17 messages
A place for non-work-related flimflam, faffing, hodge-podge or jibber-jabber you’d prefer to keep out of more focused work-related channels.
Archive: https://archive.sweetops.com/random/
Erik Osterman (Cloud Posse)9 months ago
Didn't realize he retired: https://en.wikipedia.org/wiki/Kelsey_Hightower
Jessica9 months ago
Hi everyone,
Hope you are all well! I’m Jessica, field marketing specialist at Mirantis. We’re excited to be hosting a joint webinar with our partner Nirmata, creators of Kyverno on June 18. If you know anyone who might be interested in policy-driven multi-cluster Kubernetes management, please feel free to share this with them.
• Webinar title: Policy-Driven Multi-Cluster Kubernetes for AI Workloads Made Simple with k0rdent and Kyverno
• Date/time: Jun 18 2025, 9am PT/12pm ET/5pm BST
• Registration page: https://bit.ly/4dJPrf2
• Posts to share on social media
◦ X (Twitter)
◦ LinkedIn
Hope you are all well! I’m Jessica, field marketing specialist at Mirantis. We’re excited to be hosting a joint webinar with our partner Nirmata, creators of Kyverno on June 18. If you know anyone who might be interested in policy-driven multi-cluster Kubernetes management, please feel free to share this with them.
• Webinar title: Policy-Driven Multi-Cluster Kubernetes for AI Workloads Made Simple with k0rdent and Kyverno
• Date/time: Jun 18 2025, 9am PT/12pm ET/5pm BST
• Registration page: https://bit.ly/4dJPrf2
• Posts to share on social media
◦ X (Twitter)
Erik Osterman (Cloud Posse)9 months ago
So this past week I've been using Willow Voice on my Mac and it's been freaking awesome. In Cursor (editor), I kept looking for the missing microphone feature. Well, no more.
On my Mac, every time I'm in Slack or writing an email, it's so frustrating that there's no great dictation. Let's face it, Siri sucks ass.
Like a lot of you, I've been using just ChatGPT to transcribe, then copy and paste. Well, you don't need to do that anymore. I can just press a hot key and it'll perfectly transcribe what I'm saying and paste it right in—even with rich text formatting. Just map the function key or map some key on your keyboard and when pressed it'll transcribe it perfectly, just like this message I'm sending right now.
Get one month free with this link (and give me one too!) I've already paid and signed up our team.
https://willowvoice.com/?ref=EOSTERMAN6
On my Mac, every time I'm in Slack or writing an email, it's so frustrating that there's no great dictation. Let's face it, Siri sucks ass.
Like a lot of you, I've been using just ChatGPT to transcribe, then copy and paste. Well, you don't need to do that anymore. I can just press a hot key and it'll perfectly transcribe what I'm saying and paste it right in—even with rich text formatting. Just map the function key or map some key on your keyboard and when pressed it'll transcribe it perfectly, just like this message I'm sending right now.
Get one month free with this link (and give me one too!) I've already paid and signed up our team.
https://willowvoice.com/?ref=EOSTERMAN6
E
erik8 months ago
Can’t wait to try this
Awantika Nigam8 months ago
Hey folks 👋
Refact.ai Agent has just become the #1 open-source AI Agent on the SWE-bench Leaderboard! 🔥
• SWE-bench Verified → 70.4% solved - 500 real, production-grade issues
• SWE-bench Lite → 60% solved - 300 tasks
Read the full technical breakdown on our blog: https://refact.ai/blog/2025/open-source-sota-on-swe-bench-verified-refact-ai/
We also open-sourced our SWE-bench pipeline. Check it on GitHub: https://github.com/smallcloudai/refact-bench
Give Refact.ai Agent a try and tell me what you think 🙌
Refact.ai Agent has just become the #1 open-source AI Agent on the SWE-bench Leaderboard! 🔥
• SWE-bench Verified → 70.4% solved - 500 real, production-grade issues
• SWE-bench Lite → 60% solved - 300 tasks
Read the full technical breakdown on our blog: https://refact.ai/blog/2025/open-source-sota-on-swe-bench-verified-refact-ai/
We also open-sourced our SWE-bench pipeline. Check it on GitHub: https://github.com/smallcloudai/refact-bench
Give Refact.ai Agent a try and tell me what you think 🙌
paulm8 months ago
This is AWS-related but unofficial, so hopefully #random is right.
