Ask Experienced Devs Weekly Thread: A weekly thread for inexperienced developers to ask experienced ones
Posted by AutoModerator@reddit | ExperiencedDevs | View on Reddit | 78 comments
A thread for Developers and IT folks with less experience to ask more experienced souls questions about the industry.
Please keep top level comments limited to Inexperienced Devs. Most rules do not apply, but keep it civil. Being a jerk will not be tolerated.
Inexperienced Devs should refrain from answering other Inexperienced Devs' questions.
sushiEnthusiastt@reddit
Hi all, I'm currently struggling to choose between two internships. The first one (let's call it Company A) is SWE at a large, legacy tech company. It's onsite and I don't know what my team is yet. Company B is DevOps at a small, modern SaaS company. It's remote and I'm gonna be on the cloud infrastructure team.
Company B looks good on paper (remote WLB, smaller teams = more ownership? and cloud infra) but I'm worried about a few things. First, it's devops not swe - I don't think I'm gonna be happy if I'm not writing code. The name might also just not be as recognizable as Company A. Finally, I'm not sure how I feel about it being remote. I'm worried I'm not gonna learn as much (and because its remote + devops, I technically could get stuck on some mickey mouse project). Remote is also just so isolating.
On the other hand, company A does have a pretty established internship program (that is known for being kind of flexible?), but I'm worried I might not learn as much because of potential legacy code/teams or because its slow moving. I think I should note I also have an upcoming SWE internship in the fall, so ideally I'd like to be writing as much code to prep for that one? Idk
Anyways, I'd appreciate any insights or advice or thoughts. Thanks
Sea-Activity9179@reddit
Hi , I am currently working as backend developer in spring boot. i am about to complete 2 and 1/2 year in the company .I want to switch now .But as growing AI and seeing market want to know from experienced dev .What all skills or tech I must know or how interview is taken ?
navlaan0@reddit
Interviewing is a skill in itself. When I decided it was time to change jobs, it was the greatest barrier for me, so maybe start with that.
Are you interested in switching jobs but staying in the spring boot ecosystem? If so, i believe it's easier, just review the fundamentals, practice some leetcode, and do all the interviews you can, even the ones you don't want to. The market changes from place to place, and the only way to really grasp what companies in your country want is to actually do those interviews (talking with people who have done them helps too).
If you want to change languages or frameworks, that is totally okay too. You just have to find a company that is not attached to specific language experience. In my country, those are not that hard to find.. i started my career as a .NET developer and after 5 years i switched to a company where i work with various repos in Python, Node, and Kotlin. Also, companies that don't care about the language usually care a lot about the engineering part: system design, problem-solving, etc.
Good_Skirt2459@reddit
Hello all! I'm a newbie programmer with my first job out of college. I'm having trouble with a few assignments which require modifying 1000-1500 line long SQL stored procedures which perform data export for a vendor. They do a lot, they handle dispatching emails conditional on error/success, crunching data, and enforcing data integrity. It doesn't do these things in steps but through multiple passes with patches/updates sprinkled in as needed (I think: big ball of mud pattern).
Anyways, working on these has been difficult. First off, I can't just "run the procedure" to test it since there are a lot of side-effects (triggers, table writes, emails) and temporal dependencies. Later parts of the code will rely on an update make 400 lines ago, which itself relies on a change made 200 lines before that, which itself relies on some scheduled task to clean the data and put it in the right format (this is a real example, and there are a lot of them). I try to break it down for testing and conceptual simplicity, but by the time I do I'm not testing the code but a heavily mutilated version of it.
Anyways, does anyone have advice for being able to conceptually model and change this kind of code? I want to avoid risk but there is no documentation and many bugs are relied upon (and often the comments will lie/mislead). Any advice, any tools, any kind of mental model I can use when working with code like this would be very useful! My instinct is to break it up into smaller functions with clearer separation (e.g.; get the export population, then add extra fields, then validate it, etc. all in separate functions) but the single developer of all of this code and my boss is against it. So the answer cannot be "rewrite it".
