<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:wfw="http://wellformedweb.org/CommentAPI/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
	xmlns:slash="http://purl.org/rss/1.0/modules/slash/"
	>

<channel>
	<title>Berk Ulsoy &#187; free software</title>
	<atom:link href="http://blog.ulsoy.org/lang/en-us/tag/free-software/feed/" rel="self" type="application/rss+xml" />
	<link>http://blog.ulsoy.org</link>
	<description>Completely Personal Thoughts</description>
	<lastBuildDate>Mon, 19 Oct 2009 18:28:09 +0000</lastBuildDate>
	<generator>http://wordpress.org/?v=2.8.4</generator>
	<language>en-us</language>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
			<item>
		<title>Why zealots are extremely harmful for Desktop Linux</title>
		<link>http://blog.ulsoy.org/lang/en-us/2009/06/26/why-zealots-are-extremely-harmful-for-desktop-linux</link>
		<comments>http://blog.ulsoy.org/lang/en-us/2009/06/26/why-zealots-are-extremely-harmful-for-desktop-linux#comments</comments>
		<pubDate>Fri, 26 Jun 2009 13:49:30 +0000</pubDate>
		<dc:creator>berk</dc:creator>
				<category><![CDATA[Linux]]></category>
		<category><![CDATA[free software]]></category>

		<guid isPermaLink="false">http://blog.ulsoy.org/?p=52</guid>
		<description><![CDATA[Zealots are everywhere..Trying to do things in their very own unique way..But are they harmful to for the software they like ? Are they pushing people away from Linux involuntarily ? Let's see...]]></description>
			<content:encoded><![CDATA[<p>Like most other software, there are serious number of zealots who obsessively love Linux. So far ok, the obsession is something personal and does not affect no one directly. However, these people are often considered as good users and most people judge <a href="http://en.wikipedia.org/wiki/Linux">Linux</a> experience by looking at how they use it.</p>
<p><span id="more-52"></span></p>
<p>Consider a typical zealot and his/her friends. (Are any female Linux zealot anyway ?) I will write some cases and they do happen somewhere in the world.</p>
<p>Our command prompt obsessive zealot comes to a friend&#8217;s home and asks for wireless network parameters. Once he gets, he runs series of iwconfig, wpa_supplicant and ifconfig commands to connect. He writes a profile for this network by hand and tests it. It takes about 10-15 minutes. He does not use simple network connection applets of Gnome/Kde. He does this because he is obsessed with running minimal applications on a 4GB laptop. Do you know what his friends think about now ? &#8220;It takes running some cryptic commands and 15 minute just to connect to a network using Linux&#8221;. This is the only experience they have ever seen. They will judge Linux by this from now on..</p>
<p>A friend of our zealot is curious about Linux and wants to try. As a trusted and renown friend, our zealot is happy to get a chance to win another user. They met at home at night, order some pizza and our zealot starts installing <a href="http://www.gentoo.org">Gentoo</a> or <a href="http://www.slackware.com">Slackware</a>. (Or some other enduser disoriented distribution)</p>
<p>- Hey, what&#8217;s that Slackware/Gentoo thing ? I always hear about <a href="http://www.ubuntu.com">Ubuntu</a>, why do not we just install it ?<br />
- This is the one I use. By using this, you will have the absolute control on your system. You will make sure that it runs only what you want and they will be compiled for the best performance<br />
- Cool</p>
<p>Now our zealot runs cfdisk and some other text based tools to organize disk. He carefully setups partitions and language configurations. He selects every package individually and 3 hours has already passed. Now the new user thinks that he can never do this by himself. Our lovely zealot has prevented his friend to experience nice and short gui installers of Ubuntu or Fedora based distributions. </p>
