<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	>
<channel>
	<title>Comments on: Web browsers are limiting the benefits of multi-core CPUs.</title>
	<atom:link href="http://pinderkent.blogsavy.com/archives/134/feed" rel="self" type="application/rss+xml" />
	<link>http://pinderkent.blogsavy.com/archives/134</link>
	<description>Just another Blogsavy.com weblog</description>
	<pubDate>Thu, 20 Nov 2008 21:49:31 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.5</generator>
		<item>
		<title>By: spenser</title>
		<link>http://pinderkent.blogsavy.com/archives/134#comment-289</link>
		<dc:creator>spenser</dc:creator>
		<pubDate>Wed, 26 Dec 2007 00:13:19 +0000</pubDate>
		<guid isPermaLink="false">http://pinderkent.blogsavy.com/archives/147#comment-289</guid>
		<description>The worst offenders seem to be wordpress plugins.

It's bad enough to go to a blog and find just a tiny snippet quoting some other blog as it's entire content.

It's worse when the blog uses wordpress plugins that lock up the browser. Of course, this happens while browsing the sixth page of search results requiring the user to end task the browser and then try to find the search term again.

Recently I went through the exercise of banning all offsite javascript content by using the hosts file while looking at a particular wordpress blog using plugins. No, it wasn't the two hit counter scripts, or the three ad servers. It was a wordpress plugin hosted on the same server. Once the plugin was blocked, no more lockups. If it had been the third party servers, it would have been easy to ban them, but since wordpress plugins are normally hosted on the same server, it would require disabling javascript globally.

Javascript developers: please test on all browsers, because your users won't!</description>
		<content:encoded><![CDATA[<p>The worst offenders seem to be wordpress plugins.</p>
<p>It&#8217;s bad enough to go to a blog and find just a tiny snippet quoting some other blog as it&#8217;s entire content.</p>
<p>It&#8217;s worse when the blog uses wordpress plugins that lock up the browser. Of course, this happens while browsing the sixth page of search results requiring the user to end task the browser and then try to find the search term again.</p>
<p>Recently I went through the exercise of banning all offsite javascript content by using the hosts file while looking at a particular wordpress blog using plugins. No, it wasn&#8217;t the two hit counter scripts, or the three ad servers. It was a wordpress plugin hosted on the same server. Once the plugin was blocked, no more lockups. If it had been the third party servers, it would have been easy to ban them, but since wordpress plugins are normally hosted on the same server, it would require disabling javascript globally.</p>
<p>Javascript developers: please test on all browsers, because your users won&#8217;t!</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Ev</title>
		<link>http://pinderkent.blogsavy.com/archives/134#comment-288</link>
		<dc:creator>Ev</dc:creator>
		<pubDate>Fri, 21 Dec 2007 17:50:23 +0000</pubDate>
		<guid isPermaLink="false">http://pinderkent.blogsavy.com/archives/147#comment-288</guid>
		<description>Look, it is very hard to sound polite commenting on this post. Web browsers do not NEED multi-core CPUs. In fact web browsers must not even consume more than 1% of a single core CPU, because they are what they are - freaking browsers.

I do not want browsers to introduce "parallelism" - it will be one more excuse for dumb-ass web-"developers" to produce sacks of useless JavaScript loops that consume my computing power while I am trying to read news or purchase a movie ticket.

Pentium III 800Mz used to be a powerhouse 5 years ago and it was more than enough for web surfing. Guess what? Nothing has changed in Web technology since then - same download speeds, same HTML, same CSS, same JavaScript. Why are you preaching for multi-core CPUs?</description>
		<content:encoded><![CDATA[<p>Look, it is very hard to sound polite commenting on this post. Web browsers do not NEED multi-core CPUs. In fact web browsers must not even consume more than 1% of a single core CPU, because they are what they are - freaking browsers.</p>
<p>I do not want browsers to introduce &#8220;parallelism&#8221; - it will be one more excuse for dumb-ass web-&#8221;developers&#8221; to produce sacks of useless JavaScript loops that consume my computing power while I am trying to read news or purchase a movie ticket.</p>
<p>Pentium III 800Mz used to be a powerhouse 5 years ago and it was more than enough for web surfing. Guess what? Nothing has changed in Web technology since then - same download speeds, same HTML, same CSS, same JavaScript. Why are you preaching for multi-core CPUs?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Tom Oehser</title>
		<link>http://pinderkent.blogsavy.com/archives/134#comment-287</link>
		<dc:creator>Tom Oehser</dc:creator>
		<pubDate>Tue, 18 Dec 2007 22:05:42 +0000</pubDate>
		<guid isPermaLink="false">http://pinderkent.blogsavy.com/archives/147#comment-287</guid>
		<description>SO MUCH BALONEY HERE!

