|
luchadornado posted:I was thinking about this earlier: what idioms or rules of thumb do people here see that other devs just don't "get", or that they tend to abuse? "let me just write the abstraction based on one use case so that it's sCaLaBlE and mAiNtAiNaBlE" a few months later: "oops wrong abstraction"
|
#
?
Nov 2, 2021 22:02
|
|
|
|
| # ? May 27, 2024 14:55 |
|
|
luchadornado posted:I was thinking about this earlier: what idioms or rules of thumb do people here see that other devs just don't "get", or that they tend to abuse? DRY, KISS, and/or YAGNI all tend to be thought terminating cliches that sharpen me towards scrutinizing some bullshit that’s soon about to be justified. Most developers ime have shiny bauble syndrome and work towards resume driven development so they can add the latest bleeding edge bullets onto their resume. I get it, because industry incentivizes you to do so, but it usually is the wrong choice to make unless you want to qualify some experience with it on a sinking ship and toss the hot potato on your way out to the next unlucky soul
|
|
#
?
Nov 2, 2021 22:32
|
|
|
lifg posted:Do you all consider any big tech companies more immoral to work for than others? I've been ducking a Facebook recruiter for a little while, but I'm not truly sure there's a huge moral difference, or if I just like don't like their products. I think Facebook's core conceit is incredibly dangerous and I would not contribute to it unless the alternative is to starve. Lots have been written about filter bubbles, but at it's core Cretin A likes the same stuff as Cretin B and Facebook links them up so they can convince each other that everyone else is the idiot and radicalize. The impacts are obvious and profound: if you lend your talents to Facebook you're complicit in the mainstreaming of antivax, Q Anon, resurgence of white power groups, and the election of Trump, among other things. I don't think that's fixable. I certainly don't think they care to fix it. On top of that, the perversion of a "public company" where voting doesn't follow the shares means a deeply amoral CEO has unchecked power. There is no mechanism by which he loses power short of either government regulation that breaks up the company or (not a threat or an implication) his natural death. Lots of people probably feel that way about the other tech companies. "You have to figure it out for yourself" is definitely true. I would add you also have to decide to what extent you can effect change from inside, because I have seen for myself that it's possible to move the needle.
|
|
#
?
Nov 2, 2021 23:28
|
|
|
on the other hand, my russian buddies uncle used to do bioweapons gain of function research on smallpox (which he got to learn as an adult, after the ussr fell) so, you know, there's that, try not to do that either there's a lotta poo poo in corporateland. pollutin poo poo, weapons poo poo, infowarfare-propaganda aka 'advertising' poo poo. finance poo poo, legal (which isn't to say ethical ofc) and nonlegal. keep your conscience sharp and cognizant, i guess bob dobbs is dead fucked around with this message at 01:04 on Nov 3, 2021 |
|
#
?
Nov 3, 2021 00:55
|
|
|
The only ethical code is much like the only code without bugs: no code at all
|
|
#
?
Nov 3, 2021 00:59
|
|
|
No ethical production under capitalism.
|
|
#
?
Nov 3, 2021 01:01
|
|
|
brezhnev period production also pretty unethical at times, apparently. they tried hard, it didn't work out, guess prolly gotta try again maybe without gulags this time
|
|
#
?
Nov 3, 2021 01:05
|
|
|
csammis posted:The only ethical code is much like the only code without bugs: no code at all https://dune.fandom.com/wiki/Butlerian_Jihad
|
|
#
?
Nov 3, 2021 01:12
|
|
|
kull wahad! gently caress the failson tho
|
|
#
?
Nov 3, 2021 01:18
|
|
|
To editorialize further, in the story, basically technology reached a point where they realized it was a net drag on society, and banned all technology more complex than a programmable microwave as a long term solution
|
|
#
?
Nov 3, 2021 01:40
|
|
|
Achmed Jones posted:it's this. i won't work for facebook, palantir, etc. i wouldn't work on weapon systems, but i'd probably be ok working for a defense contractor like raytheon in infosec type stuff depending on the specific job. i currently work for google doing blue team stuff, but there are things at google that i wouldn't be comfortable working on. Somehow we are all just low level workers in the Death Star.
|
|
#
?
