<?xml version="1.0" encoding="utf-8"?>
<rss xmlns:atom="http://www.w3.org/2005/Atom" version="2.0"><channel><title>Disqus - Latest Comments for metajack</title><link>http://disqus.com/by/metajack/</link><description></description><atom:link href="http://disqus.com/metajack/comments.rss" rel="self"></atom:link><language>en</language><lastBuildDate>Thu, 03 Jul 2014 12:04:38 -0000</lastBuildDate><item><title>Re: Choosing An XMPP Server</title><link>http://metajack.im/2008/08/26/choosing-an-xmpp-server/#comment-1466887569</link><description>&lt;p&gt;If I were looking again today I would start by playing with MongooseIM which is a modern ejabberd fork by the folks at Erlang Solutions. Depending on your needs, Prosody could also be a good choice, and that's what I run for &lt;a href="http://metajack.im" rel="nofollow noopener" target="_blank" title="metajack.im"&gt;metajack.im&lt;/a&gt;.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Thu, 03 Jul 2014 12:04:38 -0000</pubDate></item><item><title>Re: 
      Servo Update: Navigation, Scrolling, GPU Rendering, Underlines, and more
    </title><link>http://metajack.im/2013/05/26/servo-update-navigation-scrolling-gpu-rendering-underlines-and-more/#comment-985646724</link><description>&lt;p&gt;We had some discussion in IRC and came to the same conclusion. FirefoxOS also has a remote attribute for iframes, but this seems to imply additional behavior that isn't great for same-origin iframes.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Fri, 02 Aug 2013 08:50:35 -0000</pubDate></item><item><title>Re: 
      Servo Update: Navigation, Scrolling, GPU Rendering, Underlines, and more
    </title><link>http://metajack.im/2013/05/26/servo-update-navigation-scrolling-gpu-rendering-underlines-and-more/#comment-913955675</link><description>&lt;p&gt;We use MPLv2 on the Servo core bits. Non-core dependencies use the Rust license which is dual APLv2 and MIT.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Thu, 30 May 2013 12:36:01 -0000</pubDate></item><item><title>Re: 
      Servo Update: Upgrading Rust, GPU Rendering, and Automation
    </title><link>http://metajack.im/2013/04/12/servo-update-upgrading-rust-gpu-rendering-and-automation/#comment-865210403</link><description>&lt;p&gt;I don't think he's pushed a branch for this yet. Right now we switch out the stacks when calling C code, which is a performance hit. Patrick's changes are going to remove the stack switching. I don't remember exactly how. The best way to find out is probably to ask in #rust.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Tue, 16 Apr 2013 11:18:29 -0000</pubDate></item><item><title>Re: 
      Servo Update: Upgrading Rust, GPU Rendering, and Automation
    </title><link>http://metajack.im/2013/04/12/servo-update-upgrading-rust-gpu-rendering-and-automation/#comment-863145344</link><description>&lt;p&gt;Yes. There is a pull request for Rust that adds SIMD. I'm not sure if that's what Rust will get exactly, but we're already thinking about this.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Sun, 14 Apr 2013 09:31:51 -0000</pubDate></item><item><title>Re: XMPP Is Better With BOSH</title><link>http://metajack.im/2008/07/02/xmpp-is-better-with-bosh/#comment-853154110</link><description>&lt;p&gt;It would be better to ask on the punjab list. I'm sure someone there could help. I am not sure how to tell punjab how to use proxies.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Fri, 05 Apr 2013 11:13:16 -0000</pubDate></item><item><title>Re: Which BOSH Server Do You Need?</title><link>http://metajack.im/2008/09/08/which-bosh-server-do-you-need/#comment-669296004</link><description>&lt;p&gt;It appears to be working fine here. Do you have any more information about the failure?&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Tue, 02 Oct 2012 23:08:52 -0000</pubDate></item><item><title>Re: Strophe.js 1.0.2 Released</title><link>http://metajack.im/2011/06/19/strophejs-102-released/#comment-582149787</link><description>&lt;p&gt;See &lt;a href="http://metajack.im/2008/10/03/getting-attached-to-strophe/" rel="nofollow noopener" target="_blank" title="http://metajack.im/2008/10/03/getting-attached-to-strophe/"&gt;http://metajack.im/2008/10/...&lt;/a&gt; for more information on session attachment, which allows you to detach and reattach to an underlying BOSH session. Many people have used this for keeping XMPP around across page loads. I must warn you though, it's not a panacea especially if you're application has a lot of state.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Mon, 09 Jul 2012 09:52:06 -0000</pubDate></item><item><title>Re: Which BOSH Server Do You Need?