Date: Fri, 12 Feb 2021 08:13:48 +0100 From: Gary Jennejohn <gljennjohn@gmail.com> To: John-Mark Gurney <jmg@funkthat.com> Cc: Lutz Donnerhacke <lutz@donnerhacke.de>, "freebsd-arch@freebsd.org" <arch@freebsd.org> Subject: Re: adding a sysctl man section Message-ID: <20210212071348.42f61156@ernst.home> In-Reply-To: <20210212001209.GI31099@funkthat.com> References: <20210211001505.GF31099@funkthat.com> <CANCZdfrD3aUS9YQQkA9yZxvUDvBdbfmMcuteqPp8yeDHoA_grQ@mail.gmail.com> <20210211075456.GA7928@belenus.iks-jena.de> <20210212001209.GI31099@funkthat.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, 11 Feb 2021 16:12:09 -0800
John-Mark Gurney <jmg@funkthat.com> wrote:
> Lutz Donnerhacke wrote this message on Thu, Feb 11, 2021 at 08:54 +0100:
> > On Wed, Feb 10, 2021 at 06:37:00PM -0700, Warner Losh wrote:
> > > On Wed, Feb 10, 2021 at 5:15 PM John-Mark Gurney <jmg@funkthat.com> wrote:
> > > > Then, any page that describes a sysctl, would add an MLINK to it:
> > > > MLINK+= xhci.4 hw.usb.xhci.debug.s
> > > >
> > > > This section would be added to the default search, and then users
> > > > would simply be able to type: man <sysctl> and get directed to the
> > > > page that has information about it.
> > > >
> > > > Any objections?
> >
> > I'd oppose, it will clutter the filesystem for a "man -k" (or similar)
> > shortcut.
>
> Can you expand exactly how/why this would happen? I assume you mean
> clutter the output from "man -k", but as "man -k" uses a db for it's
> data, it'd only slightly expand that db, but it wouldn't clutter the
> file system anymore, unless I'm missing something...
>
> So, in one test case, this is what you'd see (or maybe section 9,
> which I'm find with):
> # man -k xhci
> xhci, hw.usb.xhci.ctlquirk, hw.usb.xhci.ctlstep, hw.usb.xhci.debug, hw.usb.xhci.dma32, hw.usb.xhci.streams, hw.usb.xhci.use_polling, hw.usb.xhci.xhci_port_route(4, s) - USB eXtensible Host Controller driver
>
> yes, if you do a search for a generic term like debug, you'll end up
> a lot more extra output, BUT, "man -k debug" already outputs a wall
> of text that you're likely to need to filter anyways, so I don't see
> a major difference/problem here.
>
> I will say I'm a bit surprised there isn't an option to apropos to allow
> only displaying the first matching man page to shorten the lines a bit,
> but we already have this problem for some man pages:
> # find . -type f -ls | awk '{ print $1 }' | sort | uniq -c | sort | tail -n 5
> 95 409059
> 98 405230
> 102 408573
> 116 397987
> 118 396855
>
> 409059 is nv(9)
> 405230 is snmpmod(3)
> 408573 is bhnd(9)
> 397987 is libusb(3)
> 396855 is curs_sp_funcs(3X)
>
> for example.
> # man -k libnv
> nvlist_create, nvlist_clone, nvlist_destroy, nvlist_dump, nvlist_empty, nvlist_error, nvlist_exists, nvlist_fdump, nvlist_flags, nvlist_free, nvlist_next, nvlist_pack, nvlist_recv, nvlist_send, nvlist_set_error, nvlist_size, nvlist_unpack, nvlist_xfer, nvlist_add_binary, nvlist_add_bool, nvlist_add_bool_array, nvlist_add_descriptor, nvlist_add_descriptor_array, nvlist_add_null, nvlist_add_number, nvlist_add_number_array, nvlist_add_nvlist, nvlist_add_nvlist_array, nvlist_add_string, nvlist_add_string_array, nvlist_add_stringf, nvlist_add_stringv, nvlist_append_bool_array, nvlist_append_descriptor_array, nvlist_append_number_array, nvlist_append_nvlist_array, nvlist_append_string_array, nvlist_exists_binary, nvlist_exists_bool, nvlist_exists_bool_array, nvlist_exists_descriptor, nvlist_exists_descriptor_array, nvlist_exists_null, nvlist_exists_number, nvlist_exists_number_array, nvlist_exists_nvlist, nvlist_exists_nvlist_array, nvlist_exists_string, nvlist_exists_type, nvlist_free_b
in
> ary, nvlist_free_bool, nvlist_free_bool_array, nvlist_free_descriptor, nvlist_free_descriptor_array, nvlist_free_null, nvlist_free_number, nvlist_free_number_array, nvlist_free_nvlist, nvlist_free_nvlist_array, nvlist_free_string, nvlist_free_string_array, nvlist_free_type, nvlist_get_binary, nvlist_get_bool, nvlist_get_bool_array, nvlist_get_descriptor, nvlist_get_descriptor_array, nvlist_get_number, nvlist_get_number_array, nvlist_get_nvlist, nvlist_get_nvlist_array, nvlist_get_parent, nvlist_get_string, nvlist_get_string_array, nvlist_move_binary, nvlist_move_descriptor, nvlist_move_descriptor_array, nvlist_move_nvlist, nvlist_move_nvlist_array, nvlist_move_string, nvlist_move_string_array, nvlist_take_binary, nvlist_take_bool, nvlist_take_bool_array, nvlist_take_descriptor, nvlist_take_descriptor_array, nvlist_take_number, nvlist_take_number_array, nvlist_take_nvlist, nvlist_take_nvlist_array, nvlist_take_string, nvlist_take_string_array, libnv, nv, nvlist, nvlist_in_array, nv
li
> st_add, nvlist_append, nvlist_get, nvlist_move, nvlis
> t_take(9) - library for name/value pairs
>
I tend to agree with John-Mark here. In my experience with man -k
most of the clutter comes from man pages installed by ports.
--
Gary Jennejohn
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20210212071348.42f61156>