn't be something of a
last look that sex.
Speaker 4 (28:06):
Yeah, that was also my reading that I remember is that, Yeah,
you have to kind of know your screen size or.
Speaker 2 (28:11):
Else you could be in a world of pain right there.
It's just no way to change.
Speaker 1 (28:15):
It, yeah, I will say. I mean one one thing
that's also interesting is how many inroads Nerves is made
in the IoT community, because it it makes a dent
and it is a lot nicer to use than a
lot of the other IoT library libraries that are out there.
And and that may be a place where that things
crack wide open because a lot more things are becoming
(28:36):
programmer enabled. And so as you open that up and
again you you spin up another process, you spin up
another thing that is talking to your your stuff on
your network, right, and and you have multiple cores and
everything plays nicely together to share resources on your controlling machine.
(29:01):
I mean it, it's cool stuff. I've been tempted for
years to just figure out nerves and then to figure
out how to make it do a big Christmas light
display on my front lawn, you know, and just just
program the crap out of it, right, and I think
it'd be way fun And I don't know, but you
can do all kinds of things with that, and I've
(29:22):
I remember going to code Beam and going to some
of the other conferences and seeing talks at some of
the Elixir related conferences and Erlining related conferences where they're
using things like nerves for commercial applications where they're literally
speaking over modems, you know, sending data back and forth
(29:42):
and doing really interesting stuff with that, and anyway, all
kinds of different directions that this could go. It's it's
really really cool.
Speaker 6 (29:51):
Speaking of NOSE, I know there's like a company Very
They exclusively work on embedded apps with Alix. They're a
consulting company and they I think they have that big
enough that it's like surprising that there are and so
many projects going on that use nerves.
Speaker 1 (30:06):
Yeah, one of the past panelists on this show wasn't Mark.
I think Mark was the one that was running the
Nerves online meetup group and they they were doing all
kinds of cool stuff. And Yeah, I think we're gonna
keep going more and more into a more enabled world
and it'll just be cool to see where that can
take us to and then if it gets opened up
(30:28):
into a machine learning world, what can you do with
all that data? Right?
Speaker 3 (30:32):
Yeah, now that we've talked about it like it's it's
interesting to see in how many directions is this going?
Speaker 5 (30:37):
Right?
Speaker 3 (30:37):
I mean, if we have life few, we have an X,
machine learning, we have nerves, we have scenic. There's a
lot of motion in a lot of areas. That's it's interesting,
true of surfs.
Speaker 1 (30:49):
We just need to get into mobile and then we'll
be the new JavaScript's like Life few native that I'm
just saying, Yeah, there we go. Oh well, should we
start wrapping it up? Use some picks, all right, Audi,
why don't you start us off picks?
Speaker 5 (31:05):
Great?
Speaker 6 (31:05):
So I don't really have like picks traditional picks this time.
I have a couple really cool companies I know that
are hiring for Elixir developers.
Speaker 5 (31:14):
Do we do that?
Speaker 2 (31:15):
Yeah?
Speaker 1 (31:16):
Okay, go ahead.
Speaker 6 (31:17):
Yeah. So first one is a simple bet. We spoke
about NX. Unfortunately they don't use that NX yet for
their machine learning, but they are using machine learning and
Alixler and Rust in some capacity and they sound very exciting.
And the other one is the company where I first
started working and Cassam. They are in Boston and they're
looking for junior mid or Senior Elixir developer. Both sound
(31:40):
like great places to work and links will be in
the description.
Speaker 1 (31:43):
Awesome, Alan, Do you have some picks?
Speaker 4 (31:45):
I definitely have one pick that I'm very happy about recently.
It's been out for some time, but I don't think
you guys ever play with it. But I've been trying
to use Tailwind more and more these days, and they
came out with this jig compiler that is really really
interesting and really really fast compared to normal no JAS stuff.
And the cool thing about is that you can dynamically
(32:06):
create new classes as you're actually coding up your web pages.
Speaker 8 (32:11):
Have you guys seen this before or it's really really
cool effort that they're doing. The sustain time compilation. I
mean I used to the version one and like that
was the thing like where you had to specify what
you needed to brand like and then I had to
include everything there to sustan time stuff is really really cool.
Speaker 4 (32:27):
So normally you'd have to extend using this kind of
confake JS file and you have to kind of, you know,
say okay, this is my color. So that's what I
usually do is I put a bunch of brand colors
in there, and then I just extend the current colors
that they offer, so I've used them in the background,
the text color, et cetera. But what you can do
now is, say you have like one element you need
to do something very specific, a very weird number or
(32:49):
something that's not in the default library, you can always extend,
but then you have to wait for to compile. What
you can also do is you can say, like this
thing is going to be like sixty four REM high,
So you could say like, okay, h dash and then
use left left square bracket sixty four RAM and then
ending square bracket, and it'll dynamically create the class for you,
(33:12):
so you don't actually have to go in and even
have like a CSS filling more. You could basically do
most of the work yourself just by adding in these
arbitrary files. And what THEGIT does is it'll actually scan
through your files. And I am doing this with the
l e X and also the ex templates, so it's
actually going through my Phoenix files and actually creating classes
for me, which is really really cool. And then they
(33:33):
could also purge anything I don't use, so it also
knowses all that too, so the CSS file that comes
out is super super small and very specific to my project,
and it's really really really cool. I really like this
thing because there's always some elements you always have to
be very specific with, and I think this is definitely useful.
Now I could just look at my htmail my CSS
(33:54):
together at one place, I don't have to think about
where this class is being used, where it's not being used.
So that's that's really my big pick for this really.
Speaker 1 (34:00):
Cool Sasha, what are your picks?
