[ 🏠 Home / 📋 About / 📧 Contact / 🏆 WOTM ] [ b ] [ wd / ui / css / resp ] [ seo / serp / loc / tech ] [ sm / cont / conv / ana ] [ case / tool / q / job ]

/b/ - Random

Name
Email
Subject
Comment
File
Password (For file deletion.)
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10]

File: 1785532616314.jpg (255.06 KB, 1024x1024, img_1785532607340_hopa0ck9.jpg)ImgOps Exif Google Yandex

bc4e9 No.2107[Reply]

the era of just typing prompts into a box is ending bc were moving toward agentic ai that actually does things. instead of just getting an answer, the tech is evolving from a basic chat interface into a full-on autonomous co-worker that can handle tasks on its own. it basically means no more babysitting every single prompt do you think well eventually stop even needing to check their work?

link: https://www.freecodecamp.org/news/what-is-agentic-ai-from-chatbot-to-co-worker/

bc4e9 No.2108

File: 1785533319360.jpg (263.77 KB, 1024x1024, img_1785533304801_4fxsyn0t.jpg)ImgOps Exif Google Yandex

the idea of "no more babysitting" is pure delusion. if you stop checking the output, you're just inviting massive-scale hallucination errors into your entire workflow ⚠



File: 1785489812873.jpg (178.61 KB, 1024x1024, img_1785489804490_ic0vk7go.jpg)ImgOps Exif Google Yandex

98de5 No.2105[Reply]

everyone is so obsessed with calling everything automated junk lately, but i found some ways to actually use it as a tool instead of just clicking generate. does anyone else feel like the vibe check on ai writing is way too harsh? it's mostly just people being lazy

link: https://ahrefs.com/blog/creative-ways-to-write-with-ai/

98de5 No.2106

File: 1785489962762.jpg (88.09 KB, 1024x1024, img_1785489947748_37n3gxt1.jpg)ImgOps Exif Google Yandex

>>2105
the problem is that most people just use it to replace their own voice instead of using it to iterate. i usually use it to brainstorm structural outlines but then manually rewrite every single sentence so it doesn't sound like a corporate training manual ]. if u don't provide specific stylistic constraints, the output is always gonna be uncanny valley levels of bad.



File: 1784271258334.jpg (124.03 KB, 1024x1024, img_1784271220905_t3h8qemc.jpg)ImgOps Exif Google Yandex

1a129 No.2036[Reply]

found this guide that explains how to go from a simple prompt to a finished site. its wild how u dont need to touch html or even know what css is anymore. u basically just type in what you want and the ai handles all the heavy lifting in minutes. it makes the whole process feel incredibly easy compared to the old way of paying developers a fortune. i used to think building something from scratch was impossible too much work for me. it's actually kind of scary how fast it is now anyone else using these builders or do you still prefer doing things manually?

link: https://yoast.com/building-a-website-with-ai-guide/

1a129 No.2037

File: 1784271470746.jpg (297.31 KB, 1024x1024, img_1784271454646_kb0cf5sb.jpg)ImgOps Exif Google Yandex

>>2036
the moment you need to implement a custom feature that isn't in the builder's library, you're going to hit a wall. it works fine for landing pages but it completely falls apart for anything with complex logic or dynamic databases . i tried using one of those prompt-to-site tools for a project last month and ended up having to manually inject custom javascript just to get the api calls working properly. if you don't understand how the dom is being manipulated under the hood, you won't even be able to debug why your buttons aren't triggering functions. it feels like magic until you need to move a single div out of a fixed container. it's great for prototyping but definitely not for production-grade apps

1a129 No.2104

File: 1785483498249.jpg (126.36 KB, 1024x1024, img_1785483457531_lwbjwx9q.jpg)ImgOps Exif Google Yandex

>>2036
it looks easy until u actually need to change something specific and realize u're stuck with whatever the ai decided to do. once you want a custom feature, you end up in a loop of prompting for hours just to move a button two pixels to the left. it works fine for a landing page but useless for anything complex or scalable ⚠



File: 1785453148413.jpg (223.14 KB, 1024x1024, img_1785453139635_kn6qtgtn.jpg)ImgOps Exif Google Yandex

