I worked at an insurance company as an engineer in our data science team. As part of exploring how to make the company more money we found that they should just fire everyone, shut down the insurance business and take the inconceivable cash reserves they were burning through and dump it into the S&P500 or a hedge fund.
This is me too lol. We aren't a public company though so we have a "mission". But the real post-tax returns would be 25x what we actually make on our capital. We could just pay someone else to do what we exist to do for people :(
That company exists and it's called berkshire hathaway. Seriously, that company probably could have bought a successfully run smaller insurance company and move all the clients over.
"I asked for some compute to do the data science" hit hard. Spent a full year at a company waiting for compute. Not even a lot. We're talking a lot cheaper than my salary was. They gave me a new laptop. They hired new team members for me. And I sat there in front of my new laptop with my new team members waiting for the compute. The compute never came.
that `tee` problem literally happened at my workplace a couple weeks ago and it took me 20 seconds to fix it. Turns out we've been shipping untested builds for months
@@LudicityHackernews I haven't seen that bad, but it's what happens when everyone is told to "use the existing tools" so builds what they need on top of the previous crappy layer.
I keep a script in CI/CD that enforces that all bash scripts start with ``` #!/bin/bash set -euo pipefail error_handler() { echo "Error on line ${BASH_LINENO[0]}: ${BASH_COMMAND}" } trap 'error_handler' ERR ```
Big companies make me laugh. My current employer (major US pharmacy brand) accidentally disabled my credentials September 30th and they're still trying to get em re-enabled (now November 5th)...I'm paid to do literally nothing. Maybe I'm just unassigned
i feel you. working for a uk company, my developer buddy being told to turn up but not do any work because the project could not get the funding even though HR was paying the salary. he said well you are paying me anyway so why dont i just get on with it while you sort it out. he was told he would be fired if he did so.
@@aerocodes nah, they realize and have me call the IT "no support" line daily.... But they can't figure out the difference between Mac & PC and just tell me to reset my password lol
Ever since i joined the project in this multi-billion dollar company, I regularly find AWS misconfigurations that allow me to get admin access to the whole AWS account and I try to save the company from going bankrupt over high schoolers that could hack it... it would be easier if they actually cared and let me do my job instead of discussing whether we should fix it for 3 months because there are other services that actually depend on our vulnerabilities...
I once worked in one company where delivery boxes were constantly being left in corridors and one manager actually said "We should start a PROJECT to deal with this problem". 🤣 After that, if a toilet wasn't flushed or someone's shoelace was untied "we should start a project" became the running gag. They had projects for everything but there were more meetings than things actually getting done.
I see a bright future, like 3 developers and on top 50 managers creating projects.... and then they will schedule meetings to block the 3 devs to get any shit done.
Taking boxes out is no simple task. You need to research the issue to get basic understanding. Then you need to reach out any relevant people which involves cross team - maybe someone needs those boxes? You need to learn where containers are located which will involve booking a meeting with office manager who's been working from home and is visiting his parents so will only be in office back in two weeks. You need to check up with companies policy to meet the decarbonization program to see if there is any ways the boxes can be recycled or not. And finally once all the prep was done you still need to get someone to do the task and you're short staffed so until you get a new hire you cannot add any new tasks on your team's calendar.
My guess: Because the company was still on aws free tier and they thought @dread was in cohoots with this other cloud so they booted him. Afterwards, they got a 45k bill just on RDS
I just think @dread went against the "senior cloud architects" better judgement. A lot of companies have entitled cloud engineers and architects with no experience and a huge ego
@@cmelgarejo makes no sense. First, he said he came up with a plan, not that he did something. And if you save 500k, your spending more than that. And I don’t think aws give you 500k+ for free
The project I work on is a fairly insignificant app, no huge sales, and nothing technically difficult at all. Very straightforward iOS app talking to a DB and showing some content, that's it. We have about 30-40 engineers working on this app, which is complete insanity -- three would be enough, and all that engineers actually do is writing documentation that will not be kept up to date, discussing how we can change our architecture to improve productivity (lol), or refactor some piece of code for the nth time because there's nothing else to do. Yet somehow, this app has terrible performance, it's littered with bugs, and the development hardly moves forward. I feel this dude's pain.
Guess that's what happens when you just play-pretend being a business by uncritically imitating your incestuous perception of what a business is. The best and most efficiently made projects always come from people who genuinely want to create the product and who do whatever they need to in order to make it work. Not this ritualistic child's play.
15:50 we're working on the PowerBI problem. Most of what it's used for can be skirted by hosting either custom or open source alternatives more cheaply on the same platform in the same networks.
@@marceelino Might be nice to use. I'd just prefer not to pay for it, which is apparently a popular sentiment based on the jobs floating around recently.
Problem is that if you use open source or custom alternatives, the company is stuck relying on internal manpower that may or may not know how to maintain it properly, and may or may not end up costing more for the company than just getting the damn SaaS solution. Plus using a industry-wide solution like powerBI facilitates hiring external talent without needing a fuckton of training and onboarding every time C-levels lays off half the company then hires new people. Its like using open source shit like Zabbix instead of proprietary solutions like Lansweeper on the sysadmin sphere. If you go the Zabbix role, you're just moving the costs inside your company and you dont have anyone to sue if shit goes awry.
@@Bramble20322 I can understand the concern there, but there's tradeoffs situationally. I think I'm probably speaking more to smaller projects and startups that assume they need the "industry standard" tools to get a job done and end up platform locking their company and sabotaging their payroll for an overpriced piece of software full of features they don't need.
The original article is awesome, LOL. It's very well-written, and _you can just tell_ that the content is true, because only truth can surpass fiction so convincingly. edit: autocorrect fix
Sounds legit. Big Companies always have this Management overhead, where internal political "feelings" are more important than actually being productive. The only reason why they can survive this, is because they are big and usually have big tax advantages acting internationally. Thats why big companies often then to hire smaller companies for certain aspects. Because those cant afford acting like this and need to get stuff done in order to survive.
This sounds a bit like a discord server I was moderating. Nobody ever did anything to progress things. If I did something to get anything done (naturally people aren't replying to that kind of messages for days where you ask about doing something) suddenly everyone would wake up to ask and criticise it (not that it was a bad move, just that someone had actually done something) and maybe then someone would do the change in the way that they wanted. If I suggested finally doing something instead of doing it myself to avoid the consequences, nothing happened. So basically after a while of trying to make things move, I just stopped trying and muted the server and let it descent into a tiny inside group's chat with no greater purpose for people outside the group (as was originally planned). Naming the wrong people to manage things is a quick way to kill potential. Unfortunately that server had all the people who were against any change and doing any effort in position to decide on things.