Of *COURSE* threads-per-tab is better!

I'm watching, right now, a dual-Xeon machine that shows up under Ubuntu as having 4 processors because of Hyperthreading, running Gutsy and the latest Firefox, where the machine is AT LEAST 50% idle, ONLY because of the stupid single threaded browser architecture!  If I have 8 tabs open, and 3 of them have 500 ads, stupid plugins, badly written frequently refreshing Flash crap, etc., well, if the OPERATING SYSTEM could balance the resources, then the NICE tabs could still be responsive!  It is STUPID STUPID STUPID to criticize the BAD PAGES, or that web pages """SHOULDN'T""" be like that!  THEY ARE, THEY WILL CONTINUE TO BE, THEY'RE HERE, THEY SUCK, GET USED TO IT!  If I find a browser that will do a thread per tab with no stupid cross dependencies, I'LL DROP FIREFOX IN A NEW YORK MINUTE FOR IT!  Damn, we got idiots on this thread saying that 3 threads should be enough, and other idiots saying that threads don't help desktop apps, or worry about debugging getting harder (JOIN THE 1990s!), people worried about the memory cost (HEARD OF SHARED MEMORY and COPY-ON-WRITE?), someone saying that when a tab locks up their whole browser, they get out their quill and notify the maintainer of the offending site, some IDIOT thinking that on a "desktop" you just need one thread for IO and one for UI (that is fine, PER APP or PER TAB!  O-K, I'll grant you all, we need 3 or 4 threads per tab, not just 1... PER TAB).

JEEZUZ, it has been a while since a thread irritated me this much!  All because I searched for "multi-threaded browser" in frustration.</description>
		<content:encoded><![CDATA[<p>SO MUCH BALONEY HERE!</p>
<p>Of *COURSE* threads-per-tab is better!</p>
<p>I&#8217;m watching, right now, a dual-Xeon machine that shows up under Ubuntu as having 4 processors because of Hyperthreading, running Gutsy and the latest Firefox, where the machine is AT LEAST 50% idle, ONLY because of the stupid single threaded browser architecture!  If I have 8 tabs open, and 3 of them have 500 ads, stupid plugins, badly written frequently refreshing Flash crap, etc., well, if the OPERATING SYSTEM could balance the resources, then the NICE tabs could still be responsive!  It is STUPID STUPID STUPID to criticize the BAD PAGES, or that web pages &#8220;&#8221;"SHOULDN&#8217;T&#8221;"&#8221; be like that!  THEY ARE, THEY WILL CONTINUE TO BE, THEY&#8217;RE HERE, THEY SUCK, GET USED TO IT!  If I find a browser that will do a thread per tab with no stupid cross dependencies, I&#8217;LL DROP FIREFOX IN A NEW YORK MINUTE FOR IT!  Damn, we got idiots on this thread saying that 3 threads should be enough, and other idiots saying that threads don&#8217;t help desktop apps, or worry about debugging getting harder (JOIN THE 1990s!), people worried about the memory cost (HEARD OF SHARED MEMORY and COPY-ON-WRITE?), someone saying that when a tab locks up their whole browser, they get out their quill and notify the maintainer of the offending site, some IDIOT thinking that on a &#8220;desktop&#8221; you just need one thread for IO and one for UI (that is fine, PER APP or PER TAB!  O-K, I&#8217;ll grant you all, we need 3 or 4 threads per tab, not just 1&#8230; PER TAB).</p>
<p>JEEZUZ, it has been a while since a thread irritated me this much!  All because I searched for &#8220;multi-threaded browser&#8221; in frustration.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Samus_</title>
		<link>http://pinderkent.blogsavy.com/archives/134#comment-286</link>
		<dc:creator>Samus_</dc:creator>
		<pubDate>Wed, 12 Dec 2007 20:10:38 +0000</pubDate>
		<guid isPermaLink="false">http://pinderkent.blogsavy.com/archives/147#comment-286</guid>
		<description>"So it? quite unfortunate that we have all of this processing power available to us, but it can only be exploited to a limited degree because developers are choosing to write JavaScript-based applications that execute within parallelism-unfriendly Web browsers."

