Why Our Builder Will Become a Heavily Modified GhostBSD Builder
One quote by John Muir I found on the internet goes "Don't try and reinvent the wheel - just work on making it better than anyone else." After working on the last iteration builder and finding so much difficulty due to a month's hiatus, where the code was confusing, everything irritating, and I failed to do the one thing I aimed to do with that builder: Make it easy enough to work with as a newcomer. I fell for the common issue of a programmer not understanding the code they wrote months previously, so with that I knew I needed a new strategy.
So, we are moving it to a highly modified version of GhostBSD, cool. Why? As in why the change, and why specifically GhostBSD? Well, to the first question, I answered. The current build system is confusing, and annoying to work with. Trust me, I made it. Why GhostBSD? Well because it's the closest to what we want, and it's already well designed for what we want to add. Remember that quote about not reinventing the wheel? GhostBSD is why I added it. They are already working on some amazing stuff, and using their work and merging it into mine means I don't have to focus on all the little details I was working on, and I can more or less guarantee a functional system sooner, and with proper modifications, the way I want it to be.
Will it be the EXACT SAME as GhostBSD? No. First, we plan to port and massively modify Calamares for our installer, or perhaps using a different custom installer. Second, we will have a few custom applications. We also will use MATE with custom themes as the default desktop, not XFCE. Third, we intend on porting the AppCenter, as the folk at elementary Inc, who also make elementary OS, have made a perfect appstore for the future of open-source. Lastly, the inner workings of the GhostBSD ISO use ZFS, while we intend on using UFS. The reason why is that making ZFS work in the build script originally was a nightmare, and I also just prefer using UFS. Also, their build script is not really professional, it kinda feels all over the place. We intend on solving that by using a custom project directory design, with most of the same files.
I know this project is taking a lot of time, but I am busy, and I set really high standards for myself. I want the release of Potabi 1.0 Luna to be strong, when the beta which will release soon enough will still hopefully be comparable to Trisquil, or a FerenOS release, the latter I love. There is a quote from the last lecture Randy Pausch of Carnegie Mellon University, a video my grandmother showed me and later a book I would have read that just sticks with me. "Wait long enough, and people will surprise and impress you." Potabi and side projects like LiSoSy and KISPM are my passion projects, but I really aim to release something by the end of the year, but certainly by the beginning of the next.
Comments
Post a Comment