I think there are surprisingly many of us who have similar stories, that most of us aren't allowed to talk about. Love hearing them though when they're available, thanks!
Working for one client where everybody is just a manager and it's bad. Everything, even a version bumping of an outdated package or framework version, requires meetings and etas and God knows WTF! All the while they keep talking about low performance, deadlines, etc.
At the workplace from the video, we once had a 3-day card which just involved me changing all our container images to say :python3.9 instead of :latest. It took like five minutes but people would REFUSE to vote for less than three days when we scoped the card.
@@CHR1SZ7 Yes, but then they don't enforce the times being missed, so we just vote 3 days on everything because it makes the meeting end faster. I wish I was joking.
@ThePrimeTime This article was written by the same person that mentioned PowerBi in a previous video. This person also wrote an interesting follow-up article to the first reaction video you did. It'd be interesting to hear your take on his take.
I've been there, when everything was so red-taped, that we eventually had to have a secret chat and a secret way of doing things, like 10x-20x times faster, going under the radar
16:12 Yeah TJ, I second this. Access was cool at the time. Making it easy to add masks and creating a frontend on top. I guess dbase III and IV was similar but text-based.
28:00 onwards is so real!! My company had a 3hr meeting, including all the devs, QA, customer support, HR, the CTO and the CEO, to discuss in great detail 25 QA bugs and if it would be a good investment of resources to address them...
Agile doesn't specify much except to lower the amount of intermediaries common in waterfall and enable realtime communication. Yes most people do scrum, but to be honest the goal of scrum is not efficiency it's inefficiency management. At the end of the day most large teams rely on the efficiency of a handful of high performers to make up the difference and the management strategy is just there to ensure .1 production rate rather than 0 for the rest of the team.
@@jeremiedubuis5058imo, the goal of SCRUM is to improve communications... at the cost of productivity. A very appealing idea to the non-technical employees, I've found.
@@jeremiedubuis5058Management outside of big tech is so incompetent generally that they don't realize 1 dev =/= 1 dev. You can't butts in seats programming. You need competence. Why the hell would I spend $500k on 10 $50k "devs" when I could spend the same $500k on two top 5% deva for $250k each. It's like, they talk all about "80-20 and pareto" but then conveniently forget it's WAY more applicable to people, too. 😊
The first time I saw a database it was at a month long summer job as a teen. My job was to look, in one table the age of a building, then go look on another table the location of its address and then manually color the building in a map based on age. It took me couple of weeks of trial and error and learning to write a script that did the job. (The scripting wasn't just about database but also about map projections and the coloring commands, and I had to learn everything from scratch with over thousand page manual having never seen those things before.)
I, and teams I have been a part of, have been punished for major successes in improving companies’ functionality and bottom lines so many times, I have lost count. Many companies have busted-ass toxic cultures like described here. Not sure, but I have a hunch; I may have interviewed with this company yesterday.
11:56 i’m surprised that pipefail isn’t enabled by default in ci systems. those are the lines of code you *should* blindly copy and put at the top of all your bash scripts even if you don’t understand them. (i mean it’s better to actually learn what they do, but blindly copying them is better than not having them, since they prevent footguns like that bug)
I doubt someone teeing stderr into stdout would know to set pipefail first, when even putting it back into stderr would have been enough already. Or maybe using PIPESTATUS if an non 0 is required by your CI runner (most already mark failure on stderr). Plus pipefail will actually mark a failure when you do want something to fail silently (like a grep piped into smth as an optional operation or something, idk). Well, I'm saying that but it turns out I do have pipefail set in my runners config directly, and prefer explicit || true when needed.
@@Crow-EH yeah, that’s why i expected it to be the default, so users wouldn’t need to be aware of it. the reverse bug caused by pipefail is generally easier to debug, since it’s an unexpected failure instead of an unexpected success. you can always unset it manually if you don’t like it. as to my second point, it’s actually kind of something i’ve done: blindly copying the `set -pipefail` etc that i’ve seen at the top of bash scripts, so it’s possible for people to accidentally solve such issues through blind copypasting of other scripts as templates
Most of my career has been in analytics platforms. All of it has either been pointless busywork or a shameless upsell on a core product that nobody asked for. They're all solutions looking for problems. Instead start with the problem and like this guy says, put a couple of smart people on it instead of hiring scores to build huge complex systems with no idea of the end goal, except that it should use React, Spring and microservices.
Author here. I spoke to a director recently and he posited that it is extremely difficult to approach things this way due to the incentive to "ship projects". The conversation then moved to other things, but my interpretation of that is that as an executive, you could either ship one project that works well by giving a team plenty of time to scope it properly, or you could ship four projects that barely limp over the line that the board wont understand anyway and look four times as good for your next million dollar gig. So... they just do that, since analytics falling over isn't a horrific bug that they'll even notice. Heck, one of our pipelines stopped working for two months and we noticed last Friday. Solution? Turn it back on and don't tell anyone, because no one noticed for two months anyway.
sounds something that I managed to prevent at my job. every single time i was constrained by a specific tech, it was a failure, and every single time i went rogue and used whatever i felt that was needed for the project to actually be productive. and it'd have been even more productive if i'd been allowed to use whatever it fits from the start.
@@LudicityHackernews Jesus fucking christ, at that point, just stop working and lie about everything happening, you're burning mountains of money anyway.
@@hacktor_92 Problem with that approach is that whenever you fuck off to somewhere else, whoever's left has to figure what the fucking fuck you've done every time you went rogue. Not a good approach unless you treat your job like a hobby.
So, in college I was informed of a study on Agile that showed it only worked for teams of experienced programmers, which tells me that realistically "as written" agile doesn't really work well
Jiminy Christmas, Prime. Stop doing this to me. Now I can't write a simple email without opening Neovim to do it. They hand me forms at the DMV, and I ask them, "can I fill this out with Vim motions?". I'm told that the Emacs military junta has my name on a list of undesirables, and that I have a smooth baby bottom.
9:00 It becomes more realistic if you assume that they probably didn't do it on purpose, and didn't spend any effort on making it work _reliably._ Something this intricated and pointless must have grown organically, like a tumor.
