On 07-Sep-2014 02:45, alvinpenner wrote:

two comments: the pre-existing bug has been reported at Bug 307656 and Bug 1366434. It appears to be related specifically to the rectangle tool F4. However there are two new regressions which are not related to F4.

First one: rev 13544 has broken one of the fixes for Bug 1240455, comment 24: https://bugs.launchpad.net/inkscape/+bug/1240455/comments/24

using Extensions->Render->Grids->Grid and using the default spacing of 10 px.

• case 1 : open Inkscape with default units of px, draw default Grid
• case 2 : open Inkscape, change units to mm, draw default Grid

the expectation is that these two grids will look the same relative to the page size. Rev 13534 shows the expected behavior, rev 13546 does not.

Second one:

• open Inkscape, draw two objects, a rectangle and a star, or ellipse
• select both of them
• change document units
• the expectation is that the shapes should not change size
• Rev 13534 shows the expected behavior, rev 13546 does not.

I will revert that change in a little bit. Actually I'm going to add a parameter to the call, so that it can work one way (so that it works like it does now) for a call from EMF import, and as it did before when called from elsewhere (only one other place that I know of.) This is a short term hack until the actual problem is sorted.

And I think the actual problem is in those parts of Inkscape which fail when the scale is applied to a top group. It is perfectly valid SVG to do:

<g (scale to whatever conversion one wants)> (bunch of objects in completely arbitrary units. </g>

All transform operations up the stack must be applied to an object. If some part of Inkscape only shows units properly when the top level scale is not present, then it is not correctly applying those transforms. How is it not that part of Inkscape that is broken?

Regards,

David Mathog mathog@...1176... Manager, Sequence Analysis Facility, Biology Division, Caltech