home.social

#developement — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #developement, aggregated by home.social.

fetched live
  1. I tried to install PHP from source this morning on my Mac. I had several issues, so I wrote a blog post to explain how I fixed them.

    🔗 maximepinot.com/en/blog/php-in

    #PHP #Mac #macOS #Web #Developement #Programming

  2. I tried to install PHP from source this morning on my Mac. I had several issues, so I wrote a blog post to explain how I fixed them.

    🔗 maximepinot.com/en/blog/php-in

    #PHP #Mac #macOS #Web #Developement #Programming

  3. I tried to install PHP from source this morning on my Mac. I had several issues, so I wrote a blog post to explain how I fixed them.

    🔗 maximepinot.com/en/blog/php-in

    #PHP #Mac #macOS #Web #Developement #Programming

  4. I tried to install PHP from source this morning on my Mac. I had several issues, so I wrote a blog post to explain how I fixed them.

    🔗 maximepinot.com/en/blog/php-in

    #PHP #Mac #macOS #Web #Developement #Programming

  5. I started a big _big_ BIG refactoring of my pet project and... holy crap, I am rewriting most of it.

    But that's a learning process, right? I am getting a clearer and clearer image of how to build what I wanna build, and am slowly (very slowly) getting there, I suppose...

    #rust #rustland #developement #softwaredevelopment #floss #opensource #opensourcesoftware

  6. I started a big _big_ BIG refactoring of my pet project and... holy crap, I am rewriting most of it.

    But that's a learning process, right? I am getting a clearer and clearer image of how to build what I wanna build, and am slowly (very slowly) getting there, I suppose...

    #rust #rustland #developement #softwaredevelopment #floss #opensource #opensourcesoftware

  7. I started a big _big_ BIG refactoring of my pet project and... holy crap, I am rewriting most of it.

    But that's a learning process, right? I am getting a clearer and clearer image of how to build what I wanna build, and am slowly (very slowly) getting there, I suppose...

    #rust #rustland #developement #softwaredevelopment #floss #opensource #opensourcesoftware

  8. I started a big _big_ BIG refactoring of my pet project and... holy crap, I am rewriting most of it.

    But that's a learning process, right? I am getting a clearer and clearer image of how to build what I wanna build, and am slowly (very slowly) getting there, I suppose...

    #rust #rustland #developement #softwaredevelopment #floss #opensource #opensourcesoftware

  9. I started a big _big_ BIG refactoring of my pet project and... holy crap, I am rewriting most of it.

    But that's a learning process, right? I am getting a clearer and clearer image of how to build what I wanna build, and am slowly (very slowly) getting there, I suppose...

    #rust #rustland #developement #softwaredevelopment #floss #opensource #opensourcesoftware

  10. Making the most out of a small LLM

    Yesterday i finally built my own #AI #server. I had a spare #Nvidia RTX 2070 with 8GB of #VRAM laying around and wanted to do this for a long time.

    The problem is that most #LLMs need a lot of VRAM and i don't want to buy another #GPU just to host my own AI. Then i came across #gemma3 and #qwen3. Both of these are amazing #quantized models with stunning reasoning given that they need so less resources.

    I chose huihui_ai/qwen3-abliterated:14b since it supports #deepthinking, #toolcalling and is pretty unrestricted. After some testing i noticed that the 8b model performs even better than the 14b variant with drastically better performance. I can't make out any quality loss there to be honest. The 14b model sneaked in chinese characters into the response very often. The 8b model on the other hand doesn't.

    Now i've got a very fast model with amazing reasoning (even in German) and tool calling support. The only thing left to improve is knowledge. #Firecrawl is a great tool for #webscraping and as soon as i implemented websearching, the setup was complete. At least i thought it was.

    I want to make the most out of this LLM and therefore my next step is to implement a basic #webserver that exposes the same #API #endpoints as #ollama so that everywhere ollama is supported, i can point it to my python script instead. This way it feels like the model is way more capable than it actually is. I can use these advanced features everywhere without being bound to it's actual knowledge.

    To improve this setup even more i will likely switch to a #mixture_of_experts architecture soon. This project is a lot of fun and i can't wait to integrate it into my homelab.

    #homelab #selfhosting #privacy #ai #llm #largelanguagemodels #coding #developement

  11. Making the most out of a small LLM

    Yesterday i finally built my own #AI #server. I had a spare #Nvidia RTX 2070 with 8GB of #VRAM laying around and wanted to do this for a long time.

