This seems plausible given that I recently changed the clickable hierarchy so that that positioning sprite always has no size only the sprite with the click handler is given dimensions (drawn using the draw border primitives).
On Mar 29, 2011, at 10:23, Henry Minsky <[email protected]> wrote: > well, I tried setting to false, and IE8/DHTML context menu tests in > test/contextmenu appear to behave > the same way, so I am going to remove this quirk. > > > > On Tue, Mar 29, 2011 at 8:43 AM, P T Withington <[email protected]> wrote: > I asked this a while ago, see the thread starting here: > > http://www.openlaszlo.org/pipermail/laszlo-dev/2010-December/023673.html > > apparently there was no conclusion, other than "try removing it and see what > happens". > > On 2011-03-29, at 08:32, Henry Minsky wrote: > > > Does anyone remember what the size_blank_to_zero quirk was created for? > > > > I'm trying to fix http://jira.openlaszlo.org/jira/browse/LPP-9549 > > > > The only place that the zero height/width quirk is used seems to be in > > LzSprite.setWidth/Height , where > > the div's CSS width height get zeroed out. > > e.g., > > LzSprite.prototype.setWidth = function ( w ) { > > if (w < 0 || this.width == w) return; > > > > //Debug.info('setWidth', w); > > this.width = w; > > w = this.CSSDimension(w); > > if (this._w != w) { > > this._w = w; > > var size = w; > > var quirks = this.quirks; > > // set size to zero if we don't have either of these > > if (quirks.size_blank_to_zero) { > > if (this.bgcolor == null && this.source == null && ! this.clip > > && ! (this instanceof LzTextSprite) && ! this.shadow && ! this.borderwidth) > > { > > this.__sizedtozero = true; > > size = 0; > > } > > } > > *this.applyCSS('width', size);* > > -- > > Henry Minsky > > Software Architect > > [email protected] > > > > > -- > Henry Minsky > Software Architect > [email protected] > >
