290 karma · joined May 24, 2016
Only two regrets from last FOSDEM'22 were not having some large water bottle with me and not having cash on person. There's no (were no) ATMs nearby and not everyone takes cards.
Also the idea that we'll need less engineers is bogus. Technology doesn't reduce the amount of work we do, it just increases productivity and puts more strain on individuals to perform. With AI spitting out unmaintainable code nobody understands I can only see more work for more engineers as the amount of code grows.
The article otherwise looks like cheap corp. propaganda to convince the IT crowd to get back into open spaces. I guess megacaps will want to push for getting folks back so they can be controlled by middle managers as in the good ol' times. I'm not sure if that's going to work as per plan though.
Interestingly enough my mental health being out of the office has much improved over before office work. As a developer I thrive on long focus time and good written-down documents. I don't miss being interrupted all the time, being shot at with NERF guns, being a middle man between people having conversation over you, over lunch work debates, open office background noise, over the shoulder staring etc. etc. Not at all.
Web is already hostile enough nowadays with all the tracking, scams, abuses of consent and bad ux designed to sell shit nobody needs.
This makes sense since it wasn't trained on the private codebase but not using existing helpers, naming, abstractions makes the results largely useless since only a minority of output is useful. Like the author mentions what these tools do with the input is a gray area and nobody wants to publicize their code through chatbots. It'd be cool if there was a selfhosted alternative out there, trainable on private codebases.
I've also tried using the bots for answering questions about tech problems I've been having. Answers are mostly off point, have a lot of nonrelated info and most importantly are generated slowly. In the time I get an answer from a chatbot I can scan through a dozen Stack Overflow pages.
As far as I'm concerned we're not there yet but never say never.
I also feel for iranian women. One can't but wonder what life would look like if Iran dumped the derelict religious practices and just went with a more common sense approach to religion and equality.
1. What's up? (chewing the fat about any relevant life topic)
2. My frustrations with the process and possible improvements.
2.1 Followup to recent changes in this area.
3. Experience with recently developed/released features.
4. Follow-up to company wide changes (related to things such as layoffs, direction changes, etc.)
I used to not be a fan of 1on1s while working on-site but since I've started full-remote it has become more valuable time for me since that's basically the most time I spend with my lead in a sprint or two.
Here are some reasons why I use it:
- it's fast
- works consistently well over SSH/in docker containers/remote
- can use it pretty much everywhere and if it's not installed it's easy to setup
- awesome for doing quick text edits or text manipulation (I prefer Vim to sed/awk or anything else)
I don't Vim as an IDE replacement since I find it just takes too much time to maintain a working setup. I'm also not a fan of addons/plugins as over time they tend to fall apart.
Also don't stress too much over failing interviews. Especially for full remote positions companies are trigger-happy. One, two mistakes and you're out. Request a wage that's too high and you're out without any feedback. Not to mention all the companies that are just doing the interviews but not really hiring and those special cases where they simply ignore take-home assignments.
In theory all companies like adaptable people. In practice I find most job descriptions prefer specialists nowadays. Coding interviews and job posting requirements are also very targeted to a specific skillset. Looking back it'd be a better choice to pick a specialization for up-to 4 years and retrain myself every now and then to paradigm/framework/language-du-jour rather than being proficient in everything and a master of none. As a specialist I'm more marketable, have a better workday/quality of life and fit better into teams that need a particular skillset.
Sadly I don't see this transferring to other countries any time soon. Infra was built for the car and a cyclist is often a 2nd grade participant. All the steep inclinations in cities also don't help. CPH is really flat. My hometown has so many hills a lot of people (especially elderly) are not fit to even climb them. It's also really hard to drive at decent speed on narrow lanes built as an afterthought next to the main road. Things are changing though so at least biking is doable despite many downsides.
There's no magic stick effect to increasing productivity. I reckon this exercise is an excuse that will be used in near future to prune the workforce as it's unlikely to yield anything tangible.
