'scribbling'에 해당되는 글 409건

  1. 2010.11.04 QWIKI
  2. 2010.11.01 요즘 Trend
  3. 2010.10.25 Meego
  4. 2010.10.21 Richard Stevens의 죽음
  5. 2010.10.20 한국의 운영체제 정보
  6. 2010.10.07 해외 논문 보기 (공짜로)
  7. 2010.09.17 Provisioning
  8. 2010.09.01 Javascript dom 처리를 위한 API (DOM Element methods)
  9. 2010.08.31 Graph Database and products 소개.
  10. 2010.08.24 Wired-The web is dead.

QWIKI

scribbling 2010. 11. 4. 17:09

http://qwiki.com 이라는 싸이트다.



개인과 소셜 정보를 멀티미디어로 보여준다.
플래쉬로 만들어져서 멀티미디어(동영상, 이미지)를 시간 순으로 보여준다. 또한 내레이터가 글을 읽어주면서 정보를 읽는 것이 아닌 것이 보는 것으로 인지한다.

인터넷에 있는 동영상과 이미지만을 이용해서 사람들이 원하는 정보를 보여줄 수 있다고 한다.

텍스트 정보는 위키피디아를 이용한다고 한다.



Qwiki at TechCrunch Disrupt from Qwiki on Vimeo.


Qwiki에 대한 설명이다.


정확히는 모르지만, 인력공모 내용을 보니, 대략 짐작이 되는 서비스인 것 같다. •Fluent in C++, Python, Perl or other scripting language, bash scripting, crawlers, HTML parsing. Text processing a big plus.
•Experience with mapping applications to the cloud (AWS, Hadoop, Map Reduce...).

'scribbling' 카테고리의 다른 글

구글 mod_pagespeed 출시  (0) 2010.11.05
blekko  (0) 2010.11.04
요즘 Trend  (0) 2010.11.01
Meego  (0) 2010.10.25
Richard Stevens의 죽음  (0) 2010.10.21
Posted by '김용환'
,

요즘 Trend

scribbling 2010. 11. 1. 17:13


오라클에서 자바 코드를 구글이 복사했다고 주장 ; 베낀 거 같네..^^;;
http://www.zdnet.com/blog/btl/oracle-says-google-directly-copied-java-code-heres-the-line-by-line-comparison/41025

인텔에서 Open JDK 참여
http://blogs.sun.com/theaquarium/entry/ibm_and_oracle_to_collaborate

노키아의 관료주의가 노키아를 퇴행시켜 - 뉴욕타임스
http://www.nytimes.com/2010/09/27/technology/27nokia.html?_r=2&partner=rss&emc=rss

스티브 잡스가 광야에서 배운 것 - 뉴욕타임스
http://www.nytimes.com/2010/10/03/business/03digi.html?partner=rss&emc=rss

'scribbling' 카테고리의 다른 글

blekko  (0) 2010.11.04
QWIKI  (0) 2010.11.04
Meego  (0) 2010.10.25
Richard Stevens의 죽음  (0) 2010.10.21
한국의 운영체제 정보  (0) 2010.10.20
Posted by '김용환'
,

Meego

scribbling 2010. 10. 25. 11:17


Meego는  maemo (nokia)  와 moblin (intel)의 합작 Spec이며, HWC에서 meego라는 이름으로 발표되었다.
서로 아구가 맞는 것이 maemo 의 QT 기반과 mobin의 소프트웨어 플랫폼이 합쳐졌다.
범용 cpu에 포팅될 수 있다고 하며, 노키아와 intel이 적극적으로 밀고 있다.

http://meego.com/developers/meego-architecture




노트북, 핸드폰, TV, 미디어폰, 자동차 에서 쓸 수 있는 스펙을 만든다고 하지만,
실제 홈페이지에 가보면, 크게 상용화되었거나 먼가를 만든 느낌은 그리 크지 않다.
내년에  meego기반 핸드폰이 나온다고 한다....

다만 QT 진영에서 쌍수를 드는 것이 가장 고무적으라 할 수 있다.
QT 어플리케이션이 어디서나 개발이 가능한 c++ 기반의 cross platform이기 때문에 많이 지켜보기는 하지만, 어떻게 될지 모른다.