LogicRaven_@reddit
A 1000 lines stored procedure should be broken into smaller pieces. Why do they think it is not neccesary?
As a junior dev, I would suggest to be carefull with this discussion. Suggest ideas, but disagree and commit if they push back.
You could ask the other dev about how they debug and for a walkthrough of the code.
Try to adapt to the situation and casually look for other options if this place gets too difficult.
Good_Skirt2459@reddit
It was a solo dev shop for over a decade. So I assume he just has so much innate knowledge that it's conceptually not complex to him. He has a very specific way of doing things. For example I've been told when used set-based operations that I shouldn't try to be "clever" and that my code is more error prone since it operates on all the data (not sure why that's different than looping over all of the data). I try to push back but I sense frustration. Also stuff like doing caching everywhere (which is probably necessary for that imperative style in SQL but I will benchmark my stuff and it acceptably fast).
I've sat over his shoulder and worked through debugging with him, it's just a case of "read the code and change what needs to be changed then push to production and fix issues as they come up". It chugs along but users do show dissatisfaction with quality and counter-intuitive reporting. I have been privately told by others in the business they think things have gotten better since I started and that my reports are easier to understand, so I believe I could leverage that but also why even bother when many places seme to not require playing office politics to gain the ability to break 1500 line functions into parts. I say I'm a junior but I've been active in open source since early high school and I have a few years as a data analyst working at an organization with literally thousands of times more data, This is just my first official "developer" role (although it's really similar responsibilities tbh).
LogicRaven_@reddit
This is not exactly politics, because he is not trying to undermine you. He is the owner and creator of a code, and he has a firm opinion about it. That's not unheard of.
But the fact that users don't dare to speak up and endorse the new way of reporting shows that something is off at this place.
You could try to carefully turn things around. Gain credits, have constructive discussions with him. Talk on a way that provides him an escape path without loosing face. Try to understand why the others don't speak up.
In the meantime, you could also look for other projects within this company or outside.
tinycockatoo@reddit
Does it make sense to write tests by hand and then handle development to AI tools to write features that pass them? I was thinking a little about ways to use AI tools effectively as someone who isn't very experienced
casualPlayerThink@reddit
You will learn to specify the task in great detail. Also, consider asking your tool (via a skill or sub-agent) to check your tests, criticise them, and extend them, or add new tests to ensure higher coverage.
Near it, expect to have a very bloated code base, with a lack of security, common sense, or practical usage, probably will eat multiple times the resources as it should, so always review it properly.
tinycockatoo@reddit
Are you being sarcastic? I'm sorry, I don't pick up on it very well and English isn't my first language. I'm not sure if you're saying it's okay, I guess not? Thanks for answering!
BlackHumor@reddit
Yes, that definitely sounds like a plausible way you could do it.
I would note that in the process of writing the tests you will also be coming up with the most important parts of the implementation as well, in most cases.
tinycockatoo@reddit
Thanks, I will try it out!
EnderMB@reddit
I'm currently looking at the best way to handle a task to crawl a huge number of internal pages, and to run a local set of ML models and LLM's against that page to determine if it passes or fails a set of arbitrary policies (think NSFW, but more nuanced). We'll want to do this weekly/biweekly, and we'll want it cheap.
My initial thought was that there's no reason this should be an online system and to use something like a Spark cluster. Spin up a bunch of workers, split the pages across 20-100 of them, and write a set of parquet files with the results. We could even maybe use a single script to download the pages and check them completely offline.
One of our principal engineers said this was too complicated, and that we may as well just throw it at Bedrock and (maybe) Comprehend, with a cache for pages that rarely change.
This sounds far more pricey to me, but for something I wanted to throw together in a week (which we have fine-tuned models for already) I am wondering if there is benefit to just using an AWS service here.
What would you go with, or would you do this in another way?
casualPlayerThink@reddit
No matter what, the data is online, so you will never have a truly offline system.
Sounds good, until you have to add exceptions, other sources, and take care auth and other quirks. This could become a nightmare quite quickly.