Speaker 3 (34:02):
I just have one pick, and it might even be
something which a lot of people already familiar with, but
I've only recently discovered it. It's a library in Mili
called Bypass, and what it does basically, it's like you
it really opened starts a pluck server and you specify
an expectations as a testing thing. So you start a
pluck server, you specifically I'm not going to expect a
(34:23):
request with your l here with these headers and parameters,
and then I'm going to send back a response just
with user usual plug things you olso do in Phoenix.
And the cool thing there is because it really opens
a connection from local hosts. You when you do tests,
for example, for when you test your h GDP clients
sending a request on API, it goes through the whole stack,
(34:46):
right like, it really does build the ah GDP stack
and really sends it so you don't leave out like
this one thing in your test, which might then break
on production because you didn't test it. So yeah, a
lot of people might might already be familiar with it,
but it's pretty pretty nice to years for all these
its top clienty tests.
Speaker 1 (35:06):
All right, I'm gonna throw out a few picks. I
tend to pick more stuff along the lines of, Hey,
I'm running a podcast network and so it's just stuff
that I'm learning things like that. If you want more
in depth stuff, it's also kind of my personal journey
alongside that. Go to devchat dot tv Slash Premium and
you can pick up the Doing dev Chat Premium podcast,
(35:26):
and I talk every week about Hey, this is what
I'm working on, this is what I'm thinking about, this
is what I've you know, gotten done this week, stuff
like that. I've been on this who not how hiring
slash getting people more involved in certain aspects of running
the shows kick lately and anyway, so since I mentioned
it Who Not How. I think I've mentioned it like
(35:48):
every week for the last like three weeks on like
every show I've been on. So but it's it's a
terrific book if you're trying to build a team, if
you're trying to run a business, if you're trying to
figure out how to get more stuff done, if you're
kind of a mission driven or outcome driven sort of setup,
anything like that. I highly, highly, highly recommend it if
(36:10):
you run any kind of organization. It basically just talks
about how to build an organization, get people in, build
relationships to get done what you want to get done
by empowering people to get to do the things that
they're really good at. And that way they're fulfilled doing
what they're good at, and you're fulfilled doing what you're
good at. So yeah, that's a pick. I really really
(36:34):
dig these kind of growth books. Okay, And so I'm
actually rereading Who Not How right now. I just I
have the hardback version. It's sitting on my desk, and
so when I get up in the morning and I'm
going to pick another book. It's a resource. I don't
follow their prescription really religiously, but I do kind of
(36:57):
follow what they tell me to in the sense that
I up in the morning and I follow up routine.
The book's called The Miracle Morning by Hal L. Rod,
and yeah, he kind of tells you to do a
specific set other things in the morning. I do my
own things, but the idea is pretty good. So definitely
check that book out as well. The books that I've
been listening to while I'm like either driving to swim
(37:21):
the swim team or going on my runs or whatever.
The latest ones that I've listened to are psycho Cybernetics
by doctor Maxwell Maltz, which is terrific, by the way.
It just talks about how you become the person that
is or has what you want to have or be.
And then the other book is As a Man Thinketh
(37:44):
and that's kind of more conceptual version of psycho Cybernetics, honestly,
but it just talks about how your mindset determines what
you get. And those are both terrific, terrific books, and
so I'm going to pick both of those. I know
I am picking a lot of stuff, but honestly, I've
just been thinking very very deeply about a lot of
these things, and I really feel like, this is stuff
(38:07):
that we don't talk a lot about as developers. You know,
we kind of talk about it either as a job,
or we talk about how we contribute to the community,
or we talk about what's going on in the community,
and we fail to talk about what we really want
from our careers, from our contributions, from our open source,
where we want to go and then how we get there.
(38:30):
And I really want to help people get there, So
these books will all help you get there. If you're
looking for more of that, the dev Influencers podcast is
where I'm talking more about that. I'm also trying to
put out bonus episodes where I talk more about that,
but I have other people putting demands on that sometimes,
so dev influencers dot Com slash podcast where you get
(38:52):
the rest of that. And yeah, that's pretty much my picks.
I'll save the rest of it for next week. I've
got a ton of stuff that I just want to share,
but I don't want to take all your time just
saying oh, and this and that and anyway, Well that
was fun, guys, Yeah, totally
Hey everybody, and welcome back to another episode of the
Elixer Mix. Today on our panel we have Audi, Hello, Alan.
Speaker 2 (00:14):
Hello, Hello, Sasha, Hello everybody.
Speaker 1 (00:18):
Is this your first week back?
Speaker 3 (00:19):
Yes? Regular it is.
Speaker 1 (00:23):
How how cool you are why we invited you back?
Speaker 3 (00:28):
So yeah, a while ago, I as the guests joining
the podcast and talking about the library of Mind, Nigger
and like behaviors and looking. I've been using alex for
like a four years now and focus the libraries published.
So it's great to know be here and talk to you.
(00:48):
Exchange oudiens, happy to be here.
Speaker 2 (00:50):
Nice.
Speaker 1 (00:51):
I'm Charles Maxwood from dev chat dot TV, And yeah,
so we decided to talk about kind of the history
and evolution of Alixir. Yeah, I think I kind of
want to start this one off and kick it off.
So the first podcast episode that I ever did related
to Alixir was on Ruby roachs and he has no
surprise to anybody, I think if you've been in the
Elixir community for very long. We interviewed this guy who'd
(01:14):
been in the Ruby community for a long time, written
a whole bunch of Ruby libraries, ruby gems, started a
Ruby consultancy, and then he went off and was fiddling
with this airline stuff and decided to write Elixir. So
it was jose Vaalin and so yeah, we were like, well,
come come tell us about this this thing, and so
he came on and he talked about Elixir and it
(01:35):
was kind of like this funky thing that he did
on the side while he was still writing all the
ruby stuff that we all used. So anyway, it's come
a long way since then, obviously, But yeah, I like
that flavor, right, because you never know if the wacky
project is going to turn into something that's it this
ecosystem that all these people earn a living at, right.