'scribbling' 카테고리의 다른 글

QWIKI  (0) 2010.11.04
요즘 Trend  (0) 2010.11.01
Richard Stevens의 죽음  (0) 2010.10.21
한국의 운영체제 정보  (0) 2010.10.20
해외 논문 보기 (공짜로)  (0) 2010.10.07
Posted by '김용환'
,

Richard Stevens는 아래 서적을 만든 유명한 분이다..

7.UNIX Network Programming, Volume 2, Second Edition: Interprocess Communications, Prentice Hall, 1999.
6.UNIX Network Programming, Volume 1, Second Edition: Networking APIs: Sockets and XTI, Prentice Hall, 1998.
5.TCP/IP Illustrated, Volume 3: TCP for Transactions, HTTP, NNTP, and the UNIX Domain Protocols, Addison-Wesley, 1996.
4.TCP/IP Illustrated, Volume 2: The Implementation, Addison-Wesley, 1995.
3.TCP/IP Illustrated, Volume 1: The Protocols, Addison-Wesley, 1994.
2.Advanced Programming in the UNIX Environment, Addison-Wesley, 1992.
1.UNIX Network Programming, Prentice Hall, 1990.

1999년 Richard Stevens이 사망하였는데, 한국에서는 자살이라고 알려져있지만, 사실은 아닌 것 같다.
외국 어느 문서에서도 자살(suicide)라고 적혀 있는 것이 없다. 불명 아니면 잘모르는 병명 또는 음모에 의한 사망일수 있수도 있다..
너무 추측하지 말자..


http://answers.google.com/answers/threadview?id=722674


http://www.nndb.com/people/948/000023879/



그의 홈피
http://www.kohala.com/start/

'scribbling' 카테고리의 다른 글

요즘 Trend  (0) 2010.11.01
Meego  (0) 2010.10.25
한국의 운영체제 정보  (0) 2010.10.20
해외 논문 보기 (공짜로)  (0) 2010.10.07
Provisioning  (0) 2010.09.17
Posted by '김용환'
,

'scribbling' 카테고리의 다른 글

Meego  (0) 2010.10.25
Richard Stevens의 죽음  (0) 2010.10.21
해외 논문 보기 (공짜로)  (0) 2010.10.07
Provisioning  (0) 2010.09.17
Javascript dom 처리를 위한 API (DOM Element methods)  (0) 2010.09.01
Posted by '김용환'
,

해외 논문 보기 쉽지 않습니다.

이럴 때를 학술연구 정보 서비스가 있는데 RISS (http://www.riss.kr/) 에 접속하세요

아래 해외 DB 전자 정보서비스중 일부 괜찮은 논문을을 볼 수 있습니다..


'scribbling' 카테고리의 다른 글

Richard Stevens의 죽음  (0) 2010.10.21
한국의 운영체제 정보  (0) 2010.10.20
Provisioning  (0) 2010.09.17
Javascript dom 처리를 위한 API (DOM Element methods)  (0) 2010.09.01
Graph Database and products 소개.  (0) 2010.08.31
Posted by '김용환'
,

Provisioning

scribbling 2010. 9. 17. 10:42


출처
http://kldp.org/node/42505


제가 통신회사 프로그램 모듈을 짜던 시절에, 저는 provisioning을 이렇게 이해했습니다.

인용:
고객은 통신 회사가 제공하는 여러가지 서비스 중에 어느 것이 나에게 최적일지 참 고르기 어렵다.
통신 회사의 상담원은 고객의 서비스 요청시 미리 준비된 질문들을 함으로써 고객의 요구에 최적인 서비스 품목을 찾을 수 있다.
이렇게 무언가 여럿 중에 최적인 것을 찾기 위해 필요한 지식을 미리 준비해 놓고 요청에 맞게 공급하는 절차와 행위를 provisioning이라 한다.

응용 프로그램이나 장비쪽에 적용시켜보면, 목적에 맞는 최적해를 구축/공급하는 것이라고 풀어 쓸 수 있겠는데... 딱히 한 단어로 이거다 하지는 못하겠네요



. 미리 정의된 정책이나 서비스를 사용자에게 내려(지원, 서포트) 주는것을 말합니다.