Sounds typical leadership @ startup kind of decision: don't wanna think about it, do not want to manage anything, just throw the data into a managed service and pay for it. Easy-peasy, even though with that, you will train other GPTs as well, charged greatly for something that could be done in-house.
Also, people underestimate young titans, who don't know if something is not possible, and they solve things in a way you never thought of.
I like your enthusiasm, but I think you undervalue it (or maybe I am wrong)
You should put it together, then you can compare it with paid services. You might already have a business idea, model, and MVP that you can make money from. (If so, I ask for 0.1% of profit :D ). By comparison, you should compare the finetune time, accuracy, outage percentages, results, speed, consumed power (e.g. costs) and maintainability.
EnderMB@reddit
You might be surprised to find this is Big Tech, rather than a startup. If anything, my startup experiences were more to the tune of "I don't care, just do it".
You're right on it still being an online system, but we can be constrained on it being a simple public capture, just by the nature of the pages themselves. It's unlikely we'll ever hit auth problems, and for specific instances where we'd have to worry about exceptions if we're at least acknowledging errors somewhere (even as simple as "we hit a 500 on x pages") we'd probably be fine if we're running on a cadence.
In terms of cost, I just don't see how throwing some HTML at some offline models in a rented cluster on AWS would cost more than just going straight to Bedrock + the infra you'd need to build around it. For maintainability, I am concerned about both, but it feels like the concern on Spark is less to do with it being the best solution and more to do with it being "different" to the usual lambda/step function flow we have.
casualPlayerThink@reddit
Why don't you measure both, and present the findings for the lead? Might be just need some hard fact to make better deicions.
EnderMB@reddit
That was my plan, but this was shot down as "taking too much time". Alongside this, Bedrock use is fairly easy to come up with costs for, whereas the costs for Spark clusters vary depending on how you set them up, what you need to plug in to your workers, etc.
mdt516@reddit
Hello all. I’m a 22 year old CS student in college. I started in high school in an AP class learning Java so I remember browsing stack overflow with a hope someone had answered my question. I’m lucky to say I really enjoy what I’m doing and I like programming in my free time just because it’s fun for me to learn about. I saw a post saying something about how one of y’all was disappointed that a newly hired junior didn’t know what a stack trace was (I’m assuming you mean call stack please correct me if I’m wrong). This reminded me: what skills do you think are really essential for developers that isn’t taught in a degree plan? For example (until I had a course that used GitHub classroom) I had never needed to use git for a CS class. I had already learned how to use it (and thank god I did I love git), but I wasn’t taught that the same way I was taught about linked lists. Obviously the way of problem solving and thinking is good. I’ve been going through leetcode to try and “work out” that problem solving process. I mostly know C++ since my courses have used that or Python. I know there’s a lot of AI hype right now and honestly I kinda wish we could turn back time to before it happened. I miss deeply understanding my code so I started going back to reading docs and just looking up what I don’t know. The only real application of AI that I feel like actually serves me is asking for help when I don’t yet have the vocabulary to really describe a problem, like when debugging multithreaded apps in c++ (also if anyone has any advice on that please I’d love that too)
casualPlayerThink@reddit
This differs from company-to-company and from leader to leader, as to what is focused on, usually based on their needs. The generic answer should be: the basics. You should at least hear about base concepts and should have a limited experience with some super basics stuff, and know how to read the given language-related reference documents. Everything else is just plus. You are supposed to learn them by doing.
> ...AI that I feel like actually serves me is asking for help when I don’t yet have the vocabulary to really describe a problem...
So, at the moment, the situation is that, the new wave of LLM/GPT will aim to being "orchestrated", so many provider start to sell CLI tools and agents, that can work behind the scene as well not just one but dozens of agents doing predefined jobs (thanks to ai documentations, skills, and other context shaping stuff). So right now, yes, using it to brainstorm is good, also you can ask to explain concepts, or read through a large amount of code or logs, and help to debug typical issues. Do not trust the output, always review it, and expect very verbose, over-complicated solutions that can be done with 1/3 of code with better perf, but still, could be a good MVP for a problem.