Afters years of saying we need more engineers to work on some super ambitious projects and I get assigned three power point engineers. I understand this so much. Like hey we need to do this, we need x amount of resources and about x number of qualified engineers. That means we need to take a few years to build the planning team to build the plans to build org charts and charters to build the budgets to build the teams that will build the systems.
I find this story hilarious how one employee (and maybe a group of others) are actively trying to save the company in secret from the company so the company wouldn't stop them from being rational and saving the company. And the managers are trying to kill the whole secretiveness.
People tend to attach others that succeed for a number of reasons. They get upset that others are getting credit and they aren't. They worry people above them will see this person doing good and wonder what they have contributed. I've had people try and take credit or get in on credit for work I've done. Usually I've made documents prior to hand to ensure that no one else can do that. I made people aware of the impossibility to stealing or leech credit that caused a number of ripples with it.
Sounds like a client i was on. Project got canned due to cloud costs getting out of control, turned out later some “consultants” set up an EMR cluster for a daily pipeline processing 200mb of data with 30 m4.xlarge instances
As a former contractor that got hired by my client for being a half decent software developer compared to the rest from my company, the answer is money. Companies pay for contractors to implement a project quick and cheap. That project then transfers to another contractor team to support / maintain the project in production because gotta keep the company developers free to work on other planned work. This team will not question why / how it was built like this; often saying 'it's the requirement' and scared to change it in anyway. It will be ~2 years before a software developer from the company peaks into the code base and discovers the massive technical debt created by the implementing contractors.
@@josephgonzalez9342 The last company I worked at had a team of contractors make the frontend. They did not write a single word of documentation nor use comments. They didn't even write what component library the front end used. The only comments you found in that code base were written by employees.
@@josephgonzalez9342 Same experience, different industry, its scary how close we are to outsourcing to the level of incompetence. Middle middle middle middle management disagrees
What probably happened was Big firm: oh we've done this hundreds of times. Heres a super vanilla MVP to get you guys started. Stakeholders: great but also....
If one thousand of requests were spread evenly on a 40h work week that would give a request every 2.5 minutes, so that might explain the idle time, if there is indeed a start up cost.
I remember when I was new on a small company where they advertise some rural houses, the houses publish their photos to show on the web like AirBNB, they were storing the photos in like a direct hdd on a hosting where they charged them like 500€ a month, I was like bruh wtf 500€ a month for what, it was 500gb of hdd on a random bareback metal server so I changed it to an s3 bucket and next month came an invoce for like 0,03€, so you can see the incompetence and scammy tactics of some people.
if the data driven and analytical comes then same issue. we had a system that cost 750K per year 1% cpu usage and nobody was willing to press the button.
I saved a fortune 500 company 650 direct man hours annually by fixing an Excel Macro. Lord only knows the real value. This was my first technical task for my second formal software job out of college. By all signs this macro had been in use and expanding for a decade. Some of the administrative staff were being required to come in at 6:00 a.m. to pull a report and run a macro. They would then distribute to department heads for daily processing. When it failed it may not be available for hours or until the next day. What was this amazing change? I instructed the macro to not draw to the screen while it was running. Cut the time for the macro running from 3 hours to 30 minutes. I don't know specifically how much less it crashed. Where how much it was crashing? I just know that that was a major win for the administrators coming in in the morning and the department heads actually getting their report first thing in the morning.
Last year, I was told by a high up IT manager they were replacing some hardware. They were using the total number of profiles created for this hardware vs actually actively used hardware. I had done the most basic SQL, but in a few hours, was able to create a search that found online devices. Ended up saving us 7-10 million after I showed my work to him. Most of this hardware was replaced by software while the profiles were never deleted or updated.
As I literally sit here writing a PowerPoint to show the people in my department that they didn’t have to rewrite core infrastructure had they done these couple things like my team did 😂
A million dollars a year. Nice how about a million dollars a month? I've looked at some crazy amounts of data that was being stored but not used. "What's the business reason for this?" - *crickets*
21:30 If an intern is able to do something, I haven't been able to, I'm just happy about it. It's something I (hopefully) don't need to do anymore. I as the only (almost) full time programmer at my company, so maybe I should feel bad when an intern accomplishes something, I haven't been able to, maybe for weeks.
Me screaming at the screen that he misread the bold 4th bullet point. “AFTER!!! It’s not a f-ing keep alive to stop it from spooking down, it forces them to constantly spool up new servers!!!!”
great companies begin dying when they start creating presentations. When did being creative become "create a powerpoint". Let's talk about what we could do, will do, and might've done... but never do.
6:50 I think what the letter is trying to say that Employees who were not a "fit" were given impossible tasks with the intention of letting them go. But yeah, hard to say with actual clarity.
I wrote a cost accounting system for a company in Access based on the idea of linked spreadsheets. It was a nightmare but worked. Reduced month end accounting to hours versus days and allowed them to reduce their accounting staff. It was a bad idea on multiple fronts.
These vids bring so me much validation to this field. Clever engineers just drown in the politics of business. Going rogue is how I have delivered my best tools. My motto is it is better to ask for forgiveness than permission. Design by committee is a sure fire way to stall a project and kill it dead. I believe in a balance. Have a week to crush bugs and grind on the main product. Then the next week, explore/tinker/play with ideas and try new things. I find the youngest engineers often have the best ideas due to fresh eyes and no legacy cruft in their heads.
I’m with TJ on Access. It was my first exposure to relational databases and my mind was blown. At this point I work with SQL and would never touch Access, but it was so cool to pre-programmer me.
I used to maintain an old Java system. We had several random print() functions, yes empty print functions, throughout this monolithic 8,000 line main file. When I deleted them, weird things started happening, timings were off, race conditions were triggered, etc. So if I see some weird stuff on a codebase, let alone one that's older than me, I'm leaving it there because I'm sure there's a method to that madness. No methodless madness can survive for that much time.
I saved SPACE SYSTEMS LORAL $500,000 in 30 minutes in 2002... They gave me a $500 bonus that year. I saved NASA tens of millions of dollars. Congress gave me a certificate of appreciation. The certificate is PROBABLY worth more than the $500.
I'm using tee for exactly this reason and without any issue :-/. Though, I'm using also using a FIFO to not hide the status for the build command (because piping into tee would never fail).
I hope it's not the one I'm thinking of but I literally know of a company that had and idle timer set in by a dev when they were a little startup and had limited resources. Turns out nobody removed it until recently.
