Steamline hardware detection in Hardy

Registered by Martin Pitt

In the last years, upstreams have provided vast improvements in hardware detection infrastructure. We review which parts of Debian/Ubuntu specific hardware detection bits are obsolete and how to consolidate and update them.

Blueprint information

Status:
Complete
Approver:
Colin Watson
Priority:
Medium
Drafter:
Martin Pitt
Direction:
Needs approval
Assignee:
Scott James Remnant (Canonical)
Definition:
Approved
Series goal:
Accepted for hardy
Implementation:
Implemented
Milestone target:
None
Started by
Martin Pitt
Completed by
Martin Pitt

Related branches

Sprints

Whiteboard

keybuk, 2007-10-22: this has too much work by too many people, removing an assignee - those who have work to do here know who they are (or I'm about to subscribe them)

Pitti, 2007-11-13: Ready for review, please give me some feedback. Timo, Bryce, can you please take a look at the X.org bits?

keybuk 2007-11-20: approved in its current state, if Bryce has changes to make, knock it back to drafting

bryce 2007-11-20: I've fleshed out the implementation section with more specifics about how the xorg hardware detection has and will change (including keyboard/mice and hal). I don't have the means to bump it back to 'drafting', but perhaps pitti can do so after having a review of it.

pitti 2007-11-20: reviewed, fixed formatting; looks good to me, I set it back to pending-approval for final inspection from Scott/Colin.

bryce 2007-11-21: I asked gravity to review this, as he's been working on implementing these changes in Debian's xorg. "The spec is right about the input device issues (at least the ones I'm aware of) but not about video driver issues. We use the output of discover to set specific options based on the type of card in cases where the drivers don't support certain things. If it weren't for that, I'd have ripped our discover usage out of the scripts long ago. So basically, if we don't want to break the ability to install on certain hardware, we need to move that logic in to the drivers. The spec should talk about that. Other than that, it looks really good, and matches what I've been thinking."
I've updated the spec to include his recommendations about changes needed for addressing the video driver issues.

bryce 2007-12-19: The xorg portions have had a pretty huge chunk completed as of yesterday's sync with Debian. xresprobe and laptop-detect are now no longer used. Monitor refresh rates, depths, and resolutions are no longer being listed in xorg.conf. The massive xserver-xorg.postinst.in script has been cut in half.

pitti, 2008-01-02: device ownership: hal, sane-backends, libgphoto2 changes uploaded, cameras and scanners use dynamic ACLs; plugdev can be dropped from udev now; hplip and udev still use scanner, I don't have hardware to test

pitti, 2008-12-15: this was implemented in time for hardy, wasn't marked so for some reason.

(?)

Work Items