> don’t yet have the vocabulary to really describe a problem
Thats the art part. Right now, you should learn how to describe the problem and have the basic words for the given problem. If you don't know, it will lead to false solutions or extra credits. Use your native language in a normal discussion. You can even state for an LLM, you have no idea what this stuff is called, but this is the environment, scenario, input and output data, some logs, and some code, then ask to clarify and name the technologies/syntaxes/namings, etc. You aren't a codex; it is okay to not remember names, algorithms, or a given language's syntax or quirks. When I am switching between C++/Arduino/Python/JavaScript/PHP, I am ften check the references, because I have no clue how they are designed under the given interpreter/toolchain version (I am switching often between 3 different versions of each, dueto microservices and different projects... life of a consultant ;) )
mdt516@reddit
Thank you so much for the comprehensive response! I appreciate your advice. It’s easy to doubt myself in this period of my life so I appreciate you emphasizing the idea of learning being ok. I hope you have a wonderful Easter with your family, and if you don’t celebrate I hope you have a nice day. I appreciate your time!
bobbinssobbin@reddit
At what point in your career do you begin to feel secure?
BlackHumor@reddit
Financially, after I had a few months of expenses saved up in the bank. Which I would say was about a year in.
Professionally, the first time I took prod down, because I was able to recover quickly and not panic.
boomer1204@reddit
Moneywise - immediately I almost doubled my highest salary every 38k a year to 72k a year
imposter syndrome wise - about 1.5 years. NOW a lot of this is my fault. I didn't ask enough questions, didn't ask for enough help cuz I didn't want them to know how inexperienced I was (and this team would have embraced that and helped me alot). I only bring this up cuz year 3 when we brought on a new Jr and always had his notebook taking notes and asking for help and questions and about 6 months in the guy was smoking hot at work
job security - never. i'm going on 7 years at my second job and I have upgraded a couple of things in my life but big purchases that aren't necessary do not happen that frequently
diablo1128@reddit
Secure in what way?
Job wise you can always get fired at any time if you are in the USA. You just don't worry about it. Keep up your skills, learn at work, and generally you should be able to find a new job.
In terms of money, that's a personal thing and based on the lifestyle you want to live. I make 110K in a MCOL area with 15 YOE and I'm living just fine with money to spare every paycheck.
--AL3X--@reddit
I feel insecure about asking to clarify. It's normal?
BlackHumor@reddit
It's normal and also important.
i-eat-cement99@reddit
Better to feel insecure doing it than actually mess things up because you did not get it
krom_pir@reddit
Yes it's normal, but the right people will respect you for that.
idontevenknowwhats@reddit
Yeah, it’s fine
takemetogreenwich@reddit
Are there any devs left who use no LLM tools for their work at all?
LogicRaven_@reddit
I would think there are less and less devs not using LLM tools.
There could be scenarios where the LLM bubble pops, pricing gets closer to the cost and some use cases become economically not feasible.
But in general, everyone should experiment with and learn from others about how to use LLM.
kitatsune@reddit
How can I better manage/tolerate those more junior than me?
I currently have 3 years of experience and I have a colleague who has a little more than 1. This is both our first job out of school, and we are now currently on the same project.
I have noticed that she has little to no confidence in her abilities as a developer and doubts every single one of her actions. She has begun to ask for my confirmation on absolutely everything she is doing. She trusts me and others, but she doesn't trust herself. This behavior is really starting to annoy me.
I've tried to be patient and help her, but she's not learning anything. Or rather not at the pace desired. She can't debug effectively, cannot fully work independently, nor can she fully explain her code changes. There is another junior newer than her that's already learning much faster. This new junior also seems a lot more motivated.
My behavior towards her I admit has been demotivating. I am getting annoyed with her incessant questions and asking for my confirmation, and I don't always sugarcoat it. I've made a mistake recently where I was too blunt with her and she took it personally and began apologizing to me. I don't want that to happen again.
What can I do to make this situation better? To make her have more self-confidence in her skills? How can I be more tolerant and patient with a slow learner? How can I 'level up' someone who is relatively underperforming?