Speaker 4 (01:58):
Yeah, It's it's kind of crazy how fast things have
kind of been rolled along, right that this language just
came out what came out quite some time ago, but
it still feels like quite new to me, although it's
been changing a lot, especially with this recent one at
twelve release.
Speaker 1 (02:11):
Yeah, I'm kind of wondering what are kind of the
big milestones or sign posts I guess in Elixir's history
that you guys see, right, I mean some of them
are going to come in I guess after you discovered Elixir,
But what what were the turning points I guess that
got you in in the first place.
Speaker 3 (02:31):
Right, It's actually something which I'd say that got me in.
When I started with Elixir, it was just because my
employer used the Lixir, so I basically had to. But
when I started, I think that was one point four
or one point three, we didn't have a formata back then,
so it was always like we always had these discussions about, Okay,
how does the code look like superformata. It is like
(02:52):
the superf like that. So it was pretty cool when
I'm not sure which version accepted, but when Lyxi finally
got the opinionfomata and a lot of these discussions just
died down because the formata did this for you.
Speaker 1 (03:06):
Right. I have a question for you, Sasha, regarding that.
So you said your employer was using it. Were they
using it with Phoenix or were they using it for
something else like Nerves or some other project.
Speaker 3 (03:18):
Yeah, at that point it was Phoenix. I think Nerves
wasn't even in its existence back then, so I mean
that was probably about four years ago. Maybe it was
already there, but that definitely not a speaks today. So yeah,
it was. The company I used to work back then
was doing a lot of I WoT stuff. So the
real time capabilities of Felix. They were pretty nice in
(03:39):
that area, which is why are you using it?
Speaker 1 (03:42):
Right? The other question I have related to that is,
so you came to it through a job, And it
seems like a lot of people when they go look
for their next job, it's like, Okay, I've got fifty
zillion years doing JavaScript or Ruby or GHP or Python
or whatever, right, And so they go and they just
(04:03):
kind of stick with that technology because they've got so
much seniority and they can kind of talk their way
into whatever kind of job they want. So what made
you take a job where you were going to be
working in a different technology.
Speaker 3 (04:17):
The main reason was actually because I was switching over
from a mobile developer job with like occasional backend what development,
to like a full time back and development job, like
junior side. Because I'm a previous employer, I did mostly
iOS development, but I did the occasional PHP back end
and that was always something which I enjoyed more So
(04:38):
I was looking for an opportunity to work more in
that area. And then I came across this other company
and they used the licks here and the functional language,
and I was like, oh, it sounds interesting. Enough and
now I'm here today, so things, I guess things work out.
Speaker 1 (04:52):
That makes sense.
Speaker 5 (04:53):
Yeah, so me.
Speaker 6 (04:54):
It's one of the things I remember is so something
very simple and trivial like the with statement that we
now take for granted ALEXA. When it came out, it
did not have that, and I remember even using I don't.
Speaker 5 (05:05):
Think that was even unless in all those things.
Speaker 6 (05:08):
And I know even using IF was like, don't use
it unless they really want to use it, because it
IF kind of uses case within like cases like the
main statement. And it was actually my first job too,
doing elexir in twenty fifteen, and so it looks it
was like just one pointer at that time. And when
they introduced with statement, it really kind of cleaned up
(05:29):
our code quite a bit, because you know, now you're
thinking in terms of the railway pattern, right, like returning
okay result or error error, and that just cleans up
your entire controller.
Speaker 5 (05:40):
And then Phoenix kind of.
Speaker 6 (05:41):
Adopted that pattern with the fallback controller, right, like cleaning
up that error as well, and like standardizing what error
is your APIs or return and stuff. So I think
that with kind of catalyzed that entire list of tools.
Speaker 5 (05:55):
I was a really big fan.
Speaker 3 (05:56):
Of with Ford was basically nested case statements all the
way down, right, like a lot of case statements, and
there wasn't much anything else you could do. Like that
was the way things were written it which also made
sense when you think where Alexi came from from what's
of Berlin, because that's how you do handling and arline.
You do the case statements, or you pet on if
you match explicitly on the okay with radio when you
(06:19):
say okay, I want to raise you, but yeah, right.
Speaker 6 (06:22):
Or use the pipe operator, but the functions kind of
can take okay or error energy like your entire pipeline.
Speaker 3 (06:31):
Definitely you have written code like that. I've just looked
at the release notes from the LICKS and like the
format actually came in one point six, but there's a
lot of good stuff in one point six. Also, a
dynamic supervisor came in there, so that was something we
didn't have. I mean, I remember that that the supervisor
still has the simple one for one strategy, right, which
is like kind of akin to the dynemic supervisor. But
(06:53):
there's one thing about the simple one for one strategy
that it's like a lie. It's not simple. It's it's
definitely not simple understand that. I mean, we had to
use it in one project, and like, I mean, it
was pretty new to addix. Yeah, so I was like
trying to read the documentation and I grabbed my head
around and I just I just failed. Just went to
(07:14):
a colleague at some point and please explain to me.
Speaker 1 (07:17):
Yeah, I'm kind of curious because I can't like pick
out the features or anything because I've just kind of
been watching the community more than participating in it. And
over the years, like it was kind of this thing
that was just kind of out there, you know, and
and the Airline community was also another community. It was
just kind of I just kind of watched it and
(07:37):
it's kind of steadily grown. And then I don't know
what it was, but at some point it seems like
the Elixir community just really took off. And I'm not
sure where, I'm not sure why, I'm not sure what,
but it really took off. And about I want to say,
about four or five years ago was when it really
kind of kind of caught up with me that a
(08:00):
lot of the friends of mine that I had connected
with over the years in the Ruby community, it all
moved to it a lot of the people that I
knew from some of the other communities that moved to it.