Nov 3, 2021 02:06
|
|
|
luchadornado posted:I was thinking about this earlier: what idioms or rules of thumb do people here see that other devs just don't "get", or that they tend to abuse? This is also the one that's wasted my time the most. It gets worse once Kubernetes becomes involved and you end up with miles of tooling and something that would have been a single no-hassle Rails server back in the day broken up into like 8 Docker images plus all the Kubernetes setup overhead. Monoliths are fine if you're not literally Google.
|
|
#
?
Nov 3, 2021 02:08
|
|
|
lifg posted:Do you all consider any big tech companies more immoral to work for than others? I've been ducking a Facebook recruiter for a little while, but I'm not truly sure there's a huge moral difference, or if I just like don't like their products. Yes and no? I'm not interested in working for any of them, so no (this may be one of those convenient stances you describe). But if you forced me to rank them, I could, so yes?
|
|
#
?
Nov 3, 2021 02:45
|
|
|
luchadornado posted:I was thinking about this earlier: what idioms or rules of thumb do people here see that other devs just don't "get", or that they tend to abuse? This is how I feel about strewing a bunch of single-purpose code across many different files.
|
|
#
?
Nov 3, 2021 02:47
|
|
|
|
|
#
?
Nov 3, 2021 04:05
|
|
|
The next time a coworker catches me hitting the random Wikipedia link endlessly I’ll say I’m in mentat training
|
|
#
?
Nov 3, 2021 04:06
|
|
|
Roadie posted:
Or if your company never scales beyond two teams of six: front end and back end, sure We are trying to rewrite our payment system, and it's tightly coupled with our delivery system, which is a mess because the founder decided to cut the shopping cart corner and make the delivery the cart Our warehouse inventory team is pulling their hair out on a near daily basis because they can't make any changes to their stuff without inadvertently making changes in our payment and delivery modules and inventory is impossible because some 10 year old just added $50,000 worth of 200 widgets to their "cart" for fun and until that cart clears at midnight, they're potentially backordered 150 widgets for six weeks This is just with a team of 40-50 engineers Once you get to $10mm ARR is the correct time to build major new features as seperate services so that you have a pattern for others to follow and your monolith at least somewhat supports services on some level By the time you hit $100mm ARR your monolith should have been split into at least 4 services with a plan/architecture of/for doubling within the next year Monoliths are great until they aren't
|
|
#
?
Nov 3, 2021 06:37
|
|
|
"Monoliths are good, actually." *develops something that can never be changed significantly, then quits* edit: beaten
|
|
#
?
Nov 3, 2021 11:10
|
|
|
lifg posted:Do you all consider any big tech companies more immoral to work for than others? I've been ducking a Facebook recruiter for a little while, but I'm not truly sure there's a huge moral difference, or if I just like don't like their products.
|
|
#
?
Nov 3, 2021 16:51
|
|
|
Hadlock posted:Or if your company never scales beyond two teams of six: front end and back end, sure I think good monoliths are still well architected with loose coupling between systems, it just means everything lives in a single repo and builds/ships together. Obviously by putting things in separate repos/services you’re kind of forced to loosely couple on one level, but it’s not automatic and a team that is not thinking carefully about dependencies will run into a bunch of other (exhaustively discussed) problems.
|
|
#
?
Nov 3, 2021 16:58
|
|
|
Splitting things into services makes tight coupling a lot more painful but a lot of people have discovered the hard way that microservices don't actually make it impossible. People have also very happily scaled monoliths to much larger sizes than 50 people working on them. It's always tempting to look at a code base that's seen growing pains and think that if you'd just made one big decision better a few years ago everything would be great, but if the problem is that the person in charge thought that it would be a good idea to conflate several different concepts that turned out to actually be very different, they still could have made that mistake if they'd instead wrote a separate service that handled all of the things they considered the same. They also may have instead made the inverse mistake of treating two things that need to be the same as separate, leaving you with two services that each want to be the source of truth for the same thing.
|
|
#
?
Nov 3, 2021 17:14
|
|
|
Eggnogium posted:I think good monoliths are still well architected with loose coupling between systems, it just means everything lives in a single repo and builds/ships together. Obviously by putting things in separate repos/services you’re kind of forced to loosely couple on one level, but it’s not automatic and a team that is not thinking carefully about dependencies will run into a bunch of other (exhaustively discussed) problems. I worked on a monolith with very good separation between systems, but we still did massive every-other-week releases, so we ran into problems when just one feature of a release needed to be rolled back. We could have built the tooling to fix it, but it was just never a priority. These days I care about mono-v-micro a lot less than I care about a healthy release process.
|
|
#
?
