Comments on: Twitter developer: Rails performance blows http://rc3.org/2007/04/12/twitter-developer-rails-performance-blows/ Strong opinions weakly held Mon, 15 Mar 2010 23:37:05 +0000 http://wordpress.org/?v=2.9.1 hourly 1 By: Will SXSW 2008 Be the Death of Twitter? | Medialoper http://rc3.org/2007/04/12/twitter-developer-rails-performance-blows/comment-page-1/#comment-2455 Will SXSW 2008 Be the Death of Twitter? | Medialoper Thu, 06 Mar 2008 13:54:30 +0000 http://rc3.org/wordpress/?p=7725#comment-2455 <p>[...] succeeded where its competitors have failed is all the more amazing considering the downtime and performance issues the application has [...]</p> [...] succeeded where its competitors have failed is all the more amazing considering the downtime and performance issues the application has [...]

]]> By: DOH http://rc3.org/2007/04/12/twitter-developer-rails-performance-blows/comment-page-1/#comment-1594 DOH Fri, 13 Apr 2007 16:37:15 +0000 http://rc3.org/wordpress/?p=7725#comment-1594 <p>@Jonas,</p> <p>I worked in a Ruby on Rails shop and I'm going to pick up PHP and dump Ruby soon.</p> <p>Why? Well... why not? There are tons of PHP jobs out there, tons of PHP freelance jobs, tons of "please drupal'ed my website", tons of "please fix my php"</p> <p>Flickr, Last.fm, Yahoo!, Wordpress, Joomla, Drupal, Apache+MySQL support toward PHP.</p> <p>One common theme I heard a lot in my shop was "Well.. whatever DHH thinks...<em>chuckles</em>... ".</p> @Jonas,

I worked in a Ruby on Rails shop and I’m going to pick up PHP and dump Ruby soon.

Why? Well… why not? There are tons of PHP jobs out there, tons of PHP freelance jobs, tons of “please drupal’ed my website”, tons of “please fix my php”

Flickr, Last.fm, Yahoo!, Wordpress, Joomla, Drupal, Apache+MySQL support toward PHP.

One common theme I heard a lot in my shop was “Well.. whatever DHH thinks…chuckles… “.

]]> By: Rebort http://rc3.org/2007/04/12/twitter-developer-rails-performance-blows/comment-page-1/#comment-1593 Rebort Fri, 13 Apr 2007 01:23:17 +0000 http://rc3.org/wordpress/?p=7725#comment-1593 <p>@Jonas -- good points all around and I agree.</p> <p>(I also laughed out loud at your "Not to mention that someone who works in Ruby will never ever switch to the pseudo-OOP-no-namespaces-mess that is PHP." comment. It's too true. =D )</p> @Jonas — good points all around and I agree.

(I also laughed out loud at your “Not to mention that someone who works in Ruby will never ever switch to the pseudo-OOP-no-namespaces-mess that is PHP.” comment. It’s too true. =D )

]]> By: Jonas http://rc3.org/2007/04/12/twitter-developer-rails-performance-blows/comment-page-1/#comment-1592 Jonas Fri, 13 Apr 2007 00:23:04 +0000 http://rc3.org/wordpress/?p=7725#comment-1592 <p>@Rebort: as PHP applications without an engine optimizer have to pull in the full script code at each request, even if you leverage the OS' hdd-io-cache the script's parsing time becomes a major performance factor for each request, especially with "massive" CMSs like Drupal or eZ. PHP's design choices force this problem and I've seen a lot of performance-problems because of it.</p> <p>Granted, the <em>execution-speed</em> after the code is pulled in is highly optimized by now. Essentially PHP is today where Ruby will certainly be in two years time.</p> <p>So: you're right. If someone had to make a decision based solely on performance of an implementation of his system, he'd be better off going with PHP.</p> <p>Still, I think that the expectation of an uptake of Java or PHP from this discussion can only be maintained if you limit the discussion to performance. Because if you factor in all the problems that <em>these</em> platforms have, Ruby will come out ahead by a mile.</p> <p>Not to mention that someone who works in Ruby will never ever switch to the pseudo-OOP-no-namespaces-mess that is PHP.</p> <p>Anyway... it's fun to watch ;-)</p> @Rebort: as PHP applications without an engine optimizer have to pull in the full script code at each request, even if you leverage the OS’ hdd-io-cache the script’s parsing time becomes a major performance factor for each request, especially with “massive” CMSs like Drupal or eZ. PHP’s design choices force this problem and I’ve seen a lot of performance-problems because of it.

Granted, the execution-speed after the code is pulled in is highly optimized by now. Essentially PHP is today where Ruby will certainly be in two years time.

So: you’re right. If someone had to make a decision based solely on performance of an implementation of his system, he’d be better off going with PHP.

Still, I think that the expectation of an uptake of Java or PHP from this discussion can only be maintained if you limit the discussion to performance. Because if you factor in all the problems that these platforms have, Ruby will come out ahead by a mile.

Not to mention that someone who works in Ruby will never ever switch to the pseudo-OOP-no-namespaces-mess that is PHP.

