Re: Setting font to Lucida Grande on Mac OS X

[Luc Teirlinck <teirllm@dms.auburn.edu>]

Richard Stallman wrote:

	 However, matching is
       ! implementation-dependent, and often inaccurate, when wildcards match
       ! dashes in a long name.  It works best if you supply all 14 dashes.

   The first sentence seems to imply that wildcards are so unreliable
   that you should never use them.  It is clearer to say

	 However, matching is
	implementation-dependent, and can be inaccurate when wildcards match
	dashes in a long name.  For reliable results, supply all 14 dashes
	and use wildcards only within a field.

and Stephen J. Turnbull replied:

   The second sentence is a distinct improvement (especially as it
   implies partial matching in a field, which often is useful, such as
   ...-jisx0208*-0 to catch names which specify the year of the
   standard), but I recommend that the first sentence be left as is.

I believe the difference of opinion is limited to:

"often inaccurate" (Stephen's version) vs "can be inaccurate"
(Richard's version).

I believe that we all agree with what is meant to be communicated,
namely:

1.  It is OK to try out, say:

M-x set-frame-font RET *-courier-medium-r-*-14-*

Why type fourteen dashes if you can get what you want with the above?
(This works for me.)

2.  If you do

M-x set-frame-font RET -abisource-courier-bold-i-normal--17-*

and get the error message:

Font `-abisource-courier-bold-i-normal--17-*' is not defined

(as I do), then this does _not_ mean that there really is no such font.

In fact, after supplying the fourteen dashes I get the font:

-abisource-courier-bold-i-normal--17-135-90-91-p-100-iso8859-1

3.  Always supply fourteen dashes for Elisp code, except when only
    meant for personal use.  All XLFD's I found in the Emacs
    source code actually adhere to this rule.

To me, both versions suggest 1-3 above, even though Richard's puts
somewhat more emphasis on (1) and Stephen's on (2) and (3).  To me the
difference does not seem terribly big.  My patch currently contains
Richard's version, but I will wait at least 24 more hours before
committing it, so everybody can make further comments.  (Of course,
further changes can also be made after I commit.)

Sincerely,

Luc.