I worked with a big firm for a media project. They told u they could match roughly 25% of impressions to sales so we needed to divide by .25 for all our reported sales. we had such low numbers that we almost lost vendors. When asked to look at it we found that of the 2 main channels we advertised through one of them was roughly a 25% match rate, but the other was only 5%. After the change to 5% we roughly 1.8x our total sales. Vendors that only did campaigns through the channel with the 5% match rate saw 5x increases in their sales. I am not sure we will be working with the consulting company again.
This vid makes me want to go back to Borland Paradox. Dude stares at code... every array[0] is empty, every loop starts w 1. A thousand security checkpoints in the datapipe using way too many frameworks. When you interview, ask for the tech stack. You have to believe in it, not just work the job. Don't just get interviewed, hire your boss.
I'm on team TJ here. The entire reason I got into software development was because of Access. It was the first time I could actually see what the data was like back there behind the screen. It can do some pretty impressive things. I am going to have a hard time not calling it PowerBy now though ha! (I'm currently a Business Intelligence Engineer)
Same in the company I left as well, I also don’t really see the product being used after years of development. The directors all were fired and moved lol
Yeah this much more common than people think at larger organizations. MBAs come in, "lower expenses" (salaries) then have to "manage" the worse human capital to get an inch of productivity. Processes that ensure everyone is at least 10% productive while kneecapping and ensuring nobody is more than 20% productive without breaking rules isnt a recipe for success in the modern technical world.
i had an idea for increasing productivity from "do the thing takes 3 months" to "do the thing takes 2 weeks" but it would take 2 months to implement tops. it was 8 years ago i brought up the idea and its still not implemented, and we did the thing 27 times since then, so a poor soul (not me) spent 75% of the last 8 years doing the thing, instead of doing it in one
You know who still uses COVID as an excuse? The US Embassy in London. It's _still_ fucking impossible to get appointments without getting up at the crack of dawn every day in the hopes of magically catching the one morning when they release appointments and snagging one before they go.
This is painful because just about everybody has been there. Maybe not to the scale, but to the point. It's like it's either "no planning" or "no doing" with nothing in between. In a way, it's like those products where you build layer upon layer of "functionality" by adding MORE technical debt at each layer. Then, at some point, some team gets a greenfield initiative to redo the product from the ground up. Year(s) later it comes out and it's amazing, 20x performance. Then you look at the team and they have that sad look. So you dig into the docs and find out the 20x comes from just substituting a base layer that everyone knew was piss poor. It could be 100x, but it's 20x because they had to build THE EXACT same shit that was built before, because the people calling the shots were still calling the shots and they had to have what they wanted to have.Then you grep the sad looks...
I am the author, and my guy, this hurts me to read because it's so true. The thing that upset me about this platform is that it actually was a green field - so it was a total own goal - but even if they did it correctly, the enterprise databases they were pulling information from are just worthless because of what you described above.
@@LudicityHackernews Imagine a table that is basically a key-value store. Imagine it was supposed to hold configurations. Then someone said html templates are configurations. Then someone said, why not put json in there, and have configurations inside configurations. Then... you know where it's going. Sad part if that it's not even the worst part. Hell is real ;) Last thing i "hinted" at a team doing greenfield was "you know they will if they can, so either you put the roadblocks in NOW or they'll shot themselves in the foot sooner rather than later".
I also accidentally saved half a million. Then they complained I was doing 'tax fraud'. Always the envy!
This hits way too close to home. I’ve seen finance shut down engineering initiatives because it would “save too much money and we’d be audited”
@jorderon what the crap, really?!
@@jorderonI got my company audited and sued by federal tax service once. The company won and kept the money.
How can taxes be real if our eyes aren’t real
How can taxes be real if our eyes aren’t real
I worked at an insurance company as an engineer in our data science team. As part of exploring how to make the company more money we found that they should just fire everyone, shut down the insurance business and take the inconceivable cash reserves they were burning through and dump it into the S&P500 or a hedge fund.
This is me too lol. We aren't a public company though so we have a "mission". But the real post-tax returns would be 25x what we actually make on our capital. We could just pay someone else to do what we exist to do for people :(
but the the serfs wont have jobs...
dear lord that's actually mental.
how to save money 101: "do nothing"
10% of the global companys are "dead" companys. If interest rates wouldn't be this low and investors so many, then they would go bankrupt in seconds.
That company exists and it's called berkshire hathaway. Seriously, that company probably could have bought a successfully run smaller insurance company and move all the clients over.
"I asked for some compute to do the data science" hit hard. Spent a full year at a company waiting for compute. Not even a lot. We're talking a lot cheaper than my salary was. They gave me a new laptop. They hired new team members for me. And I sat there in front of my new laptop with my new team members waiting for the compute. The compute never came.
I can see your line manager sitting there and thinking "a new laptop! That makes perfect sense!" and then getting you a chromebook.
Can't complain too hard you were being paid
that `tee` problem literally happened at my workplace a couple weeks ago and it took me 20 seconds to fix it. Turns out we've been shipping untested builds for months
Ah well, I guess each of the things mentioned is fine - people make mistakes - but taken altogether it's a horrendous lasagna of badness.
@@LudicityHackernews I haven't seen that bad, but it's what happens when everyone is told to "use the existing tools" so builds what they need on top of the previous crappy layer.
set -eo pipefail
Oh is that really you from the article? If yes then great article btw @@LudicityHackernews
I keep a script in CI/CD that enforces that all bash scripts start with
```
#!/bin/bash
set -euo pipefail
error_handler() {
echo "Error on line ${BASH_LINENO[0]}: ${BASH_COMMAND}"
}
trap 'error_handler' ERR
```
Big companies make me laugh. My current employer (major US pharmacy brand) accidentally disabled my credentials September 30th and they're still trying to get em re-enabled (now November 5th)...I'm paid to do literally nothing. Maybe I'm just unassigned
i feel you. working for a uk company, my developer buddy being told to turn up but not do any work because the project could not get the funding even though HR was paying the salary. he said well you are paying me anyway so why dont i just get on with it while you sort it out. he was told he would be fired if he did so.
@@mrrolandlawrence that's nutty....like they're paying him so their competitors can't hire him
Mandatory training week!
just sheesh... just... sheesh. Maybe they won't realise
@@aerocodes nah, they realize and have me call the IT "no support" line daily.... But they can't figure out the difference between Mac & PC and just tell me to reset my password lol
Ever since i joined the project in this multi-billion dollar company, I regularly find AWS misconfigurations that allow me to get admin access to the whole AWS account and I try to save the company from going bankrupt over high schoolers that could hack it... it would be easier if they actually cared and let me do my job instead of discussing whether we should fix it for 3 months because there are other services that actually depend on our vulnerabilities...