'scribbling' 카테고리의 다른 글

한국의 운영체제 정보  (0) 2010.10.20
해외 논문 보기 (공짜로)  (0) 2010.10.07
Javascript dom 처리를 위한 API (DOM Element methods)  (0) 2010.09.01
Graph Database and products 소개.  (0) 2010.08.31
Wired-The web is dead.  (0) 2010.08.24
Posted by '김용환'
,

http://www.javascriptkit.com/domref/elementmethods.shtml


Properties Description
addEventListener(eventType, listener, useCapture)

Not supported in IE, which uses attachEvent() instead.

Associates a function with a particular event and binds the event to the current node. addEventListener() accepts the following 3 parameters:

1) EventType: A string representing the event to bind, without the "on" prefix. For example, "click", "mousedown" etc.
2) listener: The object or function to fire when the event fires. The actual parameter entered should be a reference to the function or object (ie: "dothis" instead of "dothis()".
3) useCapture: Boolean indicating whether to bind the event as it is propogating towards the target node, (event Capture), or as the event bubbles upwards from the target (event bubble). Set to true or false, respectively.

The advantage of using the DOM to bind an event is that you can assign multiple functions to a node for the same event (ie: window.onload) without running into event handler conflicts.

Example(s):

function statusreport(){
 alert("document has loaded")
}

if (window.addEventListener)
 window.addEventListener("load", statusreport, false) //invoke function
window.onload=statusreport() //function invoked again, since no event handler conflicts

Since listener must be a function reference, a common question is how to specify the listener so it can receive parameters of its own. The solution is just to wrap the listener in an anonymous function reference:

function dothis(what){
 alert("Mom says to do " + what)
}

if (window.addEventListener)
 window.addEventListener("load", function(){dothis('Your homework')}, false) //invoke function

For events such as "click" or "mousemove" that populate the Event object with additional info such as where the mouse is at the time, addEventListner() passes this object silently to the listener. To receive it, define a single parameter in the anonymous function:

function getcoord(evt){
 alert("You clicked at coordinates [" + evt.clientX + "," + evt.clientY + "]")
}

if (document.addEventListener)
 document.addEventListener("click", function(evt){getcoord(evt)}, false) //invoke function

attachEvent(eventType, function)

IE 5+only function

The IE5+ proprietary equivalent of addEventListener(). Note that for the parameter eventType, the even string should include the "on" prefix (ie: "onload", "onclick" etc).

Example(s):

if (window.attachEvent)
window.attachEvent("onload", statusreport) //invoke function

appendChild(node) Inserts the specified node at the end of the current node object. A frequently used method for dynamically appending a new element or text to the document.

Example(s):

<div id="test"></div>

<script type="text/javascript">
var newdiv=document.createElement("div")
var newtext=document.createTextNode("A new div")
newdiv.appendChild(newtext) //append text to new div
document.getElementById("test").appendChild(newdiv) //append new div to another div

</script>

An interesting feature of appendChild() is that it can be used to directly move an existing element on the page so it becomes the child of a different parent. There's no need to remove the element from the previous parent first, as the method always checks to see first if the element beind appended currently already has a parent, and remove it from that parent first if found to be true. The below example moves a H1 element from its current location to the very end of the document:

<h1 id="myheader">Welcome to JavaScript Kit</h1>
<p>Hope you enjoy the site!</p>

<script type="text/javascript">
document.body.appendChild(document.getElementById("myheader")) //moves "myheader" to end of the page

</script>

blur() Removes keyboard focus from the current element. Used for example to fire the onBlur event handler of an element via scripting.
click() Executes a click on a element as if the user manually clicked on it. In most browsers, click() only works on form INPUT elements that's non "submit" or "reset". It can't be used to simulate a click on a link or form submit button.
cloneNode(deepBoolean) Duplicates and returns a copy of the current node as a standalone node (not part of document tree). Cloning a node copies both the original's attributes and values, including the ID attribute, so be sure to alter the cloned ID attribute's value so it's unique before introducing it to the document tree. This method supports a single Boolean parameter, "deepBoolean" that when set to true, clones all the sub nodes of the current node as well, such as any text contained within.