I also find people are twisting definition of done to make it look as-if it's over. Something being coded usually isn't done. It still needs to be tested which is the labour intensive task. Something being tested isn't done, it must be shipped and released to production. Code in production still isn't done. It must work and be performant enough, might need more logs, might need more monitoring. Last but not least it must get into the hands of the users who give the final word. But even then it might not be completely finished.
From deanonymization in crypto to Corona pass and tight border control and lockdowns people are being treated as the lowest common denominator - terrorists, money launderers and criminals.
First of all it functions as a currency which inherently brings with it a lot of value. You can exchange and hold virtual currency with minimal fees. This applies primarily to proof-of-stake blockchains (Algorand, Tezos, Cardano). Ethereum/Bitcoin have due to proof-of-work become less a currency and more an investment. There's more value just holding them than actually converting them.
Blockchain tech is trustless and decentralized both of which is very valuable to me personally. I've lost trust in the modern banking system. I guess the turning point at which I became more attent to what the financial salarymen are doing was the 2008 financial crisis. But I'd say that's just the tip of the iceberg. Between all the tax havens hiding billions in plain sight, various scandals such as Wirecard, disappearing money balance on my bank accounts, evaporating loans for large defunct companies, there is just so much crime based around hidden centralized systems. I see blockchain as having the potential to level the field in a system where (centralized) banks are just trying to ripoff the average Joe.
Then we have the stablecoins which work as a safe haven for hyperinflation. While I'm not affected I find this a very serious use case. There's a constant set of countries that are failing to stabilize their local currency and who knows when this is going to hit the reader. Being able to keep value to your monthly wage is a use case in a failing economy. Stablecoins have been found to be an essential part for decentralized finance and pretty much every smart contract-capable cryptocurrency is/has added them.
Also note that having a wallet for a cryptocurrency does not incur monthly charge. Banks charge you for just keeping a balance if you have more than X amount of currency on you account. There's also no maintenance costs which are common for some types of accounts. And plenty of crypto currencies allow you to stake and get a passive income that is higher than what you get on any bank account.
Then there's the DeFI loans. You can basically take a loan as long as you lock a collateral. That's the only condition. Now feel free to go read any of the contracts you signed for a leasing/loan and compare the user experience. DeFI loans are risky due to possibility of liquidation but I see this tech improving and maturing moving on.
While this may sound like I'm the greatest fan I'm more of realist. Blockchain does have a lot of issues. Account recovery is a impossible. Scams are rampant. Transaction speeds are a lot slower than non-blockchain tech can do. Also the immature community praying for a 10x is a pain to follow. But saying it's just a bad DB is plain false.
Mistakes happen in general to everyone in SW development - seniors will make less of them but that doesn't mean they won't happen and that they won't be costly. Generally I'd say if there were serious setbacks to the entire project I'd say something is wrong either with the development process or with communication. Lack of leadership can be a pain but although not ideal I've worked and thrived in such an environment as well.
Ideally your team should catch and prevent bad things from happening before production. This begs the question what kind of testing you do both yourself and as a team (Do you have QA? Do you do automated acc. testing? Do you do integration testing? Do you do code reviews? What about unit testing?). Lack of testing will generally lead to a variety of bugs. Having a ton of untestable legacy code on board also makes development hard in general.
Second thing I'd ask is whether you verify the approach you take to solving the task at hand with your teammates? It's easy to take a few wrong decisions especially if hard legacy is involved. It's in general better to address the approach to solving the problem early and involve seniors even if they don't have experience with the codebase.
Also regarding dealing with production it sounds to me like they simply unleashed you into prod. environment. Giving a junior prod. credentials to me is like throwing him under the bus. I mean what could possibly go wrong there? :) It's relatively easy to break prod. without experience. I suggest getting somebody to assist with production deploys if you don't feel confident. Setup a videocall with somebody from operations/senior dev., share screen and make the relevant changes. Also if somehow you managed to easily mixup QA/prod. that would also imply separation between environments is not that great.
Also if you want to improve the process consider going back and writing some post-mortems. Understanding what and how went wrong and addressing the issues means it's less likely they will happen again to you or another unlucky junior.
