- 26 Nov, 2013 3 commits
-
-
David Schnur authored
Don't add padding when there's no last tick.
-
David Schnur authored
Default lineHeight based on the font size.
-
David Schnur authored
Distinguish between the first and innermost axis.
-
- 25 Nov, 2013 3 commits
-
-
David Schnur authored
Flot 0.8.0 used the default font size, typically derived from the placeholder, as the basis for the default lineHeight. This produced incorrect results when a font.size was provided explicitly, and it differed from the placeholder’s CSS size. Fixed by waiting to default lineHeight until the actual font size has been resolved. Fixes #1131.
-
David Schnur authored
Flot 0.8 added logic to account for the size of axis tick labels and add padding around the edges of the plot, to prevent long labels from sticking out. But it padded both sides equally, which is incorrect if the right/top side has no last axis label. Fixed by allocating padding per-side, and checking whether the last label would be shown before padding the top or right. Fixes #1048.
-
David Schnur authored
The fix for #1056 caused a regression where grid lines were drawn for the innermost axes on both sides instead of just the first axis. Fixed by properly distinguishing the first axis in each direction from the innermost one on each side. Fixes #1075.
-
- 24 Nov, 2013 1 commit
-
-
David Schnur authored
Added plot.destroy() method, to properly destruct and release memory of plots.
-
- 14 Nov, 2013 1 commit
-
-
David Schnur authored
-
- 10 Nov, 2013 1 commit
-
-
David Schnur authored
:not(selector , selector) isn't actually a valid CSS selector
-
- 05 Nov, 2013 3 commits
-
-
Anthony Ryan authored
jQuery's Sizzle is the only thing that this actually works in so jQuery will fail when it passes this directly to querySelectorAll()
-
David Schnur authored
-
David Schnur authored
Add Ruby examples
-
- 04 Nov, 2013 4 commits
-
-
Mike Połtyn authored
-
Mike Połtyn authored
-
David Schnur authored
-
David Schnur authored
Fixed a bug where plotting a chart crashes if the placeholder doesn't have a font size
-
- 03 Nov, 2013 2 commits
-
-
David Schnur authored
-
David Schnur authored
Fixed Issue #1159 - Bug Redrawing Legend when using custom container
-
- 02 Nov, 2013 1 commit
-
-
David Schnur authored
-
- 30 Oct, 2013 1 commit
-
-
David Schnur authored
fix label's maxWidth calculation
-
- 29 Oct, 2013 1 commit
-
-
Benjamin Gram authored
-
- 14 Oct, 2013 3 commits
-
-
David Schnur authored
-
David Schnur authored
Also stop at root when extracting CSS color
-
Luis Silva authored
-
- 11 Oct, 2013 1 commit
-
-
Luis Silva authored
-
- 23 Sep, 2013 2 commits
-
-
David Schnur authored
-
David Schnur authored
-
- 21 Sep, 2013 1 commit
-
-
execjosh authored
This change adds an additional check for whether the parent element is `null` or `undefined` in `$.color.extract`. This can happen when working with elements that have not yet been added to the DOM under `<body>`. Consider the following example pie chart. var elm = $("<div />") .css({ width: "240px" , height: "320px" }) var data = [ {label: "One", data: "33"} , {label: "Two", data: "33"} , {label: "Three", data: "33"} ] var opts = { legend: { show: true } , series: { pie: { show: true } } } $.plot(elm, data, opts) elm.appendTo($("body")) When flot inserts each legend row, it tries to use the same color as the corresponding graph part, unless it was explicitly specified in the options. However, in this example, `$.color.extract` runs into an unexpected `null` reference because `<body>` is not an ancestor of `elm`. Specifically, a `TypeError: Cannot read property 'nodeName' of undefined` would be thrown.
-
- 15 Sep, 2013 2 commits
-
-
David Schnur authored
The pie plugin was a little too clever in its use of closures. In processDatapoints it set canvas, target, and options for use in other functions. Since options was not declared this meant that it became global. Pages containing multiple pie plots therefore saw a range of weird effects resulting from earlier plots receiving some of the options set by later ones. Resolves #1128, resolves #1073, resolves #1055.
-
David Schnur authored
If an explicit numeric offset was provided, we should not override it. The clipping is only meant to apply to the case where the center is moved to make room for the legend in 'auto' mode, anyway.
-
- 13 Sep, 2013 1 commit
-
-
Craig Oldford authored
-
- 11 Sep, 2013 1 commit
-
-
David Schnur authored
Link to PLUGINS.md
-
- 09 Sep, 2013 1 commit
-
-
Thodoris Greasidis authored
-
- 04 Sep, 2013 1 commit
-
-
Brian Peiris authored
-
- 01 Sep, 2013 1 commit
-
-
David Schnur authored
Instantiate barLeft/barRight in Bar Highlighting Section
-
- 29 Aug, 2013 1 commit
-
-
Eric Byers authored
-
- 09 Aug, 2013 1 commit
-
-
David Schnur authored
The autoscale check was too broad; it included the case where autoscale was undefined. This resulted in axes not expanding correctly when coordinates at the end of a series had null x or y values. Fixed by narrowing the check to !== false; resolves #1095.
-
- 22 Jul, 2013 3 commits
-
-
David Schnur authored
The drawBar method was always called with an offset of zero, and I see no other way in which it is currently used. Resolves #382.
-
David Schnur authored
This is up to 2x faster and appears to work around issues in Chrome's canvas implementation that sometimes result in bars not being filled. Resolves #915.
-
David Schnur authored
Fixing Issue 686: Tooltip bug
-