"because there are other services that actually depend on our vulnerabilities" lmao
Teej saying “what if they’re just removing sleep(10)s” was so funny to me
I once worked in one company where delivery boxes were constantly being left in corridors and one manager actually said "We should start a PROJECT to deal with this problem". 🤣 After that, if a toilet wasn't flushed or someone's shoelace was untied "we should start a project" became the running gag. They had projects for everything but there were more meetings than things actually getting done.
Managers doing those things is a very clear indicator that they don't have enough work to do.
If the only tool you've got is a SCRUM, everything looks like a project
I see a bright future, like 3 developers and on top 50 managers creating projects.... and then they will schedule meetings to block the 3 devs to get any shit done.
Taking boxes out is no simple task. You need to research the issue to get basic understanding. Then you need to reach out any relevant people which involves cross team - maybe someone needs those boxes? You need to learn where containers are located which will involve booking a meeting with office manager who's been working from home and is visiting his parents so will only be in office back in two weeks. You need to check up with companies policy to meet the decarbonization program to see if there is any ways the boxes can be recycled or not. And finally once all the prep was done you still need to get someone to do the task and you're short staffed so until you get a new hire you cannot add any new tasks on your team's calendar.
Boxes are complicated...really complicated.
I came up with a plan that would save my company 500k / year on AWS costs. I got laidoff instead.
What? Why? That’s fucking insane
@@greyshopleskin2315 Must have stepped on the toes of some manager with that one. And the terrible manager prioritized saving face over saving 500k.
My guess: Because the company was still on aws free tier and they thought @dread was in cohoots with this other cloud so they booted him.
Afterwards, they got a 45k bill just on RDS
I just think @dread went against the "senior cloud architects" better judgement.
A lot of companies have entitled cloud engineers and architects with no experience and a huge ego
@@cmelgarejo makes no sense.
First, he said he came up with a plan, not that he did something. And if you save 500k, your spending more than that. And I don’t think aws give you 500k+ for free
Rewrite this for product managers and you have a conference talk you could live of.
I have thought about this, but then I'd be rolling the dice on whether my employer friggin' guillotines me lmao.
The project I work on is a fairly insignificant app, no huge sales, and nothing technically difficult at all. Very straightforward iOS app talking to a DB and showing some content, that's it. We have about 30-40 engineers working on this app, which is complete insanity -- three would be enough, and all that engineers actually do is writing documentation that will not be kept up to date, discussing how we can change our architecture to improve productivity (lol), or refactor some piece of code for the nth time because there's nothing else to do. Yet somehow, this app has terrible performance, it's littered with bugs, and the development hardly moves forward. I feel this dude's pain.
sounds like the Dave and busters app
Sounds like a dream job to moonlight on
Guess that's what happens when you just play-pretend being a business by uncritically imitating your incestuous perception of what a business is. The best and most efficiently made projects always come from people who genuinely want to create the product and who do whatever they need to in order to make it work. Not this ritualistic child's play.
It's terrible BECAUSE it has 40 engineers working on it.
It is good for those at the end of their careers, bad for new hungry ones who actually want to make a difference and improve in their careers
15:50 we're working on the PowerBI problem. Most of what it's used for can be skirted by hosting either custom or open source alternatives more cheaply on the same platform in the same networks.
If you know what you doing, it's a great tool.
@@marceelino Might be nice to use. I'd just prefer not to pay for it, which is apparently a popular sentiment based on the jobs floating around recently.
Problem is that if you use open source or custom alternatives, the company is stuck relying on internal manpower that may or may not know how to maintain it properly, and may or may not end up costing more for the company than just getting the damn SaaS solution.
Plus using a industry-wide solution like powerBI facilitates hiring external talent without needing a fuckton of training and onboarding every time C-levels lays off half the company then hires new people.
Its like using open source shit like Zabbix instead of proprietary solutions like Lansweeper on the sysadmin sphere. If you go the Zabbix role, you're just moving the costs inside your company and you dont have anyone to sue if shit goes awry.
@@Bramble20322 I can understand the concern there, but there's tradeoffs situationally. I think I'm probably speaking more to smaller projects and startups that assume they need the "industry standard" tools to get a job done and end up platform locking their company and sabotaging their payroll for an overpriced piece of software full of features they don't need.
The original article is awesome, LOL.
It's very well-written, and _you can just tell_ that the content is true, because only truth can surpass fiction so convincingly.
edit: autocorrect fix
Strong agree
I've worked at places that had elements of this or all of this. If this isn't real for him it definitely is real somewhere.
@@disguysn I've also seen such organizational absurdity.
The part about deleting prod from CI is sus
Sounds legit.
Big Companies always have this Management overhead, where internal political "feelings" are more important than actually being productive.
The only reason why they can survive this, is because they are big and usually have big tax advantages acting internationally.
Thats why big companies often then to hire smaller companies for certain aspects. Because those cant afford acting like this and need to get stuff done in order to survive.
wait a minute... are you saying that big corporations are as inefficient as goverment...
This sounds a bit like a discord server I was moderating. Nobody ever did anything to progress things. If I did something to get anything done (naturally people aren't replying to that kind of messages for days where you ask about doing something) suddenly everyone would wake up to ask and criticise it (not that it was a bad move, just that someone had actually done something) and maybe then someone would do the change in the way that they wanted. If I suggested finally doing something instead of doing it myself to avoid the consequences, nothing happened.
So basically after a while of trying to make things move, I just stopped trying and muted the server and let it descent into a tiny inside group's chat with no greater purpose for people outside the group (as was originally planned). Naming the wrong people to manage things is a quick way to kill potential. Unfortunately that server had all the people who were against any change and doing any effort in position to decide on things.
@@VictorSamuelMosqueraArtamonov The size of an entity is positively correlated with its similarity to a government across all domains.
I think there are surprisingly many of us who have similar stories, that most of us aren't allowed to talk about. Love hearing them though when they're available, thanks!
Working for one client where everybody is just a manager and it's bad. Everything, even a version bumping of an outdated package or framework version, requires meetings and etas and God knows WTF! All the while they keep talking about low performance, deadlines, etc.
But they are working with agile, fast moving teams, how could it be slow and low performance?
At the workplace from the video, we once had a 3-day card which just involved me changing all our container images to say :python3.9 instead of :latest. It took like five minutes but people would REFUSE to vote for less than three days when we scoped the card.