</title><link>http://metajack.im/2008/09/08/which-bosh-server-do-you-need/#comment-440550496</link><description>&lt;p&gt;It depends on the BOSH connection manager you are using. Typically port 5280 is used, but Openfire uses 7070 by default. Consult the documentation for your server or ask on their mailing list.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Thu, 16 Feb 2012 00:15:01 -0000</pubDate></item><item><title>Re: Which BOSH Server Do You Need?</title><link>http://metajack.im/2008/09/08/which-bosh-server-do-you-need/#comment-440529610</link><description>&lt;p&gt;It can uses https to encrypt the client to BOSH server traffic. The BOSH server can also use TLS to encrypt BOSH server to XMPP server traffic.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Wed, 15 Feb 2012 23:30:39 -0000</pubDate></item><item><title>Re: Cross-Domain AJAX for XMPP HTTP Binding Made Easy</title><link>http://metajack.im/2010/01/19/crossdomain-ajax-for-xmpp-http-binding-made-easy/#comment-432233197</link><description>&lt;p&gt;The comments section here is not the best place for tech support. Please try the mailing list at &lt;a href="http://groups.google.com/group/strophe" rel="nofollow noopener" target="_blank" title="http://groups.google.com/group/strophe"&gt;http://groups.google.com/gr...&lt;/a&gt;.&lt;/p&gt;&lt;p&gt;To answer your question, Facebook chat is not federated, so your personal XMPP server will never be able to communicate with it until they support federation.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Tue, 07 Feb 2012 15:08:56 -0000</pubDate></item><item><title>Re: Introducing Snack Words</title><link>http://metajack.im/2011/03/28/introducing-snack-words/#comment-430542760</link><description>&lt;p&gt;A non-trivial amount of the reason was swift development, but I'll try and write about this at some point.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Sun, 05 Feb 2012 22:43:08 -0000</pubDate></item><item><title>Re: Building Ejabberd Modules: A Build Tool Battle Won By Autotools</title><link>http://metajack.im/2008/10/07/building-ejabberd-modules-a-build-tool-battle-won-by-autotools/#comment-430508127</link><description>&lt;p&gt;That's good news. I'll give SCons another try at some point in the future.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Sun, 05 Feb 2012 21:53:04 -0000</pubDate></item><item><title>Re: Migrating To Ejabberd: The Gory Details</title><link>http://metajack.im/2008/08/27/migrating-to-ejabberd-the-gory-details/#comment-430393181</link><description>&lt;p&gt;I fixed the links in this and other posts. Thanks for the heads up.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Sun, 05 Feb 2012 18:40:57 -0000</pubDate></item><item><title>Re: Writing Ejabberd Modules: Presence Storms</title><link>http://metajack.im/2008/08/28/writing-ejabberd-modules-presence-storms/#comment-430190125</link><description>&lt;p&gt;The new link is: &lt;a href="http://anders.conbere.org/blog/2008/07/17/building_ejabberd_modules_-_part_2_-_generic_modules/" rel="nofollow noopener" target="_blank" title="http://anders.conbere.org/blog/2008/07/17/building_ejabberd_modules_-_part_2_-_generic_modules/"&gt;http://anders.conbere.org/b...&lt;/a&gt; &lt;br&gt;I'll update the post content as well.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Sun, 05 Feb 2012 12:05:36 -0000</pubDate></item><item><title>Re: The Numbers Behind the Twitter Data Silo</title><link>http://metajack.im/2012/01/30/the-numbers-behind-the-twitter-data-silo/#comment-425603572</link><description>&lt;p&gt;If they do that, then they must continually rescrape that page, and often there won't be any new tweets. So some requests return no new data, but some return multiple tweets. They could scale the crawl rate of course to try and optimize, but over hundreds of millions of users, that's a lot of crawls that return nothing new.&lt;/p&gt;&lt;p&gt;I don't think Twitter's crawl allowance is increasing over time, and Google has to be thinking ahead. Even if they keep up now, how long can they sustain it?&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Tue, 31 Jan 2012 13:21:30 -0000</pubDate></item><item><title>Re: The Numbers Behind the Twitter Data Silo</title><link>http://metajack.im/2012/01/30/the-numbers-behind-the-twitter-data-silo/#comment-425601405</link><description>&lt;p&gt;If you don't index retweets then you lose one of the biggest indicators of social relevance that exists. Knowing the retweet count is probably important.&lt;/p&gt;&lt;p&gt;As for spammers, Twitter supposedly combats them as well, so I'm guessing while Google may be better at this, the margin of difference is probably not huge.