On Mon, Sep 22, 2003 at 05:08:16PM +0300, Martin Vermeer wrote: > On Mon, Sep 22, 2003 at 02:53:54PM +0200, Andre Poenitz spake thusly: > > What's the real world problem you are currently trying to solve? > > The one the (tested, working) solution of which is the following: > > (lyx_sty.C) > > string const boxwidth_def = > "\\newlength\\w\n"; > > string const fboxwidth_def = > "\\newcommand\\setfboxwidth[0]{%\n" > "\\setlength\\w{\\columnwidth}\n" > "\\addtolength\\w{-2\\fboxsep}\n" > "\\addtolength\\w{-2\\fboxrule}}%\n"; > > string const ovalboxwidth_def = > "\\newcommand\\setovalboxwidth[0]{%\n" > "\\setlength\\w{\\columnwidth}\n" > "\\addtolength\\w{-2\\fboxsep}\n" > "\\addtolength\\w{-0.8pt}}\n"; // 2x \thinlines width > > and (insetbox.C): > > if (pt == "Boxed") { > os << "%\n\\par\\setfboxwidth\\fbox{\\parbox{\\w}{%\n"; > } else > os << "%\n\\par\\setovalboxwidth\\ovalbox{\\parbox{\\w}{%\n"; > > ... > > void InsetBox::validate(LaTeXFeatures & features) const > { > features.require("boxwidth"); > if (params_.type == "Boxed") { > features.require("fboxwidth"); > } else { > features.require("fancybox"); > features.require("ovalboxwidth"); > } > inset.validate(features); > }
I still don't see the problem (except that a global length '\w' is a bit non-descriptive, '\lyxboxwidth' etc might be better. In fact I'd think something along the lines of \usepackage{fancybox} \usepackage{calc} \par\fbox{\parbox{\columnwidth-2\fboxsep-2\fboxrule}{xxxx}} \par\ovalbox{\parbox{\columnwidth-2\fboxsep-0.8pt}{xxxx}} might be better. Andre' -- Those who desire to give up Freedom in order to gain Security, will not have, nor do they deserve, either one. (T. Jefferson or B. Franklin or both...)