<?xml version="1.0" encoding="UTF-8"?><!-- generator="wordpress/2.0.2" -->
<rss version="2.0" 
	xmlns:content="http://purl.org/rss/1.0/modules/content/">
<channel>
	<title>Comments on: Narrowing the focus</title>
	<link>http://tech.puredanger.com/2008/07/08/narrowing-the-focus/</link>
	<description>Alex Miller's technical blog</description>
	<pubDate>Thu, 08 Jan 2009 22:44:45 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.0.2</generator>

	<item>
		<title>by: Taylor Gautier</title>
		<link>http://tech.puredanger.com/2008/07/08/narrowing-the-focus/#comment-63437</link>
		<pubDate>Wed, 09 Jul 2008 15:20:24 +0000</pubDate>
		<guid>http://tech.puredanger.com/2008/07/08/narrowing-the-focus/#comment-63437</guid>
					<description>@Guillame - we definitely do have a some interest in Grails.

Is there anything you or your community can do here - we would be more than happy to help out on an app that someone else owned but as Alex mentions, right now that wouldn't be our focus.</description>
		<content:encoded><![CDATA[<p>@Guillame - we definitely do have a some interest in Grails.</p>
<p>Is there anything you or your community can do here - we would be more than happy to help out on an app that someone else owned but as Alex mentions, right now that wouldn&#8217;t be our focus.
</p>
]]></content:encoded>
				</item>
	<item>
		<title>by: Alex</title>
		<link>http://tech.puredanger.com/2008/07/08/narrowing-the-focus/#comment-63396</link>
		<pubDate>Wed, 09 Jul 2008 11:27:39 +0000</pubDate>
		<guid>http://tech.puredanger.com/2008/07/08/narrowing-the-focus/#comment-63396</guid>
					<description>@Guillaume: Yep, that would be cool.  However, at the moment I believe the majority of our customers are using something other than Grails.  I'll let you work on that!  

Once we have the reference app out there, it should provide a significant infrastructure that people could use to rebuild the app with other web technologies if desired.</description>
		<content:encoded><![CDATA[<p>@Guillaume: Yep, that would be cool.  However, at the moment I believe the majority of our customers are using something other than Grails.  I&#8217;ll let you work on that!  </p>
<p>Once we have the reference app out there, it should provide a significant infrastructure that people could use to rebuild the app with other web technologies if desired.
</p>
]]></content:encoded>
				</item>
	<item>
		<title>by: Guillaume Laforge</title>
		<link>http://tech.puredanger.com/2008/07/08/narrowing-the-focus/#comment-63354</link>
		<pubDate>Wed, 09 Jul 2008 07:50:27 +0000</pubDate>
		<guid>http://tech.puredanger.com/2008/07/08/narrowing-the-focus/#comment-63354</guid>
					<description>A Grails app for the web stack? Would be cool :-)
I'd be interested in seeing how this would be configured and be working.</description>
		<content:encoded><![CDATA[<p>A Grails app for the web stack? Would be cool :-)<br />
I&#8217;d be interested in seeing how this would be configured and be working.
</p>
]]></content:encoded>
				</item>
	<item>
		<title>by: Taylor Gautier</title>
		<link>http://tech.puredanger.com/2008/07/08/narrowing-the-focus/#comment-63329</link>
		<pubDate>Wed, 09 Jul 2008 05:33:08 +0000</pubDate>
		<guid>http://tech.puredanger.com/2008/07/08/narrowing-the-focus/#comment-63329</guid>
					<description>To clarify on upgrades, the goal is 100% uptime during any upgrade.  

As Alex mentions, most app upgrades can be done easily today.  Some particular types can present some difficult issues, but we have solutions for those - just not implemented yet.  Since those kinds of upgrades are less common, we haven't focused on building features to support those particular types of upgrades.

For Terracotta upgrades, the same is true.  There are no fundamental or architectural issues at play here, we just haven't to date focused on upward compatibility.  There aren't any real major technical issues, other than it is something we will focus on when we have to - for now we are highly focused on the other parts of the user experience(s) as Alex mentions.</description>
		<content:encoded><![CDATA[<p>To clarify on upgrades, the goal is 100% uptime during any upgrade.  </p>
<p>As Alex mentions, most app upgrades can be done easily today.  Some particular types can present some difficult issues, but we have solutions for those - just not implemented yet.  Since those kinds of upgrades are less common, we haven&#8217;t focused on building features to support those particular types of upgrades.</p>
<p>For Terracotta upgrades, the same is true.  There are no fundamental or architectural issues at play here, we just haven&#8217;t to date focused on upward compatibility.  There aren&#8217;t any real major technical issues, other than it is something we will focus on when we have to - for now we are highly focused on the other parts of the user experience(s) as Alex mentions.
</p>
]]></content:encoded>
				</item>
	<item>
		<title>by: Alex</title>
		<link>http://tech.puredanger.com/2008/07/08/narrowing-the-focus/#comment-63303</link>
		<pubDate>Wed, 09 Jul 2008 02:59:10 +0000</pubDate>
		<guid>http://tech.puredanger.com/2008/07/08/narrowing-the-focus/#comment-63303</guid>
					<description>Yep, exactly the kind of stuff we want to have more experience with.  

I think the answer above depends on whether you are upgrading your app or Terracotta itself.  With your own app, you can roll the upgrade through (assuming your shared data structures are not evolving much) without ever having a loss of service.  

Upgrading Terracotta versions generally requires a full shutdown although there are some ways to partition your servers to even do that without a loss of service.</description>
		<content:encoded><![CDATA[<p>Yep, exactly the kind of stuff we want to have more experience with.  </p>
<p>I think the answer above depends on whether you are upgrading your app or Terracotta itself.  With your own app, you can roll the upgrade through (assuming your shared data structures are not evolving much) without ever having a loss of service.  </p>
<p>Upgrading Terracotta versions generally requires a full shutdown although there are some ways to partition your servers to even do that without a loss of service.
</p>
]]></content:encoded>
				</item>
</channel>
</rss>
