Ask Experienced Devs

Title, mainly aimed at a Blackberry 9800

hope this question doesn't sound too vague.

I've absolutely grown to hate everything AI/LLM related.

It's sole marginal benefit is to senior devs,
with generating some boiler plate code,
which you usually still need to adjust.

For the rest it's been a waste of:

• Time during development, with wrong answers.
• Time during reviews, with garbage generated PRs from junior devs.
• Energy, contributing to global warming.

But lately M$ has been intrusively shoving Copilot down our throats on Github, which I'm quite unhappy with.

So if any of you have some uBlock Origin filters,
or any other ideas on how I can block this Copilot slop out,
please do enlighten me!

Edit: Did some searching of my own.

Go to: https://github.com/settings/copilot
Block + disable everything you can under there.

Then go to uBlock Origin => Open the Dashboard => My Filters => Add:

github.com##.copilotPreview__container
github.com##.AppHeader-CopilotChat
github.com##li.ActionListItem:has-text(Copilot)
github.com##li.ActionList-sectionDivider:has-text(Copilot)
github.com##li.TimelineItem:has-text(Copilot)
github.com##div.pb-4:has-text(Copilot)
github.com###copilot_free_global
github.com###copilot-button-container
github.com###blob-view-header-copilot-icon
github.com##a[href*="/resources/articles/ai"]
github.com##a[href*="/settings/copilot"]
github.com##a[href*="/features/copilot"]
github.blog##a[href*="/features/copilot"]
github.blog##a[href*="/ai-and-ml"]
github.blog##article.changelog-label-copilot
github.blog##article.changelog-label-models

github.com##article.js-feed-item-component:has-text(LLM)
github.com##article.js-feed-item-component:has-text(OpenAI)
github.com##article.js-feed-item-component:has-text(ChatGPT)
github.com##article.js-feed-item-component:has-text(GPT)
github.com##article.js-feed-item-component:has-text(Llama)
github.com##article.js-feed-item-component:has-text(Gemini)
github.com##article.js-feed-item-component:has-text(Grok)
github.com##article.js-feed-item-component:has-text(DeepSeek)

It's not perfect, but at least it's a start of getting rid of the unwanted content.

For context, "no hello website" is a term I'm using to describe websites like https://nohello.com. The gist is that it's trying to get people to no only say "hello" but also ask whatever it is they're trying to ask you about. (In my opinion, this can even extend to generic conversations, like "hey, how are you?" versus just "hey!")

The problem is, many of these websites seem pretty rude. At the end of the day, I don't think it's possible to make one not seem at least a little rude because telling someone not to say hello just sounds rude lol. That said, even my favorite one (https://nohello.club/) has phrases like

1. Unnecessary pleasantries
2. useless phrase like "Hello"

But I like it because it doesn't say things like this that the original http://nohello.com says

please be prepared to be ignored if you only say "Hello!".

Some context, I'm not necessarily looking to include this on any corporate messaging app bio (unless it was 100% polite, which as I said I don't find possible). But the closer to something that polite, the more useful I think it is.

I'm a Senior Principal Engineer at a large tech company. My work is to define the 1-3 year future for a $1B slice of the company's revenue, that 1% of it's employees work on.

There is nearly zero code in my daily work. The closest I get is being occasionally tagged into a pull request to resolve a dispute or clarify the long term vision, but even then that's usually API contracts or data schemas that cross code bases-- and not actual code. I spend about 1/5th of my time with the engineers who actually do things and write code, but it's to talk through math, CS fundamentals, algorithms, or system design on their current project -- not code.

Day to day the bulk of my time is trying to convince non technical people, mostly in our Product organization, that we can actually make certain evolutions. Or to drop innovators dilemma type assumptions that add a year to our time to market. I get them to agree and then collaborate with them on large documents that define these futures. I vouch that we can do it, at what cost and on what timeline, and they vouch for how much it will make. We spend months writing documents like this and then pitch them up to the stratospheric leadership of our company to ask for 10-20% more resources to make 20-50% more money 2-3 years from now. I simultaneously do system design for these evolutions and sketch out the high level details of new components or what needs to change in existing components, to make the beginnings of a workable execution plan if we do get resources.