Would this also be an issue to bring up with my manager or my team lead? I don't think this is something I can shoulder myself.
mr_sudo@reddit
I'm working on legacy codebase, and I'm losing my interest in software development. It's the health tech code base, and we don't even know how the business logic works. We are using Claude to understand the logic but it doesn't cover all business logics. It's pain whenever I touch that codebase and stressed out if irrelevant things break after deploy in one of customers.
positivelymonkey@reddit
A couple of layoffs and learned apathy.
NickW1343@reddit
I feel like at some point you realize you've left legacy code behind at other jobs just like the people who wrote the current legacy code you're looking at have. Once that hits, then it started to feel like hating on people who wrote the legacy code was almost an indictment against myself, which made me feel bad and so now I avoid that by not minding legacy stuff as much.
casualPlayerThink@reddit
Instead of vibecoding - which exposes the trade secrets already - you can actually ask for the business logics. Make diagrams about the data flow, then ask for clarifications. Communication is golden.
Also, there are no senior/lead/CTOs who know these? If so, then the problem is bigger than the codebase itself...
Is there no senior/lead/CTO who knows
Constant imposter syndrome, roller coaster (have/love the industry), and lingering burnout. Never fun to clean up after others, who tend to be dumber than you might think at first glance.
Please also define unit tests, e2e, and integration tests. Define testing scenarios for QA. Have a staging/demo/dev server where everything is validated before release. Have backups, be able to restore those backups. Have a disaster recovery plan (rollback, migrations, etc).
mr_sudo@reddit
Unfortunately, noone knows all details. We even surprised when customer reports the issue, and noone knows that workflow. I think the people who know that already left company.
casualPlayerThink@reddit
Oh, a black box situation. Then I am doubling down on generic stuff that you can do: document the flows for yourself, write tests and QA scenarios, and enforce some kind of documentation over it (and/or tickets). Also, it might be worth considering rewriting the whole thing into testable, smaller chunks.
Freerrz@reddit
It definitely always sucks no matter what. What i recommend (if you don’t already) is to make unit tests as you work on tickets. It’s more work but being able to test things as you make changes will make things so much more relaxing once there’s a lot of coverage.
baked_doge@reddit
Hello! I'm a software developer with 2 years of professional experience. My team doesn't do a lot of merge reviews. I'm now heading the development of a small CLI tool for documentation and currently working with a junior we just added to the project. I've got a lot of tasks on our large client projects, hence they are doing nearly all the coding and I'm mostly reviewing merge requests and guiding the development (which features, broad setup...).
My question: do you have advice for merge requests? I want to avoid being too nitpicky, but I also see things I would consider important.
Very basic example: checking if a user input is a path or a wildcard expansion. The current code checks for the existence of * or a ?. I feel they should've used the python glob.is_magic function or at least used the correct regex:
([*?[]]).Am I being anal?
SirClueless@reddit
My advice is to come up with a concrete way the suggestion makes things better, and if you can't come up with one don't bother making the suggestion. Don't give comments on arbitrary style choices.
For your concrete example: There is a clear way in which the current code can be improved on real input so it's a worthwhile comment. For example: "This code doesn't handle input with character ranges, like '[a-z][0-9].txt'. There is an undocumented function glob.has_magic() you could use, or you could borrow the regex
'([*?[])'that it uses." (By the way, the regex in your reddit comment is not correct as there is an extra])baked_doge@reddit
Thank you, this makes sense! Have a great day
Key-Interaction-4797@reddit
I dont know what to learn anymore. What i wanna do is programming that will get me money that will help me massively. I undertsand i have to put in work but i have other responsibilities and yet am willing to put in 3 hours everyday to learn. I see everywhere that i need to put in 6 months to be entry level but thats abit long, is there any rewarding tech stack i could learn where after 3 months i could atleast put in money (dont mean to stop learning as this industry keeps changing)... and keep growing? even if its the security side its fine, i see most only focus on building and thats what almost everyone talks about. Not web dev pls, the market is saturated and full stack is in demand which will take me a long while to learn (learning is not the issue, i want something that will put in money after 3 month of learning)
No_Comedian7332@reddit
Has anyone had any good experience using one of those AI apply tools to find jobs? Like https://aiapply.co/ ? Is it actually useful? I'm looking for a new job, but as everyone knows, the market rn is terrible.
idontevenknowwhats@reddit
No don’t fall for it
No_Comedian7332@reddit
Thanks! I had the gut feeling they were not good, but yeah. Have you tried it?
