Re: Patches to findutils 4.9.0-1's updatedb to do locking, allow filenames with spaces & progress monitoring, exclude /dev on Cygwin, etc.

[Dan Harkless <cygwin-list21@harkless.org>]

On 2/27/2022 3:54 AM, Bernhard Voelker wrote:
> On 2/24/22 17:32, Dan Harkless wrote:
>> I'm finally getting around to sending in a patch (to bug-findutils and
>> the Cygwin list, to which I'm currently subscribed) to address these
>> issues, along with some others, a few of which represent small changes
>> in behavior:
> Thanks for the patch ... but:

Thanks for taking a look at it.

> a) The patch does not cleanly apply:
>
>    ~/findutils/locate> patch -t < /tmp/updatedb.patch
>    patching file updatedb
>    Reversed (or previously applied) patch detected!  Assuming -R.
>    Hunk #2 succeeded at 47 with fuzz 2.
>    Hunk #7 FAILED at 167.
>    Hunk #8 FAILED at 202.
>    Hunk #9 succeeded at 217 (offset -2 lines).
>    Hunk #10 succeeded at 276 (offset -2 lines).
>    2 out of 10 hunks FAILED -- saving rejects to file updatedb.rej

Ah.  As I mentioned, my patch was against Cygwin's findutils 4.9.0-1, 
and since my Linux systems use a different version of locate, I hadn't 
tested there (nor did I have time to look at the original 4.9.0 
source).  I'd been hoping any Cygwin patches wouldn't invalidate it; pity.

> b) The patch changes the file 'updatedb' which is created at build time
> instead of the file 'updatedb.sh' which is under version control.

Gotcha.

> c) The description says that there are 10 more or less non-trivial
> changes in it.  A squashed diff of 500 lines on a file with 342 lines
> makes reviewing and discussing of each topic impossible.

Impossible?  Since the bulk of the changes are spacing changes to 
standardize code indentation across the file, I was hoping the separate 
'diff -uw' listing would be sufficient to enable easy review and discussion.

> Would you mind re-sending as separate Git patches?

It'll be awhile before I'll have time to get set up to do that, but will do.

> Have a nice day,
> Berny

Thanks again,
Dan Harkless
http://harkless.org/dan/