Re: Minutes of the Nov 8 2007 teleconference

[Eric Blake]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

This came up on the Austin group:

According to Andrew Josey on 11/9/2007 4:44 AM:
> XCU ERN 107 touch Accept as marked below
> 
> We revised the response as follows:
> 
>  On XCU page 920 line 35645, change:
> 
>          touch [-acm] [-r ref_file | -t time] file...
> 
>  to:
> 
>          touch [-acm] [-r ref_file | -t time | -d date_time] file...

It's nice that the next revision of POSIX will be adding -d support to
touch, but it means that we need to fix getdate.y to conform (the Austin
proposal is for touch(1), but fixing getdate.y will also benefit date(1)):

> 
>  After line 35674 add:
> 
>     -d date_time
> 
>        Use the specified date_time instead of the current time.
>        The option-argument shall be a string of the form:
> 
>          YYYY-MM-DDThh:mm:SS[.frac][tz]
>        or
>          YYYY-MM-DDThh:mm:SS[,frac][tz]
> 
>        where:
> 
>          YYYY are at least four decimal digits giving the year,
> 
>          MM, DD, hh, mm, and SS are as with -t time,
> 
>          T is the time designator, and can be replaced by a single space,

This is the biggest problem.  To date (pardon the pun), getdate.y parses T
as a single-letter military timezone, rather than trying to disambiguate
whether it could also represent the ISO date/time separator.  Someone's
going to have to figure out how to fix the parser to support the new POSIX
parsing rules.

> 
>          [.frac] and [,frac] are either empty, or a period ('.') or comma
>          (',') respectively followed by one or more decimal digits,
>          specifying a fractional second,
> 
>          [tz] is either empty, signifying local time, or the
>          letter 'Z', signifying UTC.  If [tz] is empty the resulting
>          time shall be affected by the value of the TZ environment
>          variable.
> 
>        If the resulting time precedes the Epoch, the behavior is
>        implementation-defined.  If the time cannot be represented as
>        the file's timestamp, 'touch' shall exit immediately with an
>        error status.
...


[I also found the following part of the proposal rather amusing, where the
examples picked on some of the more frequent Austin group contributors -
the open source community will probably recognize a few of these sample
file names]

> 
> Add to EXAMPLES
> 
...
> 
> Create or update a file called "drepper"; the resulting file has both the
> last data modification and last data access timestamps set to 
> November 12, 2007 at 10:15:30 local time :
> 
> touch -t 200711121015.30 drepper
> 
> Create or update a file called "ebb9"; the resulting file has both the
> last data modification and last data access timestamps set to 
> November 12, 2007 at 10:15:30 local time :
> 
> touch -t 0711121015.30 ebb9
> 
> Create or update a file called "eggert"; the resulting file has the
> last data access timestamp set to the corresponding time of the file named 
> "mark" instead of the current time. The last data modification time is set
> to the current time :
> 
> touch -a -r mark eggert
> 

- --
Don't work too hard, make some time for fun as well!

Eric Blake             address@hidden
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (Cygwin)
Comment: Public key at home.comcast.net/~ericblake/eblake.gpg
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFHNh3384KuGfSFAYARAo9GAJ4xhe1HZ8ba7C8ucmB8xxevvJydCwCfVQ2S
YN4oEp8p07W4mipuv9nqpCw=
=ax1L
-----END PGP SIGNATURE-----