This is all fun and I'm grateful my career has kept changing and being different. Though I do miss the pure joy I find in coding all day, alone. But I am worried that the longer I stay this high level the more I will be full of shit, technically. I'm only effective at finding this intersection of customer desire and engineering reality if I know what is easy, or hard, to do. And I have a lot of examples of high level "technical" people at my company being full of it. It's usually odd details, like they don't know that certain problems were solved 10 or 20 years ago. So they twist designs and projects into knots to avoid or re-solve problems that the modern cloud or whatever language or library just has out of the box. Or they prevent profitable market opportunities because they think something is impossible, based on a gap in their knowledge of algorithms or newer tools or realities about how fast computers are these days.

I don't want to be full of it. People like me eventually get fired if our Big Bets can't actually be built, or enough people realize we are often wrong about details.

And I also worry that, as an IC, the longer I stay so non-technical the less likely I am to pass the technical portion of any interview. I also think it's unlikely folks are directly hired into this kind of high level position, anyways, so I need to be able to slip back into a more technical role.

So how do I keep up? How can I be less full of it next year than I am now? I can leetcode, and I should, but that's far removed from what I feel is the important part of my atrophying technical skills: the small details in tools that determine how large systems fit together and evolve.

They seem to be from 80', and somehow made their way into a french university. Has anyone here gotten a chance to read them? Some of the chapters seem interesting, lots about DOD stuff.

Hello fellow devs.

I'm in need of a new machine as I had a little mishap with my notebook. For a long time, I thought on buying a gaming notebook as normally they should have the best hardware for my personal (gaming, light video edit) and professional (full stack web dev) needs.

Next week, Asus will launch the ROG Ally X officially on my country. So, I'm wondering if it could be a viable alternative.
The other possible devices would be ROG Ally (Z1 Extreme) and Lenovo Legion Go, as Steam Deck is not available here.

I work from home for a foreign company. I have a monitor, a wireless keyboard, and a usb trackball already. I bought them to use with a mac mini my previous company lend to me. I do not have a desktop and do not intend to buy one right now.

So... My questions: Does any of you have experience using a handheld device as a main dev machine? Are there any cons I'm not considering?

Thank you!

I work in research (uni) and am writing a framework for heat processes to optimize their costs. This goes both for private houses and industrial processes. The goal is to enable industry players to see that/when renewable energies and heat reuse with heat pumps are cheaper than fossil fuels. I do this using digital twins for components and on a system level.

My boss hesitates because he thinks this must already exist. I want to pursue that path with my research, so any insights there are welcome, too, but this is primarily about the open-source project.

I searched GitHub and came up empty, but that is only a subset of the search. Do you have any idea how I can find this, one way or another? It would also be great if I could show that it likely does not exist.

I'm looking to set up a server of some kind that I can use to store more or less arbitrary files on demand.

While I have quite a bit of programming experience, I have little-to-no experience in the server-space, so I don't really know where I should be getting started/what kind of pitfalls I should be looking out for/what kind of design choices I should be making early on.

In short: I want some system that allows me to take more or less arbitrary files, send them from either my laptop or phone, and have them stored on a drive that I can have lying around somewhere hooked up to some setup. I don't need any automatic backing up, sending files manually is sufficient. The individual files I'll be sending probably won't be exceeding the MB range of sizes. Remotely downloading files from the storage is not an immediate requirement, if I need to retrieve them I can plug directly into the disk. What I want to protect myself against is the "freak accident" type of thing where all the devices I currently have copies of a file on are lost in a fire, while travelling, or something like that.

Does anyone here have any tips for where I should be looking to get started?