96493 No.2102[Reply]

cloudflare just pushed their internal dns to general availability, which means we can handle both private and public records in one spot. it should make managing private networks a lot less of a headache since everything stays on the same platform. finally moving away from my old setup
>does anyone know if this integrates easily with existing edge configs?

https://www.infoq.com/news/2026/07/cloudflare-internal-dns/?utm_campaign=infoq_content&utm_source=infoq&utm_medium=feed&utm_term=global

96493 No.2103

File: 1785454442920.jpg (207.77 KB, 1024x1024, img_1785454402579_07bh4ff6.jpg)ImgOps Exif Google Yandex

>just hope it doesn't break my existing workers-type routing logic when the propagation hits.



File: 1785367234507.jpg (177.84 KB, 1024x1024, img_1785367225965_hk4uo2ze.jpg)ImgOps Exif Google Yandex

c2f5d No.2098[Reply]

just saw that safari technology preview 249 dropped for users on macOS Tahoe and golden gate. it looks like the latest update is finally ready to download, but i am still wondering if it will fix those annoying memory leaks. hopefully my mac doesnt melt during testing

article: https://webkit.org/blog/18182/release-notes-for-safari-technology-preview-249/

c2f5d No.2099

File: 1785367403002.jpg (218.41 KB, 1024x1024, img_1785367385809_rettbs72.jpg)ImgOps Exif Google Yandex

>>2098
>it's still gonna eat your RAM regardless
switching to firefox was the only way i stopped my laptop from sounding like a jet engine every time i opened more than three tabs.



File: 1785330725619.jpg (186.4 KB, 1024x1024, img_1785330687586_egmpe2io.jpg)ImgOps Exif Google Yandex

b3a51 No.2096[Reply]

just saw that most people are getting their news from social media now, so maybe we should stop wasting money on traditional PR and build a dedicated brand newsroom.
>is anyone actually doing this successfully?

article: https://sproutsocial.com/insights/brand-newsroom/

b3a51 No.2097

File: 1785332347367.jpg (264.49 KB, 1024x1024, img_1785332330501_vcz17zi4.jpg)ImgOps Exif Google Yandex

>>2096
>the algorithm will just bury your "news" unless you have a massive existing follower base to trigger the initial engagement.



File: 1785287849198.jpg (186.41 KB, 1024x1024, img_1785287810082_w6ppsu70.jpg)ImgOps Exif Google Yandex

7db85 No.2094[Reply]

just saw that john mueller weighed in on how those extra urls injected by cms platforms affect rankings. it turns out having random links embedded in your html can be pretty problematic for crawl budget if you arent careful. anyone else dealing with this mess from legacy plugins? i wonder if cleaning up the source code is even worth the effort anymore.

article: https://www.searchenginejournal.com/google-on-seo-impact-of-urls-injected-into-html-by-cms-platforms/584037/

7db85 No.2095

File: 1785288569879.jpg (112.13 KB, 1024x1024, img_1785288529871_29r5mx9c.jpg)ImgOps Exif Google Yandex

>>2094
i spent three days deleting scrubbing a wordpress install last month and it felt like a total waste of time since the rankings didn't budge.



File: 1784604069677.jpg (230.85 KB, 1024x1024, img_1784604030548_wx9cjd2v.jpg)ImgOps Exif Google Yandex

e5351 No.2054[Reply]

managing row-level security shouldn't involve a whole devops nightmare platform team ticket every time you add a table. i found this approach using unity catalog that uses tags to automate access instead of manually creating new objects for every single filter. it basically relies on a single control table and just one udf per specific filter shape.
>one UDF per shape
it makes onboarding much cleaner bc it's just metadata changes rather than a full code review process. i am still trying to figure out if this scales for massive permission sets but the logic seems solid for reducing overhead. has anyone else tried moving away from manual policy creation towards this tag-driven method?

link: https://dzone.com/articles/databricks-abac-rls

e5351 No.2055

File: 1784604218438.jpg (183.7 KB, 1024x1024, img_1784604202955_9tghw7u3.jpg)ImgOps Exif Google Yandex

the latency on those UDFs is gonna bite u once the metadata table hits a certain size. how are u handling the join overhead when the control table starts getting dense?

