I just got my first kubernetes cluster configured and damn it get good. Like Bette then any code I’ve written in a long time. Am I...... am I turning to devops..?

The internet made going from sql to mongo so easy. We all already had sq concepts embedded so a Google search for “how to run explain in mongo” turns up hundreds of results. The concepts are there, it’s just translating them to a similar tool

One important lesson I've re-learned too many times: If your splicing, your approach is wrong.

I think this might be my managers first MVP hes worked on. We're 2 days away from go live and hes still stressing for us to push large changes up to UAT (staging). Core calculations are being changed right now and being stressed to deploy.

At the same time though hes learning that these things almost ALWAYS have some type of cascading affect so hes adding "but lets not break anything prior to going live". My response is pretty clear: "Shits going to break or experience adverse effects. Its inevitable". I can hear it in his voice. Hes struggling to learn the weights and balances between these decisions. His green fielding cherries popped and hes seeing what real product development is about.

I am leaving the .Net ecosystem.

I got into with one intention. To be an easy target for large companies and easily marketable employee. .Net is everywhere, I means absolutely everywhere. It is a great solution for large infrastructure and that is the very reason I despise it. It's big and its clunky despite all modern attempts with core to reduce setup.

More importantly, Im learning more about what kind of work I thrive in and want to perform in my career. I dont want to build large servers that uphold entire Saas platforms. I want to experiment with light weight solutions that are easily perceived and used in conjunction with other solutions. I dont want to be the .NET/microsoft stack guy, I want to be the solutions guy. These .Net roles rarely involve anything outside of the MS realm of things and its not a good look.

Im finding my niche in what we call today corporate incubators. Taking solutions that we can roll with quickly like django/laravel/symfony/Express and building out rapid prototypes that can get the job done. I guess its considered the more experimental side of development. All I know is anything beats the kind of slave labor required to just keep an oversized product running.

I know most of us help maintain and im not undermining that kind of work but FUCK its definitely not for me.

Is running a jenkins instance in kubernetes a paradox ?

I want someone to tell me I'm wrong but In my experience, Im finding all deeply experimental work to be inherently imperative in nature.

When im try to tackle a problem that I dont yet know if is possible, I don't first think through how the methods will be broken up or where composition is needed. I dont think through all the different classes that are needed to represent the domains or what abstractions could help. Its usually just alot of if/elses, a ton of variables holding on to everything in context of the train of thought I need to keep together.

Maybe its just that my mind still works in an imperative way and that paradigms have to intentionally be implemented, thus being an afterthought. Personally I think its better to abstract last anyway once you know what it is your working with. Why build something out in the most complex way just to find out that your architecture doesn't represent the domain layer in the way you had first perceived it ?

Well thats my thought for the day as I begin building out this CLI framework scaffolding tool.

I use to obsess over client work and being proficient with all the diff js libraries. Now I just want to focus on the data. Server work has this sort of structure to it that is satisfying compared to heaps of react digging day by day. Unfortunately most of the work out there today is SPA work. Fat clients are whats in. Even when hired on as a server specialist or dev ops specialist, we all somehow get sucked into front end stuff.Its really the biggest seperator Id say in this generations web technology, SPAs, fluid interfaces, NEXT/NUXT, all that flutter

I refuse to work anywhere where I am forced into becoming a professional in navigating their mess and understanding their business logic through legacy systems.

If you treat and pay your people shitty and because of that they’ve left their years of mounted shit behind, that doesn’t mean everyone else wants to pick it up and there’s not a dollar sign you can put on that.