I'd like actual examples instead of "I work faster", something like "I can move straight to the middle of the file with 7mv" or "I can keep 4 different text snippets in memory and paste each with a number+pt, like 2pt", things that you actually use somewhat frequently instead of what you can do, but probably only did once.

I'm just so exhausted these days. We have formal SLA's, but its not like they're ever followed. After all, Customer X needs to be notified within 5 minutes of any anomalous events in their cluster, and Customer Y is our biggest customer, so we give them the white glove treatment.

Yadda yadda, bla bla. So on and so forth, almost every customer has some exception/difference in SLAs.

I was hired on to be an SRE, but I'm just a professional dashboard starer at this point. The amount of times I've been alerted in the middle of the night because CPU was running high for 5 minutes is too damn high. Just so I can apologize to Mr. Customer that they maybe had a teensy slowdown during that time.

If I try to get us back to fundamentals and suggest we should only alert on impact, not short lived anomalies, there is some surface level agreement, but everyone seems to think "well we might miss something, so we need to keep it".

It's like we're trying to prevent outages by monitoring for potential issues rather than actually making our system more robust and automate-able.

How do I convince these people that this isn't sustainable? That trying to "catch" incidents before they happen is a fools errand. It's like that chart about the "war on drugs" where it shows exponential expense growth as you try to prevent ALL drug usage (which is impossible). Yet this tech company seems to think we should be trying to prevent all outages with excessive monitoring.

And that doesn't even get into the bonkers agreements we make with customers to agree to do a deep dive research on why 2 different environments have a response time that differs by 1ms.

Or the agreements that force us to complete customer provided training - while not assessing how much training we already committed to. It's entirely normal to do 3-4x HIPAA / PCI / Compliance trainings when everyone else in the org only has to do one set of those.

I'm at a point where I'm considering moving on. This job just isn't sustainable and there's no interest in the org to make it sustainable.

But perhaps one of y'all managed to fix something similar in their org with a few key conversations and some effort? What other things could I try as a sort of final "Hail Mary" before looking to greener pastures?

cross-posted from: https://lemm.ee/post/33211685

Hello Lemmy,

I am the author of bluetuith, an open-source TUI-based bluetooth manager for Linux only. I have been working on this project for over 2 years on and off, and I was wondering about extending support to other platforms as well.

To begin with, the Bluetooth Classic (BR/EDR) implementation on Linux is fairly standardized (via bluez APIs), but on other platforms, especially windows, Bluetooth APIs are finicky, and tricky to deal with, and also there is no standardized management in general.

I would like to start creating a centralized Bluetooth server or a daemon for other platforms (natively maybe), mainly Windows and Linux, which can expose relevant APIs so that clients can use them to handle Bluetooth-based operations. I know this is quite an uphill task, but I would like suggestions on how to implement it, or if anyone has a better idea, please do suggest that as well.

To summarize, my current plan is this:

• Create bluetooth servers natively for each platform, utilizing the platform's proven APIs to handle bluetooth-based functions and expose a standard API to clients
• Adapt clients to use said APIs provided by the daemons to allow the user to control bluetooth in general.

For the server implementation (mainly to other platforms), I will require contributors, so contributors are highly welcome to be involved in the project. I am in the process of securing an NLnet grant to invest into this project and mainly pay contributors to implement this platform-wise (the proposal has been accepted, and the negotiation call will be hosted in a few weeks, more details about this can be further published if anyone has questions about this. If contributors are confirmed, maybe the budget could be adjusted as well).

I apologize if the post is naive or does not fit this community's guidelines, and if it doesn't, a comment on where to redirect this question would be great.

Constructive feedback is appreciated. Thank you.

Note: By Bluetooth operations, I mainly mean Bluetooth Classic based operations.

The problem is I am using beam.ParDo to call api, to enrich my data.

pipeline | "read_csv" >> beam.io.ReadFromCsv("...") 
                | "get_info_api" >> beam.ParDo(GetApiDoFN()) 
                | "write_output" >>  beam.io.WriteToBigQuery(...)

