芦荟和桔子们
2
Monthly Archives: October 2009
Restore Nautilus to open gnome menu after installed PCManFM
After installed PCMan File Manager with LXDE, the gnome “places” menu items were opened by PCManFM as default. To switch back to nautilus, follow these steps:
- check /usr/share/applications/nautilus-folder-hanlder.desktop
make sure the line “Exec=nautilus” is not modified. (There are lots of hack article that teach user to change default file manager by modify this) - check /usr/share/applications/defaults.list
grep the line “x-directory” by
cat /usr/share/applications/defaults.list | grep x-directory
make sure the value was kept as “nautilus-folder-handler.desktop”, if not, restore it.
Then grep the line “inode” by
cat /usr/share/applications/defaults.list | grep inode
also be sure about the result “nautilus-folder-handler.desktop”. if not, restore it again. - So it must be your own configuration that has been changed. check ~/.local/share/applications/defaults.list
grep the two lines described above, it must be changed to “pcmanfm-folder-handler.desktop”, so replace them by “nautilus-folder-handler.desktop”
To check the result, click menu item at once (no restart or logout needed), again, you see the slow file manager…
Yan Captcha Service
I’d like to announce my recent works, a project called Yan Captcha Service written in Java which is aimed to provide whole solutions of captcha for your websites. It will be very easy to use the service because 1. interfaces are based on plain http url; 2. different kinds of usage are supported to fit your requirements; 3. the architecture is open so you can add your own solid implementation of captcha; 4. less coupling with your system. And designed for scalability, currently, it applies JGroups to share sessions (memcached support will be added soon), thus you can setup a cluster for the service.
As you may know, the open-source project is split from my current work-time project because it is more closed to my idea. However, soon we will have the product opened to our thousands (millions ?) of users. I can gain feedback from the challenge and improve the open-source edition.
The code is maintained by open-source scm, mercurial (also known as hg). The project is now hosted on bitbucket.org. You can clone the code to local via:
$ hg clone https://sunng@bitbucket.org/sunng/yan/
To build the project, run this command in root of project directory:
mvn install
To run in a develop environment:
mvn jetty:run
Issue reporting and patches are always welcomed.
Check the wiki pages for more information about the project:
http://bitbucket.org/sunng/yan/wiki/Home
Open Source Charting Toolkits in JavaScript
项目的下一阶段，需要做一些数据展现方面的工作，处于对Adobe产品的抵触情绪，我个人还是倾向于用JavaScript来完成。今天找了几个画Chart的库，比较一下功能和编程接口的使用，为接下来的开发做一些准备。
dojox.charting
(BSD License)
没什么悬念，第一个想到的就是曾经用过的dojo。之前实习时候用dojo画chart也算是积累了一些心得。dojox.charting的最主要优点是编程接口完善、全面，可以配置的项目很多，接口易于编程，易于动态地生成、操作Chart。而另一方面最主要的问题就是文档比较匮乏，不了解的人可能问了，doc.dojotoolkit.org好强大，怎么会文档匮乏呢。其实是dojo和mootools类似，都很喜欢用option object来传递可选参数，但是这些option object在文档中没有任何涉及，很多关键的属性都在其中但是却没有办法查到，算是美中一大不足。
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<script src="dojo-release-1.3.2/dojo/dojo.js" djConfig="isDebug: true"></script>
<title>Dojo</title>
<script type="text/javascript">
dojo.require("dojox.charting.Chart2D");
</script>
</head>
<body>
<h1>Dojox.charting</h1>
<div id="charting" style="width:400px; height:250px;">
</div>
<script type="text/javascript">
（function charting(){
dojo.empty("charting");
var data = [2,4,5,7,7,4,4,7];
var chart = new dojox.charting.Chart2D("charting");
chart.addPlot("default", {type: "Lines", markers: true, shadows: {dw:2, dy:2, dx:2}});
chart.addPlot("additional", {type: "Areas"});
chart.addPlot("other", {type: "ClusteredColumns"})
chart.addAxis("x");
chart.addAxis("y", {vertical: true, min:0, max: 10});