Nov 3, 2021 17:21
|
|
|
lifg posted:These days I care about mono-v-micro a lot less than I care about a healthy release process. That's where I'm at too. Can you get stuff out to production with some reason to believe that it's sane? Can you roll it back quickly if something's wrong? If the answers to both of those questions are "yes", then whatever you're doing is probably good enough.
|
|
#
?
Nov 3, 2021 17:25
|
|
|
lifg posted:I worked on a monolith with very good separation between systems, but we still did massive every-other-week releases, so we ran into problems when just one feature of a release needed to be rolled back. We could have built the tooling to fix it, but it was just never a priority. Yes, I think a flighting/killswitch system for changes is a requirement at a certain size. I don’t know your circumstances but usually this should be an easy sell to management, “hey your pet feature was delayed by a bug in this completely unrelated feature, we should probably prevent that in the future.”
|
|
#
?
Nov 3, 2021 17:25
|
|
|
Mono (that works fine, even if it's an over-engineered mess) with a hearty collection of micros (that work fine, even if they introduced substantial dependency complexity) on K8's (that doesn't work fine, sometimes doesn't work at all) checking in. In the end, we all should have something that supports review apps. Mono or micro, if you have review apps, you can test in isolation and not have to worry about blocking deploy. e: To not rebump the thread with a monologue... Had an applicant apply with their medium blog linked, which I thought was cool. Until I read a blog post of theirs, in which they expressed a dev opinion that ran completely counter to how our team feels (and, imo, shows that they lack experience with any mature codebase). So anyway, they're no longer an applicant... kayakyakr fucked around with this message at 22:13 on Nov 3, 2021 |
|
#
?
Nov 3, 2021 18:58
|
|
|
I took a job at Google, one of the less evil companies (compared to FB and AMZN at least), and I still don’t trust it wholeheartedly. Professional respect and trust have to be earned for corporations, not just freely given. Thankfully I work on a boring but integral part of the whole stack so I lucked out on Not Being Evil so far.
|
|
#
?
Nov 4, 2021 14:23
|
|
|
I just met an ex-Apple guy who did acoustic engineering for them (how the noise/vibration of capacitors and whatnot impact form factor) and he's retired at my age. Major FOMO of figgies going on now, even though I'm pretty well paid.
|
|
#
?
Nov 4, 2021 18:15
|
|
|
https://twitter.com/Firr/status/1456324664628846599 I highly recommend reading through that posted Twitter thread. It involves going to interview at his old work (while wearing a fake mustache), getting lots of snacks, and actually getting through most of the interview process before being escorted out of the building.
|
|
#
?
Nov 5, 2021 23:13
|
|
|
Early retirement sounds scary. Imagine something big happening and you have to re-enter the workforce except everyone’s got a decade or two of experience over you.
|
|
#
?
Nov 6, 2021 17:24
|
|
|
Pollyanna posted:Early retirement sounds scary. Imagine something big happening and you have to re-enter the workforce except everyone’s got a decade or two of experience over you. I feel this, but spending at least 40 hours a week of my healthiest next couple of decades continuing to work for the surveillance state on widgets and gizmos for extremely wealthy speculators is also scary
|
|
#
?
Nov 6, 2021 17:47
|
|
|
Pollyanna posted:Early retirement sounds scary. Imagine something big happening and you have to re-enter the workforce except everyone’s got a decade or two of experience over you. Based on historic data and sequence return risk you're far more likely to have your investments massively outgrow the pessimistic estimates SWR relies on or realize failure within ~5 years.
|
|
#
?
Nov 6, 2021 19:37
|
|
|
I would simply only early retire if I had a substantial amount of money and planned on moving to a very low CoL area.asur posted:Based on historic data and sequence return risk you're far more likely to have your investments massively outgrow the pessimistic estimates SWR relies on or realize failure within ~5 years. Also, this.
|
|
#
?