@@LudicityHackernewsoof they make you estimate on actual time??
@@CHR1SZ7 Yes, but then they don't enforce the times being missed, so we just vote 3 days on everything because it makes the meeting end faster. I wish I was joking.
@ThePrimeTime This article was written by the same person that mentioned PowerBi in a previous video. This person also wrote an interesting follow-up article to the first reaction video you did. It'd be interesting to hear your take on his take.
I've been there, when everything was so red-taped, that we eventually had to have a secret chat and a secret way of doing things, like 10x-20x times faster, going under the radar
these 2 are funnier than any twitch streamer group I have ever watched.
The second one pretty noises for listening.
The guest host was amazing on this, very charismatic. Looking forward to more!
Do you love Microsoft Access or something?
Agreed. I wish I knew where to find more content from teej_dv.
He has a great mustache too!
TJ is a an actor, he can’t code, but he’s nice. MS Access(!?) for the love of SQLing, teej, bruv..shake my head 😢😂
@@kiikohI see what you did
16:12 Yeah TJ, I second this. Access was cool at the time. Making it easy to add masks and creating a frontend on top. I guess dbase III and IV was similar but text-based.
28:00 onwards is so real!! My company had a 3hr meeting, including all the devs, QA, customer support, HR, the CTO and the CEO, to discuss in great detail 25 QA bugs and if it would be a good investment of resources to address them...
"actually trying it" is what agile is.
People hate agile because they don't do agile, they do scrum or safe or waterfall and call it agile.
Nothing says "self organizing" like mindlessly following the rigid structure prescribed by some document.
Agile doesn't specify much except to lower the amount of intermediaries common in waterfall and enable realtime communication. Yes most people do scrum, but to be honest the goal of scrum is not efficiency it's inefficiency management. At the end of the day most large teams rely on the efficiency of a handful of high performers to make up the difference and the management strategy is just there to ensure .1 production rate rather than 0 for the rest of the team.
@@jeremiedubuis5058imo, the goal of SCRUM is to improve communications... at the cost of productivity. A very appealing idea to the non-technical employees, I've found.
@@jeremiedubuis5058Management outside of big tech is so incompetent generally that they don't realize 1 dev =/= 1 dev. You can't butts in seats programming. You need competence. Why the hell would I spend $500k on 10 $50k "devs" when I could spend the same $500k on two top 5% deva for $250k each.
It's like, they talk all about "80-20 and pareto" but then conveniently forget it's WAY more applicable to people, too. 😊
Faux leather bagpipes
The first time I saw a database it was at a month long summer job as a teen. My job was to look, in one table the age of a building, then go look on another table the location of its address and then manually color the building in a map based on age. It took me couple of weeks of trial and error and learning to write a script that did the job. (The scripting wasn't just about database but also about map projections and the coloring commands, and I had to learn everything from scratch with over thousand page manual having never seen those things before.)
I, and teams I have been a part of, have been punished for major successes in improving companies’ functionality and bottom lines so many times, I have lost count. Many companies have busted-ass toxic cultures like described here. Not sure, but I have a hunch; I may have interviewed with this company yesterday.
11:56 i’m surprised that pipefail isn’t enabled by default in ci systems. those are the lines of code you *should* blindly copy and put at the top of all your bash scripts even if you don’t understand them. (i mean it’s better to actually learn what they do, but blindly copying them is better than not having them, since they prevent footguns like that bug)
I doubt someone teeing stderr into stdout would know to set pipefail first, when even putting it back into stderr would have been enough already. Or maybe using PIPESTATUS if an non 0 is required by your CI runner (most already mark failure on stderr).
Plus pipefail will actually mark a failure when you do want something to fail silently (like a grep piped into smth as an optional operation or something, idk).
Well, I'm saying that but it turns out I do have pipefail set in my runners config directly, and prefer explicit || true when needed.
@@Crow-EH yeah, that’s why i expected it to be the default, so users wouldn’t need to be aware of it. the reverse bug caused by pipefail is generally easier to debug, since it’s an unexpected failure instead of an unexpected success. you can always unset it manually if you don’t like it. as to my second point, it’s actually kind of something i’ve done: blindly copying the `set -pipefail` etc that i’ve seen at the top of bash scripts, so it’s possible for people to accidentally solve such issues through blind copypasting of other scripts as templates
Most of my career has been in analytics platforms. All of it has either been pointless busywork or a shameless upsell on a core product that nobody asked for.
They're all solutions looking for problems. Instead start with the problem and like this guy says, put a couple of smart people on it instead of hiring scores to build huge complex systems with no idea of the end goal, except that it should use React, Spring and microservices.
Author here. I spoke to a director recently and he posited that it is extremely difficult to approach things this way due to the incentive to "ship projects". The conversation then moved to other things, but my interpretation of that is that as an executive, you could either ship one project that works well by giving a team plenty of time to scope it properly, or you could ship four projects that barely limp over the line that the board wont understand anyway and look four times as good for your next million dollar gig.
So... they just do that, since analytics falling over isn't a horrific bug that they'll even notice. Heck, one of our pipelines stopped working for two months and we noticed last Friday. Solution? Turn it back on and don't tell anyone, because no one noticed for two months anyway.
That’s the name of the game! 😂 Consultants don’t solve problems, they sell solutions lol.
sounds something that I managed to prevent at my job. every single time i was constrained by a specific tech, it was a failure, and every single time i went rogue and used whatever i felt that was needed for the project to actually be productive. and it'd have been even more productive if i'd been allowed to use whatever it fits from the start.
@@LudicityHackernews Jesus fucking christ, at that point, just stop working and lie about everything happening, you're burning mountains of money anyway.
@@hacktor_92 Problem with that approach is that whenever you fuck off to somewhere else, whoever's left has to figure what the fucking fuck you've done every time you went rogue. Not a good approach unless you treat your job like a hobby.
So, in college I was informed of a study on Agile that showed it only worked for teams of experienced programmers, which tells me that realistically "as written" agile doesn't really work well
it's not even about experience. It's a company culture problem.
Jiminy Christmas, Prime. Stop doing this to me. Now I can't write a simple email without opening Neovim to do it. They hand me forms at the DMV, and I ask them, "can I fill this out with Vim motions?". I'm told that the Emacs military junta has my name on a list of undesirables, and that I have a smooth baby bottom.
9:00 It becomes more realistic if you assume that they probably didn't do it on purpose, and didn't spend any effort on making it work _reliably._
Something this intricated and pointless must have grown organically, like a tumor.