114sbavert@reddit
How do I make my work count? I find that the kind of impact I make is very important, and my technical manager appreciates them but my product managers don't notice them. Building an aho-corasick based system to replace linear search, creating CI jobs to enforce code quality standards and outdated package checks, adding strict type validation instead of using string everywhere (like some others in my team had been doing before me), creating an automated logging system with granular Logging control over the previous tools, these things aren't visible to product managers. How do I make these kinds of contributions count? I am worried my impact isn't felt and I may get included in an inevitable layoff round.
rhd_live@reddit
Which of those are important?
Automated logging: who does it help, how does it help your product?
Linters: is this actually helping or limiting velocity with flagging capitalization and code comments that should end with periods?
Replace linear search: has this improved your product/product behavior in a meaningful way?
The most important job is to empower our products in terms of reliability, capability, and scalability. If what you’re working on doesn’t or only very slightly improves this, it’s not that important and you’re kind of wasting your time
positivelymonkey@reddit
Those contributions don't count and get you labelled an engineer's engineer.
Do them for fun as part of visible work. Don't talk about it.
roodammy44@reddit
Be careful about these fears. This exact situation led me to burnout recently, and it didn’t matter how much of an impact I had anyway because 2/3 of the company got laid off anyway (catching me in it). Your role has a far bigger say than your work when it comes to layoffs, so my advice is not to care about your impact for your own mental health.
Doctuh@reddit
You are looking at your "impact" through an engineering lens. So yes a technical manager may appreciate it. Product people generally think in terms of the actual product. And, no, not the stuff that long-term keeps the codebase healthy.
If you spend 80% on your time on infastructure and 20% of time on the actual feature you will be beaten on visibility every singly day by that other son-of-a-bitch one cube over who has the reverse ratios.
You know, the guy who's bugs you keep fixing?
He is product impactful and visible. You are not. Sorry.
If you have a good manager and they understand the true need of your work then you may be ok. If you do not, then you will not be. Also sorry.
Ashken@reddit
Yeah I find myself in this position. And unfortunately the inverse is also true.
Because I have so much system level knowledge and I’m see as not being impactful. But if I have a sick day or take time off there’s contention because I’m not around to hold anybody’s hand. I’m like “there’s my impact”. Lol
vanit@reddit
And make sure you get close with that product person that the CEO is always bugging, because your stuff will get noticed.
114sbavert@reddit
yes that's exactly what's happening right now btw
ericmutta@reddit
> How do I make these kinds of contributions count?
Start by determining whether those contributions count in the eyes of the people you want to notice you. You can do all the right things and still fail if those were not the things that matter most to the people who have the power to lay you off.
If you can, find something that really irritates your customers, work back to the find the cause and fix that. If customers notice, your product managers will too.
IronWombat15@reddit
I found myself in a similar position previously. The trick is often finding another employer who values these things. It's hard to get internal recognition for these sorts of projects, but interviewed LOVE them.
Loosely speaking, it should be your manager's problem to get you promoted. If you're not currently aligned, you likely either need to align or move on.
For getting recognition for your existing efforts, I find the main things are measurement and visibility. How many bugs did CI prevent from reaching production in the last 90 days? How much compute savings or latency reduction did search optimization bring? Being able to send a newsletter style email to all of ~engineering with clear and compelling metrics demonstrating recent wins is a great way to shine a spotlight on your (or ideally, your team's) contributions.
