On Wed, Apr 24, 2019 at 3:58 PM Pierre Labastie <[email protected]> wrote: > > You're all but annoying! If refactoring is what you are interested in, please > go ahead. I've maintained the code by adding what I needed at a given time. > But the code is certainly in big need of some more modular approach. I've > never taken the time to do refactoring, except when modifying big chunks of > code (mainly in the BLFS part).
OK - I'm considering two options for how to approach this, and I've actually already started a little bit on both. 1. Clean up, improve, refactor the current code, keeping more or less its current structure. 2. Move the current jhalfs script to a new name (jhalfs.sh or something) and add a new 'jhalfs' entrypoint in python. To start, it just sets up the menu options and then execs the original jhalfs.sh to continue from there. After that, little by little add the existing shell features into the python code until the whole thing is handled in python (and possibly still make for actual execution). It's probably obvious given my other comments that my preference is #2 because of the testing and packaging advantages it brings. But if you prefer it to stay with shell, I can help refactor that as much as possible, too. > Also, I've stopped testing CLFS when I realized that the book was not updated > often enough to be buildable on more recent platforms. So I would have spent > more time trying to fix the book than on the jhalfs code. > > CLFS has recently moved to github. I'm not sure the development has restarted > though. > > Do you want to do the update for the CLFS git urls? I'm rather busy with the > move: need to figure out how to send notifications to the linuxfromscratch > lists, and how to couple with trac on the wiki... Yeah, I will see how viable CLFS is. I think there is an 'embedded' CLFS that is still pretty active. What about HLFS? JH -- http://lists.linuxfromscratch.org/listinfo/alfs-discuss FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