it's always easier to blame those damn goofy developers to choose to work on a technology that is demanded from the market and has lots of advantages as well as the same amount of problems that any other non-web app can have.</description>
		<content:encoded><![CDATA[<p>&#8220;So it? quite unfortunate that we have all of this processing power available to us, but it can only be exploited to a limited degree because developers are choosing to write JavaScript-based applications that execute within parallelism-unfriendly Web browsers.&#8221;</p>
<p>it&#8217;s always easier to blame those damn goofy developers to choose to work on a technology that is demanded from the market and has lots of advantages as well as the same amount of problems that any other non-web app can have.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Montoya</title>
		<link>http://pinderkent.blogsavy.com/archives/134#comment-285</link>
		<dc:creator>Montoya</dc:creator>
		<pubDate>Tue, 11 Dec 2007 15:42:49 +0000</pubDate>
		<guid isPermaLink="false">http://pinderkent.blogsavy.com/archives/147#comment-285</guid>
		<description>A thread per tab is both: a better solution than executing all tabs in a single thread, and, not nearly enough to improve the current browser problem. Browsers either have to be the most difficult programs to develop, or just the realm for the worst programming of our day. My computer can handle a high-detail, dynamic 3D game with many other players on the screen, particle effects, etc., I can run multiple applications at the same time such as a raster editor, vector editor, spreadsheet app, etc., but once I have 10 tabs open in my web browser with nothing else loaded and a heavy web app like Gmail starts throwing errors or executing large loops, everything comes to a halt.

Maybe it's just my fault for depending on Firefox so much, which is memory hungry and full of memory leaks. The point is, *something* needs to be done about the way browsers are developed, because they clearly are not taking advantage of present hardware and they are not handling these popular runtimes (JS, Flash, etc) well enough.</description>
		<content:encoded><![CDATA[<p>A thread per tab is both: a better solution than executing all tabs in a single thread, and, not nearly enough to improve the current browser problem. Browsers either have to be the most difficult programs to develop, or just the realm for the worst programming of our day. My computer can handle a high-detail, dynamic 3D game with many other players on the screen, particle effects, etc., I can run multiple applications at the same time such as a raster editor, vector editor, spreadsheet app, etc., but once I have 10 tabs open in my web browser with nothing else loaded and a heavy web app like Gmail starts throwing errors or executing large loops, everything comes to a halt.</p>
<p>Maybe it&#8217;s just my fault for depending on Firefox so much, which is memory hungry and full of memory leaks. The point is, *something* needs to be done about the way browsers are developed, because they clearly are not taking advantage of present hardware and they are not handling these popular runtimes (JS, Flash, etc) well enough.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Chris</title>
		<link>http://pinderkent.blogsavy.com/archives/134#comment-284</link>
		<dc:creator>Chris</dc:creator>
		<pubDate>Tue, 11 Dec 2007 00:24:48 +0000</pubDate>
		<guid isPermaLink="false">http://pinderkent.blogsavy.com/archives/147#comment-284</guid>
		<description>This is so true. It's beyond frustrating when I'm browsing the web on my linux box (Ubuntu 7.10, Athlon 3800x2, 2Gb DDR and 10k SCSI HD) which should rip through most applications fine and does but hit a website with alot of flash or javascript? Watch the system...just...start...to......die.

