them all with one regexp.
Fix: correctly detect android platform.
Fix: prevent minimizer from messing up index settings code.
… skip fallbacks for figures that have already been used.
Many devices set navigator.platform to "Linux armv7l", so that they end up having class os-linux instead of os-android.
[#12] Support for webfonts.
Issue 298 one more fix
…oader to prevent treesaver from unbooting if font loading takes too much time.
The purpose is to have a reliable way to detect which fonts are ready when treesaver starts, and to prevent fonts loaded afterward from messing the measurements up. The actual font loading is delegated to an adapter; currently, the only provided one is googleadapter, which uses WebFont Loader.
Fix octal literals to work with recent versions of CoffeeScript.
Fire a custom event right after showing the loading message.
The Grid's line height is also passed to Page, while the Content's line height is also inherited by Block's. As long as Grid and Content agree on how much rounding they apply, we should be golden. More testing needed on IE9.
Fix issue 298: support browsers that report sub-pixels
…an unpredictable way; let's round them off. For example, when you have two columns with the same (fractional) width and different (fractional) margins, Chrome will report widths that may differ by less than 10^-5.
…, since we use it as the basic unit we base all calculations on. Just round it up to the next whole pixel. This may not be ideal, maybe we do want fractions; but it's better to be safe than sorry, at least for the time being.
…strating the failure on fractional numbers.
…lso affects Metrics, since it uses getOffsetWidth(). element.getBoundingClient originated on IE, and is implemented by every browser we are compatible with.
* consistently use treesaver.dimensions.Metrics to compute the column width, as well as the content width (since the two must be consistent); * store it on the column itself; * read it from there when validating that they are consistent. The main goal here is to a single method that knows how to properly compute the width.
… fails at this time on browsers that support sub-pixels. While here, also add a sanity test for expected validation errors.
Tested on: * Safari 6 Mac (no subpixels); * Chrome 21, 22 and 23 Mac (subpixels); * Firefox 14 Mac (subpixels)
[#295] Fix blocks traversal in atEnd() to account for block children.
[#290] When looking for the next block for keepwithnext we need to exclude fallback content for figures that have already been placed.
…clude fallback content for figures that have already been placed.