1. I would like to determine from the community if there is worth in continuing the project, and to determine what direction it should move in.
2. Hopefully gain some community backing, as it is not something I (currently and in the future) would be able to achieve on my own.
The Pi-Kitchen project was born around a year ago by myself and Kevin Hill, with the aim to provide an easy way to automate deployment of Raspberry Pi setups.
__________________The Pi-Kitchen Concept
The overall concept for using the Pi-Kitchen is to provide a wide selection of “ingredients” from which you can select and include in your very own special “recipe” for your own perfect SD-Card.
In theory, you will be able to mix and combine the various ingredients (=a particular feature/setup) with other ingredients, to create that perfect blend of special spices to go into your recipe (=your desired setup).
The key difference to the standard method of building custom images is the that source files (the ingredients) are kept separate to the installation (the recipe). This means we are free to mix and combine them how we wish.
Since NOOBS supports the installation of flavours we can (if we want to) create a range of flavours each using a different recipe, all drawing from a common set of ingredients.
The Pi-Kitchen works using standard files making it easy to edit, adjust and re-configure your setup (even from OSX/Windows or directly from your Raspberry Pi).
Another aim of the configuration to is to allow off-line installs by storing/updating the required files locally, all without network and/or a display.
More details on my site: https://pihw.wordpress.com/guides/pi-kitchen/ and GitHub: https://github.com/PiHw/Pi-Kitchen (my own copy of the project has more Recipes to get pushed on there - but I've had very little time to test the newer ones).
What state is the project?
Since its start, I've developed multiple recipes and every time I use a Raspberry Pi I use the Pi-Kitchen to deploy it (I forget about the pain manually configuring a new setup has). On the whole, it works great...for me. I understand how it works and how to make it work, but it is no where near as user friendly as it should/could be. My focus has been on developing the function and testing the principles behind it to see if they make sense before implementing the user-facing aspects (I'm now happy the concept works very well).
It has been very useful to deploy setups just by writing the SDCard via the PiKitchen and applying power. The resulting RPi appearing on the network ready to go shortly after. I have a simple "Bake" script which has been perfect for quickly dropping the files on to a SDCard and deploying (it takes me a few mins to setup, the rest is install/copy time). If I need to blitz a card it is back and running very quickly, and since I setup network links as part of the setup I can work from a network share (so nothing important is ever lost!).
What are the issues?
1. A big issue is that once a Recipe is written, unless I use it in my setups it doesn't get tested on new releases. Plus even the recipes I do use, a new release of NOOBS can often break it and as the only user, I have to find time to solve the issue, fix the recipe and retest it (which I don't often get time to do). Additional recipes, the more recipes I add the more testing and fixes that need to be made as new NOOBS/OS releases occur. The community support is needed to achieve this, as I simply don't have the time to do that on my own. Imagine recipes for LCD screens, hardware keyboards, Pi Camera animation setups etc so everything gets deployed and works, just from a few clicks!
Broken recipes just ruin the whole idea! - So more users hopefully = issues getting fixed by users!
2. After developing the Pi-Kitchen, I became more aware that there may be better solutions out there in the Linux world which would achieve the same solution. Clearly it would be a waste to reinvent something which is already able to do the job and the effort better spent on improving that instead.
This is something I would like to canvas the community on... Would Puppet / Chef do this or is there something else which achieves a similar goal?
3. Although the "Bake" script makes it super simple for me to deploy setups, it is the sort of tool where a GUI (which I have in mind) would make it very powerful indeed (and should be usable for anyone). Creating a GUI would take the tool to the next level and allow easy creation of recipes and flavors.
If this tool is determined to be useful (and there are no better alternatives) adding a GUI would be worth the effort...If it becomes a community tool then a GUI should reflect the needs and use-cases of the community so needs input from it.
So there you have it...Now over to you, I would love to hear what the community has to say about it and general thoughts. I'm prepared to hear that there is a better solution already, if so please let me know so I can use that instead (or that it is just an impossible dream).