<p>Now the system has been installed and our zealot carefully checks for hardware components. He grabs the latest vanilla kernel and select modules one by one. It takes another 3 hour to make a fitting kernel image for that computer. He assures that this will give the best performance and he does not tell what will happen if the user tries to connect a different hardware. He gains may be 3-5% in performance but he loses flexibility of having thousands of hardware drivers which are ready to plug and play.</p>
<p>When it is 3 A.M, the system is ready to run, at last. Thanks to our zealot. Now the users asks for some applications. Our zealot grabs source files and compiles by setting configure flags according to his own taste. He quickly mentions about the compilers, source files etc to his friend, who has absolutely no idea what a programming language is..Our zealot thinks he can install better software although there are tens of thousands of testers and developers who create one-click-install packages which are tuned and tested.</p>
<p>Our unlucky user tries to use Linux for a week. He can not change display properties because most of them were handwritten to xorg.conf although it is not required in most cases. He barely remembers how to compile a program and fails trying to install something. At last, he does not access his external disks easily because our zealot does not let hal and udev run because he thinks they are not necessary..</p>
<p>This is completely harmful to free software society. These kind of people should stop and think about what they are doing. If you want to help, do not try to consider end users like yourself. Consider them as babies and give what they want. Give them a system which can use most of the connected hardware, easily play media, and watch videos on YouTube. Ordinary users do not need to run tiny window managers. They do not need to know about source code or build system. </p>
<p>Just show them installing the system by clicking next. Then explain that backing up is as easy as archiving home directory. Lastly, instruct to use package manager to search, install and uninstall applications by clicking their name.</p>
<p>If you will not do this, just let those users use other operating systems. Creating a negative image of desktop Linux experience is the last thing we want. Right ?</p>
]]></content:encoded>
			<wfw:commentRss>http://blog.ulsoy.org/lang/en-us/2009/06/26/why-zealots-are-extremely-harmful-for-desktop-linux/feed</wfw:commentRss>
		<slash:comments>0</slash:comments>
		</item>
		<item>
		<title>Free Software vs. Microsoft in the Cloud of Web Servers</title>
		<link>http://blog.ulsoy.org/lang/en-us/2009/04/29/free-software-vs-microsoft-in-the-cloud-of-web-servers</link>
		<comments>http://blog.ulsoy.org/lang/en-us/2009/04/29/free-software-vs-microsoft-in-the-cloud-of-web-servers#comments</comments>
		<pubDate>Wed, 29 Apr 2009 11:14:33 +0000</pubDate>
		<dc:creator>berk</dc:creator>
				<category><![CDATA[Linux]]></category>
		<category><![CDATA[apache]]></category>
		<category><![CDATA[free software]]></category>
		<category><![CDATA[microsoft]]></category>
		<category><![CDATA[nginx]]></category>

		<guid isPermaLink="false">http://blog.ulsoy.org/?p=32</guid>
		<description><![CDATA[There are some web sites providing reports on web server software usage in the internet. One of the most populer is Netcraft with their infamous web server survey. In this post, I would like to talk more about decisions made by highest traffic sites and countries rather than cumulative results.
