I tried gettext-tiny when it was still a stub library and didn't really provide any internationalization support.  Nice to see that's changing.  I wasn't aware that they've added more support beyond just the stub functionality.  Personally, I've used code originally based on BSD gettext ( https://www.mmnt.net/db/0/18/ftp.khstu. … /distfiles ) with some updated code from the BSD citrus project.

For rar extraction, I like the unarr library.  There's source code for it here:  https://github.com/sumatrapdfreader/sum … /ext/unarr
I seem to remember reading that it was a clean room implementation.  This is the license:
https://github.com/sumatrapdfreader/sum … rr/COPYING
I use my own makefiles with it, but it looks like someone's created cmake files for it:
https://github.com/selmf/unarr
I use the library in conjunction with a program that can display CBR and CBZ files.

I'm finding it very interesting reading through the Hyperbola web site and wiki and seeing which packages were included and which are being replaced.  Hyperbola looks like a really interesting distribution and I really like the idea of the possibility of a BSD based kernel in the future.

gettex
As you noted already the project is not allowing a complete different packaging free from Java and C#. We have chosen therefore gettext-tiny as working alternative and all packages in our repositories in need for internationalization are using this.

unrar
We have unar therefore working so far, but would you mind share the alternative mentioned?

Was surprised to see gettext listed as a removed package on the wiki.  I thought gettext was GPL licensed and that it was used to provide internationalization in several C/C++ packages.  Was not familiar with any connection to promoting Java and C#.  I'd be curious to know more about that.  I don't use gettext in my own package builds.  I use a fairly compatible alternative (with a BSD 2 clause license) based on code by some PostgreSQL developers and the BSD citrus project.  Was wondering, what does Hyperbola use for internationalization of text in packages in place of gettext?

Also noticed mention in the wiki FAQ of no support for unrar.  Wasn't sure if you were aware of some of the other free implementations of it.  There's one library I  use with some packages I build (including a CBR viewer).  It's GNU LGPLv3 licensed and not directly based on the original unrar code.  It works well enough to be able to view CBR files.  Most Free alternatives seem to fail when trying to decompress CBR files.

Found it very interesting reading about the reasoning behind removing Xorg and about the Xenocara project.  Did not about the Xenocara project.  Looking forward to investigating that further.

Appreciate your descriptions of why you avoided or chose specific packages.  It's interesting to read about.