I was actually getting asked on a regular basis when
I was going to start an Elixer show, so we
started this one. I mean, was there something around then
that really just gelled? Was it that they finally there
(08:25):
was Alixer comp and there were some of these other things.
Those tend to be more the symptoms of the community
growing rather than the cause. I'm trying to put my
finger on, you know, Was there any one or several
things that happened around then that that kind of came
together for it?
Speaker 6 (08:40):
I guess must be like the bigger companies kind of
adopting it, like Pepsi e commerce and I know, cars
and you know, like the like who were in like
more legacy kind of a programming language, just them adopting it.
I actually remember like a more quantifiable like difference, the
difference in the number of responses, the number of people.
In alexir conf twenty seventeen and twenty eighteen, they announced
(09:03):
it at the keynote that the number of sponsors they
have more than quadrupled in one year.
Speaker 5 (09:10):
And it's just that what happened during that time.
Speaker 6 (09:13):
I don't know.
Speaker 1 (09:14):
Yeah, I guess that makes sense. I mean, I think
another thing that I saw at the time was and
I guess this isn't a technical milestone, but so I
got invited to CODEBAM in twenty twenty eighteen or twenty nineteen.
But there were a whole bunch of talks for Yeah,
these bigger companies that were talking about and they weren't
like talking. Some of them were talking about how they
(09:35):
had built their infrastructure on the Beam and how they
had built a lot of it on a elixer. But
more than that, they were talking about how they had scaled,
how they had lowered the response times, and a lot
of the concurrency setups that they were doing and things
like that. And so they had actually solved a lot
(09:57):
of the issues that existed in some of the systems
that they it had before by migrating to Elixir. And
some of them were like these huge setups like WhatsApp
I think was one of them. I think Discord does
a whole bunch of stuff in the back end with Elixir.
There are a whole bunch of other ones and they
were like these massive massive back end systems that either
(10:20):
run in like dedicated data centers or running the cloud,
and I mean they just like crush the amount of
bandwidth and data that just just course through them. And yeah,
I think people really saw what elixor systems were capable
of that you couldn't do with say a no JS
or a Ruby on rails, and people started saying, Okay, well,
(10:45):
if I need say data overforms, then dot net Ruby
on rails, that makes sense, right, But if I need
if I'm sending raw data, audio data, if I'm sending
gobs and gobs and gobs of data, it makes a
lot of sense to have a system that can handle
the volume of data, but not just the volume, but
(11:08):
the sheer number of discrete messages at a time. And
that's that's where I think people figured out, oh, yeah,
I need something that can do this. And Elixir was
easier for the developers to digest than earline.
Speaker 3 (11:24):
I think a huge part of that is also how
easy it is to pick up Phoenix, because I mean
you just also, yeah, to like submit the form right
safer form. It's super easy in Rails, but it's also
super easy in Phoenix, so you can get started pretty
easily there too. But then the skating story is a
completely different one. So while earline, yeah, it is certainly
(11:48):
something which could also super easily handle these these these
skating scenarios, it's not as easy to pick up as
serious So I think a lot of success came from
that too, that that people could see, Okay, yeah, this
can scale two different in different ways and maybe better
than than other technologies out there, but it's also not
as super hard to pick up as some of these
(12:09):
more let's say these things totally.
Speaker 6 (12:12):
Yeah, that's definitely I think a huge point. And as
you guys are talking about it, I just like it
kind of reminded me of my older daid Alixa, and
like why did I like Elixer, And I just like
I also realized, like everyone around me seemed to just
leve Alixer and the involvement of you know, Jose and
Christma code in the community and like listening to feedback
(12:32):
and advertising, and Jose was being involved in like Phoenix
and writing marks and like kind of driving the community's
vision forward while keeping it fun for developers, right, Like
I mean, like everyone who's worked with Alixa wants to
work in the Elixer. So like companies who've adopted a
lix a little bit have no choice but to adopt
it more because people who work there want to keep
(12:54):
working in Alixer. So yeah, like in twenty fifteen, my
first job project we built with the company's first Elixa app,
And in twenty twenty we're probably building it's thirty thirty
of the Elixa app.
Speaker 5 (13:08):
So just like because I didn't want to build Ruby
on rails app.
Speaker 1 (13:13):
So yeah, it's interesting too. A friend of mine, he's
actually the one that recruited me at my current employer.
He just let everybody know he was leaving, and he's
going to a company that was founded by the folks
that founded the company that I guess I'm it's kind
of the company I'm at that was acquired by the
company that I'm at, So they they're his former bosses,
(13:34):
not mine, but because he'd been there for several years.
But yeah, they're doing rails and Elixer, and I'm wondering
if it was that same kind of thing right where
they started with rails, because this company uses rails pretty
much exclusively, and then that somebody brought in the Elixer
and it's just kind of taken over most of the
new stuff it sent me.
Speaker 3 (13:57):
Something I observed with my current employee because also use Elixir,
and we use Rails, and we have some very very
long term Rails developers. We're like really attached to Rails
and Ruby. But we are also some colleagues who like
don't didn't have a lot of Rails or Elixir background
before they joined. And what ends up happening is that
especially the people who don't have like a strong attachment
(14:20):
one way or another after a while, prefer to write
things in elix heir. I can't like point to specific
things and say, okay, this is one reason why. But
it's an interesting trend. I've observed that, like if people
basically don't don't come with pre attached notions, they tend
to like cover Okay, like maybe I write this additional
(14:41):
service in elix Hea just because I like it to
do that that way. That's certainly something I've observed for myself.
