020 — Keeping Momentum

A quick, long one about trying to maintain momentum as challenges and inconsistencies present themselves.

To follow along, you can find me at ryanhefner.com, follow me on Twitter @ryanhefner, and keep up with the show on allplay.fm and @allplayfm.

Help yourself, while supporting the show, by trying some of the services that I use, and highly recommend:
Transistor FM
Fathom Analytics

Episode Transcript

Ryan (00:00):
Hey. How's it going? This is Ryan Hefner, and this is the
All Play podcast. So this weekwell, this week, this episode.
I've I say this most weeks.
I think I'm gonna do multipleepisodes, but really, the
reality is, probably just gonnado one. I definitely could
probably do more, but, you know,it's Wednesday at 4:30 AM. I

(00:30):
just got done taking the dogoutside, and then I was laying
in bed and realized that I got alot of stuff to do. You know?
And my head wasn't letting me goback to sleep, so I was like,
well, might as well just wake upand start getting at it.
And while I was doing that, Iwas just thinking about all the
stuff. You know, as the as themind does, it was racing at

(00:53):
night, and I was thinking aboutall the stuff I had to get done.
So with that, keeping momentumis really tough. You know? I
mean, I guess I can only reallyspeak from my own perspective,
but being a software developer,your job is, you know, depend I

(01:15):
guess, depending on whatprojects you're on or whatever,
your job is pretty much likebanging your head against a wall
until you get their eurekamoment, you know, and solve,
like, whatever the the technicalissue is or or, you know, making
the proper long term decisionbased off of, you know, maybe

(01:42):
limited knowledge or or rathershort term knowledge and hoping
that that plays out later ondown the road.
And then if not, having to, youknow, figure out ways to course
correct, when it doesn't. So,yeah, keeping momentum is just
really tough. I mean, the someof the client project stuff that
I'm working on I mean, most ofmy stuff is going okay. There's

(02:05):
just inherently, you just runinto these challenges where
you're just like, man, how am Igonna figure this out? You know,
how am I gonna take, like, theserandomly named files and then
try to cross reference them withexisting things and somehow
magically pair them up throughsome script and upload those to

(02:26):
some CDN provider and swap out,you know, those, those files
being, like, relationallymapped, you know, to another
entity or something like that.
That's just a for instance,which is one of the things
that's racing through my headright now. On the personal

(02:48):
front, you know, I I've, youknow, I've stated before. I've
been kinda going back and forthon stacks. And, with that, I've
been trying to, like, evaluatedifferent, you know, like,
hosting providers because I'mjust not a DevOps person. I've
never done it.

(03:09):
I mean, I've done it enough justto to to get stuff live, but, my
knowledge isn't really deepthere, and nor do I really want
my knowledge to be deep there. Ijust wanna pay someone to do it,
but at the same time, I don'thave infinite cash to just fork
over to services where, themoney invested isn't really

(03:33):
necessarily being, like,reciprocated, or I'm not
basically making money off ofthe money being invested. So I
don't know. I mean, I've beenthere's a a period like, I think
Vercel is actually a kind of anamazing, product because I think
they do mask some of thecomplexities of spinning out
stuff. I was just I was tryingto do a little test on render.

(03:57):
I'm basically, like, dualpublishing to both render and
Vercel. Obviously, Vercel has noissues, spinning up like a a
next step and doing the doingthe build process. But for
whatever reason on render, thatsame app, obviously, I'm just on

(04:18):
one of their free little dynos.So I I'm I'm sure I could
probably fix this by just payingthe $7 or $14 or whatever it's
gonna be to get it to work, butit's I maxed out the memory just
building the app, you know, the5 12 megabytes of RAM. It's
like, okay.

(04:39):
Well, that doesn't give me muchconfidence in the service. The
fact that I can't just, like,generate a static, website via
their build script. And thenalso running into similar kind
of, like, issues even justgetting a nest, like, web

(05:04):
service a you know, basically,an express app, you know, but
it's it's using Nest JS, for thethat's the framework. And then
just running into things where,you know, it's failing to build
because I don't I have to pay torun the, you know, the the post

(05:30):
build predeploy hook, command sothat way I can actually, like,
spin up the Prisma instance onthe server and stuff. And great.
Like I said, I'm like, I'm I'mnot unwilling to pay, but at the
same time, I guess I I lose alittle faith in a service when I

(05:51):
can't even just, like, spin upthe static Next. Js, app. And
it's like, well, do I reallywanna pay the $7 to see if this
is gonna be there? Because longterm, is this is this what I'm
gonna be dealing with if I kindaget this kinda, like, bristled
first, experience? And what'sinteresting is, like, you know,

(06:16):
I think Vercel does a great, isdefinitely masking this, but I'm
sure they also are throwingjust, like, more RAM so that way
people don't kind of run intothis kinda, early turn offs, of
the of the service.
And I think, you know,obviously, Heroku and, Netlify,

(06:39):
and I haven't done much stuffwith any of the cloud foot
Cloudflare, like, workers orpages and stuff like that, but,
I'm guessing those are probablya little bit smoother
experiences. And so, yeah, it'sjust like all this friction, and
how do you keep momentum withthe friction that just presents

(07:00):
itself? And that's where mythat's where my head's at right
now. But, so it's, yeah, it's436 now. I'm gonna try to knock
out, this client thing that Ijust need to get done because we
we need to get this and this isthis is not even the the file

(07:22):
mapping thing that I was justreferring to.
I need to get this other clientthing done so that way I can get
back to doing that clientmapping. So then I can actually
figure out what's going on withthe deploys to render, and
whether I even wanna go thatroute or not or, or something
else. But, I guess I and I guessthe only reason why I'm really

(07:45):
thinking render is because Ilike the idea of their managed
postgres service potentially.But I don't know. Maybe I just
go SQLite, and I do thereplicash thing or something or
have some other way of beingable to manage, you know, SQLite
at scale.

