@asomers wrote:

The only thing I wanted different is a more detailed commit message. Of course, Mark might disagree.

I gave a lighting talk about it at BSDCan

Thanks for sharing, haven't seen this video clip before. I fully agree and like the useful, concise mtw(4) example, please keep it up!

Interesting, although it's still unclear what's going on. Perhaps the error should be reset upon some condition, like in commit 0c38e3dbbf6e where a similar pattern could be seen?

@des wrote:

OpenPAM does not look for a security subdirectory and never has.

@ziaee wrote:

[vi, cat, and ed take] 8 characters on the left side of the screen anytime lines are numbered.

Neither of these programs have numbered lines enabled by default, that would be preposterous. Any decorations you might have enabled in your text editor shall not be accounted for in this argument or assumption thereof be reflected in the documentation.

@ziaee wrote:

Vi puts an 8 character gutter [or "takes exactly 8 characters off the side", per D50410]

How do you mean (takes for what, from which side)? Mine doesn't take anything from 80 chars. Could it be that you misconfigured it?

@imp wrote:

72 is needlessly restrictive. There's no reason for that number. 80 might make sense (its how I wrap my man pages).

What's usually implied is that 72 should be the lower bound and 78 be the upper. I usually play with values in that range and select the one which yields the most appealing look (less ragged right edge). Wrapping strictly at 80 doesn't look very neat and is ambiguous, which may cause mouse copy-pasting issues on some buggy terminal emulators, e.g. I've seen eating a letter at the border or copying two adjacent words as one.

@mhorne wrote:

There are a dozen solutions to your self-inflicted choice of editor that don't involve a policy change and its enforcement.

Sorry if I've missed all the drama and thus being out of context, but is this about wrapping lines before 80 chars (72 to account for reasonable amount of slack)? Then I don't see what's wrong with this suggestion, we already [try to] adhere to it in src and ports, and I certainly welcome a polite nudge to do so elsewhere (also, I don't see how it's being enforced with current wording).

The mitshm compile is needed for multiple screen. When copying the whole image, it takes 10 secs to take the screenshot

@kbowling wrote:

@danfe indicated support for moving this to some quorum, maybe we need an nvidia@ alias or can discuss with x11@ shared ownership of all the ports.