route.8: Describe weight modifier

[pouria]

Summary

Add and describe -weight argument
to the manual.

Test Plan

mandoc -Tlint

[markj]

sbin/route/route.8

309

(I am not familiar with how this actually works.) What does "twice as likely" mean? Where does the weight actually get evaluated?

[pouria]

I want to work on metric or administrative distance too.
But I'm waiting on @melifaro response to some of my questions.

sbin/route/route.8

309

sys/net/route/nhgrp_ctl.c:233:

224 /*
225  * Compile actual list of nexthops to be used by datapath from
226  *  the nexthop group @dst.
227  *
228  * For example, compiling control plane list of 2 nexthops
229  *  [(200, A), (100, B)] would result in the datapath array
230  *  [A, A, B]
231  */

[glebius]

IMHO, the description is good enough and clear. Later, native-speakers-doc-committers can improve.

[pouria]

kind ping @ziaee

[ziaee]

Thanks for the ping!

sbin/route/route.8

306

I don't know how this works. It's rendering as:

-gateway -3fff::1 -weight -100

Is that correct? Seems like the IP addresses aren't usually a flag?

[pouria]

@ziaee done.

[pouria]

sbin/route/route.8

306

You're right. I should prefix it with Li tag.
Now it should be:

-gateway 3fff::1 -weight 100

[ziaee]

sbin/route/route.8

306

Prefixing with Li makes it look like this when rendered to things other than terminal. I think they are arguments to the flag, so should use Ar probably.

[pouria]

@ziaee done.

[ziaee]

sbin/route/route.8

306

What do you think of if we did them like this?

My rule is, if the page uses Ql, then use that, and if it uses Dq, then I try to use that. We have some other Dq's on the page. It bloats the lines a bit, but I think it kinda looks nice.

[pouria]

sbin/route/route.8

306

Definitely better.

[pouria]

Use .Dq.