I&#8217;ve collected the list of the [...]]]></description>
			<content:encoded><![CDATA[<p>There are some web sites providing reports on web server software usage in the internet. One of the most populer is <a href="http://www.netcraft.com">Netcraft</a> with their infamous web server survey. In this post, I would like to talk more about decisions made by highest traffic sites and countries rather than cumulative results.</p>
<p>I&#8217;ve collected the list of the top sites from <a href="http://www.alexa.com">Alexa</a>. They provide a csv file ordered by ranking. Then I used a tiny script to collect Server header from top 15000 sites. After that, I used libgeo-ipfree-perl package to guess geographic location of sites. The numbers may not be exact but enough to give very strong idea.</p>
<p><span id="more-32"></span></p>
<p>There are some points that I should say:</p>
<ul>
<li>Geographical location has been found based on ip/country match. It may not be 100% correct</li>
<li>Some sites does not include Server header and some of them just give incorrect information. (about 900 in total) It does not make the site more secure, and  I do not think it will make any difference in percentages. It is possible to modify Server header in all important web server software</li>
<li>Only for the Top 100, I used Netcraft for the information about missing 8 servers.</li>
<li>Of course there are cache engines, other web servers and proxies but their numbers are just neglectable</li>
<li>Apache-Coyote has also some usage but I omit to make graphs simpler. It does not have affect on the idea of this work</li>
<li>The numbers in the charts are not percentages but the actual server numbers. For the country pie charts, I also omit unknown servers if there are many (not much indeed)</li>
<li>For the countries, I only checked those who has at least 5 sites in Top 15000</li>
</ul>
<h3>Top 100 Sites</h3>
<p>We see a clean win for <a href="http://httpd.apache.org">Apache</a>. This is the most important and marginal output. If you look at only Netcraft&#8217;s deployed server counts, <a href="http://www.microsoft.com">Microsoft</a> IIS seems only 15% behind Apache. However when we check the best of the best, the busiest web sites in the Internet, we see that Microsoft IIS is far behind Apache. Besides, <a href="http://nginx.net">nginx</a> is only a few sites behind of IIS. The names of the sites running IIS are also notable:</p>
<p><a href="http://www.live.com">www.live.com</a> (Microsoft)<br />
<a href="http://www.msn.com">www.msn.com</a> (Microsoft)<br />
<a href="http://www.myspace.com">www.myspace.com</a><br />
<a href="http://www.microsoft.com">www.microsoft.com</a> (Microsoft ..ehh)<br />
<a href="http://www.doubleclick.com">www.doubleclick.com</a><br />
<a href="http://www.go.com">www.go.com</a><br />
<a href="http://www.conduit.com">www.conduit.com</a><br />
<a href="http://espn.go.com">espn.go.com</a></p>
<p><img src="http://chart.apis.google.com/chart?chtt=Web%20server%20software%20for%20Top%20100%20Sites&#038;cht=p3&#038;chd=t:45,8,18,6,3&#038;chds=0,45&#038;chs=500x200&#038;chl=Apache%20%20%2045&#124;IIS%20%20%208&#124;GWS%20%20%2018&#124;nginx%20%20%206&#124;lighthttpd%20%20%203" alt="Top 100 Sites" /></p>
<p>In the previous year, Google has changed their Apache server headers to GWS. GWS is a modified version of Apache and if we combine this number with plain Apache  deployments, the difference increases.</p>
<p><img src="http://chart.apis.google.com/chart?chtt=Web%20server%20software%20for%20Top%20100%20Sites%20(Apache%20forks%20combined)&#038;cht=p3&#038;chd=t:45,8,6,3&#038;chds=0,63&#038;chs=500x200&#038;chl=Apache%20%20%2063&#124;IIS%20%20%208&#124;nginx%20%20%206&#124;lighthttpd%20%20%203" alt="Top 100 Sites" /></p>