Nov 6, 2021 19:42
|
|
|
Good Will Hrunting posted:I would simply only early retire if I had a substantial amount of money and planned on moving to a very low CoL area. And had no emergency medical expenses, and are comfortable living on a fixed income, etc Retirement can get expensive really quickly and plenty of older folks who are retiring at 65 are running into dwindling savings because life expectancy is longer than it used to be. If you retire at 65 you have (optimistically) another 20 years of paying for your lifestyle where a portion of that will likely have some kind of long-term medical expenses. Now consider what that looks like if you decide to retire at 40. Health insurance has to be purchased separately and it can get expensive for the same coverage the average tech company has. CoL will continue to rise regardless of where you're living. Vacations, entertainment, major purchases, and many other things that might take a chunk out of a replenishing savings now have to be accounted for in a much longer timeline. You can't liquidate your retirement accounts without penalty until 65 so that's a big chunk of deferred money. None of this is even accounting for having a family or taking care of aging parents, etc. Obviously risk tolerance, etc but I'd personally have to have at least $5 million in the bank before even thinking about early retirement and given the uncertain state of the world I probably wouldn't consider it unless I had at least 10 and even then idk probably not full retirement.
|
|
#
?
Nov 6, 2021 20:01
|
|
|
yeah, just gently caress off for a year like i do sometimes and set up an llc that you pretended to work for
|
|
#
?
Nov 6, 2021 20:16
|
|
|
lol at needing 10 million to retire it's absolutely imperative that i spend 250k a year for 40 years while leaving my retirement fund in my checking account
|
|
#
?
Nov 6, 2021 20:19
|
|
|
it is a good point re: health insurance though. that shits mad expensive
|
|
#
?
Nov 6, 2021 20:20
|
|
|
What type of investments are you making thinking you'll need 10 mil?! Unless you're investing at a fixed very, very low rate and not adjusting risk, reinvesting, a number of factors.... 10 million is an insane amount.
|
|
#
?
Nov 6, 2021 20:22
|
|
|
Blinkz0rz posted:Obviously risk tolerance, etc but I'd personally have to have at least $5 million in the bank before even thinking about early retirement and given the uncertain state of the world I probably wouldn't consider it unless I had at least 10 and even then idk probably not full retirement. This is a little crazy, IMHO, but it would depend on your expenses. Assuming your housing situation is sorted (really low mortgage, or paid for entirely), a 3% SWR should be like 99.9% likely to work, even for a long retirement/FIRE. If you have $3MM, that's 90K a year, which is a lot almost anywhere, considering you don't have childcare or commuting costs. Healthcare is for sure expensive, but ACA subsidies are generally based on income, so playing with withdrawal rates could easily net you very cheap plans through that marketplace. Numbers like 5MM/10MM make me think I'm on bogleheads.org where posters are unaware of the bubble they live in. The average amount of savings most people retire with is less than $250K, so even talking about FIRE, throwing numbers around like 5MM (top 3% of wealth) or 10MM (top 1%) is a little out of touch with the rest of the country/world.
|
|
#
?
Nov 6, 2021 20:26
|
|
|
Achmed Jones posted:lol at needing 10 million to retire It's not 250k because you're still paying taxes on your "income", managing insurance and immediate medical expenses, maintaining your residence (or buying property, etc), replacing deprecating assets like cars, and plenty of other stuff I'm sure I'm forgetting without active income. Also, if you have a family you have to think about college for your kids and combined retirement expenses for your partner. But lol, if you or someone in your family gets cancer your plans are hosed because unless you're paying a bananas amount for insurance (and even if you are) that's gonna be a massive bill. Also lol if you have to care for an elderly parents who doesn't have enough money socked away for medical expenses and/or hospice care. Anyway, $10 million is my tolerance because of my particular situation.
|
|
#
?
Nov 6, 2021 20:28
|
|
|
|
| # ? May 27, 2024 14:55 |
|
|
Good Will Hrunting posted:What type of investments are you making thinking you'll need 10 mil?! Unless you're investing at a fixed very, very low rate and not adjusting risk, reinvesting, a number of factors.... 10 million is an insane amount. B-Nasty posted:This is a little crazy, IMHO, but it would depend on your expenses. Assuming your housing situation is sorted (really low mortgage, or paid for entirely), a 3% SWR should be like 99.9% likely to work, even for a long retirement/FIRE. If you have $3MM, that's 90K a year, which is a lot almost anywhere, considering you don't have childcare or commuting costs. Healthcare is for sure expensive, but ACA subsidies are generally based on income, so playing with withdrawal rates could easily net you very cheap plans through that marketplace. You're still thinking about 65-death as retirement (and whether 250k is actually enough to cover that depends on a lot of luck). Add another 20+ years and it becomes a different proposition, especially because lifestyles change as you age. Even more so if you're looking to retire young enough that you might not have even thought about having a family.
|
|
#
?
Nov 6, 2021 20:32
|
|