/job/ - Job Board

the gap between entry-level and senior management is widening. recruiters are mostly looking for people who can hit the ground running with zero training required. it feels like companies are cutting out the middle layer to save on overhead, leaving junior talent in a very difficult position .
>no more mentorship programs
instead, we just see much higher expectations for immediate impact on every single task. many roles that used to be standard mid-level positions are now being rebranded as senior specialist roles to justify the workload. it is becoming harder to find a natural stepping stone in a career path

im heading to plano next week for the commit your code lineup and realized i need a real plan. instead of just wandering around, im trying to focus on one specific goal like meeting two new devs. it's way better than just grabbing free swag and makes the whole networking grind feel less random intentional. does anyone else go to these events w/ a strict checklist?

https://dev.to/amandamayfield/the-art-of-intentional-networking-at-tech-conferences-39n7

just realized how easy it is to leave old iam access keys active forever. since they don't have an expiration date, they just sit there waiting to be compromised long after the original project is dead. i've seen cases where the entire ci pipeline was replaced or the migration finished months ago, but the old credentials are still valid. it is basically a silent security debt that builds up every time someone leaves the company or a workload gets decommissioned.
>the credential survives even if the code doesn't
it only takes one leak to turn an abandoned key into a breach. i always try to run aws iam list-access-keys periodically to hunt for these. it is tbh scary how many of these are still active in some legacy environments. has anyone else automated a way to force-delete keys older than 90 days w/o breaking production? i am trying to avoid the manual cleanup nightmare next time.

found this here: https://dev.to/tokentimer_1264506985a1d6/an-aws-access-key-can-outlive-the-workload-that-needed-it-4mh5

just found a solid way to manage notifications by storing the next due instant in UTC and using a short dispatcher to feed a worker pool. it keeps everything aligned during dst shifts as long as you treat local calendar time as pure schedule data rather than relying on the delivery system itself but don't forget to rate-limit that queue or your downstream services will melt . anyone else using a separate checkpoint layer for this?

full read: https://dev.to/nikitachristensen2691/rate-limited-worker-queues-for-timezone-aware-user-reminders-in-nodejs-34nl

just found this piece about how we way too much importance on being busy. it traces back to when everything was slow, like using pagers and fax machines for everything. now that the internet is basically integrated into our nervous systems, we just expect more output 24/7. it makes me think we've traded the beauty of inefficiency for a constant state of burnout. maybe being "slow" was actually the secret to not losing your mind. do you guys feel like the ability to work from anywhere has just made us permanently tethered to our desks?

link: https://dev.to/isurogi/we-saved-time-then-filled-it-up-22pb

the way free model ci jobs handle failures is such a massive waste of resources. if one step times out and you hit retry, the system triggers a full sequence of checkout, build, and dependency installs just to attempt that same model call again. it feels like re-running the whole world for a tiny error. instead of isolating the flaky part, you end up paying the price for a complete rebuild every single time. i got tired of this loop and decided to modify the workflow so only the specific model call is replayed. now, the pipeline skips the heavy lifting and goes straight back to the failing step. it basically turns a massive headache into a simple one-step retry . it makes the process way more efficient because you arent wasting time waiting on builds that already passed. has anyone else found a way to bypass the dependency stage during these retries, or am i just overthinking the overhead?

found this here: https://dev.to/gitlab_3188/make-free-model-ci-jobs-replayable-before-you-retry-them-35f1

use

display: grid; place-items: center;

to avoid the old way of using margins. it makes responsive layouts much easier when managing nested flex containers .

found this list of 15 questions u can use during interviews to sniff out bad leadership before u sign anything. it even pulls in some useful tips from the glassdoor community to help you spot those red flags. don't let them gaslight you with vague answers has anyone else found that asking about turnover rates is a total game changer?

https://www.glassdoor.com/blog/spot-a-toxic-boss-in-an-interview/

fr stop treating sms timeouts as failures because that just triggers a loop of duplicate resets. instead, you should save the idempotency key and expiry before dispatching so the worker can reconcile the mess later. it's basically about keeping an audit trail for compliance rather than just guessing trusting the retry logic. does anyone else use a specific lock_id strategy to prevent these double-sends?

link: https://dev.to/liamfoster1844/preventing-duplicate-password-reset-notifications-under-sms-timeout-and-retry-pressure-50d6

fr just found a breakdown comparing vllm, sglang, and llama. cpp using an rtx pro 6000. the dev even included an FP8 pass on the top performer to see if the speedup holds up. it's pretty wild how much difference the serving stack makes when you have raw csvs and a repro script to verify everything. does anyone else think llama. cpp is still too slow for production workloads like this?

https://dev.to/conatusai/qwen3-8b-on-workstation-blackwell-vllm-vs-sglang-vs-llamacpp-plus-an-fp8-pass-325c