I'd like to share two cost-cutting tools.
💰️
• 🏷️📆⏱️ github.com/sqlxpert/lights-off-aws stops and starts EC2 instances and RDS/Aurora databases based on
• 🛑 github.com/sqlxpert/stay-stopped-aws-rds-aurora re-stops RDS and Aurora databases when AWS starts them after the 7-day maximum.
Confession: I use CloudFormation (because it's available to all AWS users, with nothing extra to set up). Each tool supports multi-region, multi-account installation as a StackSet and single-region, single-account installation as an ordinary stack. I provide references for wrapping CloudFormation in HCL, since most people here use Terraform.
Even if we don't manage EC2 instances directly anymore, lots of money is wasted on dev/test RDS and Aurora databases that sit idle all weekend. Lights Off is unusually flexible because schedules, not references to schedules, go in each resource's tags.
As for Stay-Stopped, you'll find alternatives all over re:Post, AWS blogs, and Medium, but they suffer from timing bugs/race conditions, bound to bite us eventually in large environments. The Stay-Stopped ReadMe ends with a perspective on the two best alternatives I found, plus AI-generated solutions from Amazon Q Developer.
This work is a labor of love. We heard in Office Hours yesterday that, according to the New York Times, programming is devolving into menial work as AI code generation spreads within Amazon. Stay-Stopped was an eye-opener because it showed that poor-quality AI training data (solutions on the Internet), plus Amazon Q Developer's poor awareness of AWS API documentation, can contribute to low-quality AI-generated code. There is no substitute for human understanding and effort.
Apart from the pleasure of learning, trying to make simple but reliable software, and documenting it so that others can learn from it, I enjoy hearing from users. I've worked on Lights Off and its predecessor since 2017 but I wrote Stay-Stopped recently, based on feedback from a user. Please try these tools and let me know how I can improve them. Thank you!
I'd like to share two cost-cutting tools.
💰️• 🏷️📆⏱️ github.com/sqlxpert/lights-off-aws stops and starts EC2 instances and RDS/Aurora databases based on
cron schedules in their tags. You can also tag these resources, plus EBS volumes, to trigger AWS Backup without writing backup plans. Last but not least, the tool can delete and re-create any resource specifiable in CloudFormation, based on schedules in stack tags. (My example deletes AWS Client VPN associations overnight, reducing VPN cost to $1 a day.)• 🛑 github.com/sqlxpert/stay-stopped-aws-rds-aurora re-stops RDS and Aurora databases when AWS starts them after the 7-day maximum.
Confession: I use CloudFormation (because it's available to all AWS users, with nothing extra to set up). Each tool supports multi-region, multi-account installation as a StackSet and single-region, single-account installation as an ordinary stack. I provide references for wrapping CloudFormation in HCL, since most people here use Terraform.
Even if we don't manage EC2 instances directly anymore, lots of money is wasted on dev/test RDS and Aurora databases that sit idle all weekend. Lights Off is unusually flexible because schedules, not references to schedules, go in each resource's tags.
As for Stay-Stopped, you'll find alternatives all over re:Post, AWS blogs, and Medium, but they suffer from timing bugs/race conditions, bound to bite us eventually in large environments. The Stay-Stopped ReadMe ends with a perspective on the two best alternatives I found, plus AI-generated solutions from Amazon Q Developer.
This work is a labor of love. We heard in Office Hours yesterday that, according to the New York Times, programming is devolving into menial work as AI code generation spreads within Amazon. Stay-Stopped was an eye-opener because it showed that poor-quality AI training data (solutions on the Internet), plus Amazon Q Developer's poor awareness of AWS API documentation, can contribute to low-quality AI-generated code. There is no substitute for human understanding and effort.