I mean, I'm just add the unicorn who did Ruby
after I did elix heir, so I never wrote a
line of Ruby before I've already written a bunch of
lines in elix heir, So my my justment's probably bit
(15:02):
clouded there. Yeah, it's interesting to see how people act
to these technologies, even my dad experience before.
Speaker 2 (15:11):
Yeah.
Speaker 1 (15:11):
I try not to be too precious about it, but
at the end of the day, I get in and
I eventually just want to get crap done. And Rails
is just the thing that I'm most familiar with, right,
and so it's like, oh, I just banged out four features,
right instead of I had to go figure out how
to do this in Phoenix too, because I didn't know
how to do it before. But I can definitely see,
(15:34):
you know it, it has some characteristics and I see
the performance metrics on Alixer and I'm just going, holy crap,
that is just fast. So another milestone that I am
fully aware of and was around when we interviewed Chris
record for or when it came out, was live view
and that was I mean that that was a big deal.
(15:59):
And I have some friends that have kind of duplicated
that for Rails with stimulus right, stimulus reflex, But that
was amazing. That was freaking amazing.
Speaker 2 (16:09):
Yeah.
Speaker 3 (16:10):
I was actually on very elixy, kind of like a ELIXACONPUA.
When when Chris mccle also did a demo and that
was like super impressive because it did like it did
this demo with like the way animated things. By the way,
you also said like, okay, it's a super bad idea
to do the animation with live View, but just to
show you like you can, like it works. It's fast
enough to do like an animation.
Speaker 2 (16:31):
It's smooth.
Speaker 3 (16:32):
So yeah, that was insane. Just did to see like
this this thing on stage live by Chris and then
it really took up steam. I mean they you just
mentioned stimulus Reflex, but there's also like a PHP kind
of version of it, and I'm not the one ninety
percent sure how it's called that.
Speaker 1 (16:47):
No, it's it's rails. It runs over action cable on
the rails.
Speaker 3 (16:54):
Yeah, I know, like as soon as we fixed it's
all raised. But that's also a p spe clone of
Life View. Like I'm not more on hundred percent.
Speaker 1 (17:01):
Yeah, there are other plans of it.
Speaker 4 (17:03):
Yeah, that one's called live Wire And actually, if you
know Alpine Alpine actually came out as from the same
guy and now we're using Alpine within Live View, right,
So there's a lot of things that actually come outside
the community that also inspire or help out us, to
which I think is something that we should talk about
too because or something that needs to be mentioned at
(17:24):
least because it seems like we always inspire people, right,
Live View came out and it seems like an explosion
of people doing similar things like live Wire, and I
think the wine and Rails, if you guys, is that
culled stimulus or is that steamus is something else similar?
And then there's yeah, now we got Alpine, right, and
that came from the PHP community and that's part of
the table, that's part of the pedal stack, right.
Speaker 2 (17:44):
Which is pretty huge.
Speaker 1 (17:46):
Yeah. I keep wanting to do Alpine on jobs for Jabber.
We just haven't managed to get them around yet. But yeah,
it's brought about kind of these mini frameworks bringing them back.
Speaker 3 (17:55):
And I think Alpine was also like a big prompt
for Christmas call too at these script hooks into by
a few. So it was interesting to see like all
these ideas bounced back and forth between like these like
different communities to inspire of it to build something that
bigger than the parts of this the governments, some of
the products. I was curious to see, like, well, what's
(18:16):
going to be the next big thing? Right? Like I mean,
we just talk Life View. It's like thing X yeah yeah,
yeah probably yeah.
Speaker 1 (18:28):
Well that's interesting too because NX is the mathematics library, right, yeah.
And what's interesting to me is that, so I host
a whole bunch of the shows on the chatout TV,
including adventures and machine learning. And what's interesting is is
that we keep having conversations about some of the limitations
(18:49):
of Python and JavaScript, because because those are the two
primary languages, mostly Python and and specifically in their ability
to parallelize, and Alix are laying beam languages kind of
have I figured out yeah, just saying right, And so
I keep thinking that, yeah, if you can train your
(19:11):
your models and you can just here's the data, go go,
go right, your limitation is your number of fours.
Speaker 5 (19:18):
Yeah.
Speaker 7 (19:19):
Hey, folks, if you love this podcast and would like
to support the show, or if you wish you could
listen without the sponsorship messages, then you're in luck. We're
setting up new premium podcast feeds where you can get
all of the episodes released after Christmas twenty twenty without
the ads. Signing up will help us pay for editing
and production, and you can go sign up at devchat
dot tv slash Premium.
Speaker 5 (19:38):
It sounds like a great fit for Aleksir.
Speaker 6 (19:40):
Yeah, but when when I when I learned about that,
I was like this, I thought that that would like
break a lot of things, like bigger than live view
at it. Yeah, it's just it's almost revolution and the
amount that Jose is invested to see us through like
they I know that within seven days of announcing n
(20:00):
they had started and like kind of gotten like a
neural network's library action on like workable.
Speaker 5 (20:07):
Like that's the amount of hours they're putting into it.
It's crazy.
Speaker 3 (20:12):
It's super cool to see how Jose is also contributing
to like stuff beyond the core language, right, Like I mean,
what's Broadway and stuff like a lot of a lot
of really cool things an ex accent also have like
just has Today's do they has his hands in there?
Like it's pretty cool to see that. Yeah, I was
super impressed by an accent, but that they did most
(20:33):
like pretty much all of that was done without getting
new stuff to the language, like having these what's it
called death and when the computation happens on the GPU
if I remember correctly. To be honest, that's not my area.
Speaker 5 (20:46):
Of expert I think so, I think so yeah, I
think you're right.