<h3>Top 1000 Sites</h3>
<p>We still have same aggressive percentage of Apache and its derivatives</p>
<p><img src="http://chart.apis.google.com/chart?chtt=Web%20server%20software%20for%20Top%201000%20Sites&#038;cht=p3&#038;chd=t:487,131,72,64,29&#038;chds=0,487&#038;chs=500x200&#038;chl=Apache%20%20%20487&#124;IIS%20%20%20131&#124;nginx%20%20%2072&#124;GWS%20%20%2064&#124;lighthttpd%20%20%2029" alt="Top 1000 Sites" /></p>
<h3>Top 10000 Sites</h3>
<p>As the number increases to 10000, Microsoft IIS gains some more points. </p>
<p><img src="http://chart.apis.google.com/chart?chtt=Web%20server%20software%20for%20Top%2010000%20Sites&#038;cht=p3&#038;chd=t:5475,1714,717,107,246,128&#038;chds=0,5475&#038;chs=500x200&#038;chl=Apache%20%20%205475&#124;IIS%20%20%201714&#124;nginx%20%20%20717&#124;GWS%20%20%20107&#124;lighthttpd%20%20%20246&#124;IBM%20HTTP%20Server%20128" alt="Top 10000 Sites" /></p>
<h3>Top 15000 Sites</h3>
<p>The scheme does not change very much. Even we include rather low traffic sites, the winner is Apache </p>
<p><img src="http://chart.apis.google.com/chart?chtt=Web%20server%20software%20for%20Top%2015000%20Sites&#038;cht=p3&#038;chd=t:8347,2642,1077,116,356,178&#038;chds=0,5475&#038;chs=500x200&#038;chl=Apache%20%20%208347&#124;IIS%20%20%202642&#124;nginx%20%20%201077&#124;GWS%20%20%20116&#124;lighthttpd%20%20%20356&#124;IBM%20HTTP%20Server%20178" alt="Top 15000 Sites" /></p>
<h3>Choices based on countries</h3>
<p>It is also interesting to check which countries prefer which software. Although open source software dominates in the world, we can notice some local differences.</p>
<p><img src="http://chart.apis.google.com/chart?chtt=Web%20server%20software%20in%20US%20for%20Top%2015000&#038;cht=p3&#038;chd=t:4098,1118,341,137&#038;chds=0,4098&#038;chs=500x200&#038;chl=Apache%20%20%204098&#124;IIS%20%20%201118&#124;nginx%20%20%20341&#124;lighthttpd%20%20%20137" alt="" /></p>
<p>This is one of the most important graph since most of the high traffic web sites are originated from the US. Although IIS seems closing the gap in the world scale, Apache is certainly preferred much more than IIS in here.</p>
<p>The picture changes when we move to Europe. Finland, Norway, Ukraine, Slovakia, Austria, Bulgaria, Romaina, Hungary, Ireland, Luxembourg, Portugal seem to use only open source web server software in Top 15000. Some other notable countries are here:</p>
<p><img src="http://chart.apis.google.com/chart?chtt=Web%20server%20software%20in%20Sweden%20in%20Top%2015000&#038;cht=p3&#038;chd=t:38,12,10,7&#038;chds=0,38&#038;chs=500x200&#038;chl=Apache%20%20%2038&#124;IIS%20%20%2012&#124;nginx%20%20%2010&#124;lighthttpd%20%20%207" alt="" /></p>
<p><img src="http://chart.apis.google.com/chart?chtt=Web%20server%20software%20in%20Netherlands%20in%20Top%2015000&#038;cht=p3&#038;chd=t:299,29,94,29&#038;chds=0,299&#038;chs=500x200&#038;chl=Apache%20%20%20299&#124;IIS%20%20%2029&#124;nginx%20%20%2094&#124;lighthttpd%20%20%2029" alt="" /></p>
<p><img src="http://chart.apis.google.com/chart?chtt=Web%20server%20software%20in%20Germany%20in%20Top%2015000&#038;cht=p3&#038;chd=t:562,49,75,47&#038;chds=0,562&#038;chs=500x200&#038;chl=Apache%20%20%20562&#124;IIS%20%20%2049&#124;nginx%20%20%2075&#124;lighthttpd%20%20%2047" alt="" /></p>
<p><img src="http://chart.apis.google.com/chart?chtt=Web%20server%20software%20in%20France%20in%20Top%2015000&#038;cht=p3&#038;chd=t:376,60,10,13&#038;chds=0,376&#038;chs=500x200&#038;chl=Apache%20%20%20376&#124;IIS%20%20%2060&#124;nginx%20%20%2010&#124;lighthttpd%20%20%2013" alt="" /></p>