Its insane that we have browsers that aren't multi-threaded really at all.</description>
		<content:encoded><![CDATA[<p>This is so true. It&#8217;s beyond frustrating when I&#8217;m browsing the web on my linux box (Ubuntu 7.10, Athlon 3800&#215;2, 2Gb DDR and 10k SCSI HD) which should rip through most applications fine and does but hit a website with alot of flash or javascript? Watch the system&#8230;just&#8230;start&#8230;to&#8230;&#8230;die.</p>
<p>Its insane that we have browsers that aren&#8217;t multi-threaded really at all.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Samori Gorse</title>
		<link>http://pinderkent.blogsavy.com/archives/134#comment-283</link>
		<dc:creator>Samori Gorse</dc:creator>
		<pubDate>Mon, 10 Dec 2007 21:48:51 +0000</pubDate>
		<guid isPermaLink="false">http://pinderkent.blogsavy.com/archives/147#comment-283</guid>
		<description>It seriously doubt that each browser tab should have its own thread. It would be too many threads. Three of them should ease the pain enough.

One solution is to use Lazy Loading, freezing is usally caused by long loops, thoose loops will go from start to end no matter what, and nothing can run aside them. By using Lazy you actually cut long loops into smaller ones, and you're adding delay between them, so another part of the application can take the control.</description>
		<content:encoded><![CDATA[<p>It seriously doubt that each browser tab should have its own thread. It would be too many threads. Three of them should ease the pain enough.</p>
<p>One solution is to use Lazy Loading, freezing is usally caused by long loops, thoose loops will go from start to end no matter what, and nothing can run aside them. By using Lazy you actually cut long loops into smaller ones, and you&#8217;re adding delay between them, so another part of the application can take the control.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Tonto</title>
		<link>http://pinderkent.blogsavy.com/archives/134#comment-282</link>
		<dc:creator>Tonto</dc:creator>
		<pubDate>Mon, 10 Dec 2007 21:33:47 +0000</pubDate>
		<guid isPermaLink="false">http://pinderkent.blogsavy.com/archives/147#comment-282</guid>
		<description>Provide evidence for your claims, data and analysis.

Talk about it with some of the mozilla developers.

&#62;&#62; The solution is quite clear

Is that so? Did you just start learning about parallelism or something?</description>
		<content:encoded><![CDATA[<p>Provide evidence for your claims, data and analysis.</p>
<p>Talk about it with some of the mozilla developers.</p>
<p>&gt;&gt; The solution is quite clear</p>
<p>Is that so? Did you just start learning about parallelism or something?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Q</title>
		<link>http://pinderkent.blogsavy.com/archives/134#comment-281</link>
		<dc:creator>Q</dc:creator>
		<pubDate>Mon, 10 Dec 2007 21:27:11 +0000</pubDate>
		<guid isPermaLink="false">http://pinderkent.blogsavy.com/archives/147#comment-281</guid>
		<description>Hi, a little off-topic, but you should really change your captcha format.  Use Re-Captcha, it's much better, plus it helps translate books.  The captcha you use is trivial for computers to solve.</description>
		<content:encoded><![CDATA[<p>Hi, a little off-topic, but you should really change your captcha format.  Use Re-Captcha, it&#8217;s much better, plus it helps translate books.  The captcha you use is trivial for computers to solve.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: PENIX</title>
		<link>http://pinderkent.blogsavy.com/archives/134#comment-280</link>
		<dc:creator>PENIX</dc:creator>
		<pubDate>Mon, 10 Dec 2007 21:21:01 +0000</pubDate>
		<guid isPermaLink="false">http://pinderkent.blogsavy.com/archives/147#comment-280</guid>
		<description>Per tab as a first step maybe. The entire page really should be multi-threaded. The GUI MUST be separated from the rendering processes. It's happened so many times that a page renders incorrectly or a runaway Javascript application becomes a CPU hog and the main interface stops responding.</description>
		<content:encoded><![CDATA[<p>Per tab as a first step maybe. The entire page really should be multi-threaded. The GUI MUST be separated from the rendering processes. It&#8217;s happened so many times that a page renders incorrectly or a runaway Javascript application becomes a CPU hog and the main interface stops responding.</p>
]]></content:encoded>
	</item>
</channel>
</rss>