Can confirm that working with it does feel cancerous.
Afters years of saying we need more engineers to work on some super ambitious projects and I get assigned three power point engineers.
I understand this so much. Like hey we need to do this, we need x amount of resources and about x number of qualified engineers. That means we need to take a few years to build the planning team to build the plans to build org charts and charters to build the budgets to build the teams that will build the systems.
I find this story hilarious how one employee (and maybe a group of others) are actively trying to save the company in secret from the company so the company wouldn't stop them from being rational and saving the company. And the managers are trying to kill the whole secretiveness.
People tend to attach others that succeed for a number of reasons. They get upset that others are getting credit and they aren't. They worry people above them will see this person doing good and wonder what they have contributed.
I've had people try and take credit or get in on credit for work I've done. Usually I've made documents prior to hand to ensure that no one else can do that. I made people aware of the impossibility to stealing or leech credit that caused a number of ripples with it.
What were your strategies to ensure that others couldn't take credit for your work?
The main take of the article, said by the Prime is on 25:58
It seems the company i work for is actually pretty damned good, i enjoy reading these horror stories as it makes me feel better
The ability to recognize that your solution is becoming a Rube Goldgerg machine is a top tier skill.
At 14:00 in the chat: _"No better cure for impostor syndrome than reading stuff like this..."_
For real.
Sounds like a client i was on. Project got canned due to cloud costs getting out of control, turned out later some “consultants” set up an EMR cluster for a daily pipeline processing 200mb of data with 30 m4.xlarge instances
Wow, I am speechless. I actually believe this to be true. How can such incompetence exist??
As a former contractor that got hired by my client for being a half decent software developer compared to the rest from my company, the answer is money. Companies pay for contractors to implement a project quick and cheap. That project then transfers to another contractor team to support / maintain the project in production because gotta keep the company developers free to work on other planned work. This team will not question why / how it was built like this; often saying 'it's the requirement' and scared to change it in anyway. It will be ~2 years before a software developer from the company peaks into the code base and discovers the massive technical debt created by the implementing contractors.
@@josephgonzalez9342 The last company I worked at had a team of contractors make the frontend. They did not write a single word of documentation nor use comments. They didn't even write what component library the front end used. The only comments you found in that code base were written by employees.
@@josephgonzalez9342 Same experience, different industry, its scary how close we are to outsourcing to the level of incompetence. Middle middle middle middle management disagrees
@@markm1514 damn i did not consider that other industries would be doing this. This workers as a service trend is weird.
Lolll this is not uncommon.
Thanks for these streams,man. Gave me a good laugh or two.
What probably happened was
Big firm: oh we've done this hundreds of times. Heres a super vanilla MVP to get you guys started.
Stakeholders: great but also....
If one thousand of requests were spread evenly on a 40h work week that would give a request every 2.5 minutes, so that might explain the idle time, if there is indeed a start up cost.
I remember when I was new on a small company where they advertise some rural houses, the houses publish their photos to show on the web like AirBNB, they were storing the photos in like a direct hdd on a hosting where they charged them like 500€ a month, I was like bruh wtf 500€ a month for what, it was 500gb of hdd on a random bareback metal server so I changed it to an s3 bucket and next month came an invoce for like 0,03€, so you can see the incompetence and scammy tactics of some people.
"Access is kind of cool" -> browsing for a pitchfork at amazon, craigslist for a flamethrower
if the data driven and analytical comes then same issue. we had a system that cost 750K per year 1% cpu usage and nobody was willing to press the button.
I work at Small Firm Consulting, where we love the messes that Big Firm Consulting leaves behind which we fix.
I saved a fortune 500 company 650 direct man hours annually by fixing an Excel Macro. Lord only knows the real value.
This was my first technical task for my second formal software job out of college. By all signs this macro had been in use and expanding for a decade.
Some of the administrative staff were being required to come in at 6:00 a.m. to pull a report and run a macro. They would then distribute to department heads for daily processing. When it failed it may not be available for hours or until the next day.
What was this amazing change?
I instructed the macro to not draw to the screen while it was running. Cut the time for the macro running from 3 hours to 30 minutes. I don't know specifically how much less it crashed. Where how much it was crashing? I just know that that was a major win for the administrators coming in in the morning and the department heads actually getting their report first thing in the morning.
Last year, I was told by a high up IT manager they were replacing some hardware. They were using the total number of profiles created for this hardware vs actually actively used hardware. I had done the most basic SQL, but in a few hours, was able to create a search that found online devices. Ended up saving us 7-10 million after I showed my work to him. Most of this hardware was replaced by software while the profiles were never deleted or updated.
As I literally sit here writing a PowerPoint to show the people in my department that they didn’t have to rewrite core infrastructure had they done these couple things like my team did 😂
This sounds like also someone else has generated someone else's big fat bonus.
So they used that half mil to cancel the next set of layoffs or give some engineers bonuses, right? Right?
A million dollars a year. Nice how about a million dollars a month? I've looked at some crazy amounts of data that was being stored but not used. "What's the business reason for this?" - *crickets*
i love teejay's chuckle...it always makes me want to laugh with him LOL.
This was too real and too painful.
21:30 If an intern is able to do something, I haven't been able to, I'm just happy about it. It's something I (hopefully) don't need to do anymore.
I as the only (almost) full time programmer at my company, so maybe I should feel bad when an intern accomplishes something, I haven't been able to, maybe for weeks.
Me screaming at the screen that he misread the bold 4th bullet point. “AFTER!!! It’s not a f-ing keep alive to stop it from spooking down, it forces them to constantly spool up new servers!!!!”
I'm now at a security company
without any real security whatsoever.
Get me out of consultingland
I can't even imagine what a high paced, high red tape environment would look like. Sounds way too contradictory lol
It's basically the army.
We had a saying of "hurry up and wait"
How do you always select text without the leading or trailing char every time?
great companies begin dying when they start creating presentations. When did being creative become "create a powerpoint". Let's talk about what we could do, will do, and might've done... but never do.
6:50 I think what the letter is trying to say that Employees who were not a "fit" were given impossible tasks with the intention of letting them go. But yeah, hard to say with actual clarity.
This video is like the Superbowl. Commentary and commercials interrupted by the actual thing you are interested in.
"WOW, YOU CAN SEE WHAT'S IN THERE" 🤣😂🤣
9:40 couldn't they have used AWS CloudTrail for auditing?
do you have a link to the guy you are doing this show with?