We thrive from making things happen, not plucking little fuck ups from your rusty cogs

I use to want to understand how to build a compiler more in depth and eventually build a little language of my own for something niched. Today I realize languages are a dime a dozen and that I’m better as an aid to help businesses get where they need through tooling I don’t have to understand the source code of to use. I’m working towards becoming a proficient devops specialist so that would be my wish cause the learning around it all is frustrating. Lot more theoretical concepts required then there was to become a decent programmer. I learn things in reverse order most the time so I’m tackling dev ops subjects and learning networking that is implied knowledge as I go

Is anybody like actually been capable of implementing atomic design successfully? The whole concept sounds fantastic and too good to be true. I think the analogy well overused is creating Legos to later stack as you please. I found this to be somewhat reachable if you know every single business requirement upfront but we all know that is never the case. These components that had such simple intentions soon need this and that and functionality here and the ability to provide this back to a parent and this other component know when you’re molecule component does this. Soon enough you have a bunch of dispatches and a molecule that was originally supposed to be a clean component. Seriously where do I learn how to manage the scope creep of what is meant to be a clean atom or molecule component?

Oversized projects always make me want to just give up development entirely. They always leave this defeating feeling even though its not your/my fault. It snot our fault that this 2500 line react component (of 200+ other 1500+ line components) is impossible to reuse. its not your fault that the server is so overly abstracted that no one can makes heads or tails of it.

If theres no other reason for anyone to like microservices both client and server side, having a reasonably sized project to work on in peace is that reason.

You guys ever let your colleague carry the heavy logic some days and you just sit on the slack/zoom call replying with “yeah” and “makes sense” all day while they do the logic lifting.

Keep this one on your radars. Sounds promising. React in Deno

https://link.medium.com/WU9h6uDohgb

Learning kubernetes is like pulling teeth. Maybe the dev ops path is just difficult for me but I’m finding it all really difficult

Dont worry what about what comes from naysayers. The Javascript/Typescript route is a great choice. Ive been trying to get away from node for years and back into a more sTeAdY side of things with C# or PHP but nope, as soon as anyone gets whiff that your generally great with JS, theres an offer on the table. The grass is green over here. It might be a bunch of 25 yr old hipsters eating it but its green.

BBQ Traveling pit master. I’m trying to do it on the side right now but got a large family and will have to set the dream aside till the kids are a bit older. But I love smoking meat and crushing beers all day and then watch people indulge on my creation come evening. I love making fires and tending to them. After working with computers in my little fuckin chair 50 hours a week, I really like the caveman feel of cooking with fire all day

Me and my one other colleagues that are on this project are literally pushing each sprint deadline hour for hour. We’re working nights and weekends putting in combined 110 hours a week on sprints that were estimated at 70 hours a week.

And yeah we estimated them but the facts given look like X plus Y plus B minus Z equals end goal and then when we get into the sprint our manager calls as in for three hours to elaborate on just how complex the problem actually is.

So here we are about one week behind on the predetermined schedule which I feel alone is a gigantic feat to only be one week. It’s kind of funny we look at every sprint in advance in the description given and think OK this will be the sprint that we knock out sooner than expected and get caught up on time but it never fails that we go into the sprint and get absolutely skull fuck by the hidden details and requirements.

I’ve worked for a lot of different places with different styles. The agencies tend to be similar to this where they push as much into a short time frame as they can and write the requirements up so that the managers are putting in the least amount of thought and are easily able to convince us of the time frame. Then I’ve worked for big SAAS companies that make habits of underestimating I just want to keep their developers happy at all costs.

But coming on at the large corporation that I’m at now I had not expected things to be so hectic and demanding. Luckily I love self deprivation and hard requirements. I’ve worked on enough boring projects with boring companies to realize that the rapid prototyping feel of things is my go to

Asked my manager at apple what my odds of going from contract to perm looked like. Pretty much said the budget wasn’t originally there and I’d have to compete with the other few contractors and move to Austin and that’s only if a spot even becomes available. I’m pretty beat up over it. I love this position. It’s perfect for me long term. I need this.

