On Wed, Aug 30, 2023 at 11:21 PM Linus Torvalds
<torvalds@xxxxxxxxxxxxxxxxxxxx> wrote:
> I've pulled this, but I think the new NLS_UCS2_UTILS config option
> shouldn't be something that is asked about. The filesystems that want
> it already select it, and users shouldn't be asked about a module with
> no use.
>
> The way to do that is to simply not have a user query string for it,
> ie instead of
>
> config NLS_UCS2_UTILS
> tristate "NLS UCS-2 UTILS"
>
> it could be (an dI think should be) just
>
> config NLS_UCS2_UTILS
> tristate
>
> which tells the config system not to ask users about it.
>
> Because users really shouldn't be asked questions that there is no point in.
Agree. Will fix
> And then, on a purely visual commentary about your pull request -
> lines like these are just noise:
>
> > fs/{smb/server/uniupr.h => nls/nls_ucs2_utils.c} | 156
> > +++++-------------------------------------
> > fs/nls/nls_ucs2_utils.h | 285
> > ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>
> and the problem seems to be that you generate the diffstat in a very
> wide terminal (where git tries to be helpful and give you lots of
> detail), and then you cut-and-paste the result.
>
> If you pipe it to a tool instead (xsel, perhaps), git will limit the
> width of the diffstat to something sane.
Yes - today my terminal was set at about 180 character width. Just
tried an experiment and piping to xsel did fix the output. Will
update my scripts.
--
Thanks,
Steve