<p><img src="http://chart.apis.google.com/chart?chtt=Web%20server%20software%20in%20UK%20in%20Top%2015000&#038;cht=p3&#038;chd=t:318,123,35,11,13&#038;chds=0,318&#038;chs=500x200&#038;chl=Apache%20%20%20318&#124;IIS%20%20%20123&#124;nginx%20%20%2035&#124;lighthttpd%20%20%2011&#124;Zeus%20%20%2013" alt="" /></p>
<p><img src="http://chart.apis.google.com/chart?chtt=Web%20server%20software%20in%20Russia%20in%20Top%2015000&#038;cht=p3&#038;chd=t:101,34,258&#038;chds=0,318&#038;chs=500x200&#038;chl=Apache%20%20%20101&#124;IIS%20%20%2034&#124;nginx%20%20%20258&#124;" alt="" /></p>
<p>Japan has also notable Apache usage:</p>
<p><img src="http://chart.apis.google.com/chart?chtt=Web%20server%20software%20in%20Japan%20in%20Top%2015000&#038;cht=p3&#038;chd=t:617,57&#038;chds=0,617&#038;chs=500x200&#038;chl=Apache%20%20%20617&#124;IIS%20%20%2057" alt="" /></p>
<p>Also, there are some countries which prefer Microsoft IIS over Apache and other open source solutions<br />
Argentina, Chile, Bermuda, Egypt have only Microsoft based sites in Top 15000.</p>
<p>China is also an important Microsoft user although it has the highest rate for software piracy along with Russia and Thailand.</p>
<p><img src="http://chart.apis.google.com/chart?chtt=Web%20server%20software%20in%20China%20for%20Top%2015000&#038;cht=p3&#038;chd=t:328,453,120,16&#038;chds=0,453&#038;chs=500x200&#038;chl=Apache%20%20%20328&#124;IIS%20%20%20453&#124;nginx%20%20%20120&#124;lighthttpd%20%20%2016" alt="" /></p>
<p><img src="http://chart.apis.google.com/chart?chtt=Web%20server%20software%20in%20Israel%20in%20Top%2015000&#038;cht=p3&#038;chd=t:11,18,8&#038;chds=0,18&#038;chs=500x200&#038;chl=Apache%20%20%2011&#124;IIS%20%20%2018&#124;nginx%20%20%208" alt="" /></p>
<p><img src="http://chart.apis.google.com/chart?chtt=Web%20server%20software%20in%20Vietnam%20in%20Top%2015000&#038;cht=p3&#038;chd=t:27,40&#038;chds=0,40&#038;chs=500x200&#038;chl=Apache%20%20%2027&#124;IIS%20%20%2040" alt="" /></p>
<p>Here is Iran, where everything from US is considered to be demonic, except Microsoft. Business is business anyway..</p>
<p><img src="http://chart.apis.google.com/chart?chtt=Web%20server%20software%20in%20Iran%20in%20Top%2015000&#038;cht=p3&#038;chd=t:10,23&#038;chds=0,23&#038;chs=500x200&#038;chl=Apache%20%20%2010&#124;IIS%20%20%2023" alt="" /></p>
<p>And lastly, my own country Turkey</p>
<p><img src="http://chart.apis.google.com/chart?chtt=Web%20server%20software%20in%20Turkey%20in%20Top%2015000&#038;cht=p3&#038;chd=t:42,62&#038;chds=0,62&#038;chs=500x200&#038;chl=Apache%20%20%2042&#124;IIS%20%20%2062" alt="" /></p>
<h3>Conclusion</h3>
<p>After all these, there is not much need for a conclusion but I will write anyway</p>
<ul>
<li>Open source platforms (mainly Linux) and Apache (with derivatives) are the most preferred combination for high traffic web sites.period.</li>
<li>nginx performed relatively well and it is a rising star. In some countries, it is used more than Microsoft IIS</li>
<li>Usage of Microsoft IIS increases as we move away from Top 100</li>
<li>Free software in web services has well adopted in most European countries and United States</li>
<li>Some countries develop technology and some consume it. It is noticable that, countries which are well-known for their technology leadership also prefer using Free Software in their busiest web services. Consumers lack vision, knowledge, qualified engineers and in my opinion, this affects the platform decision.</li>
</ul>
]]></content:encoded>
			<wfw:commentRss>http://blog.ulsoy.org/lang/en-us/2009/04/29/free-software-vs-microsoft-in-the-cloud-of-web-servers/feed</wfw:commentRss>
		<slash:comments>0</slash:comments>
		</item>
	</channel>
</rss>