    The problem is that most #LLMs need a lot of VRAM and i don't want to buy another #GPU just to host my own AI. Then i came across #gemma3 and #qwen3. Both of these are amazing #quantized models with stunning reasoning given that they need so less resources.

    I chose huihui_ai/qwen3-abliterated:14b since it supports #deepthinking, #toolcalling and is pretty unrestricted. After some testing i noticed that the 8b model performs even better than the 14b variant with drastically better performance. I can't make out any quality loss there to be honest. The 14b model sneaked in chinese characters into the response very often. The 8b model on the other hand doesn't.

    Now i've got a very fast model with amazing reasoning (even in German) and tool calling support. The only thing left to improve is knowledge. #Firecrawl is a great tool for #webscraping and as soon as i implemented websearching, the setup was complete. At least i thought it was.

    I want to make the most out of this LLM and therefore my next step is to implement a basic #webserver that exposes the same #API #endpoints as #ollama so that everywhere ollama is supported, i can point it to my python script instead. This way it feels like the model is way more capable than it actually is. I can use these advanced features everywhere without being bound to it's actual knowledge.

    To improve this setup even more i will likely switch to a #mixture_of_experts architecture soon. This project is a lot of fun and i can't wait to integrate it into my homelab.

    #homelab #selfhosting #privacy #ai #llm #largelanguagemodels #coding #developement

  12. Making the most out of a small LLM

    Yesterday i finally built my own #AI #server. I had a spare #Nvidia RTX 2070 with 8GB of #VRAM laying around and wanted to do this for a long time.

    The problem is that most #LLMs need a lot of VRAM and i don't want to buy another #GPU just to host my own AI. Then i came across #gemma3 and #qwen3. Both of these are amazing #quantized models with stunning reasoning given that they need so less resources.

    I chose huihui_ai/qwen3-abliterated:14b since it supports #deepthinking, #toolcalling and is pretty unrestricted. After some testing i noticed that the 8b model performs even better than the 14b variant with drastically better performance. I can't make out any quality loss there to be honest. The 14b model sneaked in chinese characters into the response very often. The 8b model on the other hand doesn't.

    Now i've got a very fast model with amazing reasoning (even in German) and tool calling support. The only thing left to improve is knowledge. #Firecrawl is a great tool for #webscraping and as soon as i implemented websearching, the setup was complete. At least i thought it was.

    I want to make the most out of this LLM and therefore my next step is to implement a basic #webserver that exposes the same #API #endpoints as #ollama so that everywhere ollama is supported, i can point it to my python script instead. This way it feels like the model is way more capable than it actually is. I can use these advanced features everywhere without being bound to it's actual knowledge.

    To improve this setup even more i will likely switch to a #mixture_of_experts architecture soon. This project is a lot of fun and i can't wait to integrate it into my homelab.

    #homelab #selfhosting #privacy #ai #llm #largelanguagemodels #coding #developement

  13. Making the most out of a small LLM

    Yesterday i finally built my own #AI #server. I had a spare #Nvidia RTX 2070 with 8GB of #VRAM laying around and wanted to do this for a long time.

    The problem is that most #LLMs need a lot of VRAM and i don't want to buy another #GPU just to host my own AI. Then i came across #gemma3 and #qwen3. Both of these are amazing #quantized models with stunning reasoning given that they need so less resources.

    I chose huihui_ai/qwen3-abliterated:14b since it supports #deepthinking, #toolcalling and is pretty unrestricted. After some testing i noticed that the 8b model performs even better than the 14b variant with drastically better performance. I can't make out any quality loss there to be honest. The 14b model sneaked in chinese characters into the response very often. The 8b model on the other hand doesn't.

    Now i've got a very fast model with amazing reasoning (even in German) and tool calling support. The only thing left to improve is knowledge. #Firecrawl is a great tool for #webscraping and as soon as i implemented websearching, the setup was complete. At least i thought it was.

    I want to make the most out of this LLM and therefore my next step is to implement a basic #webserver that exposes the same #API #endpoints as #ollama so that everywhere ollama is supported, i can point it to my python script instead. This way it feels like the model is way more capable than it actually is. I can use these advanced features everywhere without being bound to it's actual knowledge.