114sbavert@reddit
I would have no way to measure the number of bugs prevented by my CI pipelines because usually I'm not looking at which MR is getting resolved and merged by my manager at which time. Also, for some reason even with all checks and balances, my manager ends up giving force push access to some guys while trusting they won't break production.
As for the search one, it's currently awaiting production deployment because of other things of higher priority, but I'm sure my tech manager would love it.
IronWombat15@reddit
Measurement is often hard, but worth doing. Have metrics on how many MRs we're blocked by (correctly-failing) CI failures. Maybe keep note of a few handpicked examples of major errors that would have slipped through if not for CI. (Or examples of where skipping CI led to a preventable breakage.)
If you don't have any measurements, then your impact is assessed solely on feelings and competitive storytelling.
114sbavert@reddit
Can you please explain what you mean by "aligned'?
Did you mean interviewers? Could you please expand on this?
IronWombat15@reddit
Alignment mostly meaning that you're working on the things your manager (or their manager) place priority on. If your manager doesn't see the value in CI, you either need to convince them (e.g. with compelling metrics), or prioritize what the business cares about.
Yes, "interviewers." Being able to say in interviews that your team had no CI, and you spearheaded the effort to add it shows that you place value in modern dev practices and that you had the skill/initiative/leadership to implement them in a semi-hostile environment. "If they care that much about code quality and system improvement, I want them working here!"
SaVaGe19765@reddit
4th-semester CS student from Afghanistan here. I got an internship opportunity with Afghan Wireless Communication Company (AWCC)—a major telecom provider in mobile, internet, and enterprise services.
I’d love advice on:
Thanks in advance for your guidance!
LogicRaven_@reddit
What is the other option, not doing anything?
Experience is good. Use all your chances to learn practical things.
sorrymylovee@reddit
I am looking for a developer to build a custom PvP Target-Sharing Tool for a Silkroad Online private server. The server is protected by Maxiguard, so the solution needs to be stealthy and bypass signature/packet checks.
The Vision: Game disabled the target support option, it was available before where you just select a character then press a hot key and it gets you the target selected by this character, so in fights you just select the leader character then press that hotkey so all players attack the same chatacter.
Technical Requirements: DLL Injection pro to create it and run smoothly without issues, I've seen some players using it but it's very rare to find and they never share it.
Filter Bypass: The tool must work alongside Maxiguard. It should ideally be a DLL injected into the client or a very sophisticated external tool that mimics legitimate player input to avoid detection.
Initial Build: Willing to pay a significant flat fee for a working prototype.
Profit Sharing: If you can make this stable, i can promise to make you really rich by helping you selling this tool.
bvdevvv@reddit
If I get a rejection email 2 weeks after applying that says they're moving forward with candidates, whose background and experience are a better match for the role, is this an HR rejection or the hiring manager? Just trying to understand if I need to either tweak my resume or I don't fit the job requirements that I initially thought I did.
positivelymonkey@reddit
Try asking. Make it super clear you're open to feedback and not looking to reapply or change the decision in any way.
Usually they won't respond.
In my experience it has nothing to do with you and everything to do with who else they found.
Melodic-Flatworm-596@reddit
Hi there! I work at a company that is fairly large in scale, but is just now starting to believe in in-house software development (which is where I come in). I am a self-taught programmer, but am about to start my final semester in an online B.S. degree program in Cloud Computing specializing in AWS. This is a grass roots movement from the VP of my department, and I pretty much just report to him. Anyways, I'm basically a Junior dev without anyone working over me. I spend a lot of my time researching best practices and trying to make sure I get things in line with industry standards, but in a lot of ways, I work in a vacuum. I've written and maintain 2 Python projects ( 1 of which is bundled and packaged using pypackage, and I learned only later on that Python really was not the right language for the job, but it works well enough for what we need now—perhaps I'll rewrite it the future). Now that I see all of the blind spots I had when I first started I take great painse to make sure everything is best practice coding, follow PEP 8, document everything well, etc. etc. so that if this department ever expanded, other devs could jump right in and continue to maintain the codebase. What this means is I end up spending more time reading and teaching myself than I do coding.