(08:06):
But at the same time, I just Ilistened to, the recent episode
of database school with, AaronFrancis and taking listening to
them, or him discuss, Postgreswith, I guess it's called Xata
or xata, the company that'sactually gonna be sponsoring his
Postgres course that he'sworking on. I mean and that

(08:29):
stuff sounds super cool. Youknow? Like, the the idea of, you
know, using their extensions,pgstream, and I'm blanking on
the other one, but it allsounded really great. And it's
like, okay.
Well, you know, maybe Ireevaluate the provider that I'm

(08:55):
thinking about using fordatabase stuff. And I guess also
my I guess, the wrestling thereis I've definitely you know,
I've I've I've either you know,I've done, like, the $5

(09:15):
DigitalOcean, spin up a droplet,have, you know, the data like,
whether it's MySQL or Postgres,you know, all just bundled in
there. And it works fine untilyou wanna start kind of, like,
breaking up resources and makingsure that one service, if it's

(09:38):
running a little too hot,doesn't take down the other. And
that's kind of where myreservations lie with just going
with the $5 potato serverapproach. I feel like I've been
kind of burnt by that a littletoo much in the past.
And, also, I mean, the tworeferences that I'm thinking
about, one was, a service that Iwas that I have, like, kind of

(10:04):
just running in the backgroundcalled the optimized toolset,
which runs lighthouse test inthe background on the server.
And you could basically, youknow, throw it in a bunch of
URLs to different projects thatyou're working on and then
monitor the lighthouse scores.And And that was kind of, you
know, something that spun outfrom working on a client project

(10:26):
for an agency. I was just seeingwildly inconsistent, results
when I was trying to do sometests for a specific client that
we were working on. I figured Iprobably shouldn't be running
these tests here locally andthen also just, like, dealing
with these, like, JSON files andthen trying to cross compare

(10:46):
different runs from the previousone to the next and and managing
all these files locally.
Let's just throw it up on theserver. And so that's, you know,
that's definitely a resourcewhere because, you know,
Lighthouse isn't a non it's notjust a database transaction.
It's actually, you know, runningsome stuff that's doing some

(11:10):
heavier lifting. And granted, itwas on a queue and everything
else, but, you know, that wasall on the same server. So once
those lighthouse tests startrunning, everything else starts
to drop down, and that's justnot a great experience.
I think I tried to do a, like,connect like a worker, instance

(11:31):
that I would use to then throwoff all that queue stuff over
there. But I think I was runninginto some weird connectivity
issues either withinDigitalOcean or how I was trying
to set that up in Forge. Not a100% sure. So, you know, those
are the kind of things that haveare in the back of my mind when
I'm trying to make thesedecisions on how I break this

(11:53):
stuff up, because I just don'twant to be frustrated with that
kind of stuff. And I wanna focuson the real things of building.
But you know what that means?That's just the reality. They
these kind of, like,frustrations or, these hurdles

(12:15):
that you gotta jump over andfigure out. This is what makes
or breaks, you know, someonefrom actually breaking out and
having something that they can,get out there and sell and is
useful. And, also maybe kindagoing back to, something that
Justin Jackson had, I think,threw out there a couple years

(12:38):
ago.
But building stuff when you havesmall kids is just tricky. I
mean, the one thing I've beentrying to do is just really work
on my reps and just make surethat I am consistently trying to
do something in a in a forwardmoving fashion. But trying to do
that with kind of like aninconsistent, lifestyle and,

(13:05):
inconsistent, opportunities todo different stuff. Like, you
know, every every single day iskinda slightly different.
Thankfully, this last last weekor so has been kind of
consistent ish, but my kids aregoing on fall break.
We have conferences coming uptomorrow and Friday, and they're
on, like, a, kinda like, a,like, a mini fall break. And

(13:26):
that's just throwing in somedifferent stuff that I gotta,
kinda, deal with. We're gonna doa little trip, which is gonna be
fun, but it's just, you know,just another thing that's gonna
kind of, like, make it hard toget in those consistent reps. So
not sure where the I mean,basically, just trying to keep
momentum, trying to keep, like,a positive outlook on this

(13:49):
stuff, trying to make the bestdecision that I can within the
the period of time that I'm ableto make the decision. And
hopefully, those decisions,don't cause too much friction
in, in for my future self.
And I guess I'm just gonna tryto keep it there. I'm I'm over
on this one big time, but, justgot a lot of thoughts here at

(14:11):
4:43 AM in the morning on aWednesday. And gonna try to
knock out some client stuff so Ican get to some personal stuff
and try to build some margin inmy life. So with that, have a
great one, and I'll talk to youlater.