The problem is that api have limit. I need a way to control so it do not overload api or get 429 error.

I would like to set api call should not exceed 4 call / second.

How should I do it ?

cross-posted from: https://sh.itjust.works/post/19440902

I want to document my debugging sessions in a text file but I don't know if anyone did this before.

I came up with this kind of "language" that is a mix between Markdown and C++, but I still wonder if something equivalent exists already.

// When you click on the button
# [click button]
- A::f()
// - ... other method calls, don't document if you don't need to

# A::f()
// "..." for "parameters" where you don't need the details
- Stuff::g(...)
- Stuff::h(...)

// <Class> is a fake template thing to show the possible types of an object
# <SubStuffA | SubStuffB> Stuff::g(...)
- Stuff::g() {} // empty but I use v/=> for virtual call
  v/=> SubStuffA::g()
  v/=> SubStuffB::g()

# SubStuffA::g()

# SubStuffB::g()

# Stuff::h(...)

I document methods in the order of appearance in the code.

If you have any good idea about a reliable way to document a list of function calls, I'm interested!

Sorry for the burner account.

I have to figure out what to do with my life right now. I really enjoy programming, and honestly, of any kind. Haven't really found a kind of developing I dislike yet. I have been doing stuff for around 4-5 years by now, so I have confidence that I'm a good programmer, with the huge caveats that I've never finished any presentable project and I've never done anything with a team, I've only done solo stuff.

It seems like the logical thing to pick for a job. However, I've heard experiences of people with programming jobs and CS degrees that they're absolute hell to be in. Super long work days, absurd deadlines, crunch, and that doing a CS degree means you have absolutely zero time for anything else in your life.

Having a life like that really scares me. I'm not really a strong, disciplined person. I know I can't handle living like that. I'm scared I'll just realize I want to quit and end up having wasted years of money and work on a degree I don't want to use for anything - and that's even assuming CS college isn't that awful.

My biggest dream is doing indie game development, and it has always been that since I was a little kid, but I know that's not a safe prospect for a reliable living wage. At the same time, abandoning that dream completely would make me feel awful. So I NEED to have time to work on my own stuff.

I wouldn't go to a CS degree purely for more job opportunities, I'm sure there's a lot of things I'd be able to learn in one that I need. I just don't want to end up living just to work. I'm really only going off on rumours and experiences of other people I know though - and I don't have much of a chance of visiting a campus or talking to professors. Because of life reasons it'd have to be abroad and I'd have to do at least the first year online.

So... yeah. I'd appreciate hearing some experiences in CS degrees and in programming jobs. Is it really that bad time-wise? Is it something enjoyable?

I'm really enthusiastic about anything involved with Unix/Unix-like operating systems and their ecosystems (Mostly Linux and a bit of BSD variations). I also know a couple of programming languages including C, C++23 and OCaml. But other than doing a couple of tiny projects mainly to practice my programming skills in the languages mentioned above, most of my experience is theoretical and it comes from reading books, blogs and watching conferences.

I'm interested in gaining "actual" experience by doing systems programming related projects but I'm not sure how to get started. How do I decide on which topic to choose? A topic which would not be too overwhelming and actually achievable? And how do I gain the background knowledge needed to implement a project without ending up copying and pasting everything from existing implementations?

Thank you!

Referring more to smaller places like my own - few hundred employees with ~20 person IT team (~10 developers).

I read enough about testing that it seems industry standard. But whenever I talk to coworkers and my EM, it's generally, "That would be nice, but it's not practical for our size and the business would allow us to slow down for that." We have ~5 manual testers, so things aren't considered "untested", but issues still frequently slip through. It's insurance software so at least bugs aren't killing people, but our quality still freaks me out a bit.

I try to write automated tests for my own code, since it seems valuable, but I avoid it whenever it's not straightforward. I've read books on testing, but they generally feel like either toy examples or far more effort than my company would be willing to spend. Over time I'm wondering if I'm just overly idealistic, and automated testing is more of a FAANG / bigger company thing.

