Nithin Bose's picture


This is kind of my first post here.  I have been using various TKL images for a long time now. I was trying to develop some tkl-patches recently. The process of doing so is simple enough but i believe it could be made much simpler. An easier approach that IMHO would be the following:

1. Run the image to be patched in a VM

2. Perform the various activities like installing packages, changing configurations etc to get an image that meets our requirements.

3. Some how create the patch ie add required contents in the deb/overlay folders and generate the conf file automatically based on the activities in step 2. 

4. Tweek/optimize the automatically generated patch files if necessary

5. Build, publish and apply the patch to similar appliances using the existing tkl-patch commands.

This might not be a good process for all scenarios but since i wanted to have something like this, I have attempted to do so with a python project "tkl-patcher" hosted on github.

I have tried to create a shell where a user can type in linux commands, which will create the patch files after running the command on the system.

There are lots of things to be done including proper documentation. However, I have created  a "help" command to show all supported commands.

Let me know of your opinions.




Maybe just the thing to change my life. Again. Kudos for working productively with TKLPatch workflow.

Jeremy Davis's picture

Awesome idea. I'll have to take it for a spin when I get a chance and give you some feedback.

Nithin Bose's picture

The script is really very crud now... so your feedbacks would help me a lot in improving it. 

I need a clarification though, in what scenareo is the debs folder used? I read, in the docs that packages in the debs folder are installed before processing overlays or conf file, what is the reason for this?? 

Add new comment