Apart from the pleasure of learning, trying to make simple but reliable software, and documenting it so that others can learn from it, I enjoy hearing from users. I've worked on Lights Off and its predecessor since 2017 but I wrote Stay-Stopped recently, based on feedback from a user. Please try these tools and let me know how I can improve them. Thank you!
shiv8 months ago
👋 Hey team!
I just put together a GitHub repo covering real-world Docker scenarios that are super useful for interview prep and day-to-day DevOps work. It includes practical commands with scenarios and clear explanations — like multi-stage builds, resource limits, health checks, logging, cleanup, and more.
🔗 Check it out here: https://github.com/SHIVANIUM-GIT/Docker_interview_question
Whether you're brushing up for an interview or want to sharpen your Docker skills, I think you'll find this helpful! Feel free to star or fork it if it helps you too. 💡
#docker #devops #tools #interview-prep
I just put together a GitHub repo covering real-world Docker scenarios that are super useful for interview prep and day-to-day DevOps work. It includes practical commands with scenarios and clear explanations — like multi-stage builds, resource limits, health checks, logging, cleanup, and more.
🔗 Check it out here: https://github.com/SHIVANIUM-GIT/Docker_interview_question
Whether you're brushing up for an interview or want to sharpen your Docker skills, I think you'll find this helpful! Feel free to star or fork it if it helps you too. 💡
#docker #devops #tools #interview-prep
Sri Nikitha8 months ago
hey all,
Been hacking on a small tool to make 3 a.m. pages less painful.
It bundles three things I always wanted:
• squashes duplicate/flapping alerts through noise rules
• grabs logs/metrics/past context and drafts a quick RCA
• auto-links the alert to the runbook that fixes it
It’s rough but running. If you’re on call and have ten minutes to kick the tires, DM me. You can signup too—just looking for some really blunt feedback.
Thanks 🙂
Been hacking on a small tool to make 3 a.m. pages less painful.
It bundles three things I always wanted:
• squashes duplicate/flapping alerts through noise rules
• grabs logs/metrics/past context and drafts a quick RCA
• auto-links the alert to the runbook that fixes it
It’s rough but running. If you’re on call and have ten minutes to kick the tires, DM me. You can signup too—just looking for some really blunt feedback.
Thanks 🙂
Samson8 months ago
🔥🔥End-to-End Static Website Auto Deployment Using Bash Scripts🔥🔥
Tired of manually copying and pasting files every time you deploy your application? You're in the right place. In this tutorial, we'll learn how to write a Bash script that automates the deployment of static files to a server simply by providing the path to a zipped file. This will work on every server.
Prerequisite(s)
✅️ Basic knowledge of bash script
Objective(s)
✅️ Create Bash Script that auto deploys a static website
Click on the link
https://devsecops-blog.hashnode.dev/end-to-end-static-website-deployment-using-bash-scripts
Are you interested in cloud/devops. I have videos for you. Everything simplified for everyone (whether you are just starting you have been in the game)
https://www.youtube.com/@africacloudpapa
Don't forget to subscribe🙏🙏. "Make we see food chop"🤣🤣
Follow us on our SM platforms. "Make we self happy" 🙏🙏🙏
https://x.com/africacloudpapa
https://www.instagram.com/africacloudpapa/
Tired of manually copying and pasting files every time you deploy your application? You're in the right place. In this tutorial, we'll learn how to write a Bash script that automates the deployment of static files to a server simply by providing the path to a zipped file. This will work on every server.
