Permalink
Commits on Sep 5, 2011
Commits on Aug 17, 2011
  1. Opera support.

    joseph committed Aug 16, 2011
Commits on Aug 16, 2011
  1. Minor test nav tweak.

    joseph committed Aug 16, 2011
Commits on Aug 7, 2011
  1. Retiring framer, since Monocle now solves that problem a different way.

    joseph committed Aug 7, 2011
    (That is, because Monocle loads components into iframes, there isn't as
    much of a need to load Monocle into an iframe. If you do need to do
    this, it's probably better to do it yourself.)
Commits on Aug 4, 2011
Commits on Aug 1, 2011
Commits on Jun 5, 2011
  1. Recover reasonably gracefully if component fails to load.

    joseph committed Jun 5, 2011
    A component is considered to have failed to load if getComponent returns
    false (or the callback is called with false as the argument).
    
    Monocle will dispatch a "monocle:componentfailed" event and try to load
    the last good component instead. Listen for "monocle:componentfailed" on
    the reader to alert the user to the problem.
  2. JavaScript load strategy for components.

    joseph committed Jun 5, 2011
    If you return content from your bookdata getComponent like this:
    
        { javascript: "... a string ..." }
    
    ... it will be loaded much like if you had just returned a string or
    a html result. The difference is that regexes won't be run over the
    string to make it 'safe' in a JavaScript statement before it is loaded
    into the component frame each time. This will speed things up if you are
    storing components locally — clean them once before storing them, so
    that Monocle doesn't have to clean them up every time.
Commits on May 12, 2011
  1. Initial styles are a separate stylesheet, not appended to defaults.

    joseph committed May 12, 2011
    This is so you can get a reference to the initial stylesheet
    (reader.properties.initialStyles) and update it without affecting
    Monocle's required, essential styles.
  2. Not everything (notable Android) has XPath support.

    joseph committed May 12, 2011
    For these user agents, an xpath locus always points to page 1 of the
    component.
Commits on May 2, 2011
  1. README explains the edge branch.

    joseph committed Oct 29, 2010
Commits on Mar 27, 2011
  1. New controlType 'hud' allows multiple controls to use the overlay.

    joseph committed Mar 27, 2011
    ...which is what 'modal' used to do. Now 'modal' only allows one control
    using the overlay at a time. Reader#showControl returns true if
    allowed, and false if disallowed.
Commits on Mar 24, 2011
Commits on Mar 23, 2011
  1. A fix for Firefox4 beachballs on the first page turn.

    joseph committed Mar 23, 2011
    Only seems to arise if a locus has been given to the place argument
    during reader initialisation.
  2. Monocle.Place#percentAtTopOfPage

    joseph committed Mar 22, 2011
    Part of twin methods:
    
       place.percentAtTopOfPage()
       place.percentAtBottomOfPage()
    
    As before, place.percentageThrough() returns the percent at the bottom
    of the page, and is now just an alias to that method.
    
    Some additional comments in src/place.js for clarity.
Commits on Mar 22, 2011
Commits on Mar 12, 2011
  1. A workaround for the Chrome 10 blank pages bug.

    joseph committed Mar 12, 2011
    Discussion of bug is here:
    
       https://github.com/joseph/monocle/issues#issue/25
    
    The bug affects WebKit nightlies since December 2010 and Chrome since
    version 10. Reasonable expectation of fix in Chrome 11, and that it
    won't ever appear in Safari.
    
    The workaround is problematic because:
    
    * it significantly modifies the DOM of the component
    * it hides all scrollbars within the component
    
    But it's a lot better than blank pages. Thanks to madfarmer for
    identifying the workaround.
Commits on Mar 8, 2011
Commits on Mar 5, 2011
  1. Make the default style rules harder to accidentally override.

    joseph committed Mar 5, 2011
    Encountered a style: .bodycenter { margin: 25% !important; }, that broke
    the body element because it overrode the default style body { margin: 0
    !important; }. By making our rule more specific, it's less likely that
    this sort of thing will happen.
    
    Of course, it remains an arms race. Applying inline styles to every
    single element is the nuclear option -- there's no intention of doing
    that though.
Commits on Mar 3, 2011
  1. Stencil default click behaviour affects links in interactive mode too.

    joseph committed Mar 3, 2011
    This was temporarily disabled so that any click listeners on the
    original link could perform the handling. But of course then you don't
    get nice skipToChapter overrides.
    
    So now the links in the original documents have a stencilClickHandler
    property added to them, which you can deafen if you don't want the
    default stencil click behaviour in interactive mode.