User:Maffblaster/Raising Gentoo to prominence

This article is a dedicated location to write to improvement ideas of a broad scope for Gentoo.

Long term projects (LTPs) 
These projects will take a long time to complete (months/years), but are worth working toward.

Installer

 * Installer - A way to automate the installation of Gentoo Linux. Goals include:
 * Ability to pass configuration file.
 * Custom kernel generation based on detected devices.
 * REST API (write your own front end)
 * Support for 'Mix-ins' (support basic profiles first).

Automated testing
Need more research on this front.


 * Gentoo Ebuild Repository Test Suite (GERTS)
 * Performs quality assurance and package pre-stabilization across various ebuild repositories.
 * Works with Qemu to perform packages testing across stable arches (needs packaged up in a container).
 * Parses atomic ebuild information and prioritizes security related atoms via Bugzilla API.
 * Automated package testing based on Docker and a test harness. This should be relatively simple. Idea based upon this talk.

Automated package manager porting

 * Automated ebuild generation for language-specific package managers (such as, , , , etc.):
 * g-sorcery (backend)
 * gs-elpa (lisp)
 * gs-pypi (python)

Automated Portage error reporting
Ubuntu uses Apport along with an opt-in feature that automates error reporting (crash report data, even includes the core dump!) for their user base. I believe Apport only catches and reports on runtime crashes. It should be possible for Gentoo to do something similar with Portage failures.

Tinderbox already exists for a similar purpose, but it does not catch many real-world failures (or at least not packages/USE combinations that are rarely used together).

This new user utility would accept a users Bugzilla credentials, would need a application key, but then would automatically create a directory containing metadata from compilation (or other issues), then upload them to Bugzilla. It would be able to weed out invalid cases (out of disk space or out of memory failures). I'm sure there could be some other useful features to add to it given enough thought.

Mix-ins

 * Bring 'Mix-ins' (which should really be renamed to USE flag sets or something) to Gentoo.
 * Mix-in news announcement on Funtoo.org.
 * Mail thread (on gentoo-dev?) by.

Every developer should have a blog!
STATUS: IN PROGRESS!

These projects will take a short amount of time to complete (days/weeks), but are worth working toward. This initiative I'm calling "Every developer should have a blog" project. :)


 * Templates for static developer blogs:
 * Enable each developer to quickly start writing entries on his/her statically generated blog.
 * Write article about hosting developer static blogs/sites on https://dev.gentoo.org/~${USER}/
 * Make it easy to RSS tag/subpage side-projects

Generate InfoBoxes from ebuilds using SMW properties
Use data from Portage database to generate URLs on the wiki for each package...