I'm thinking of ways to help people move from established software to more open, flexible forms that don't lock them to another organization.

I am professionally a software developer for 8 years and I simply don't have ideas for personal projects (Can't find any problem that I can fix with programming). At times I feel like that's natural and I shouldn't worry about it. But on the other hand, I do like to imagine having something personal that I can work on so that even if some days on my main job are not satisfying, I can always work on my hobby project and find that missing satisfaction.

End goal here is obviously to get better sleep as sometimes my mind feels dissatisfied with the day's work.

Funnily, I day-dream about the idea of already having done the boring parts (simply manifesting a project that already exists) of some personal project and only solving exciting problems in relation to adding a new feature or exciting aspects.

This creates a problem as I hate staring at a blank file not knowing what to write.

Imagine, your are a Java developer with multiple years of experience in the job. You really like working with the language.

Your employer kind of canceled most of the Java projects of the company over time and is now really focused on AI...... And AI means here: LLMs, GPT, ... Not like basic Machine learning... It's all about language models. Most of this stuff and the tools are written in Python and your employers wants you to kind of throw away your pretty good Java skills completely and start over in Python.

The new tasks would be kind of "easy"... You have to prototype "LLM bots". And that's your perspective for like at least 1.5 years. No, not real software development.. Prototyping... And that means, quick and dirty is what they want... It's also very easy to impress your employer with GPT doing things. Easy money, isn't it?

I'm in this exact situation right now and worried.. What, if I quit in 2 years and the new potential employer for a Java job asks "What have you done recently?"

I kind of liked working there and like the colleges and the salary is fine and switching the job and maybe moving away is a huge thing for me... It could get better... But maybe also worse........?

What would you do in my situation? Accepting it? Starting a rebellion? Looking for a new job somewhere else?

I've been doing some interviews lately, and my most recent one seemed to be my most promising one so far, until he started doing what he called "light tech screening" and did some trivia questions to test my knowledge on what goes on under the hood. I do admit, I should have been better prepared on the topics, and didn't do as well as I could have. At the end, he said my experience was interesting but I needed a better grasp of the fundamentals, and that we should keep in touch. Afterwards, I messaged thanking him for his time, and that I needed to review my course material, asked if he thought I should review anything specific and if I could do that for a few weeks and revisit it like he said. He suggested I do a deep dive into JS/TS to really understand what goes on and why, and would be happy to revisit it in a few weeks, and again said to keep in touch.

Does that really mean keep in touch regularly or often? Or is it just a throw away yea reach out again in a few weeks when you're ready? I'm already pretty stoked to get a second chance at this, the company seems cool and I like that they aren't weighing everything on leetcode problems. If I should keep in touch regularly, what would that even be about? I also don't want to annoy him before meeting again.

Any advice would be appreciated!

My current team runs weekly retrospectives using the Lean Coffee format. More and more, I find that the items people are bringing up aren't really important or could just be a question in Slack.

For example, someone recently made a topic for how we can test credit card payments. Another topic was navel gazing about how we use Jira and multiple team members asked "what's the problem you're hoping to solve?" to which the only answer was "That's not what I've seen elsewhere".

I'm beginning to think that there's something wrong with our format or prompts, in that we aren't identifying important issues for discussion. Perhaps the format is stale or there's no serious issues lingering each week?

Any advice on alternative formats, how to get better feedback, etc. would be greatly appreciated.

Hello! I'm making changes to a web app that involves some aria classes, and I read the docs about how they work, but I'm not real confident in my ability to construct them into something that produces a well-usable experience for someone who's using the app. Is there a good way to get direct feedback from someone who uses a screenreader, or otherwise get some evaluation on whether what I've fumbled together is actually functioning well?

Would I have to drive there myself or could I pay an on-site tech to do it for a courtesy fee? Does it depend on the company?