<?xml version="1.0" encoding="UTF-8"?><!-- generator="wordpress/2.2" -->
<rss version="2.0" 
	xmlns:content="http://purl.org/rss/1.0/modules/content/">
<channel>
	<title>Comments on: MagLev and language implementation</title>
	<link>http://collison.ie/blog/2008/06/maglev-and-language-implementation</link>
	<description>I hate self-referential taglines</description>
	<pubDate>Fri, 12 Mar 2010 12:39:23 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.2</generator>

	<item>
		<title>By: Gemstone&#8217;s MagLev presented at RubyConf &#171; The Weekly Squeak</title>
		<link>http://collison.ie/blog/2008/06/maglev-and-language-implementation#comment-3970</link>
		<author>Gemstone&#8217;s MagLev presented at RubyConf &#171; The Weekly Squeak</author>
		<pubDate>Tue, 03 Jun 2008 17:30:00 +0000</pubDate>
		<guid>http://collison.ie/blog/2008/06/maglev-and-language-implementation#comment-3970</guid>
		<description>[...] The presentation caused lots of excitement at the conference, and has sparked lots of heated debate within the Ruby development community with some very different views of MagLev from Charles Nutter, Giles Bowkett,  Obie Fernandez, and Antonio Cangiano, as well as an article at Slashdot and posts all over Reddit.  Avi has a blog-post addressing some of the discussion, as does Patrick Collison. [...]</description>
		<content:encoded><![CDATA[<p>[&#8230;] The presentation caused lots of excitement at the conference, and has sparked lots of heated debate within the Ruby development community with some very different views of MagLev from Charles Nutter, Giles Bowkett,  Obie Fernandez, and Antonio Cangiano, as well as an article at Slashdot and posts all over Reddit.  Avi has a blog-post addressing some of the discussion, as does Patrick Collison. [&#8230;]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Mark</title>
		<link>http://collison.ie/blog/2008/06/maglev-and-language-implementation#comment-3943</link>
		<author>Mark</author>
		<pubDate>Tue, 03 Jun 2008 04:13:01 +0000</pubDate>
		<guid>http://collison.ie/blog/2008/06/maglev-and-language-implementation#comment-3943</guid>
		<description>Hear hear!</description>
		<content:encoded><![CDATA[<p>Hear hear!</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Brett Morgan</title>
		<link>http://collison.ie/blog/2008/06/maglev-and-language-implementation#comment-3942</link>
		<author>Brett Morgan</author>
		<pubDate>Tue, 03 Jun 2008 03:49:13 +0000</pubDate>
		<guid>http://collison.ie/blog/2008/06/maglev-and-language-implementation#comment-3942</guid>
		<description>Thank christ someone said it. I'm a bit sick of people ignoring all the different techniques for running smalltalk like languages quickly.</description>
		<content:encoded><![CDATA[<p>Thank christ someone said it. I&#8217;m a bit sick of people ignoring all the different techniques for running smalltalk like languages quickly.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: ben</title>
		<link>http://collison.ie/blog/2008/06/maglev-and-language-implementation#comment-3940</link>
		<author>ben</author>
		<pubDate>Tue, 03 Jun 2008 01:19:43 +0000</pubDate>
		<guid>http://collison.ie/blog/2008/06/maglev-and-language-implementation#comment-3940</guid>
		<description>Who cares what some commercial software tied to an ODB is doing? Anyone?
Rails took off because it was open source, not because it was proprietary.

Honestly, do you think they will open source it? I doubt it!</description>
		<content:encoded><![CDATA[<p>Who cares what some commercial software tied to an ODB is doing? Anyone?<br />
Rails took off because it was open source, not because it was proprietary.</p>
<p>Honestly, do you think they will open source it? I doubt it!</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: links for 2008-06-03 &#171; Mike Does Tech</title>
		<link>http://collison.ie/blog/2008/06/maglev-and-language-implementation#comment-3938</link>
		<author>links for 2008-06-03 &#171; Mike Does Tech</author>
		<pubDate>Tue, 03 Jun 2008 00:33:33 +0000</pubDate>
		<guid>http://collison.ie/blog/2008/06/maglev-and-language-implementation#comment-3938</guid>
		<description>[...] MagLev and language implementation (tags: maglev) [...]</description>
		<content:encoded><![CDATA[<p>[&#8230;] MagLev and language implementation (tags: maglev) [&#8230;]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: StCredZero</title>
		<link>http://collison.ie/blog/2008/06/maglev-and-language-implementation#comment-3932</link>
		<author>StCredZero</author>
		<pubDate>Mon, 02 Jun 2008 19:58:36 +0000</pubDate>
		<guid>http://collison.ie/blog/2008/06/maglev-and-language-implementation#comment-3932</guid>
		<description>Sho has changed his tone in his blog comments.</description>
		<content:encoded><![CDATA[<p>Sho has changed his tone in his blog comments.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: markus</title>
		<link>http://collison.ie/blog/2008/06/maglev-and-language-implementation#comment-3930</link>
		<author>markus</author>
		<pubDate>Mon, 02 Jun 2008 19:38:31 +0000</pubDate>
		<guid>http://collison.ie/blog/2008/06/maglev-and-language-implementation#comment-3930</guid>
		<description>Seems there is some conflict beginning to start, which I believe is not to the best 
but here goes .... ;)

I only want to nitpick on one thing here because it is my personal pet peeve:

"For a start, MRI is neither a compiler or a VM—it’s an interpreter."

The thing is that right now no VM can "deliver" really the way MRI does. Let's put speed issues aside for now (and I think the situation with python is not sooooooo much different anyway)

As far as I understood it, every VM has a few problems, so it is no big deal to have problems here or there. It's still work in progress....
Some have less problems than others, some can run rails, some can't etc... doesn't matter that much. Future will show of improves.