Anyway… it’s fun to watch ;-)

]]> By: Mack D. Male http://rc3.org/2007/04/12/twitter-developer-rails-performance-blows/comment-page-1/#comment-1591 Mack D. Male Fri, 13 Apr 2007 00:22:57 +0000 http://rc3.org/wordpress/?p=7725#comment-1591 <p>Umm...how about ASP.NET? Wicked scalability and performance.</p> Umm…how about ASP.NET? Wicked scalability and performance.

]]> By: kellan http://rc3.org/2007/04/12/twitter-developer-rails-performance-blows/comment-page-1/#comment-1590 kellan Thu, 12 Apr 2007 23:55:00 +0000 http://rc3.org/wordpress/?p=7725#comment-1590 <p>DHH "response" reminds me of the early days of Rails development where you had to be part of the secret cabal to find out about performance/security issues, and a party line of universal positivity was strictly enforced. I'm glad those days are gone.</p> DHH “response” reminds me of the early days of Rails development where you had to be part of the secret cabal to find out about performance/security issues, and a party line of universal positivity was strictly enforced. I’m glad those days are gone.

]]> By: Rebort http://rc3.org/2007/04/12/twitter-developer-rails-performance-blows/comment-page-1/#comment-1589 Rebort Thu, 12 Apr 2007 23:45:14 +0000 http://rc3.org/wordpress/?p=7725#comment-1589 <p>@Jonas: "but performance-wise, PHP is not necessarily the better trade-off, nor has it "excelled" at performance."</p> <p>Eight or ten years ago, I never heard anyone saying PHP was particularly slow (unless it was a language war, Perl vs Python vs PHP sort of thing). I never heard that had the sort of scaling problems that Ruby/Rails seems to have.</p> <p>Whereas there have been plenty of discussions and comments made about Ruby's inherent slowness, and the difficulty of scaling Rails "out of the box." And these comments are coming from people looking at the thing as a standalone technology, without too many comparisons and language war stuff. You see enough of it, and you have to figure there's something to it, beyond the average geek with an axe to grind.</p> <p>I do agree, though, that PHP has many many other factors that will cause grief, but at the same time the discussion wasn't about that.</p> @Jonas: “but performance-wise, PHP is not necessarily the better trade-off, nor has it “excelled” at performance.”

Eight or ten years ago, I never heard anyone saying PHP was particularly slow (unless it was a language war, Perl vs Python vs PHP sort of thing). I never heard that had the sort of scaling problems that Ruby/Rails seems to have.

Whereas there have been plenty of discussions and comments made about Ruby’s inherent slowness, and the difficulty of scaling Rails “out of the box.” And these comments are coming from people looking at the thing as a standalone technology, without too many comparisons and language war stuff. You see enough of it, and you have to figure there’s something to it, beyond the average geek with an axe to grind.

I do agree, though, that PHP has many many other factors that will cause grief, but at the same time the discussion wasn’t about that.

]]> By: Jonas http://rc3.org/2007/04/12/twitter-developer-rails-performance-blows/comment-page-1/#comment-1588 Jonas Thu, 12 Apr 2007 22:29:52 +0000 http://rc3.org/wordpress/?p=7725#comment-1588 <p>I'd say that adoption of PHP or Java will not pick up, Groovy and Python/TurboGears or Python/Django might see more adoption... but performance-wise, PHP is not necessarily the better trade-off, nor has it "excelled" at performance. Anyway, PHP will cause you a lot of grief with all its problems and I think people are realizing that.</p> I’d say that adoption of PHP or Java will not pick up, Groovy and Python/TurboGears or Python/Django might see more adoption… but performance-wise, PHP is not necessarily the better trade-off, nor has it “excelled” at performance. Anyway, PHP will cause you a lot of grief with all its problems and I think people are realizing that.

]]> By: Gerald Buckley http://rc3.org/2007/04/12/twitter-developer-rails-performance-blows/comment-page-1/#comment-1587 Gerald Buckley Thu, 12 Apr 2007 21:32:37 +0000 http://rc3.org/wordpress/?p=7725#comment-1587 <p>What with Leopard and XCode 3 shipping with Ruby/Rails as one of the default frameworks... This has to be addressed sooner. Not later.</p> <p>Twitter, thanks for crossing the line. Now, something will get done or Ruby/Rails will get tossed on the rubbish heap.</p> What with Leopard and XCode 3 shipping with Ruby/Rails as one of the default frameworks… This has to be addressed sooner. Not later.

Twitter, thanks for crossing the line. Now, something will get done or Ruby/Rails will get tossed on the rubbish heap.

]]> By: Rebort http://rc3.org/2007/04/12/twitter-developer-rails-performance-blows/comment-page-1/#comment-1586 Rebort Thu, 12 Apr 2007 21:28:00 +0000 http://rc3.org/wordpress/?p=7725#comment-1586 <p>Well, that didn't take long =)</p> <p><a href="http://www.loudthinking.com/arc/000608.html" rel="nofollow">http://www.loudthinking.com/arc/000608.html</a></p> Well, that didn’t take long =)

http://www.loudthinking.com/arc/000608.html

]]>