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Tue, 31 Jan 2012 13:18:34 -0000</pubDate></item><item><title>Re: Superfeedr : Publish Subscribe for the web</title><link>http://blog.superfeedr.com/pubsub-web/#comment-410055091</link><description>&lt;p&gt;The community group is named after a particular implementation which is actually not super suited to the web at large. Is the intent for this to be a group that defines new pubsub extensions for the web, or one that promotes pubsubhubbub?&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Fri, 13 Jan 2012 15:00:08 -0000</pubDate></item><item><title>Re: The Potentially Dark Future of Search</title><link>http://metajack.im/2012/01/12/the-potentially-dark-future-of-search/#comment-409085115</link><description>&lt;p&gt;Of course they have access to Youtube, that was my point. No one else does. Google is just as guilty as anyone in regards to keeping data from others.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Thu, 12 Jan 2012 19:15:36 -0000</pubDate></item><item><title>Re: The Potentially Dark Future of Search</title><link>http://metajack.im/2012/01/12/the-potentially-dark-future-of-search/#comment-408807578</link><description>&lt;p&gt;I agree that firehoses aren't the be all end all of data access, but it's still true that targeted subscriptions take engineering effort and other resources. I wasn't trying to promote a specific kind of data access; I was just using firehoses as the most obvious example, and probably the one most relevant to large search engines.&lt;/p&gt;&lt;p&gt;Maybe we will be saved by more efficient data access methods in the future, although I don't think those will help with companies desire to hoard their data to keep it from others.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Thu, 12 Jan 2012 13:24:23 -0000</pubDate></item><item><title>Re: Cross-Domain AJAX for XMPP HTTP Binding Made Easy</title><link>http://metajack.im/2010/01/19/crossdomain-ajax-for-xmpp-http-binding-made-easy/#comment-406778525</link><description>&lt;p&gt;I don't know how to help further as both &lt;a href="http://bosh.metajack.im" rel="nofollow noopener" target="_blank" title="http://bosh.metajack.im"&gt;http://bosh.metajack.im&lt;/a&gt;:5280/xmpp-httpbind and &lt;a href="http://bosh.metajack.im" rel="nofollow noopener" target="_blank" title="http://bosh.metajack.im"&gt;http://bosh.metajack.im&lt;/a&gt;:5280/crossdomain.xml load fine for me here.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Mon, 09 Jan 2012 22:04:55 -0000</pubDate></item><item><title>Re: Cross-Domain AJAX for XMPP HTTP Binding Made Easy</title><link>http://metajack.im/2010/01/19/crossdomain-ajax-for-xmpp-http-binding-made-easy/#comment-406328106</link><description>&lt;p&gt;It was down briefly a few days ago, but should be working fine now. Please try again.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Mon, 09 Jan 2012 11:10:03 -0000</pubDate></item><item><title>Re: XMPP Questions Answered: Pubsub versus Multi-User Chat</title><link>http://metajack.im/2010/01/15/xmpp-pubsub-versus-multiuser-chat/#comment-392037748</link><description>&lt;p&gt;You can put your machine readable payload right in the &amp;lt;message&amp;gt; stanzas. There's no requirement in XMPP that those live in IQ stanzas.&lt;/p&gt;&lt;p&gt;For example, Chesspark sent out messages with chess moves in normal room directed &amp;lt;message&amp;gt; stanzas with no &amp;lt;body&amp;gt; children, just application-specific metadata.&lt;/p&gt;&lt;p&gt;It doesn't make much sense to have the request-response semantics of IQ to a broadcast group without knowing how you want responses to be processed (eg by consensus, by wait-for-all, by wait-for-first, etc).&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Thu, 22 Dec 2011 01:20:10 -0000</pubDate></item><item><title>Re: XMPP Questions Answered: Pubsub versus Multi-User Chat</title><link>http://metajack.im/2010/01/15/xmpp-pubsub-versus-multiuser-chat/#comment-280388713</link><description>&lt;p&gt;I'm not sure what you mean when you say they are added but immediately removed. What do the clients see? Are they getting disconnected? Kicked? Banned?&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Mon, 08 Aug 2011 09:12:44 -0000</pubDate></item><item><title>Re: Fastest XMPP Sessions with HTTP Pre-Binding</title><link>http://metajack.im/2009/12/14/fastest-xmpp-sessions-with-http-prebinding/#comment-204294176</link><description>&lt;p&gt;I don't know of one, sorry.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jack Moffitt</dc:creator><pubDate>Sun, 15 May 2011 22:30:49 -0000</pubDate></item></channel></rss>