References: <1713005794.1189899.1293724895711.JavaMail.fmail@mwmweb012>
            <19749.63653.743551.563511@morse.mittelbach-online.de>
MIME-Version: 1.0
Content-Type: multipart/alternative; boundary="0-1937116341-1294335455=:60246"
Message-ID:  <743624.60246.qm@web82001.mail.mud.yahoo.com>
Date:         Thu, 6 Jan 2011 09:37:35 -0800
Reply-To: Mailing list for the LaTeX3 project
              <LATEX-L@listserv.uni-heidelberg.de>
Sender: Mailing list for the LaTeX3 project <LATEX-L@listserv.uni-heidelberg.de>
From: Paul Thompson <patjah@SBCGLOBAL.NET>
Subject: Re: variable opt arg delimiters [was: Defining a ...]
To: LATEX-L@listserv.uni-heidelberg.de
In-Reply-To:  <19749.63653.743551.563511@morse.mittelbach-online.de>
Precedence: list
Status: R

--0-1937116341-1294335455=:60246
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable

The optional argument thing could be better handled by using named argument=
s.  =0ASo, for \parbox, which has optional and mandatory arguments, we curr=
ently have =0A=0A=0A\parbox[pos][height][inner-pos]{width}{text}=0A=0AThis =
is mindlessly confusing in many cases.  Why not have =0A=0A\parbox{pos=3Dt,=
height=3Dsomething,innerpos=3Dt,width=3Dsomething}{text}=0A=0A =0AThe use o=
f named arguments is vastly superior to positional arguments, =0Aespecially=
 when optional arguments can be specified.  =0A=0ADon't even get me started=
 on \newcommand.=0A=0APaul Thompson=0A=0A=0A=0A____________________________=
____=0AFrom: Frank Mittelbach <frank.mittelbach@LATEX-PROJECT.ORG>=0ATo: LA=
TEX-L@LISTSERV.UNI-HEIDELBERG.DE=0ASent: Thu, January 6, 2011 11:15:17 AM=
=0ASubject: Re: variable opt arg delimiters [was: Defining a ...]=0A=0AUwe =
Lueck writes:=0A> I wrote:=0A> > A better proposal than using different cat=
egory codes is =0A> > programming macros with optional arguments obeying =
=0A> > a convention that makes a choice of delimiters available to users.  =
=0A> > Any macro package replacing \@ifnextchar[ by some =0A> > \ifx\let@to=
ken\opt@arg@delim could be run this way. =0A> > =0A> > If no LaTeXY will ad=
opt this, I will call it L=FCTeX :-)=0A> =0A> I had thought of posting such=
 a "feature request" earlier, =0A> refrained from it for not having found a=
 good idea how =0A> to implement it. If I now found it so simple, I must ha=
ve =0A> overlooked a problem ...=0A> =0A>  [...]=0A> =0A> Maybe this alread=
y is in xparse? Sorry, I cannot afford more right now ...=0A=0AUwe=0A=0Axpa=
rse offer a fairly general snytax including variable delimiters and default=
=0Avalues per argument and more. Perhaps you should *indeed* have a look at=
 it=0Aone day :-) and it could be easily used on top of 2e if so desired=0A=
=0Awhat is not offered right now is mandatory arguments with arbitrary deli=
miters=0Aon the left or right, though one could achieve even those in an em=
ergency=0Ausing "u" argument type twice (I guess)=0A=0Afrank=0A
--0-1937116341-1294335455=:60246
Content-Type: text/html; charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable

<html><head><style type=3D"text/css"><!-- DIV {margin:0px;} --></style></he=
ad><body><div style=3D"font-family:verdana,helvetica,sans-serif;font-size:1=
0pt">The optional argument thing could be better handled by using named arg=
uments.&nbsp; So, for \parbox, which has optional and mandatory arguments, =
we currently have <br><br>\parbox[pos][height][inner-pos]{width}{text}<br><=
br>This is mindlessly confusing in many cases.&nbsp; Why not have <br><br>\=
parbox{pos=3Dt,height=3Dsomething,innerpos=3Dt,width=3Dsomething}{text} <br=
><div>&nbsp;<br>The use of named arguments is vastly superior to positional=
 arguments, especially when optional arguments can be specified.&nbsp; <br>=
Don't even get me started on \newcommand.<br><br></div>Paul Thompson<div><d=
iv style=3D"font-family: verdana,helvetica,sans-serif; font-size: 10pt;"><b=
r><div style=3D"font-family: arial,helvetica,sans-serif; font-size: 13px;">=
<font face=3D"Tahoma" size=3D"2"><hr size=3D"1"><b><span style=3D"font-weig=
ht:
 bold;">From:</span></b> Frank Mittelbach &lt;frank.mittelbach@LATEX-PROJEC=
T.ORG&gt;<br><b><span style=3D"font-weight: bold;">To:</span></b> LATEX-L@L=
ISTSERV.UNI-HEIDELBERG.DE<br><b><span style=3D"font-weight: bold;">Sent:</s=
pan></b> Thu, January 6, 2011 11:15:17 AM<br><b><span style=3D"font-weight:=
 bold;">Subject:</span></b> Re: variable opt arg delimiters [was: Defining =
a ...]<br></font><br>=0AUwe Lueck writes:<br> &gt; I wrote:<br> &gt; &gt; A=
 better proposal than using different category codes is <br> &gt; &gt; prog=
ramming macros with optional arguments obeying <br> &gt; &gt; a convention =
that makes a choice of delimiters available to users.&nbsp; <br> &gt; &gt; =
Any macro package replacing \@ifnextchar[ by some <br> &gt; &gt; \ifx\let@t=
oken\opt@arg@delim could be run this way. <br> &gt; &gt; <br> &gt; &gt; If =
no LaTeXY will adopt this, I will call it L=FCTeX :-)<br> &gt; <br> &gt; I =
had thought of posting such a "feature request" earlier, <br> &gt; refraine=
d from it for not having found a good idea how <br> &gt; to implement it. I=
f I now found it so simple, I must have <br> &gt; overlooked a problem ...<=
br> &gt; <br> &gt;&nbsp; [...]<br> &gt; <br> &gt; Maybe this already is in =
xparse? Sorry, I cannot afford more right now ...<br><br>Uwe<br><br>xparse =
offer a fairly general snytax including variable delimiters and default<br>=
values per
 argument and more. Perhaps you should *indeed* have a look at it<br>one da=
y :-) and it could be easily used on top of 2e if so desired<br><br>what is=
 not offered right now is mandatory arguments with arbitrary delimiters<br>=
on the left or right, though one could achieve even those in an emergency<b=
r>using "u" argument type twice (I guess)<br><br>frank<br></div></div></div=
>=0A</div></body></html>
--0-1937116341-1294335455=:60246--