    To improve this setup even more i will likely switch to a #mixture_of_experts architecture soon. This project is a lot of fun and i can't wait to integrate it into my homelab.

    #homelab #selfhosting #privacy #ai #llm #largelanguagemodels #coding #developement

  14. Making the most out of a small LLM

    Yesterday i finally built my own #AI #server. I had a spare #Nvidia RTX 2070 with 8GB of #VRAM laying around and wanted to do this for a long time.

    The problem is that most #LLMs need a lot of VRAM and i don't want to buy another #GPU just to host my own AI. Then i came across #gemma3 and #qwen3. Both of these are amazing #quantized models with stunning reasoning given that they need so less resources.

    I chose huihui_ai/qwen3-abliterated:14b since it supports #deepthinking, #toolcalling and is pretty unrestricted. After some testing i noticed that the 8b model performs even better than the 14b variant with drastically better performance. I can't make out any quality loss there to be honest. The 14b model sneaked in chinese characters into the response very often. The 8b model on the other hand doesn't.

    Now i've got a very fast model with amazing reasoning (even in German) and tool calling support. The only thing left to improve is knowledge. #Firecrawl is a great tool for #webscraping and as soon as i implemented websearching, the setup was complete. At least i thought it was.

    I want to make the most out of this LLM and therefore my next step is to implement a basic #webserver that exposes the same #API #endpoints as #ollama so that everywhere ollama is supported, i can point it to my python script instead. This way it feels like the model is way more capable than it actually is. I can use these advanced features everywhere without being bound to it's actual knowledge.

    To improve this setup even more i will likely switch to a #mixture_of_experts architecture soon. This project is a lot of fun and i can't wait to integrate it into my homelab.

    #homelab #selfhosting #privacy #ai #llm #largelanguagemodels #coding #developement

  15. I made a small break from the #network monitor to focus on my #youtube #music downloader and if there's one thing i can say it's that youtube is fucked up. The same playlist can return different videos (which happens a lot of times). Also reliable metadata searching is a pain in the ass. For 100% reliable metadata searching you have to strip the video title down, remove any fragments that don't belong in the original title, normalize it and then compare it against a normalized, lowercase version of the metadata search result.

    #developement #coding #go #golang

  16. I made a small break from the #network monitor to focus on my #youtube #music downloader and if there's one thing i can say it's that youtube is fucked up. The same playlist can return different videos (which happens a lot of times). Also reliable metadata searching is a pain in the ass. For 100% reliable metadata searching you have to strip the video title down, remove any fragments that don't belong in the original title, normalize it and then compare it against a normalized, lowercase version of the metadata search result.

    #developement #coding #go #golang

  17. I made a small break from the #network monitor to focus on my #youtube #music downloader and if there's one thing i can say it's that youtube is fucked up. The same playlist can return different videos (which happens a lot of times). Also reliable metadata searching is a pain in the ass. For 100% reliable metadata searching you have to strip the video title down, remove any fragments that don't belong in the original title, normalize it and then compare it against a normalized, lowercase version of the metadata search result.

    #developement #coding #go #golang

  18. I made a small break from the #network monitor to focus on my #youtube #music downloader and if there's one thing i can say it's that youtube is fucked up. The same playlist can return different videos (which happens a lot of times). Also reliable metadata searching is a pain in the ass. For 100% reliable metadata searching you have to strip the video title down, remove any fragments that don't belong in the original title, normalize it and then compare it against a normalized, lowercase version of the metadata search result.

    #developement #coding #go #golang

  19. Today i've nearly finished the transition to #Vue.

    Next steps are adding more #API #endpoints, implementing the pages in the #frontend and finally #dockerizing and publishing it 🔥

    #coding #developement #docker #linux #sysadmin #selfhosting #homelab #homeserver #server

  20. Today i've nearly finished the transition to #Vue.

    Next steps are adding more #API #endpoints, implementing the pages in the #frontend and finally #dockerizing and publishing it 🔥

    #coding #developement #docker #linux #sysadmin #selfhosting #homelab #homeserver #server

  21. Today i've nearly finished the transition to #Vue.

    Next steps are adding more #API #endpoints, implementing the pages in the #frontend and finally #dockerizing and publishing it 🔥

    #coding #developement #docker #linux #sysadmin #selfhosting #homelab #homeserver #server

  22. Today i've nearly finished the transition to #Vue.

    Next steps are adding more #API #endpoints, implementing the pages in the #frontend and finally #dockerizing and publishing it 🔥

    #coding #developement #docker #linux #sysadmin #selfhosting #homelab #homeserver #server

  23. Today i've nearly finished the transition to #Vue.

    Next steps are adding more #API #endpoints, implementing the pages in the #frontend and finally #dockerizing and publishing it 🔥

    #coding #developement #docker #linux #sysadmin #selfhosting #homelab #homeserver #server

  24. I realized that i need to separate the #frontend and #backend. I'm glad i didn't do a lot of work on the frontend yet. I will probably go with #react. Since the dashboard will display a lot of informations, which must update live, a #uiframework becomes necessary.