Example(s):

p=document.getElementById("mypara")
pclone = p.cloneNode(true)

detachEvent(eventType, function)

IE 5+only function

Removes an event handler and its function previously associated with a node in IE5+, via attachEvent() for example. The IE5+ proprietary equivalent of DOM2's removeEventListener().

Example(s):

if (window.detachEvent)
window.detachEvent("onload", statusreport) //invoke function

dispatchEvent(eventObject)

Not supported in IE, which uses fireEvent() instead.

Dispatches an event to fire on a node artificially. This method returns a Boolean indicating whether any of the listeners which handled the event called preventDefault (false if called, otherwise, true). IE's equivalent of dispatchEvent() is fireEvent().

Example(s):

<div id="test" onclick="alert('hi')">Sample DIV.</div>

<script type="text/javascript">
//Generate an artificial click event on "test". Fires alert("hi")
var clickevent=document.createEvent("MouseEvents")
clickevent.initEvent("click", true, true)
document.getElementById("test").dispatchEvent(myevent)

</script>

focus() Sets focus on the current node.
getAttribute(attributeName) Returns the value of the attribute named attribute of the current node.

Example(s):

document.getElementById("test").getAttribute("align")

getAttributeNS(namespace, localname) Returns the value of the attribute with the given local name and namespace. Applicable in XML documents.
getAttributeNode(attributename) Returns/references the attribute of the current element as a stand only node (not part of document tree).

Example(s):

var attributeobj=document.getElementById("nav").getAttributeNode("align")
attributeobj.value="center"

getAttributeNodeNS(namespace, localname) Returns/references the attribute of the current element with the given local name and namespace. Applicable in XML documents.
getElementsByClassName

Note: Supported in FF3+, Opera 9.5+, Safari 3+, though not IE8.

Gets a collection of element(s) within the invoking element that shares the designated CSS class name, for example:

//get elements within element "house" that has a CSS class of "cats"
document.getElementById("house").getElementsByClassName("cats")

You can get elements with different CSS class names within an element all in one scoop, by separating each class with a space:

//get elements within element "house" that has a CSS class of "cats" or "dogs"
document.getElementById("house").getElementsByClassName("cats dogs")

Note that getElementsByClassName() also exists on the document element to get every matching element within the document.

getElementsByTagName(tagName) Returns as an array all the child elements of the current element matching the "tagName" parameter (ie: "li"). In Firefox/ IE6+, you may enter an asterisk ("*") for the method's parameter to retrieve a list of all elements within the current.

Example(s):

var mylist=document.getElementById("navlist")
var listitems= mylist.getElementsByTagName("li")
for (i=0; i<listitems.length; i++)
//manipulate each li element

var alltags=document.getElementsByTagName("*") //returns all elements on page

getElementsByTagNameNS(namespace, localname) Returns as an array all the child elements of the current element with the given local name and namespace. Applicable in XML documents.
hasAttribute(attributename) Returns a Boolean value indicating whether the current element contains an attribute (ie: "align").

Example(s):

if (document.getElementById("mytable").hasAttribute("style"))
//manipuate the element's style

hasAttributeNS(namespace, localname) Returns a Boolean value indicating whether the current element contains an attribute with the given local name and namespace. Applicable in XML documents.
hasAtrributes() Returns a Boolean value indicating whether the current element has any explicit attributes defined.
hasChildNodes() Returns a Boolean value indicating whether the current element contains any child nodes.
insertBefore(newElement, targetElement) Inserts a new node "newElement" as a child node of the current node. The "targetElement" property dictates where "newElement" is inserted within the list of child nodes. If set to null, the new element is inserted as the last child node; otherwise, it's inserted right before "targetElement".

Example(s):

<div id="employees">
<div id="george">George Doe: Human resources department</div>
</div>

To insert a new DIV directly above "george", so the outcome becomes:

<div id="employees">
<div id='kevin">Kevin Lin: Main system administrator</div>
<div id="george">George Doe: Human resources department</div>
</div>

You would do the following:

<script type="text/javascript">

var newemployee=document.createElement("div")
var oldemployee=document.getElementById("george")
newemployee.setAttribute("id", "kevin")
newemployee.innerHTML="Kevin Lin: Main system administrator"
document.getElementById("employees").insertBefore(newemployee, oldemployee)

</script>

Important: Like many DOM methods that change the structure of the document, insertBefore() can only be called after the page has fully loaded. Doing so before will return strange errors in most browsers!
 

insertAfter(newElement, targetElement) As of DOM Level 2 there does NOT exist an insertAfter() method. However, it can be easily simulated using insertBefore() above. For the "targetElement" parameter, just use "targetElement.nextSibling" instead.