Speaker 3 (20:49):
Yeah, and just like build all of this with like
the tools which were already there, like super cool, super impressive.
Speaker 6 (20:56):
And like speaking of NX, they are, they're also building
they built. I don't know if you guys have tried
the live book stuff for like like NX demoing, and
I think I know they did like a lot of
like scripting updates and the new ALIXA release for live view,
like the mix install you know, abiliity to like without
having to do a mixed project, you can just to
mix install in a script and like pull a new
(21:18):
package and try it out. I think they did that
for live book, which I think is part of the
NX ecosystem at the very least. And that's also very exciting.
Like I actually have considered try to convince my publisher
should we should we write a live book version of
my book as well.
Speaker 3 (21:37):
Yeah, I certainly play a wrong with that. So at
least one person's interested.
Speaker 5 (21:43):
Awesome, that's all I.
Speaker 1 (21:43):
Needed there, you go. Yeah. I mean the other thing
is is that just the nature of the Beam and
Elixir to me, there's so many directions that could really go.
I mean that's what's hard on how to pin down
what the next big thing is going to be. Right,
the communities that got brought into it were kind of
traditional communities, but.
Speaker 3 (22:06):
Yeah, especially with an accent now, I mean there's there's
an influx of people from different communities now, right, So
these people are going to bring in their ideas and like,
who knows what that's gonna end up. I certainly don't
because like machine learning is something for me, Like, yeah,
that's like there he'll be dragons, magic captains here. That's
not my area of expertise. So I wouldn't have expected
(22:29):
this to work out like it did.
Speaker 2 (22:31):
I was.
Speaker 3 (22:31):
I was a bit when I had first heard about
the announcement of an accent that people thinking of this
might be a machine or anything. It wasn't bit skeptical,
but yeah, that blew me out of the water. So yeah,
it's really really hard to think what what's the next
thing which is going to pop up? Especially with people
coming in and they're new ideas.
Speaker 5 (22:51):
It sounds like whatever chose it gets into.
Speaker 1 (22:56):
That may not be so far off, but it is
it because he got into machine learning, or because he
kind of saw that that was kind of a right
field to pluck.
Speaker 5 (23:09):
I think both, he said.
Speaker 6 (23:12):
He says very casually in his demos that I was
just dabbling with machine learning, and you know, right now,
obviously he thought about its futures, and Stephanie, he won't
be like investing so much time without knowing that the
right fit, like you said, shuck.
Speaker 1 (23:26):
Yeah, we're going to have quantum elixer. That'll be awesome,
just running on some cue bits IBM, we're coming for you.
But yeah, I think it occupies a lot of space,
and I think some of it is right for disruption.
I honestly think the traditional web development is right for
disruption at this point. I don't know what's going to
(23:48):
come in and hit it, but it is. And so yeah,
you know, kind of seeing the next thing. It's exciting
to see that they're looking at it. People like Jose
You don't know.
Speaker 6 (23:59):
I know they were like box of like web assembly
and Alixir or Beam at least last year. I don't
know if anything happened with that. I don't know if
you guys know.
Speaker 3 (24:08):
Lumen, it's called SOMs like this that assembly thing written
in Rust where they try to basically get at Beam
languages run in the browser and they deliberately don't try
to like build the full beam because madness, but like
a browsery five version of it very cold. But I'm
(24:31):
not sure that's standing.
Speaker 1 (24:32):
I mean, yeah, we had a really interesting discussion on
jobascript jabber about web containers and I don't think I
could even do it justice, But they managed to get node,
like full on node running in the browser. And yeah,
the guys aren't stack blits. They were explaining how they
(24:52):
put it all together and crap like that, and it
was it was pretty wild, right, And it runs on
web assembly and you is web assembly threads and things
like that, and yeah, you're bringing up WASM. I mean,
it does. It crosses my mind. It's like, how much,
even even a slightly hampered version of Elixir running on WASM,
(25:14):
what can we do in this kind of secure, sandboxed
arena that is a web container. What kind of apps,
what kind of functionality, what kinds of things can we
do with that that may consume some of this? I
don't know. I mean, it's it's it's crazy what people
are coming up with.
Speaker 5 (25:34):
Yah.
Speaker 2 (25:34):
Yeah.
Speaker 3 (25:35):
One thing I think is also interesting there is that
when you actually in for example, look at the performance
characteristics of life, you're a lot of time to spend
like encoding and decoding jacent. So for example, if you
would then actually have something running in the browser understanding
native Elixir data types, you lose that overhead. So maybe
that also boosts performances into new areas which we haven't
(25:57):
seen before, right, like who knows? Who knows?
Speaker 1 (26:01):
Yeah, I don't know.
Speaker 6 (26:02):
Ye, speaking of like similar projects, there was an open
gl Elixir port that was happening at one point too.
Someone gave a talk about that in Alexicon twenty seventeen,
or at least like open gel like interface. I don't
want to say open gl itself, but just graphics bringing
graphics to Elixir.
Speaker 5 (26:19):
I don't know about that's it's state. Also, I'm not
sure if it's a thing anymore.
Speaker 2 (26:23):
Yeah, that was void, wasn't it. I believe the guy
came from Xbox.
Speaker 5 (26:27):
I think the name sounds familiar.
Speaker 2 (26:29):
Yeah, yeah, so that's still going on.
Speaker 4 (26:31):
Actually he created a whole company with that, and he's
kind of working a lot of stuff and I think
he presented something a year or two ago where it's
really looking quite nice and you can remotely log into
these guys and play with them like you're extra there
so he's really moving forward with that. I fact, I
think he just got a chair in the Alixir in
(26:53):
the Airline Foundation. Oh wow something if I remember. Yeah,
So it may seem like it's not going for it,
but as far as I know, like if you make
a company that's going to be selling this kind of service,
you're moving forward. You're just you're selling to businesses, right,
because what he's investing heavily into it. Business had the
cash to pay for that, not individual people and you know,
(27:15):
open source kind of communities.