    #developing #developement #coding #webdevelopment

  25. I realized that i need to separate the #frontend and #backend. I'm glad i didn't do a lot of work on the frontend yet. I will probably go with #react. Since the dashboard will display a lot of informations, which must update live, a #uiframework becomes necessary.

    #developing #developement #coding #webdevelopment

  26. I realized that i need to separate the #frontend and #backend. I'm glad i didn't do a lot of work on the frontend yet. I will probably go with #react. Since the dashboard will display a lot of informations, which must update live, a #uiframework becomes necessary.

    #developing #developement #coding #webdevelopment

  27. I realized that i need to separate the #frontend and #backend. I'm glad i didn't do a lot of work on the frontend yet. I will probably go with #react. Since the dashboard will display a lot of informations, which must update live, a #uiframework becomes necessary.

    #developing #developement #coding #webdevelopment

  28. I realized that i need to separate the #frontend and #backend. I'm glad i didn't do a lot of work on the frontend yet. I will probably go with #react. Since the dashboard will display a lot of informations, which must update live, a #uiframework becomes necessary.

    #developing #developement #coding #webdevelopment

  29. I made so much progress yesterday. I've completely restructured the #codebase, improved the #performance, added logging with loglevels, and worked a bit on the webui. I still have a lot of work to do, but the project is a lot of fun and i really want to realize this. Can't wait to release the first alpha version. Who knows, maybe this becomes a cool community project. It will be extensible and flexible, so that might be realistic.

    #coding #developement #go #security #homelab #selfhosting #opensource #freeandopensource

  30. I made so much progress yesterday. I've completely restructured the #codebase, improved the #performance, added logging with loglevels, and worked a bit on the webui. I still have a lot of work to do, but the project is a lot of fun and i really want to realize this. Can't wait to release the first alpha version. Who knows, maybe this becomes a cool community project. It will be extensible and flexible, so that might be realistic.

    #coding #developement #go #security #homelab #selfhosting #opensource #freeandopensource

  31. I made so much progress yesterday. I've completely restructured the #codebase, improved the #performance, added logging with loglevels, and worked a bit on the webui. I still have a lot of work to do, but the project is a lot of fun and i really want to realize this. Can't wait to release the first alpha version. Who knows, maybe this becomes a cool community project. It will be extensible and flexible, so that might be realistic.

    #coding #developement #go #security #homelab #selfhosting #opensource #freeandopensource

  32. I made so much progress yesterday. I've completely restructured the #codebase, improved the #performance, added logging with loglevels, and worked a bit on the webui. I still have a lot of work to do, but the project is a lot of fun and i really want to realize this. Can't wait to release the first alpha version. Who knows, maybe this becomes a cool community project. It will be extensible and flexible, so that might be realistic.

    #coding #developement #go #security #homelab #selfhosting #opensource #freeandopensource

  33. I made so much progress yesterday. I've completely restructured the #codebase, improved the #performance, added logging with loglevels, and worked a bit on the webui. I still have a lot of work to do, but the project is a lot of fun and i really want to realize this. Can't wait to release the first alpha version. Who knows, maybe this becomes a cool community project. It will be extensible and flexible, so that might be realistic.

    #coding #developement #go #security #homelab #selfhosting #opensource #freeandopensource

  34. This is just beautiful. The project uses #MariaDB to store all the collected data. It makes heavy use of #API endpoints which will be a top priority not only for the internal workings. My focus mostly lies in making this as flexible as possible so people can configure it exactly as they need it. Configuration will be in #yaml.