Anyone else here work in business intelligence ? I need someone to cry with

You guys ever had someone who obsesses over making sql do everything ? As if they would rather treat mysql as a the programmatic language over actually letting the server do the logic. Im not saying I cant be witty with my queries, especially if it saves resources and time, but this guy spent a week on this over the top query. Was really proud of this big steaming pile of "case: case: case:". It ends up having issues. Doesnt know why some of this 75 mysql variables arent being set and everything's falling apart. Ends up falling back to server implementation of the logic like it damn well should've been done.

MYsql wasnt made for this, or I at least hope it wasnt.

Is a slight loss in performance or even maybe more than slightly, maybe a second response time difference worth having more sustainable software?

If the 1200 line query that no one wants to nor knows how to touch takes half the time than having a proper set of classes that represent how the domain models interact with each other, is it justified ?

I'll always argue for adequate architecture over faster, yet unmaintainable. Not saying you cant have both, but theres gotta be fine line between the two.

For the angular devs. Should I turn this to an NPM package? It can be a little tedious to set up. I think most would rather have a plug and play package.

https://link.medium.com/F75TgMMSrfb

Anyone ever have to use an npm DI/IoC library? This is such a pain in the ass. Even teh most renown libraries out there still only give you the bare essentials. If you want something flexible and easily understood, it's all on you. Why the hell couldnt the Angular team just release their DI frame work for us all to use ?

8 years doing this and im just using
ON DUPLICATE KEY UPDATE
clause for the first time. Unbelievable

Very random, I know, but have anyone of you ever looked at the source code for styled components ?
It's the most convoluted, cryptic, pile of WTF Ive ever laid eyes on. I cant imagine the devastation of ramping someone up on that.

Pro-tip: dont be lazy, review the PR one commit at a time in chronological order. Dont scroll the whole features "file changes" thinking you'll understand the changes implemented. Your damn job is to understand what they're doing as well as what your doing. Take the time.

If a company makes a habit or policy of shutting doors on me as an innovationist, I leave.

I don't overthink it because I know that the general pursuit innovation and a placement of pursued knowledge will rarely backfire. I might be out of work sometimes, I might not find that seniorish status at many companies that are reserved for their 8 year veterans and I could care less. But the long gain is a vast amount of experience across multiple industries. You cant get that without constantly being challenged!

Consider me the guy that comes in, cleans up, guides new arch and then leaves once the dust settles because It's better then siting on a fat salary at a place for 10 years, doing little to nothing and then trying to justify to the next place how 10 years as a bug tech makes me qualified.

I'm beyond the want to justify my qualifications! It's all contextual. No I haven't worked with google cloud but have worked with AWS at 4 different companies. No im not proficient with ruby but Ive used python on a few servers and automation routines over the last few years. No im not super proficient with elixer but I've dabbled with javascript's take on the functional paradigm enough to have gained familiarity and a conceptual basis.

If you can see what I have accomplished, the risks taken to pursue them and still think that "well we'd really prefer to have someone well versed in haskell" or someone thats proven to "stick around" then I wish you the best of luck, but innovation and "sticking around" rarely come as one.

Juniors, get a few years under your belt and then begin finding your challenges. Honestly to really get your feet wet in this market I would suggest taking year long contracts or agency positions. Place yourself in positions that are constantly fluctuating and place you close to the stakeholders. These jobs will rocket you to the level of expertise needed in half the time as it would take sitting around an over cramped corporation, waiting for legals validation of an open source package needed or some <insert typical corporate time sucking bullshit here>

All day I've battled with CORS just to find that an options argument for an internal proxy image were using was removed cause "he didnt think it was being used".

My whole sprints in the shitter over this

It's been a long time sense I've been given so much to implement or have been able to offer architectural consultation. It's really refreshing and im finally remembering what I use to love so much about development.

"Surely he's at a small start up or somewhere in need of leadership right?" Nope, im 2 months in at apple on one of their internal systems teams. Its damn impressive how such a big company implements small teams to make big things happen. Im talking like 1-2 people per app.

This is the way!