<div id="employees">
<div id="george">George Doe: Human resources department</div>
</div>

To insert a new DIV directly below "george", so the outcome becomes:


<div id="employees">
<div id="george">George Doe: Human resources department</div>
<div id='kevin">Kevin Lin: Main system administrator</div>
</div>

You would do the following:

<script type="text/javascript">

var newemployee=document.createElement("div")
var oldemployee=document.getElementById("george")
newemployee.setAttribute("id", "kevin")
newemployee.innerHTML="Kevin Lin: Main system administrator"
document.getElementById("employees").insertBefore(newemployee, oldemployee.nextSibling)

</script>

 

item(index) Retrieves a node based on its index within the document tree. IE4+ and FireFox1+.

Example(s):

<div id="div1"></div>
<div id="div2"></div>

<script type="text/javascript">
var mydivs=document.getElementsByTagName("div")
alert(mydivs.item(1).id) //alerts "div2"
</script>

normalize() Normalizes the current node and its sub tree. See here for more info.
querySelector(selectors, [NSResolver])

Note: Currently supported in FF3.1+, IE8+ (only in IE8 standards mode), and Safari 3.1+

Accepts a CSS selector(s) and returns the first matching element (based on the document tree) within the invoking element, or null.

Example:

<ul id="mylist">
<li>Item 1</li>
<li>Item 2</li>
<li>Item 3</li>
</ul>

<script type="text/javascript">
var item2=document.getElementById("mylist").querySelector('li:nth-of-type(2)')
alert(item2.innerHTML) //alerts "Item 2"
</script>

You can enter multiple CSS selectors each separated by a comma (,), in which case the first matching element of any of the CSS selectors entered will be returned:

//returns either element "#leftcolumn" or "#rightcolumn", depending on which one is found first:
myelement.querySelector("#leftcolumn, #rightcolumn")

querySelector() supports an optional 2nd "NSResolver" parameter to resolve namespace prefixes in XHTML documents. Not supported in IE8.

querySelectorAll(selectors, [NSResolver])

Note: Currently supported in FF3.1+, IE8+ (only in IE8 standards mode), and Safari 3.1+

Accepts a CSS selector(s) and returns all matching elements (based on the document tree) within the invoking element as a staticNodeList, or null. A staticNodeList is a static collection of elements that are not affected by any subsequent changes occuring on the document tree.

Example:

<ul id="mylist">
<li>Item 1</li>
<li>Item 2</li>
<li>Item 3</li>
</ul>

<script type="text/javascript">
var odditems=document.getElementById("mylist").querySelectorAll('li:nth-of-type(odd)')
for (var i=0; i<odditems.length; i++)
 alert(odditems[i].innerHTML) //alerts "Item 1", "Item 3" etc.
</script>

You can enter multiple CSS selectors each separated by a comma (,), in which case all matching elements found using the entered CSS selectors will be returned:

//returns both elements "#leftcolumn" or "#rightcolumn", or one of them if only one defined:
myelement.querySelectorAll("#leftcolumn, #rightcolumn")

querySelectorAll() supports an optional 2nd "NSResolver" parameter to resolve namespace prefixes in XHTML documents. Not supported in IE8.

See "Overview of CSS3 Structural puesdo-classes" for advanced CSS selectors you can use with the query selector methods.

removeAttribute(attributename) Removes an attribute by its name.

Example(s):

document.getElementById("test").removeAttribute("href")

removeAttributeNode(attributereference) Remove an attribute by passing in as parameter a reference to the attribute object to remove. It offers an alternate way to removeAttribrute()"for removing attributes, when all you have is a reference to the attribute object in your script.