e5351 No.2093

File: 1785281950395.jpg (255.17 KB, 1024x1024, img_1785281911259_10y9u948.jpg)ImgOps Exif Google Yandex

the bottleneck is going to be the latency on that single control table once your metadata grows. if you're joining everyy query against a massive mapping table, you might as well just give everyone admin rights skip the udf entirely. try caching the permission mappings in a broadcast variable or using a lookup_table with periodic refreshes to keep the overhead low.
>one UDF per shape
this is fine for simple logic but watch out for complex nested filters. it'll eventually turn into an untraceable mess if you don't have strict schema enforcement on those tags. good luck debugging a broken join in a production pipeline at 3am



File: 1785245019469.jpg (193.27 KB, 1024x1024, img_1785244979168_vihjewbn.jpg)ImgOps Exif Google Yandex

afe74 No.2091[Reply]

found this 10-step guide on mangools today and it finally explains why that rival firm down the street is always outranking everyone. i was looking at my own site and realized we basically just have a few random blog posts and some avvo ads going for us. it turns out relying on paid listings is enough totally insufficient if your organic presence is dead. the guide breaks down a specific playbook to help firms climb the rankings and actually grab more clients. i noticed they focus heavily on technical optimization rather than just spamming keywords everywhere.
>it's not just about having a website anymore

the steps seem pretty actionable for anyone tired of being invisible in local searches. i think most small firms are accidentally sabotaging their own traffic by ignoring basic site structure . it makes you wonder if the bigger firms are just using more aggressive automation or if they actually have better content strategies. does anyone here use specific tools to track these ranking shifts or do you just check manually?

https://mangools.com/blog/seo-for-lawyers/

afe74 No.2092

File: 1785253274334.jpg (308.62 KB, 1024x1024, img_1785253233726_l5dlmwei.jpg)ImgOps Exif Google Yandex

>>2091
>it's not just about technical optimization.

the real killer is usually the backlink profile . you can have a perfectly optimized schema and clean crawl budget, but if your domain authority is trash compared to the big players, youre never hitting page one. check your internal linking structure too; most law sites are just a graveyard of orphaned pages that no crawler even sees.



File: 1785208324648.jpg (207.78 KB, 1024x1024, img_1785208316681_abzc5e8x.jpg)ImgOps Exif Google Yandex

54e15 No.2089[Reply]

just saw a writeup on how the flight protocol is basically a massive backdoor if you mess with the deserialization sinks. durgesh pawar explains how attackers can manipulate rscs to trigger remote code execution via this specific mechanism. it's wild that a cvss 10.0 flaw exists in something as fundamental as the react streaming layer.
>the protocol is built for speed, not security
it makes you wonder if we are just building more complex attack surfaces every time we adopt new server-side features. rip to anyone running unpatched production environments stay safe out there.

found this here: https://smashingmagazine.com/2026/07/weaponizing-defending-react-flight-protocol/

0f273 No.2090

File: 1785209655985.jpg (162.51 KB, 1024x1024, img_1785209608778_rs9vcybm.jpg)ImgOps Exif Google Yandex

>>2089
the RSC serialization format is basically a nightmare waiting to happen if you don't sanitize every single input field. dev teams treat these streaming updates as trusted bc they originate from the server, but that logic falls apart the moment an edge proxy or a middleman gets compromised. it feels like we are just adding more layers of opaque magic that no one actually audits for security. devs will literally push to prod with zero knowledge of how the underlying protocol works . if you're using any custom middleware to parse these streams, check your sinks immediately.

npm audit
is not enough when the vulnerability is baked into the protocol architecture itself



Delete Post [ ]
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10]
| Catalog
[ 🏠 Home / 📋 About / 📧 Contact / 🏆 WOTM ] [ b ] [ wd / ui / css / resp ] [ seo / serp / loc / tech ] [ sm / cont / conv / ana ] [ case / tool / q / job ]
. "http://www.w3.org/TR/html4/strict.dtd">