I wrote a cost accounting system for a company in Access based on the idea of linked spreadsheets. It was a nightmare but worked. Reduced month end accounting to hours versus days and allowed them to reduce their accounting staff. It was a bad idea on multiple fronts.
"Access is actually kinda cool"
"What was that, TJ??"
Spending $500k on Agile discovery, only to deny the $50 solution.
I didn't realize they weren't sitting together in the same room until the very end.
The amount of times i declare "in just going to do it" and it's done 15-30m later is great. It makes so many people shut up
These vids bring so me much validation to this field. Clever engineers just drown in the politics of business. Going rogue is how I have delivered my best tools. My motto is it is better to ask for forgiveness than permission. Design by committee is a sure fire way to stall a project and kill it dead.
I believe in a balance. Have a week to crush bugs and grind on the main product. Then the next week, explore/tinker/play with ideas and try new things. I find the youngest engineers often have the best ideas due to fresh eyes and no legacy cruft in their heads.
I’m with TJ on Access. It was my first exposure to relational databases and my mind was blown. At this point I work with SQL and would never touch Access, but it was so cool to pre-programmer me.
I used to maintain an old Java system. We had several random print() functions, yes empty print functions, throughout this monolithic 8,000 line main file. When I deleted them, weird things started happening, timings were off, race conditions were triggered, etc. So if I see some weird stuff on a codebase, let alone one that's older than me, I'm leaving it there because I'm sure there's a method to that madness. No methodless madness can survive for that much time.
I saved SPACE SYSTEMS LORAL $500,000 in 30 minutes in 2002... They gave me a $500 bonus that year.
I saved NASA tens of millions of dollars. Congress gave me a certificate of appreciation.
The certificate is PROBABLY worth more than the $500.
The secret group chat that runs the company? Yeah that's probably a legitimate reimplementation of the cigarette break for remote work.
lol I kid you not decisions by committee is how my company operates for every data transformation we do. It is a joke.
I'd rather shake an 8-ball to be perfectly honest.
Saved my old company 4 million dollars, convinced the senior managment to give me 1 million to do the work McKinsey took 5 for.
I'm using tee for exactly this reason and without any issue :-/. Though, I'm using also using a FIFO to not hide the status for the build command (because piping into tee would never fail).
Or just using `set -eo pipefail`
I hope it's not the one I'm thinking of but I literally know of a company that had and idle timer set in by a dev when they were a little startup and had limited resources. Turns out nobody removed it until recently.
this is amazing and everyone clearly relates
I worked with a big firm for a media project. They told u they could match roughly 25% of impressions to sales so we needed to divide by .25 for all our reported sales. we had such low numbers that we almost lost vendors. When asked to look at it we found that of the 2 main channels we advertised through one of them was roughly a 25% match rate, but the other was only 5%. After the change to 5% we roughly 1.8x our total sales. Vendors that only did campaigns through the channel with the 5% match rate saw 5x increases in their sales. I am not sure we will be working with the consulting company again.
Now you need the opposite of I accidentally lost half a million dollars for my company. Couple of those stories out there.
😂 cosplaying real business and the board thinks the costume is convincing. so legit lol
he said wow you can see whats in there😂
As a PDF engineer I can confirm that I'm losing my mind.
This vid makes me want to go back to Borland Paradox. Dude stares at code... every array[0] is empty, every loop starts w 1. A thousand security checkpoints in the datapipe using way too many frameworks. When you interview, ask for the tech stack. You have to believe in it, not just work the job. Don't just get interviewed, hire your boss.
I'm on team TJ here. The entire reason I got into software development was because of Access. It was the first time I could actually see what the data was like back there behind the screen. It can do some pretty impressive things. I am going to have a hard time not calling it PowerBy now though ha! (I'm currently a Business Intelligence Engineer)
I love hearing tech war stories like this.
Same in the company I left as well, I also don’t really see the product being used after years of development. The directors all were fired and moved lol
This has been the best laught I‘ve had in ages
I am with TJ on this one. As a first introduction to databases, access was in fact quite cool.
Yeah this much more common than people think at larger organizations. MBAs come in, "lower expenses" (salaries) then have to "manage" the worse human capital to get an inch of productivity.
Processes that ensure everyone is at least 10% productive while kneecapping and ensuring nobody is more than 20% productive without breaking rules isnt a recipe for success in the modern technical world.
i had an idea for increasing productivity from "do the thing takes 3 months" to "do the thing takes 2 weeks" but it would take 2 months to implement tops. it was 8 years ago i brought up the idea and its still not implemented, and we did the thing 27 times since then, so a poor soul (not me) spent 75% of the last 8 years doing the thing, instead of doing it in one
You know who still uses COVID as an excuse? The US Embassy in London. It's _still_ fucking impossible to get appointments without getting up at the crack of dawn every day in the hopes of magically catching the one morning when they release appointments and snagging one before they go.
They probably just need more tax dollars printed out of the ether.
4:23 wheel of prime
I felt the big consulting firm joke.
I think we just discovered how the mythical 10x engineer is born.
This is painful because just about everybody has been there. Maybe not to the scale, but to the point. It's like it's either "no planning" or "no doing" with nothing in between. In a way, it's like those products where you build layer upon layer of "functionality" by adding MORE technical debt at each layer. Then, at some point, some team gets a greenfield initiative to redo the product from the ground up. Year(s) later it comes out and it's amazing, 20x performance. Then you look at the team and they have that sad look. So you dig into the docs and find out the 20x comes from just substituting a base layer that everyone knew was piss poor. It could be 100x, but it's 20x because they had to build THE EXACT same shit that was built before, because the people calling the shots were still calling the shots and they had to have what they wanted to have.Then you grep the sad looks...
I am the author, and my guy, this hurts me to read because it's so true. The thing that upset me about this platform is that it actually was a green field - so it was a total own goal - but even if they did it correctly, the enterprise databases they were pulling information from are just worthless because of what you described above.
@@LudicityHackernews Imagine a table that is basically a key-value store. Imagine it was supposed to hold configurations. Then someone said html templates are configurations. Then someone said, why not put json in there, and have configurations inside configurations. Then... you know where it's going. Sad part if that it's not even the worst part. Hell is real ;)
Last thing i "hinted" at a team doing greenfield was "you know they will if they can, so either you put the roadblocks in NOW or they'll shot themselves in the foot sooner rather than later".
one thing I noticed: why does he leave out the first and last letter when he select text?
22:50 that's the best strategy ever