Example(s):

var hrefattr=document.getElementById("test").getAttributeNode("href")
document.getElementById("test").removeAttributeNode(hrefattr)

removeAttributeNS(namespace, localname) Removes an attribute with the specified namespace and localname.
removeChild(childreference) Removes the child node of the current node. The removed node can then be reinserted elsewhere in the document tree.

Example(s):

<div id="father"><div id="child">A child</div></div>

<script type="text/javascript">
var childnode=document.getElementById("child")
var removednode=document.getElementById("father").removeChild(childnode)
</script>

removeEventListener(eventType, listener, useCapture)

Not supported in IE, which uses detachEvent() instead.

Removes the specified event from being binded to the current node:

1) EventType: A string representing the event to unbind, without the "on" prefix. For example, "click", "mousedown" etc.
2) listener: The function or method to associate with the event.
3) useCapture: Boolean indicating whether to unbind the event as it is propagating towards the target node, (event Capture), or as the event bubbles upwards from the target (event bubble). Set to true or false, respectively.

 NS6/Firefox method.

replaceChild(newChild, oldChild) Replaces one child node of the current node with another child node.

Example(s):

<div id="adiv"><span id="innerspan"></span></div>

<script type="text/javascript">
var oldel=document.getElementById("innerspan")
var newel=document.createElement("p")
document.getElementById("adiv").replaceChild(newel, oldel)
</script>

scrollIntoView([Boolean]) Firefox/IE4+ proprietary method that scrolls an element into view. It accepts an optional Boolean parameter that when set to true (default), scrolls the element so its top left corner touches the top of the viewable window. If false, the element's bottom left corner touches the bottom of the window.
setAttribute(attributename, value, [iecaseflag]) Sets an attribute's value for the current element. If the attribute doesn't exit yet, it creates the attribute first. Otherwise, the existing attribute is modified with the new value. In IE, the following two pitfalls exist:
  • To set the "class" attribute, use "className" instead.
  • The "attributename" parameter is case sensitive by default in IE . This means if you attempt to set the "align" attribute and "Align" already exists on the element, both will be present as a result. To turn off case sensitivity, set the IE-only 2nd parameter of setAttribute() to 0 (instead of default, which is 1).

Example(s):

document.getElementById("test").setAttribute("title", "JavaScript Kit")

setAttributeNS(namespace, qualifiedname, value) Sets or creates an attribute for the current node with the given local name and namespace. Applicable in XML documents.
setAttributeNode(attributereference) Sets or creates an attribute for the current node. "attributereference" should be a reference to a attribute you wish to insert. If an attribute of the same name (as referenced) already exists on the node, it is replaced with the newly inserted one.

Example(s):

<div id="brother" style="border:1px solid black; padding: 2px">Brother</div>
<div id="sister">Sister</div>

<script type="text/javascript">
var bro=document.getElementById("brother")
var sis=document.getElementById("sister")
var brostyle=bro.getAttributeNode("style")
var clonebrostyle=brostyle.cloneNode(false) //clone attribute first. Required.
sis.setAttributeNode(clonebrostyle)
</script>

supports(feature, [version]) Tests to see if this DOM implementation supports a particular feature.

'scribbling' 카테고리의 다른 글

해외 논문 보기 (공짜로)  (0) 2010.10.07
Provisioning  (0) 2010.09.17
Graph Database and products 소개.  (0) 2010.08.31
Wired-The web is dead.  (0) 2010.08.24
Skype Architect의 Learnings from Five Years as a Skype Architect  (0) 2010.08.19
Posted by '김용환'
,
Posted by '김용환'
,

Wired-The web is dead.

scribbling 2010. 8. 24. 11:26

Wired-The web is dead.
http://www.wired.com/magazine/2010/08/ff_webrip/all/1

Sources: Cisco estimates based on CAIDA publications, Andrew Odlyzko

Sources: Cisco estimates based on CAIDA publications, Andrew Odlyzko



Web->App으로 변화되고 있는 것이 아니냐? 얘기한다..

@ Comment
Traffic만으로 보는 것은 어불성설.. 파이가 점점 크고 있는 상황에서 동영상 서비스가 많은 트래픽을 유발한다고 할 때, 꼭  web이 죽고 있다고 말하기는 어려운듯..
Posted by '김용환'
,