    What i got so far is:

    • packet capturing (from client, sent to the control server)
    • a webhook (which will be a drop-in replacement for #Discord's since a lot of services support sending logs to discord webhooks)

    What's planned:

    • Log file monitoring (like #Fail2Ban, but more advanced and easier to configure)
    • A fully featured dashboard which visualizes the data and gives you control and a transparent overview of your network activity.
    • IP banning (multiple ways to make it flexible)
    • Maybe even some advanced responses (like reporting all ports as open for nmap scans)

    I would be very interested to know what you think. Ideas, criticism and questions are very welcome. As soon as the base is working, i will push it to #Github.

    #developement #coding #sideproject #homelab #security #networking #monitoring #xdr

  35. This is just beautiful. The project uses #MariaDB to store all the collected data. It makes heavy use of #API endpoints which will be a top priority not only for the internal workings. My focus mostly lies in making this as flexible as possible so people can configure it exactly as they need it. Configuration will be in #yaml.

    What i got so far is:

    • packet capturing (from client, sent to the control server)
    • a webhook (which will be a drop-in replacement for #Discord's since a lot of services support sending logs to discord webhooks)

    What's planned:

    • Log file monitoring (like #Fail2Ban, but more advanced and easier to configure)
    • A fully featured dashboard which visualizes the data and gives you control and a transparent overview of your network activity.
    • IP banning (multiple ways to make it flexible)
    • Maybe even some advanced responses (like reporting all ports as open for nmap scans)

    I would be very interested to know what you think. Ideas, criticism and questions are very welcome. As soon as the base is working, i will push it to #Github.

    #developement #coding #sideproject #homelab #security #networking #monitoring #xdr

  36. This is just beautiful. The project uses #MariaDB to store all the collected data. It makes heavy use of #API endpoints which will be a top priority not only for the internal workings. My focus mostly lies in making this as flexible as possible so people can configure it exactly as they need it. Configuration will be in #yaml.

    What i got so far is:

    • packet capturing (from client, sent to the control server)
    • a webhook (which will be a drop-in replacement for #Discord's since a lot of services support sending logs to discord webhooks)

    What's planned:

    • Log file monitoring (like #Fail2Ban, but more advanced and easier to configure)
    • A fully featured dashboard which visualizes the data and gives you control and a transparent overview of your network activity.
    • IP banning (multiple ways to make it flexible)
    • Maybe even some advanced responses (like reporting all ports as open for nmap scans)

    I would be very interested to know what you think. Ideas, criticism and questions are very welcome. As soon as the base is working, i will push it to #Github.

    #developement #coding #sideproject #homelab #security #networking #monitoring #xdr

  37. This is just beautiful. The project uses #MariaDB to store all the collected data. It makes heavy use of #API endpoints which will be a top priority not only for the internal workings. My focus mostly lies in making this as flexible as possible so people can configure it exactly as they need it. Configuration will be in #yaml.

    What i got so far is:

    • packet capturing (from client, sent to the control server)
    • a webhook (which will be a drop-in replacement for #Discord's since a lot of services support sending logs to discord webhooks)

    What's planned:

    • Log file monitoring (like #Fail2Ban, but more advanced and easier to configure)
    • A fully featured dashboard which visualizes the data and gives you control and a transparent overview of your network activity.
    • IP banning (multiple ways to make it flexible)
    • Maybe even some advanced responses (like reporting all ports as open for nmap scans)

    I would be very interested to know what you think. Ideas, criticism and questions are very welcome. As soon as the base is working, i will push it to #Github.

    #developement #coding #sideproject #homelab #security #networking #monitoring #xdr

  38. This is just beautiful. The project uses #MariaDB to store all the collected data. It makes heavy use of #API endpoints which will be a top priority not only for the internal workings. My focus mostly lies in making this as flexible as possible so people can configure it exactly as they need it. Configuration will be in #yaml.

    What i got so far is:

    • packet capturing (from client, sent to the control server)
    • a webhook (which will be a drop-in replacement for #Discord's since a lot of services support sending logs to discord webhooks)

    What's planned:

    • Log file monitoring (like #Fail2Ban, but more advanced and easier to configure)
    • A fully featured dashboard which visualizes the data and gives you control and a transparent overview of your network activity.
    • IP banning (multiple ways to make it flexible)
    • Maybe even some advanced responses (like reporting all ports as open for nmap scans)

    I would be very interested to know what you think. Ideas, criticism and questions are very welcome. As soon as the base is working, i will push it to #Github.

    #developement #coding #sideproject #homelab #security #networking #monitoring #xdr