Speaker 6 (27:16):
Oh wait, I think I think we might be mixing up. Yeah,
I just I just googled it. I think what you're
talking about is Scenic, right.
Speaker 2 (27:24):
Yeah, yes, that one I think also uses open gl. Right.
Speaker 3 (27:28):
I just want to just mention scene.
Speaker 6 (27:31):
Yeah, I don't think Scenic uses open gl at least
I don't think it did. I thought it was just
a very simple way to build. I thought he's used
web gl for the desktop environments. But my understanding might
be off.
Speaker 3 (27:45):
From what I thirsd Scenic is mostly used like where
like more embedded areas, Like it's not really like BVV
reactive your eye scenario or what monogopa. I know where
I have a screen of size and x son of
some kind of the face of that. But it might
be off here, but that wasn't be something of a
last look that sex.
Speaker 4 (28:06):
Yeah, that was also my reading that I remember is that, Yeah,
you have to kind of know your screen size or.
Speaker 2 (28:11):
Else you could be in a world of pain right there.
It's just no way to change.
Speaker 1 (28:15):
It, yeah, I will say. I mean one one thing
that's also interesting is how many inroads Nerves is made
in the IoT community, because it it makes a dent
and it is a lot nicer to use than a
lot of the other IoT library libraries that are out there.
And and that may be a place where that things
crack wide open because a lot more things are becoming
(28:36):
programmer enabled. And so as you open that up and
again you you spin up another process, you spin up
another thing that is talking to your your stuff on
your network, right, and and you have multiple cores and
everything plays nicely together to share resources on your controlling machine.
(29:01):
I mean it, it's cool stuff. I've been tempted for
years to just figure out nerves and then to figure
out how to make it do a big Christmas light
display on my front lawn, you know, and just just
program the crap out of it, right, and I think
it'd be way fun And I don't know, but you
can do all kinds of things with that, and I've
(29:22):
I remember going to code Beam and going to some
of the other conferences and seeing talks at some of
the Elixir related conferences and Erlining related conferences where they're
using things like nerves for commercial applications where they're literally
speaking over modems, you know, sending data back and forth
(29:42):
and doing really interesting stuff with that, and anyway, all
kinds of different directions that this could go. It's it's
really really cool.
Speaker 6 (29:51):
Speaking of NOSE, I know there's like a company Very
They exclusively work on embedded apps with Alix. They're a
consulting company and they I think they have that big
enough that it's like surprising that there are and so
many projects going on that use nerves.
Speaker 1 (30:06):
Yeah, one of the past panelists on this show wasn't Mark.
I think Mark was the one that was running the
Nerves online meetup group and they they were doing all
kinds of cool stuff. And Yeah, I think we're gonna
keep going more and more into a more enabled world
and it'll just be cool to see where that can
take us to and then if it gets opened up
(30:28):
into a machine learning world, what can you do with
all that data? Right?
Speaker 3 (30:32):
Yeah, now that we've talked about it like it's it's
interesting to see in how many directions is this going?
Speaker 5 (30:37):
Right?
Speaker 3 (30:37):
I mean, if we have life few, we have an X,
machine learning, we have nerves, we have scenic. There's a
lot of motion in a lot of areas. That's it's interesting,
true of surfs.
Speaker 1 (30:49):
We just need to get into mobile and then we'll
be the new JavaScript's like Life few native that I'm
just saying, Yeah, there we go. Oh well, should we
start wrapping it up? Use some picks, all right, Audi,
why don't you start us off picks?
Speaker 5 (31:05):
Great?
Speaker 6 (31:05):
So I don't really have like picks traditional picks this time.
I have a couple really cool companies I know that
are hiring for Elixir developers.
Speaker 5 (31:14):
Do we do that?
Speaker 2 (31:15):
Yeah?
Speaker 1 (31:16):
Okay, go ahead.
Speaker 6 (31:17):
Yeah. So first one is a simple bet. We spoke
about NX. Unfortunately they don't use that NX yet for
their machine learning, but they are using machine learning and
Alixler and Rust in some capacity and they sound very exciting.
And the other one is the company where I first
started working and Cassam. They are in Boston and they're
looking for junior mid or Senior Elixir developer. Both sound
(31:40):
like great places to work and links will be in
the description.
Speaker 1 (31:43):
Awesome, Alan, Do you have some picks?
Speaker 4 (31:45):
I definitely have one pick that I'm very happy about recently.
It's been out for some time, but I don't think
you guys ever play with it. But I've been trying
to use Tailwind more and more these days, and they
came out with this jig compiler that is really really
interesting and really really fast compared to normal no JAS stuff.
And the cool thing about is that you can dynamically
(32:06):
create new classes as you're actually coding up your web pages.
Speaker 8 (32:11):
Have you guys seen this before or it's really really
cool effort that they're doing. The sustain time compilation. I
mean I used to the version one and like that
was the thing like where you had to specify what
you needed to brand like and then I had to
include everything there to sustan time stuff is really really cool.
