[cups.development] [RFE] STR #2844: Support for Wildcards in CUPS
Jerome Alet
alet at librelogiciel.com
Mon Jun 2 10:21:12 PDT 2008
On Mon, Jun 02, 2008 at 09:41:00AM -0700, Michael Sweet wrote:
>
> [STR Closed w/o Resolution]
>
> Given that the shell will try to interpret the wildcards first (requiring
> the user to escape any wildcards),
Yes, just like with ANY other command line tool, CUPS is not different.
This is what makes wildcards great : you can have the shell expand them
when needed, or the application if you prefer/need to.
> and given that *, ?, [, and ] are currently valid printer name characters,
That's exactly why I said this "all characters are valid" feature
has to die. Only allowing * and ? [ and ] would solve many many
problems already, while only sacrifying four characters which are
probably almost never used as part of a real print queue name.
> and (finally) given that you can
> filter the results with grep and other commands, I don't see us ever
> adding support for printer name wildcards in CUPS.
No, filtering the result with grep doesn't help at all, unless you want
and enjoy to write 4 commands pipes each time you want to remove all jobs
from only a few printers...
In addition, filtering with grep means that your request would have
to be done on all printers, then filtered on the client. This is
highly inefficient as far as network bandwidth is concerned, and
probably also would require much more CPU on the server (to act on
all printers each time and send appropriate results each time, over
tens of IPP requests) instead of filtering first and acting on a
filtered list, with a single (or a few) IPP requests.
Maybe I don't see the Big Picture (tm), but IMHO you are sacrificing
the power of the *nix command line for bad reasons : 1 : to please
*nix newbies, 2 : to not remove four probably unused characters from
the set of allowed ones, 3 : some client side actions are not
possible without extensive, and expensive, client side piped
commands, multiplying IPP requests for no valid reason.
Jerome Alet
More information about the cups-devel
mailing list