chart.addSeries("Testing Data 1", data);
chart.addSeries("Average", dojo.map(data, function(it){return (Math.sin(it)+Math.random())*4}), {
fill: "rgba(145, 213, 100, 0.5)", plot:"additional"
});
chart.addSeries("All Random", dojo.map(data, function(it){return (Math.random()*5)}), {
plot: "other", fill: "#AEC6E2", stroke:{color:"#333", width:1}
});
chart.render();
}）();
</script>
</body>
</html>
Protovis
(BSD License)
Protovis，看域名就知道是斯坦福的项目，Protovis是在prototype基础上，不仅有charting的功能，还包括其他可视化方法的实现。看了Demo会发现它真的很不错，不过Protovis的本意似乎只是借助浏览器的图形功能实现可视化，它并没有提供超出可视化范畴的其他东西，甚至看过实例的代码你会发现，作者是不折不扣的函数式编程和链式方法的爱好者，Protovis的代码算是出神入化了，如果你想看看学术的JavaScript是如何写的，那就看看这个发表了论文的JavaScript库吧。
看起来是很出神入化，可是这样的编程接口如果要用在动态的数据展现上，实在是难以想象。况且，pv.Panel居然没有传入dom对象的接口，无从控制chart的生成位置。说白了，Protovis并不是一个满足项目需要的Library，不过它的功能和代码还是值得多看几眼的。
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<script src="protovis-3.1/protovis-r3.1.js"></script>
</head>
<body>
<div id="page">
<h1>Protovis</h1>
<div style="border:1px #CCC solid;">
<script type="text/javascript">
var width = 300;
var height = 225;
var data = [1,2,3,4,6,7,5,8,9];
var dataMax = data[0];
for(var i=1; i<data.length; i++){
dataMax = dataMax > data[i] ? dataMax : data[i];
}
var x = pv.Scale.linear(0, dataMax).range(0, width);
var y = pv.Scale.ordinal(pv.range(10)).splitBanded(0, height, 0.9);
var charting = new pv.Panel();
charting.width(width).height(height).left(20).top(20);
var bar = charting.add(pv.Bar);
bar.data(data).top(function() y(this.index)).height(y.range().band).left(0).width(x);
bar.anchor("center").add(pv.Label)
.textStyle("white")
.textAlign("center")
.text(function(d) d.toFixed(1));
charting.add(pv.Rule)
.data(x.ticks())
.left(function(d) Math.round(x(d)) - .5)
.strokeStyle(function(d) d ? "rgba(255,255,255,.3)" : "#000")
.add(pv.Rule)
.bottom(0)
.height(5)
.strokeStyle("#C00")
.anchor("bottom").add(pv.Label)
.text(function(d) d);
charting.render();
</script>
</div>
</div>
</body>
</html>
Flotr
(MIT License)
不同于上面两个库通过SVG绘图，Flotr是用canvas api来完成绘制的。与dojo相比，Flotr的API比较简单，主要只有一个Flotr.draw方法，可以传入三个参数：目标DOM对象，数据和绘图选项。和dojo不同，Flotr的文档里关于这个option object有非常明确的文档。另外强大的是Flotr还提供了一些交互的功能，支持事件绑定。
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<script src="flotr-0.2.0-alpha/flotr/lib/prototype-1.6.0.2.js"></script>
<script src="flotr-0.2.0-alpha/flotr/flotr-0.2.0-alpha.js"></script>
<title>Flotr</title>
<script type="text/javascript">
</script>
</head>
<body>
<h1>Flotr</h1>
<div id="charting" style="width:400px; height:250px;">
</div>
<script type="text/javascript">
(function d(){
var series1 = {data: (function(){
var d = [];
for(var i=0; i<10; i+=0.5){
d.push([i, Math.random()-0.5+Math.sin(i)*3]);
}
return d;
})(), label:"d1", lines:{fill:true}};
var series2 = {data: (function(){
var d = [];
for(var i=0; i<10; i+=0.5){
d.push([i, (Math.random()-0.5)*2+Math.cos(i)*2]);
}
return d;
})(), label: "s2", points:{show:true}, lines:{show:true}};
Flotr.draw($("charting"), [series1, series2], {
legend: {position :"se"},
yaxis: {min: -3.5, max:3.5}
});
})();
</script>
</body>
</html>
此外，还有一个与Flotr很类似的Protochart，后者的网站上写自己是motivated by Flotr，不过我很难想象功能几乎相同是如何motivated的。
MilkChart
(Apache License 2.0)
MilkChart是一个建立在Mootools上的Charting库，采用canvas api绘图。它的特点是把HTML Table的数据绘制成Chart，如果场景允许，倒是省去了不少构建数组的麻烦。这个项目目前规模还不大，并且在作者积极的开发中，可以参考它的Wiki获得详细的使用方法。
Time Tracking and Task Management with TaskCoach
TaskCoach其实是老朋友了，wxpython的，跨平台，最初只是TODO LIST，后来又加进了Time Tracking的功能。早先给frank05推荐过，不过在学校的时候，在电脑上的时间往往是支离破碎的，没有什么tracking的价值。
现在不同了，工作以后每天在电脑前面坐至少8个小时，如果没有点控制，时间一样是像以前一样莫名其妙地溜走了。去年刚开始实习的时候，我大概有个习惯是把每天的TODO写在本子上，一来当时清楚，二来时间长了也有数。今年工作以后一开始也是这样，但是时间长了发现只用笔来记录功能还是太有限，比如事情做了一半pending了，经常是画个圈，但是时间长了还是忘掉了。而进一步想做Time Tracking就更麻烦了。
上周想了一下确实需要个工具，这周一开始就重新开始在开发用的Windows机器上用TaskCoach了。有将近一年的时间没关注这个软件，版本号虽然没有什么大的突破，不过用起来倒是更顺手了。现在可以最小化到系统栏，开机自动启动。本身的功能除了支持多层的Task管理，Time Tracking，还可以顺手记一些Notes，另外还有定时的事件提醒，也不会错过总也想不起来的培训了。
推荐给大家在Windows上用。