Do you have any advice for a Junior dev in my situation? Should I continue to rough it out and self teach myself, which has the advantage of teaching me how to be a disciplined, self-reliable engineer, but with the disadvantage of very likely missing out on many teaching moments and knowledge blind spots that I would learn from working on an established software dev team under seniors?
Thank you in advance for any and all advice. Cheers!
Fit_Economist_3966@reddit
As an unexperienced webdev, are ai ide/ ai agents usually mandatory on big companies? (I am applying for Angular and Spring-Boot). Can I make a living coding on my own and asking Claude whatever I need or will I be forced to prompt to any LLM for coding?
This is a serious question because I am thinking on a job change if it is the case
boring_pants@reddit
Two answers here:
boomer1204@reddit
I think most will be. My previous company didn't use much cuz it was still newer and we were a startup and didn't have the funding for enterprise level accounts. That same CEO started another venture that a couple of my old coworkers work at and they use AI HEAVILY
My current job is a huge finanical institute and we use AI and are encouraged to use it. Also during my interviews (which I only had like 4) all asked if I was comfortable using it so I think it will be pretty common in a lot of roles
This is obviously gonna be company dependent but the trend definitely seems to be using LLM's are more common than not using them. And i'll say they are really good at a lot of the boilerplate stuff you get sick of typing out anyways so I'm a big fan of them personally
PancakeWithSyrupTrap@reddit
why does management sometimes retaliate against subordinates ? this is an extreme example: https://www.reddit.com/r/LinkedInLunatics/comments/1eyp3wf/ceo_seduced_wife/
this example is for a former VP of engineering, but it can happen at any level.
why can't leaders discuss ideas, even if they don't like them ? why retaliate ?
yeah I'm a fucking idiot.
OtaK_@reddit
Your example is linkedin bs. Never happened.
PancakeWithSyrupTrap@reddit
how do you know it didn’t happen ? not suggesting you’re wrong
OtaK_@reddit
Dude, it's linkedin. Also he's the former VP of Engineering of Amazon. If his CEO ever did that, that would mean Mr. Bezos is the person he talks about. Do you imagine the size of the libel lawsuit?
Ok_Finish_494@reddit
I'm interviewing at the moment, and just found out the place I'm interviewing for works in their own proprietary language and uses a non-text based input via a GUI - think filling in forms, instead of writing code.
This is going to be a terrible idea for getting jobs later, right? I am getting the ick personally, but don't have any other interviews lined up.
casualPlayerThink@reddit
You can make some money until you find something better.
[TL;dr]
This kind of GUI is mostly for teaching and for make complicated/tech-heavy parts easier. It will have an incredible amount of complexity and challenges. Companies made attempts hundreds of times every year and failed spectacularly (exceptions are electronics and tech teaching apps like Sketch for kids). To make such a tool viable, you not only have to understand fully what happens under the hood but also have to flawlessly translate with any given extra circumstances. While companies pursue a happy path, it seems viable when they face the hidden errors, human factors, random outages, and partial issues, then it will most likely crumble.
Not much option you have, so try to get the most out of it (e.g:. learn stuff you don't know, face challenges you didn't know you have)
nana_3@reddit
Bird in hand. No other offers, no harm in following it and keeping it short term if it goes anywhere.
Seems potentially limiting if you stay there long term - UNLESS it’s like a whole industry like that and it’s just how they do it when you’re programming systems for mining equipment or something idk.
PancakeWithSyrupTrap@reddit
why doesn't management take criticism well ?
an EM made some code changes which resulted in an outage. I had to clean the mess up.
our process requires writing a report explaining what happened. I called out the EM. upper management was not happy about this and asked me to change the report so it doesn't point to the EM. for example instead of saying "EM made a code change ..." change to "a code change was made ..."
why defend incompetence to this extent ?
yeah I'm fucking stupid. is it likely I'm top of the layoff list now ?
114sbavert@reddit
Can someone please review my Resume: https://jumpshare.com/s/1XdvSTLCmOhDaNq44hnr
I want feedback specifically on two things: