Re: [PATCH 08/19] mkfs: getbool is redundant

[Jan Tulak <jtulak@xxxxxxxxxx>]

On Thu, Apr 7, 2016 at 7:25 PM, Eric Sandeen <sandeen@xxxxxxxxxxx> wrote:

"Many options allow for an optional argument of 0 or 1, ..."

> +disable or enable the functionality, in a forward-compatible syntax.

What does "forward-compatible syntax" mean? I'm not sure that clarifies
anything for the reader.

âYeah, I should reformulate it, I think. The meaning is that it won't matter what the defaults are now, or will be in the future. E.g., if you had a script creating a fs without crc before, when it was disabled by default, and we changed the default, you are now creating with the crc. But if you give it -m crc=0, then no matter what the default is, you have it always disabled.

How about changing the line to "Boolean options allows for optional argument of value 0 or 1, to explicitly disable or enable the functionality," and dropping the forward-compatible part?

Â

Otherwise this looks ok to me; Dave explained that it is intentional to
make every single option accept a value, whether it is now
boolean or a numeric value, so there is no such thing as a bare "--flag"
anymore; such flags are always "--flag [0|1]" now, right?

âFor options inside of -m, -d and such, yes. Top-level flags, that is -f, -q, -N, -K and -V, are still only flags, but these don't change the FS attributes. They are something different from the other. Still, I wonder whether they should accept [0|1] too...

â

Â

> @@ -312,10 +315,12 @@ and
>Â .B swidth
>Â values.
>Â .TP
> -.BI noalign
> -This option disables automatic geometry detection and creates the filesystem
> +.BI noalign[= value ]
> +If the
> +.I value
> +is 1, this option disables automatic geometry detection and creates the filesystem
>Â without stripe geometry alignment even if the underlying storage device provides
> -this information.
> +this information. 0 disables this option, that is, enables automatic detection.
>Â .RE
>Â .TP
>Â .B \-f

The realtime section of the manpage needs this same fixup now too, I guess.

Doneâ

Â

Thanks,
-Eric

_______________________________________________
xfs mailing list
xfs@xxxxxxxxxxx
http://oss.sgi.com/mailman/listinfo/xfs

âThanks,

Janâ

--

Jan Tulak

jtulak@xxxxxxxxxxÂ/ jan@xxxxxxxx