Prerequisite(s)
✅️ Basic knowledge of bash script
Objective(s)
✅️ Create Bash Script that auto deploys a static website
Click on the link
https://devsecops-blog.hashnode.dev/end-to-end-static-website-deployment-using-bash-scripts
Are you interested in cloud/devops. I have videos for you. Everything simplified for everyone (whether you are just starting you have been in the game)
https://www.youtube.com/@africacloudpapa
Don't forget to subscribe🙏🙏. "Make we see food chop"🤣🤣
Follow us on our SM platforms. "Make we self happy" 🙏🙏🙏
https://x.com/africacloudpapa
https://www.instagram.com/africacloudpapa/
C
Cloud Ref8 months ago
Link previews for AWS console links in Slack
I got tired of AWS and GCP console links in Slack - there's no useful preview, obviously! So I built something to preview AWS resources when you mention console links (and ARNs) in Slack.
I'd love to get some feedback on pricing and feature set!
https://cloudref.app
I got tired of AWS and GCP console links in Slack - there's no useful preview, obviously! So I built something to preview AWS resources when you mention console links (and ARNs) in Slack.
I'd love to get some feedback on pricing and feature set!
https://cloudref.app
Weston Platter8 months ago
I remember hearing about the context7 MCP server in a previous CP Office Hours. I have it a try writing a Dagger module and found it helped a lot with prompting cursor to work with Dagger's specific conventions around working with containers, directories, and modules.
dude8 months ago
aw, I missed office hours, but wanted to mention this: https://github.com/aws/aws-cdk/issues/34685
Awantika Nigam8 months ago
Hey everyone! We ran a session on prompt engineering + AI Agent workflow.
Based on feedback, it turned out to be more useful than expected. So I’m sharing the recap.
Not about 'magic words' but about:
• workflow prep
• proper context mngmnt
• iteration strategy etc.
Watch the recordings here : https://youtu.be/DP_yKoHeWI8
Based on feedback, it turned out to be more useful than expected. So I’m sharing the recap.
Not about 'magic words' but about:
• workflow prep
• proper context mngmnt
• iteration strategy etc.
Watch the recordings here : https://youtu.be/DP_yKoHeWI8
Jessica8 months ago
Hi everyone!
I'm Jessica, field marketing specialist at Mirantis. We’re excited to be hosting a joint webinar with our partner Tigera, creators of Calico on July 2. If you know anyone who might be interested in multi-cluster Kubernetes security and networking solutions, please feel free to share this with them.
• Webinar title: Simplify and Secure Multi-Cluster Kubernetes with k0rdent and Calico
• Date/time: July 2 2025, 9am PT/12pm ET/5pm BST
• Registration page: https://bit.ly/405Ohoj
• Posts to share on social media
◦ X (Twitter)
◦ LinkedIn
I'm Jessica, field marketing specialist at Mirantis. We’re excited to be hosting a joint webinar with our partner Tigera, creators of Calico on July 2. If you know anyone who might be interested in multi-cluster Kubernetes security and networking solutions, please feel free to share this with them.
• Webinar title: Simplify and Secure Multi-Cluster Kubernetes with k0rdent and Calico
• Date/time: July 2 2025, 9am PT/12pm ET/5pm BST
• Registration page: https://bit.ly/405Ohoj
• Posts to share on social media
◦ X (Twitter)
Erik Osterman (Cloud Posse)8 months ago
Things are moving really fast
Awantika Nigam8 months ago
Hey all 👋!
Proud to share that Refact.ai Agent is now ranked #1 on SWE-bench Multimodal — a benchmark built around real-world, screenshot-based bug reports.
Really useful for JavaScript devs working on UI issues 🤝
It also leads on SWE-bench Verified (Python, pass@1).
The best part? It’s fully open source, runs autonomously, and solves tasks end-to-end.
https://github.com/smallcloudai/refact
Feel free to try it out, share feedback, or contribute — we’d love to have your input!
Proud to share that Refact.ai Agent is now ranked #1 on SWE-bench Multimodal — a benchmark built around real-world, screenshot-based bug reports.
Really useful for JavaScript devs working on UI issues 🤝
It also leads on SWE-bench Verified (Python, pass@1).
The best part? It’s fully open source, runs autonomously, and solves tasks end-to-end.
https://github.com/smallcloudai/refact
Feel free to try it out, share feedback, or contribute — we’d love to have your input!