Speaker 4 (32:27):
So normally you'd have to extend using this kind of
confake JS file and you have to kind of, you know,
say okay, this is my color. So that's what I
usually do is I put a bunch of brand colors
in there, and then I just extend the current colors
that they offer, so I've used them in the background,
the text color, et cetera. But what you can do
now is, say you have like one element you need
to do something very specific, a very weird number or
(32:49):
something that's not in the default library, you can always extend,
but then you have to wait for to compile. What
you can also do is you can say, like this
thing is going to be like sixty four REM high,
So you could say like, okay, h dash and then
use left left square bracket sixty four RAM and then
ending square bracket, and it'll dynamically create the class for you,
(33:12):
so you don't actually have to go in and even
have like a CSS filling more. You could basically do
most of the work yourself just by adding in these
arbitrary files. And what THEGIT does is it'll actually scan
through your files. And I am doing this with the
l e X and also the ex templates, so it's
actually going through my Phoenix files and actually creating classes
for me, which is really really cool. And then they
(33:33):
could also purge anything I don't use, so it also
knowses all that too, so the CSS file that comes
out is super super small and very specific to my project,
and it's really really really cool. I really like this
thing because there's always some elements you always have to
be very specific with, and I think this is definitely useful.
Now I could just look at my htmail my CSS
(33:54):
together at one place, I don't have to think about
where this class is being used, where it's not being used.
So that's that's really my big pick for this really.
Speaker 1 (34:00):
Cool Sasha, what are your picks?
Speaker 3 (34:02):
I just have one pick, and it might even be
something which a lot of people already familiar with, but
I've only recently discovered it. It's a library in Mili
called Bypass, and what it does basically, it's like you
it really opened starts a pluck server and you specify
an expectations as a testing thing. So you start a
pluck server, you specifically I'm not going to expect a
(34:23):
request with your l here with these headers and parameters,
and then I'm going to send back a response just
with user usual plug things you olso do in Phoenix.
And the cool thing there is because it really opens
a connection from local hosts. You when you do tests,
for example, for when you test your h GDP clients
sending a request on API, it goes through the whole stack,
(34:46):
right like, it really does build the ah GDP stack
and really sends it so you don't leave out like
this one thing in your test, which might then break
on production because you didn't test it. So yeah, a
lot of people might might already be familiar with it,
but it's pretty pretty nice to years for all these
its top clienty tests.
Speaker 1 (35:06):
All right, I'm gonna throw out a few picks. I
tend to pick more stuff along the lines of, Hey,
I'm running a podcast network and so it's just stuff
that I'm learning things like that. If you want more
in depth stuff, it's also kind of my personal journey
alongside that. Go to devchat dot tv Slash Premium and
you can pick up the Doing dev Chat Premium podcast,
(35:26):
and I talk every week about Hey, this is what
I'm working on, this is what I'm thinking about, this
is what I've you know, gotten done this week, stuff
like that. I've been on this who not how hiring
slash getting people more involved in certain aspects of running
the shows kick lately and anyway, so since I mentioned
it Who Not How. I think I've mentioned it like
(35:48):
every week for the last like three weeks on like
every show I've been on. So but it's it's a
terrific book if you're trying to build a team, if
you're trying to run a business, if you're trying to
figure out how to get more stuff done, if you're
kind of a mission driven or outcome driven sort of setup,
anything like that. I highly, highly, highly recommend it if
(36:10):
you run any kind of organization. It basically just talks
about how to build an organization, get people in, build
relationships to get done what you want to get done
by empowering people to get to do the things that
they're really good at. And that way they're fulfilled doing
what they're good at, and you're fulfilled doing what you're
good at. So yeah, that's a pick. I really really
(36:34):
dig these kind of growth books. Okay, And so I'm
actually rereading Who Not How right now. I just I
have the hardback version. It's sitting on my desk, and
so when I get up in the morning and I'm
going to pick another book. It's a resource. I don't
follow their prescription really religiously, but I do kind of
(36:57):
follow what they tell me to in the sense that
I up in the morning and I follow up routine.
The book's called The Miracle Morning by Hal L. Rod,
and yeah, he kind of tells you to do a
specific set other things in the morning. I do my
own things, but the idea is pretty good. So definitely
check that book out as well. The books that I've
been listening to while I'm like either driving to swim
(37:21):
the swim team or going on my runs or whatever.
The latest ones that I've listened to are psycho Cybernetics
by doctor Maxwell Maltz, which is terrific, by the way.
It just talks about how you become the person that
is or has what you want to have or be.
And then the other book is As a Man Thinketh
(37:44):
and that's kind of more conceptual version of psycho Cybernetics, honestly,
but it just talks about how your mindset determines what
you get. And those are both terrific, terrific books, and
so I'm going to pick both of those. I know
I am picking a lot of stuff, but honestly, I've
just been thinking very very deeply about a lot of
these things, and I really feel like, this is stuff
(38:07):
that we don't talk a lot about as developers. You know,
we kind of talk about it either as a job,
or we talk about how we contribute to the community,
or we talk about what's going on in the community,
and we fail to talk about what we really want
from our careers, from our contributions, from our open source,
where we want to go and then how we get there.
(38:30):
And I really want to help people get there, So
these books will all help you get there. If you're
looking for more of that, the dev Influencers podcast is
where I'm talking more about that. I'm also trying to
put out bonus episodes where I talk more about that,
but I have other people putting demands on that sometimes,
so dev influencers dot Com slash podcast where you get
(38:52):
the rest of that. And yeah, that's pretty much my picks.
I'll save the rest of it for next week. I've
got a ton of stuff that I just want to share,
but I don't want to take all your time just
saying oh, and this and that and anyway, Well that
was fun, guys, Yeah, totally
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
CrimeLess: Hillbilly Heist
It’s 1996 in rural North Carolina, and an oddball crew makes history when they pull off America’s third largest cash heist. But it’s all downhill from there. Join host Johnny Knoxville as he unspools a wild and woolly tale about a group of regular ‘ol folks who risked it all for a chance at a better life. CrimeLess: Hillbilly Heist answers the question: what would you do with 17.3 million dollars? The answer includes diamond rings, mansions, velvet Elvis paintings, plus a run for the border, murder-for-hire-plots, and FBI busts.