BUT any bugs that are left after - let's say 2 years - would basically mean that they target _another_ language basically, not ruby (if it behaves differently from MRI).
And I say ruby, because there is more than rails in (or within) ruby. Rails is just a comparatively complex application compared to others. 

So this is my foremost scepticism because, frankly, I don't care about a nice VM that runs a few apps and forgets the rest. I am not accusing anyone of this at all, it just seems this way with all the VMS right now. If the goal is to let rails users run their rails apps on VMs, then I think this should be made clear, so that it does not generate even more hype than rails generated, and we "normal" ruby folks can continue to concentrate on other things and let the railsers play with that.

Being completely "ruby compliant" should be the number one priority.</description>
		<content:encoded><![CDATA[<p>Seems there is some conflict beginning to start, which I believe is not to the best<br />
but here goes &#8230;. ;)</p>
<p>I only want to nitpick on one thing here because it is my personal pet peeve:</p>
<p>&#8220;For a start, MRI is neither a compiler or a VM—it’s an interpreter.&#8221;</p>
<p>The thing is that right now no VM can &#8220;deliver&#8221; really the way MRI does. Let&#8217;s put speed issues aside for now (and I think the situation with python is not sooooooo much different anyway)</p>
<p>As far as I understood it, every VM has a few problems, so it is no big deal to have problems here or there. It&#8217;s still work in progress&#8230;.<br />
Some have less problems than others, some can run rails, some can&#8217;t etc&#8230; doesn&#8217;t matter that much. Future will show of improves.</p>
<p>BUT any bugs that are left after - let&#8217;s say 2 years - would basically mean that they target _another_ language basically, not ruby (if it behaves differently from MRI).<br />
And I say ruby, because there is more than rails in (or within) ruby. Rails is just a comparatively complex application compared to others. </p>
<p>So this is my foremost scepticism because, frankly, I don&#8217;t care about a nice VM that runs a few apps and forgets the rest. I am not accusing anyone of this at all, it just seems this way with all the VMS right now. If the goal is to let rails users run their rails apps on VMs, then I think this should be made clear, so that it does not generate even more hype than rails generated, and we &#8220;normal&#8221; ruby folks can continue to concentrate on other things and let the railsers play with that.</p>
<p>Being completely &#8220;ruby compliant&#8221; should be the number one priority.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Sho</title>
		<link>http://collison.ie/blog/2008/06/maglev-and-language-implementation#comment-3926</link>
		<author>Sho</author>
		<pubDate>Mon, 02 Jun 2008 18:12:51 +0000</pubDate>
		<guid>http://collison.ie/blog/2008/06/maglev-and-language-implementation#comment-3926</guid>
		<description>Well, sorry for the mindless rant! But it was really a rant against vapourware and hype.

See that quote you want to save for posterity? Notice the words in the middle? "Until conclusively proven otherwise". I'm a Ruby developer and I would love to be proven wrong. If this product fulfils these claims then I would buy it in a millisecond. No need to save my quote somewhere, I'll put it up in flashing letters as I rave about how happy I am to be proven completely wrong.

However, I have plenty of doubts about Maglev. I've been paying pretty close attention through the development of JRuby and something their experience has really rammed home for me is the complexity and nuance of Ruby as a language. It may well be similar to Smalltalk in some respects but I don't think it's as simple a matter to transmute a VM designed for one to the other. And I am well aware that MRI is not a particularly advanced interpreter but you can't possibly argue that the JRuby one isn't.

BTW, you're right that I mixed up VM and interpreter a couple of times, I'll fix that. But your "correction" on the definition of shared-nothing is ridiculous - my usage is the commonly accepted one, you're just nitpicking at that point.

Anyway, no offence intended and sorry if it was caused. I just don't like way-ahead-of-release vapourware, and the hype that followed it is inarguably unjustified based on the very incomplete demonstration given. I'll stand by that opinion, and wish the best of luck to your friends to live up to the expectations they've created.</description>
		<content:encoded><![CDATA[<p>Well, sorry for the mindless rant! But it was really a rant against vapourware and hype.</p>
<p>See that quote you want to save for posterity? Notice the words in the middle? &#8220;Until conclusively proven otherwise&#8221;. I&#8217;m a Ruby developer and I would love to be proven wrong. If this product fulfils these claims then I would buy it in a millisecond. No need to save my quote somewhere, I&#8217;ll put it up in flashing letters as I rave about how happy I am to be proven completely wrong.</p>
<p>However, I have plenty of doubts about Maglev. I&#8217;ve been paying pretty close attention through the development of JRuby and something their experience has really rammed home for me is the complexity and nuance of Ruby as a language. It may well be similar to Smalltalk in some respects but I don&#8217;t think it&#8217;s as simple a matter to transmute a VM designed for one to the other. And I am well aware that MRI is not a particularly advanced interpreter but you can&#8217;t possibly argue that the JRuby one isn&#8217;t.</p>
<p>BTW, you&#8217;re right that I mixed up VM and interpreter a couple of times, I&#8217;ll fix that. But your &#8220;correction&#8221; on the definition of shared-nothing is ridiculous - my usage is the commonly accepted one, you&#8217;re just nitpicking at that point.</p>
<p>Anyway, no offence intended and sorry if it was caused. I just don&#8217;t like way-ahead-of-release vapourware, and the hype that followed it is inarguably unjustified based on the very incomplete demonstration given. I&#8217;ll stand by that opinion, and wish the best of luck to your friends to live up to the expectations they&#8217;ve created.</p>
]]></content:encoded>
	</item>
</channel>
</rss>