Otherwise I decide ahead of time and focus on one area at a time that I know is a small chunk of functionality (like adding boilerplate for a small script, unit testing a piece of functionality, implementing an endpoint) and as long as I'm working on it I just use git add -A ; git commit --amend --no-edit . Once I move on to new self-contained batch of functionality I make a new commit and repeat the amending. It requires a bit of discipline to keep commits small enough but I like it since it's an easy process to follow.
Before I make a review request I usually do an interactive rebase and reword commit messages, sometimes reorder them and squash small stuff if it makes sense. After a review is in process I generally no longer rebase (to keep review transparent) and commit each fix separately.
I do however check out code locally. Online code review only shows what's been changed and I often find subtle issues in the code that wasn't changed. In short I'm trying to ask myself: What's missing in implementation?
I pray for WASM to eventually replace JS. NodeJS ecosystem with all the experimental features, job ads in update notes, shitload of dependencies, one-function packages, legacy JS is a special kind of pain that I avoid whenever I can.
I'd love to see better mocking support. Doing mock.On("fun name", ...) is so backwards, confusing and brittle. It's also a great source of confusion for teammates when tests fail.
I miss better transaction management. I regularly juggle db, transactions and related interfaces and it's a continuous pain.
Then there's the "workaround" for enforcing interface implementation: _ InterfaceType = &struct . This could be easily part of struct def. rather than having it in var section.
As was mentioned by others doing x := 5 only to later do JsonField: &x is just a waste of intellectual firepower. Maybe this can be alleviated by generics but the lang should be able to make this a one liner.
Since you've mentioned fullstack I guess this means you decided not to specialize. Being a little bit of everything is useful for some types of companies and work but is one source of also being inefficient. It's easier to pick frontend, backend, ops or some other specialization. Backend development is a bit more stable than FE but it depends on your taste. Having less specialities means its also easier to fit into certain job postings and crunch through the interviews but it seems to me you don't have an issue there. Being more specialized also passively makes you more efficient in that domain.
Also note that by switching jobs per year you've probably spent half of that time being onboarded. Obviously you'll be slower during the first few months at works unless you're doing something very standard and well defined. I'd also question the underperforming part. Is this due to overpromising? I've regularly worked with devs who overpromised and underdelivered. The trick here is also to properly time those estimates and take on as much work as you can carry. Note that when a company wants to get rid of you they'll always say its a competitive market or claim underdelivery but the real reasons are much more varied.
I'd say top 2 reasons are tied to either adversity in software engineering or simple economy. Devs are pressured to work fast, overtime, in understaffed teams, prioritizing features over quality. Once a project becomes old enough changes suddenly become much more expensive and issues start getting lost in the bottom of backlog. Large projects are chock full of tech debt, bad design decisions that used to be good at the time of writing, undocumented features, untested edge cases. IT workers constantly changing jobs means knowledge slowly evaporates. And lets not forget how fast tech becomes a piece of legacy and all the associated difficulties of working with old tech.
The economy part is simple. Businesses prioritize profit. If people can live with a few quirks they'll stay there indefinitely. If a company gets X cash for a project they'll maximize the profit by doing as little as possible while getting it done. If a product isn't really making that much profit it goes to maintenance mode sacrificing quality for a shiny new thing that will hopefully be more profitable. Projects once they go to maintenance mode only fix what's being payed for and that's the bare necessities.
It feels like everyone is trying to optimize for something. NBA teams optimize for points based on data which clearly affects where and how often they attempt to score. Companies optimize for profit as they always did. Academia optimizes for citations. Hackers optimize for productivity. Optimization has a clear path and there's no room there for innovation. Getting incrementally better is a priority.
The thing is there is a certain point where incremental steps no longer yield that much benefit. This is most obvious in the differences between athletes in some top sports where timings that used to be measured in seconds are now measured in milliseconds. Once you're perfectly optimized there's no more room for growth or progress. I feel like this is where everyone is congregating today. At the max peak of the growth function.